VOLUME 79

JCD
JOURNAL OF COUNSELING & DEVELOPMENT
NUMBER 3 SUMMER 2001

PRACTICE & THEORY

259 Rational Emotive Behavior Therapy Successes and Failures: Eight Personal Perspectives
Stephen G. Weinrach, Albert Ellis, Catharine MacLaren, Raymond DiGiuseppe,
Ann Vernon, Janet Wolfe, Ruth Malkinson, and Wouter Backx

269 Metaphor and Change in Counseling

William J. Lyddon, Alison L. Clay, and Cheri L. Sparks

275 FOCUS on the Family Caregiver: A Problem-Solving Training Intervention

Monica F. Kurylo, Timothy R. Elliott, and Richard M. Shewchuk

282 Where Has Our Theory Gone? Learning Theory and Intentional Intervention
Sterling Gerber

RESEARCH
292 Clinical Judgment in Case Conceptualization and Treatment Planning Across Mental
Health Disciplines
Janet E. Falvey

304 Counselors’ Attribution of Responsibility, Etiology, and Counseling Strategy

Jerry L. Kernes and J. Jeffries McWhirter

314 Risk Factors for Suicidality Among Clients With Schizophrenia

Robert C. Schwartz and Benjamin N. Cohen

320 Helping Seventh Graders Be Safe and Successful: A Statewide Study of the Impact
of Comprehensive Guidance and Counseling Programs
Richard T. Lapan, Norman C. Gysbers, and Gregory F. Petroski
331 High School Students’ Career-Related Decision-Making Difficulties
Itamar Gati and Noa Saka

341 Assessing the Impact of a Prejudice Prevention Project

Michael Salzman and Michael D’Andrea

347 Training AIDS and Anger Prevention Social Skills in At-Risk Adolescents
Melbourne F. Hovell, Elaine J. Blumberg, Sandy Liles, Linda Powell, Theodore C. Morrison,
Gabriela Duran, Carol L. Sipan, Susan Burkham, and Norma Kelley

356 Adolescent Strategies for Coping With Cultural Diversity

Hardin L. K. Coleman, Sherry B. Casali, and Bruce E. Wampold

365 Rational Suicide: An Empirical Investigation of Counselor Attitudes

James R. Rogers, Christine M. Gueulette, Jodi Abbey-Hines, Jolynn V. Carney, and
James L. Werth, Jr.

373 The Impact of Group Work on Offender Adolescents

Linda L. Viney, Rachael M. Henry, and Joanne Campbell

TRENDS
382 Finding Happiness for Ourselves and Our Clients
Geri Miller

© 2001 by the American Counseling Association. All rights reserved.

 PRACTICE & THEORY
Rational Emotive Behavior Therapy Successes and
Failures: Eight Personal Perspectives

Stephen G. Weinrach, Albert Ellis, Catharine MacLaren, Raymond DiGiuseppe, Ann Vernon, Janet
Wolfe, Ruth Malkinson, and Wouter Backx

Eight experts in Rational Emotive Behavior Therapy (REBT) provided personal examples of their own successes and failures in
applying REBT to themselves. The experts actively talked to themselves both rationally and irrationally. Understandably, there
were far more shoulds, oughts, musts, and have to’s in the narratives in which the experts described when they failed to use REBT
than when they succeeded in using REBT. Rational self-talk was more prevalent in the examples of how REBT was successfully
used by the experts.

T
o its followers, Rational Emotive Behavior
Therapy (REBT) is an excellent theory of men-
tal health and philosophy of life. If anyone could
be expected to have mastered its finer nuances,
and thus be less prone to overreacting emo-
tionally or behaviorally, it would be REBT experts. Would
it not be reasonable to expect REBT experts to “practice
what they preach?” Yet common sense dictates that even
experts are unlikely to be exempt from overreacting emo-
tionally or behaviorally some of the time. After all, even
experts are fallible human beings. The purpose of this
project was to provide a panel of REBT experts with the
opportunity to reveal where they succeeded and where they
failed at applying REBT to themselves.
GUIDELINES
In April 1999, the 11 members of the Albert Ellis Institute’s
International Training Standards Committee were invited to
write two essays each. In the first essay, committee members
were asked to provide a personal example of how they had
been successful in using REBT on themselves. In the second
essay, committee members were asked to provide an example
of how they failed in using REBT on themselves.
The Committee comprised experts in REBT from around
the world. The committee, of which REBT’s founder Albert
Ellis is a member, has an international membership with repre-
sentatives from Australia, Germany, Israel, the Netherlands, and
the United States. Former members have come from Canada,
England, and France. The committee serves several functions,
including the approval of REBT training programs around the
world. Current and past committee members are among the
world’s experts on REBT.
Specifically, each committee member was asked to do the
following:
Describe in 400 words or less, one example of where you have
successfully invoked REBT in your own life. Include, where
applicable, rational and irrational self-talk. Label your feelings.
Indicate how long it took you to be successful. Indicate the
extent to which your use of REBT in this situation has been
generalized to similar situations (elegant solution). [The term
“elegant solution” refers to the search for profound philosophic
changes in the client’s thinking.] If you had help in applying
REBT to yourself (as opposed to using REBT on yourself with-
out seeking help), so indicate. What did the helper do that you
found helpful. Please be very specific. Do not write in general
terms about how REBT has changed your life.
and,

Stephen G. Weinrach is a professor of counseling and human relations at Villanova University, Villanova, Pennsylvania, and a Fellow of the Albert Ellis
Institute in New York. Albert Ellis is the founder and president, Catharine MacLaren is a training and special projects coordinator, and Janet Wolfe is the
executive director, all at the Albert Ellis Institute in New York. Raymond DiGiuseppe is a professor of psychology at St. John’s University, Jamaica, New York.
Ann Vernon is the graduate coordinator for the School Counseling Program at the University of Northern Iowa, Cedar Falls. Ruth Malkinson is a counselor
at the Israeli Center for REBT, Rehovot, Israel. Wouter Backx is a counselor at the Institute voor RET, Haarlem, Netherlands. Correspondence regarding this
article should be sent to Stephen G. Weinrach, Department of Education and Human Services, Villanova University, 800 Lancaster Avenue, Villanova, PA
19085-1699 (e-mail: stephen.weinrach@villanova.edu).

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 259

We i n r a c h , E l l i s , M a c L a r e n , D i G i u s e p p e , Ve r n o n , Wo l f e , M a l k i n s o n , a n d B a c k x
Describe in 400 words or less, one example of where you tried but
were unsuccessful in a timely manner to invoke REBT in your own
life. Include, where applicable, rational and irrational self-talk. La-
bel your feelings. Fully describe the dysfunctional or irrational be-
havior. Indicate the obstacles that you encountered in using REBT.
What were you continually telling yourself? If you sought help in
applying REBT to yourself (as opposed to using REBT on yourself
without seeking help), so indicate. How come counseling didn’t
help, either? Please be very specific. Do not write in general terms.
Between May and December, 8 committee members gen-
erated 15 responses. Three committee members chose not
to participate. Most essays were returned to their respec-
tive authors with requests for revisions by the first author,
who also served as compiler and editor. All essays were copy
edited so that there was some semblance of uniformity and
consistency across essays. Some key words, phrases, and con-
cepts, commonly used by REBT counselors have been itali-
cized so as to sensitize readers to rational self-talk as well as
irrational self-talk. Word limits were not strictly enforced.
The order of authorship, and therefore the order in which
the responses are presented, as follows, was based on the
order in which the first drafts were received. In the past,
then current members of the committee have participated
in writing two other articles (Weinrach, 1996; Weinrach et
al., 1995). This is the third such article.
REVEALED PERSPECTIVES: COMMONALITIES AND DIFFERENCES
There is considerable emotional energy found in many of
the narratives. These are not the narratives of individuals
with very low or flat affect. To the contrary, some of the
experts displayed considerable anger and frustration in those
areas where they failed to apply REBT to themselves as
well as satisfaction and contentment where they succeeded.
The experts used many of the same words and phrases
common to REBT. They actively talked to themselves both
rationally and irrationally. It is interesting that none cited verba-
tim, or even close to verbatim, the well-known irrational
beliefs that are found in most standard counseling theory
textbooks. An example is “It is a dire necessity for adult hu-
mans to be loved or approved by virtually every significant
other person in their community” (Ellis, 1994, p. xx). Rather,
the experts, who presumably have been practicing REBT on
themselves for many years, have adapted the standard irratio-
nal beliefs to fit their own idiosyncratic belief systems or just
shortened them for the sake of convenience. Understandably,
there were far more shoulds, oughts, musts and have to’s in the
narratives in which the experts described when they failed to
use REBT than in those describing when they succeeded. Ra-
tional self-talk was more prevalent in the examples of how
the experts succeeded in using REBT—the type of rational
self-talk that contributes to positive counseling outcomes.
There are several themes that are found in both the nar-
ratives in which the experts succeeded in applying REBT
to themselves and those in which they failed. One expert’s
area of success was often another expert’s area of failure.
For example, Weinrach succeeded in using REBT to man-
age his anxiety, whereas Vernon failed to manage hers. Ellis
260 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
and Wolfe succeeded in using REBT in the area of uncondi-
tional other-acceptance, yet Malkinson and Weinrach failed
in managing their issues of unconditional self-acceptance.
DiGiuseppe cited low frustration tolerance (LFT) as an
example of how he succeeded (and also failed), whereas
Ellis and Wolfe cited LFT as an area where they both failed.
Backx’s narrative about how he was successful in applying
REBT to himself was not classified because of its general
nature. The assignment of narratives into specific themes
was not precise because some of the narratives could have
been assigned to more than one theme. The classification
of narratives by theme was made primarily on the basis of
self-talk as opposed to the experts’ description of the sur-
rounding circumstances.
Walen, DiGiuseppe, and Dryden (1992) subsumed Ellis’s
irrational beliefs into four categories, which they referred
to as “core irrational beliefs.” The categories are “demands,
awfulizing, low frustration tolerance, and global evaluations
of human worth” (p. 17). This four-category model was
applied only to the narratives in which the experts described
their failures at applying REBT to themselves. (It was not
applied to the narratives wherein the experts succeeded
because there was considerably less evidence of irrational
thinking in these examples.) Two of the experts (Weinrach
and Malkinson) dealt primarily with issues of global evalu-
ations of human worth, and three of the experts (Ellis,
DiGiuseppe, and Wolfe) experienced LFT, primarily.
MacLaren’s narrative about depression was not classified
because its origin seems to have fallen primarily outside of
REBT’s sphere. It serves as a cogent reminder that relying
solely on one approach, in this case REBT, may not always
be in the client’s best interest.
Finally, more than one core irrational belief was found
within most of the narratives, which is often the case with
clients. For example, Weinrach dealt with both global evalu-
ations of (his) human worth and the demand that his col-
league offer greater respect for the importance of his work.
Ellis displayed classic LFT and demanded that the airline
treat him better. DiGiuseppe experienced LFT (by display-
ing an intolerance for discomfort) and demanded that he
should not have to use an appointment book. Vernon de-
manded a certainty of the outcome of her husband’s sur-
gery as well as “awfulized” about what life would be like
for her son if her husband were to die. Wolfe displayed
LFT along with global evaluations of human worth of her
employees; Malkinson dealt with global evaluation of (her)
human worth and demanded that she be treated better.
SUCCESSES IN APPLYING REBT
In the following section, the experts provided examples of
how they succeeded in applying REBT to themselves.
Anxiety Reduction (Stephen G. Weinrach)
In August of 1998 I had coronary bypass surgery. It is com-
mon for bypass surgery patients to complain more about
all of the tubes in their mouth and nose than about pain

from the surgery itself. The tubes and oxygen mask are
suffocating. Although the surgery went well, there were
serious problems with my respiration immediately there-
after. Pulmonary therapists have this funny habit of want-
ing their patients to breathe on their own, and I was not
doing a very good job of it. At one point, my pulmonary
therapist, whom I shall call Debbie, after trying to moti-
vate me to breathe more independently, calmly informed
me that if my output did not improve significantly within
the next hour, she would have no choice but to insert a
nasotracheal tube. The tube is inserted in one’s nose and
wends its way down to one’s lungs. Once in place, mu-
cous is suctioned out of the lungs.
I fully understood Debbie’s directions about inhaling and
exhaling, and I was willing but unable to comply. There
was a cognitive disconnect between comprehending and
acting. For example, I was unable to imitate someone show-
ing me how to cough. Debbie, in her own charming and
nonthreatening way, calmly made it clear to me that I was
not following her directions. I fully understood the impli-
cations of her threat to insert the tube—something I did
not want to take place. I was not particularly frightened
that I was not breathing adequately because I was unaware
of its serious implications. I was frightened that inserting
the tube would be very painful.
As the hour elapsed, I found myself remarkably and un-
characteristically calm about the prospect of the tube be-
ing inserted. I was alert enough to talk to myself rationally. I
told myself that “I was powerless to change whatever the
ultimate outcome was” (tube or no tube) and that “were
the tube to be inserted, it would take only a minute or two
and that I could stand a lot of pain for that short amount of
time. That was the worst thing that could realistically hap-
pen.” The prospect of dying had never occurred to me.
Throughout the hour, Debbie coached and encouraged me
to inhale and exhale to no avail. At the appointed hour, she
told me to swallow once she got the tube halfway down. I
relaxed, swallowed, and the tube was painlessly inserted. I
was very excited that I could invoke REBT under the cir-
cumstances. Its use helped me handle a situation that I had
found very threatening. At the time, I remember being more
elated that I could rely on invoking REBT while lying in
the cardiothoracic intensive care unit than I was upset with
the necessity to insert the tube. I may have forgotten how
to breathe, but I did not forget my REBT.
A month after my initial discharge from the hospital I
was readmitted for 2 more weeks, during which time I
had two additional surgeries and the temporary insertion
of a port in my arm so that intravenous antibiotics could
be self-administered 3 times a day for the next 45 days at
home. For this procedure, I was not sedated and was even
more anxious than before, with some justification. Then,
every few weeks, I had to have blood tests to determine if
the antibiotics were damaging my liver. The blood needed
to be drawn from the top of my hands because I had an
artery removed from my left arm for the bypass surgery
and I had the temporary port in my right arm. These blood
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Rational Emotive Behavior Therapy Successes and Failures
tests were far more painful than anything I had experi-
enced previously and at times I would cry afterwards. In
all three situations, I was successful in calming myself
down to the point where I was merely appropriately ap-
prehensive. I was able to tell myself “I may not like it, but
I can stand it. I have stood it before so I know that I can
stand it again.”
Unconditional Other-Acceptance (Albert Ellis)
I used to be quite angry at people who acted stupidly or
immorally and at horrible world conditions. My use of REBT
on myself, when I started to apply it to others in January
1955, convinced me that my angry feelings were largely
created by my absolutistic demands that people must treat
me fairly and considerately and were rotten people when
they did not. I taught many of my clients unconditional
other-acceptance (UOA), to accept the sinner and not his
or her sins, and thereby refuse to feel angry.
I got a good chance to put my unconditional other-
acceptance into practice in 1993, when I was celebrating
my 80th birthday. For several years before that, I had
been collaborating on a Dictionary of Sexological Terms
with B., with whom I had maintained a close friendship
since our college days. I gave him considerable material
that I had already gathered for the Dictionary, and he
promised to complete it in, at most, a year and by doing
so get some credit on his PhD in philology. B. did little
work on the Dictionary, kept lying to me and our pub-
lisher, and completed the work he had to do on his PhD.
The publisher rightly canceled the contract for the book
after 3 years of B.’s abysmal procrastination; the Dictio-
nary was never published.
Under ordinary conditions, before I began to use REBT
on myself, I would have been incensed for several reasons:
(a) B. was slothfully procrastinating, (b) he kept lying about
the work he was doing, (c) he had me write a special letter
to his Philology Department saying that he was working
hard on the Dictionary, and (d) I had agreed to forgive him
$8,000 he owed me when he would finish his work on the
Dictionary. But after the publisher canceled our contract,
he invented an outrageous lie to the effect that I had for-
given the debt just for his starting to work on the Dictio-
nary. So he canceled the debt.
Instead of making myself incensed at B., I used REBT to make
myself feel very displeased with his behavior but not angry
with him. How? By telling myself, “Too bad that B. has lied,
cheated, and procrastinated, but that rotten behavior doesn’t
make him a bad person.” I no longer viewed him as a close
friend but was not angry at him. I invited him to my 80th
birthday party in 1993, and we talked in a pleasant manner,
reminisced about some of our mutual experiences over the
years, and semihumorously mentioned our monetary differ-
ences. No, no feelings of anger, rancor, or resentment on my
part. Using REBT and its philosophy and practice on myself, I
accepted B. with his poor behavior and I rarely think any more
of the highly immoral way he treated me.
261
We i n r a c h , E l l i s , M a c L a r e n , D i G i u s e p p e , Ve r n o n , Wo l f e , M a l k i n s o n , a n d B a c k x
Former “Love Slob” (Catharine MacLaren)
This is the confession of a former love slob. I used to be-
lieve that I could feel positive about myself only if almost
everyone else on the planet approved of me. It did not matter
what it cost me. I now know that attempting to attain that
goal is not only exhausting but impossible and only leads to
frustration, resentment, and poor self-care. Unfortunately,
it took me a little time and a lot of REBT to figure that out.
In my late adolescence and early 20s, I noticed that I was
often trying to “fix” things for people so that they would
not have to deal with them. If I was incapable of “making it
better,” or, worse yet, happened to do something that con-
tributed to someone’s stress (even if it was in my best inter-
est), then I would berate myself for days or months and feel
miserable and guilty. Phrases like I’m a terrible friend (per-
son) and I don’t deserve to (fill in the blank) would flow
through my mind endlessly.
In my mid-20s, I started to practice REBT. I would listen
to clients at the beginning of my Fellowship at the Albert
Ellis Institute and think, “that sounds a little bit like me.”
Then one night in group, after a member had been lament-
ing the sacrifices she was making for a friend, Al Ellis turned
to her and said, “You’re a love slob of the worst sort and
you’d better strongly work to get over it and start accepting
you for you instead of for what you can do for other people.”
Something clicked for me.
I poured over all the information I could find on love
slobbism, guilt, and global self-rating. I cut out particularly
relevant small sections of resources and pasted them into
my planner as daily reminders that looking out for myself
was acceptable and only made me self-interested, not self-
ish. I monitored my constant internal monologue by set-
ting up cues at home and in my office. When I noticed
these cues (small familiar articles placed in innocuous
places), I stopped myself immediately and “rewound” the
tape in my head to pinpoint unhelpful thinking.
Once I identified the negative messages I was sending
myself (without really being mindful of them), I disputed
them with REBT self-help forms and reviewed the forms ev-
ery day. I developed and consistently practiced coping state-
ments like I am allowed to say “No” and It doesn’t make me a
bad person if I can’t accommodate everyone. I deliberately
pushed my comfort limits in terms of asking for things from
others to prove that I could do it and still maintain my rela-
tionships with them. Slowly, over a period of several months
and with lots of hard work, I began to experience a shift and
feel more comfortable asking for what I wanted (and did not
want) instead of trying to provide others with what I thought
they wanted. My confidence in myself increased dramati-
cally and my relationships with others improved. I am still a
work in progress and have weak moments when I despair at
disapproval from certain people, but I am quick to take ac-
tion against that process and remind myself that others will
inevitably disapprove of me some of the time and that is O.K.
I now have significantly more energy to devote to other, more
productive endeavors.
262 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Low Frustration Tolerance (Raymond DiGiuseppe)
Presently I have two young children, ages 20 and 36 months,
and many writing obligations. Since the birth of these chil-
dren, I almost never have enough time to do my profes-
sional writing activities. Almost every time I sit down to
write, I am distracted by one or both of them, or I have
family activities that preclude my writing. But I remind
myself that “things don’t have to go my way, I don’t always
get what I want, and I can tolerate not having my own way.”
Now I can be pretty happy and content and not upset.
I did, however, spend about 6 months being upset, vacil-
lating between being angry and depressed. When I felt angry,
I was telling myself that “I should be able to do whatever I
want to do, and my wife should do the child care. It’s unfair
to me and I’ve got to have my way.” And I felt depressed when
I thought, “It’s awful that I’ll never get what I want done.
Child care is an unfair restriction on my personal choice.
Everyone will forget who I am because I’m not publishing
anything any more, and I’ll just be a mediocre nobody.”
I disputed these beliefs by telling myself “I don’t have to
have what I want! There may be other things I can enjoy.
There’s more than one way to enjoy your life, and there’s no
one absolute way you have to live your life. If I never have a
professional publication again, I’ll still enjoy Thomas and
Anna. And I guess I won’t be a mediocre nobody.”
Regret (Ann Vernon)
I employed REBT effectively in a situation that occurred
this past year. My 21-year-old niece has a very serious case
of Crohn’s Disease and was hospitalized at the Mayo Clinic
for more than 3 months, during which time she had several
operations, including a colostomy. While she was in the hos-
pital, my family and I visited several times, called regularly,
and sent cards and gifts to show our support.
About 4 months after the colostomy, my niece returned
to college, and we began to exchange e-mail. After my ini-
tial inquiry about how she was feeling and adjusting to be-
ing back in school, I shared my thoughts about all the
changes she had been through recently and specifically asked
how she was adjusting to the colostomy. I asked this ques-
tion out of concern, assuming that there would be signifi-
cant physical as well as psychological ramifications.
The next e-mail was from my sister who was furious that
I had asked such a direct question. She adamantly stated
that my niece wanted no contact with me for a long time
and that what I had done was extremely insensitive and
outrageous. I was dumbfounded and hurt by her response,
but I did not take the attack personally. I told myself that
they were overreacting because they had been through so
much with this illness and that I was probably a scapegoat
for their anger about the entire situation.
I e-mailed an apology, expressing that I had not inten-
tionally been insensitive but in fact felt that I was being
thoughtful by expressing concern about an adjustment that
could be very difficult for a 21-year-old. I also shared that I

had asked similar questions of two clients who had had the
same operation and that they had appreciated such a direct
inquiry because it helped them talk honestly about the
issues. My sister’s response to this message was that I must
be a terrible therapist.
Once again, I did not take this personally. I was very dis-
appointed by her comments but had not intended to be in-
sensitive and realized I could not control their reaction. I
knew I was not a terrible therapist, nor was I a rotten person
for making a comment that was misinterpreted. I did not
get angry, thinking that my sister shouldn’t react as she did,
even though I would have preferred a more mature response
without her personal attacks. By thinking rationally, I was
able to stop this from escalating into a bigger argument.
Because I stayed calm, I could assertively request that she
accept my apology and that we put this issue behind us so
as not to jeopardize our relationship. Although I do not
think I made a mistake by asking such a direct question, I
am not perfect and certainly had no way of predicting their
response. I can regret my behavior but not beat myself up
about it, thanks to REBT.
Unconditional Other-Acceptance (Janet Wolfe)
Having been brought up with the quaint notion that when
I got mated, I would have someone with whom to walk on
moonlit beaches, travel, and do other activities, it was quite
jolting for me to realize at age 25 that I had cast my lot
with a “workaphile” with just about zero desire for socializ-
ing and who is too busy for nature, culture that does not
come piped in through the radio or television, or other
things most so-called normal couples do together. In fact,
he and his former wife separated for that very reason, he
pointed out before I moved in with him some 30 years ago.
For the first few years—although happy when in his com-
pany and generally pleased with a lot of his behavior (for
example, his tremendous encouragement of my personal
and career goals and his utter lack of male chauvinism)—I
would periodically find myself in tears because every time
I wanted to travel or do something with him, I would wind
up having to recruit someone else or go by myself. I believe
that my mate should want to spend time with me, and it is
really awful that he does not (and, if he does not, I must not
be very worthwhile). Although busy with my own work, in-
terests and friends, I would periodically feel depressed, self-
pitying, and angry and perorate at him for being so selfish
and unloving (to which he replied, to my further furor,
“that’s the way I am—now what are you telling yourself to
get yourself so upset?”).
Many years later, I find myself with just the opposite feel-
ings. Barely a day goes by that I do not congratulate myself
for really having the best of both worlds, the single and the
mated. As an independent agent, I can go where I want,
when I want, and with whom I want (with no flack from
my mate, because he is happily working away in his own
space). Unlike most women who are more traditionally
coupled, I spend nice chunks of prime time with close
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Rational Emotive Behavior Therapy Successes and Failures
friends, which has meant that my friendships have deep-
ened and grown rather than tapered off. And I also have a
tremendous amount of time alone in the beautiful space I
have created because my partner conveniently spends most
of his waking hours in his office. And because he recog-
nizes our home as largely my space (because he mainly sleeps
or grabs a quick meal there), I am probably one of the few
women who has been able to hang 40 yards of Victorian
lace around her bed without a peep of objection from her
mate! During the 10–15 minutes a day we spend in the
same space we giggle, mime, occasionally discuss profes-
sional matters, or say nothing at all, and generally have a
fine time.
My success in this area is not completely due to my hav-
ing reframed something in my head from awful to highly
advantageous; it is also because I discovered empirically that
the way relationships are traditionally supposed to be is, in
the last analysis, really not the kind of relationship that is
best for my own growth and happiness. However, were it
not for my having done a lot of disputing, I probably would
not have stayed around long enough to find this out.
Anxiety Reduction and Shame Attacking (Ruth Malkinson)
Two years ago, I was on my way to a memorial service for
my parents. I went with my son who drove our car, and we
stopped at a nearby plant nursery to get some flowers and
plants. The nursery was located next to a gas station where
my son checked the air in the car’s tires. On my way back
to the car, holding the plants in my hands, I slipped on an
oil spill and fell into a ditch. I was in great pain. My son,
who saw me approaching the car and falling, ran quickly to
help but realized that I was badly hurt. Although I was in
great pain, I kept telling myself that I could stand the pain
and that it was not awful.
I was hospitalized with a broken elbow and had to un-
dergo surgery. During those few days in the hospital I could
not move, felt miserable, and was in great pain. Also, for the
first time in my life I was going to have surgery under gen-
eral anesthesia, of which I was terribly afraid. I heard my-
self saying that I couldn’t stand the pain and I was miserable
for having to go through the discomfort of being in bed
with a broken elbow. I immediately told myself that I could
stand the pain, in spite of the discomfort of being in bed
and unable to function. I found that rational self-talk was
very helpful in dealing with my pain and discomfort.
But the great moment of which I knew very little was yet
to be experienced. I was unaware of the effect of the anes-
thesia, especially its aftereffect. I remember hearing some-
one call my name over and over again and not understand-
ing why I was being shouted at. I felt very “frightened and
angry” with whomever it was who was shouting at me. I
remember refusing to respond to the voice that called my
name. Shortly afterward, I opened my eyes and felt a com-
forting touch—my daughter was standing next to me and
explaining that the nurse had been rather harsh in awaken-
ing me. I felt very embarrassed for what seemed to me a
263
We i n r a c h , E l l i s , M a c L a r e n , D i G i u s e p p e , Ve r n o n , Wo l f e , M a l k i n s o n , a n d B a c k x
noncompliant behavior on my part and proceeded to put
myself down for behaving childishly. But not for very long,
as I told myself that I have never before experienced any-
thing similar to that and there is nothing wrong in reacting
the way I did. Also, my anger toward the nurse subsided
when I reminded myself that she was only doing her job,
even though it was not in a particularly nice way. “It is defi-
nitely not the end of the world,” I told myself. I was able to
accept myself as a fallible human being.
Self-Help Group (Wouter Backx)
Some 8 years ago, two very good friends of my partner and
me wanted to have some psychological support but did not
know how to proceed. When we discussed their problems,
it became clear that their problems were rather similar to
those that my partner and I sometimes experienced. Some
of the difficulties my female partner had and the woman in
the other couple had were particularly similar, which was
quite understandable because they are sisters.
We started to explore solutions very tentatively. We came
up with the idea of forming a self-help group. We wondered
how to create an egalitarian group even though, initially, I
was the one who had the expertise in REBT. The four of us
developed a set of rules whose purpose was to prevent us
from becoming a group of three clients around one thera-
pist—me. I was convinced that the other three group mem-
bers would be able to learn enough REBT to make it pos-
sible to have three (nonprofessionally trained) therapists
working with one client, every time. Even better yet would
be three clients working on the problem of a fourth client.
We aspired to replicate the traditional model of REBT groups
in which members help each other.
The beginning was difficult because we did not know
what would happen and how it would go. We made some
mistakes like first having a nice dinner together followed
by our group session. That did not work. It was better to
separate our social activities strictly from the therapy.
But at some point the group really began to click. All
four of us brought to the group our problems with our
respective partner or more often with somebody or
something outside the group. In the beginning, every-
body took turns, and we came together on a regular basis.
But later, when we were more used to it, one person
could just ask for a session with the group for some
problem that was bothering him or her. As a result we
began to meet less often.
Our group has been functioning this way for the past 7
years. It has effectively helped to resolve personal prob-
lems and marital crises. As a group, we have become very
sophisticated in applying REBT. As such, the group has
become a very safe place to bring in my problems and work
on them, a luxury therapists do not always have. The group
has also served to prevent problems before they arise. By
applying what we do in the group in problematic situa-
tions we are able to reduce much of the unnecessary suffer-
ing we would experience otherwise.
264 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
FAILURES IN APPLYING REBT
In the following section, the experts provided examples of
how they failed at applying REBT to themselves.
Need for Unconditional Self-Acceptance
(Stephen G. Weinrach)
The individual faculty teaching load, where I work, is largely
controlled by the recommendations of a faculty committee to
the dean. The standard teaching load is four courses per
semester. For many years I have been granted a one-course-
per-semester reduction in load because of my scholarly
productivity. In 1997, I decided to apply for a two-course-
per-semester reduction in teaching load, which is rarely granted.
At the time, I had just completed an article titled “Publication
Patterns of the Personnel and Guidance Journal/Journal
of Counseling and Development—1978 to 1993” (Weinrach,
Lustig, Chan, & Thomas, 1998). My productivity not only
placed me in the top 1% of contributors for the 15-year
period of the study, but it had even increased appreciably
between 1993 and 1997.
My application to the committee included evidence of
past performance and an outline of what I planned to do in
the future. It is impossible to reconstruct all of the discus-
sion that took place at my meeting with the committee,
and I was not privy to their private deliberations. However,
while discussing my level of scholarly output, one member
of the committee, after I referred to my being in the top
1% as measured by the study I had conducted, moved his
thumb and index finger a few inches apart and stated: “But
at Villanova University, your performance is only average.”
Aside from the statement being patently inaccurate, I was
irrationally angry that this committee member disparaged
my work, and indirectly me. It reflected a condescending
attitude, not uncommon among some faculty toward col-
leagues in a different department.
For about a year, I spent an inordinate amount of time
unsuccessfully trying to dispute my irrational anger at the
person who made this statement. I was far more angry at
the inaccurate characterization of my life’s work than I was
upset about not having my teaching load further reduced. I
know that my worth is unrelated to my productivity. I know
that his saying something does not make it so. I know that life
is not fair. I know that I was acting like a big baby—demand-
ing that others think highly of me and my work (approval).
I know. I know. I know. But for the longest time I continued
to ruminate about it just the same.
Then one day my anger mysteriously disappeared. I freely
tell myself now that “people make inaccurate and insensi-
tive statements all of the time, and it is up to me whether I
choose to dwell on them or not.” I do not believe my being
able to let go was primarily a function of my finally using
REBT on myself (at least at any conscious level) because
my anger vanished weeks before my rational self-talk
emerged. Even if REBT could be appropriately credited with
my finally letting go, letting go took a very long time, which

when I think about it, is not all that uncommon for me.
Maybe had I sought help from an REBT counselor, I would
have been able to let go sooner. A case could be made that
I really have not yet fully let go because I am publicly bring-
ing attention to this matter, which could conceivably cause
the object of my discomfort some discomfort himself. The
issue that I have presented is not a rare event; there are
other unresolved examples that, if space permitted, I could
have used.
Low Frustration Tolerance (Albert Ellis)
Normally, I have fairly high frustration tolerance (HFT),
which I have increased over the years by using REBT on
myself. Witness my publishing over 65 books and some 700
articles, giving over 80 professional and public workshops a
year in New York and throughout the world, regularly su-
pervising 10 interns and therapists in REBT, and having ses-
sions every week with about 70 individual and group
therapy clients. Pretty good for an old man! In fact, in this—
my 86th year—I shall probably publish 4 or 5 new books
and a dozen or more articles. So I hardly give in to my
potential low frustration tolerance and procrastination.
Nonetheless, I still at times suffer from abysmal LFT—
and my use of REBT almost always removes it. Take, for
example, what happened to me in August 1999. I was sched-
uled to give workshops from 10:00 AM to 5:00 PM in Colo-
rado Springs. To make sure I arrived in time, I planned on
taking one of the largest airlines from New York directly to
Colorado Springs the afternoon before. We were scheduled
to leave New York at 5:00 PM, stop over briefly in Dallas,
and then arrive at our destination at 9:00 PM. I deliberately
took this flight because it was the only one that went, with
an interim stop, directly to Colorado Springs. So, presum-
ably, I could not miss getting there Thursday night, in good
time to be fresh for my workshops on Friday.
No such luck. The airline never told us passengers what
happened to our plane, but after 2 hours of its getting lost
somewhere, they finally got us on a substitute plane that
was to leave LaGuardia airport at 7:00 PM. A thunderstorm
occurred while we were about to take off. All plane traffic
was halted, and we finally left New York at 9:00 PM—a
good 3 hours late.
The pilots did their best, but by the time we arrived in
Dallas, it was 12:00 AM local time, so the airline—quite
unethically, I thought—canceled our continuing flight to
Colorado Springs and we were stuck for the night in Dal-
las. I explained to the airline agent that I had to start my
workshop in Colorado Springs at 10:00 AM the next morn-
ing and therefore had to have an early flight to that city.
Whereupon he lied to me, said there was no early morning
flight, and that I would have to take an 11:00 AM flight
and arrive, at the earliest, at 12:15 PM—long after my work-
shop was scheduled to begin. Actually there was a 6:30 AM
flight from Dallas to Colorado Springs, but it was on a rival
airline and he wrongly told me it did not exist. So he put
me up at a “flea-bag” hotel for the night in Dallas. Not only
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Rational Emotive Behavior Therapy Successes and Failures
was the hotel third-rate, but about a hundred small ants
kept falling on the floor of my bathroom and I spent a good
deal of the night killing them.
Fortunately, I called the people for whom I was giving
the workshops in Colorado Springs; they got on the phone
to the airlines, and I was able to get the 6:30 AM flight for
Denver and then on to Colorado Springs. With only about
3 hours of sleep on Thursday night, I arrived at my confer-
ence at 10:30 AM. I still missed some of my workshop time,
but at least I gave most of my presentations.
To make matters still worse, I was supposed to return on
Friday night, after my workshops, to New York on the same
delinquent airline that had unethically given me so much
trouble on Thursday. The airline only allowed 45 minutes
for me to connect in Denver with my New York flight,
however my flight to Denver was a half hour late. I had to
run a fantastic distance at the Denver airport to catch my
New York flight. I and my bedraggled luggage finally made
the New York flight, just as they were closing the doors to
take off.
I was convinced that this famous American airline was
reprehensibly responsible for my frustration, because (a)
my original plane got unexplainably lost; (b) we were there-
fore 2 hours late in starting; (c) we consequently sat on the
field another hour waiting for a thunderstorm to stop; (d)
we were 3 hours late to Dallas; (e) they unethically can-
celed our duly scheduled flight to Colorado Springs—for
no good reason that I could see; (f) their agent lied to me
about there being no early flight Friday morning to my des-
tination; and (g) they put me up for the night in a “flea-
bag” hotel, which had hundreds of ants in my bathroom.
Frankly, I was incensed. I obsessively awfulized. I tried to
use my best REBT—but it did not work. I foolishly vowed
never to use that infamous airline again. My usual high frus-
tration tolerance failed me, and for 2 days, including my
trip back to New York and the following day, I inwardly
seethed. As soon as I returned, I wrote the airline a scathing
letter and demanded monetary remuneration.
Finally, I went over the dismal events of the trip, saw that
the airline was wrong and its unethical agent was highly
fallible, but that those kinds of mistakes were not horrible
and sometimes inevitably occur. I then got back some mea-
sure of my HFT. I damned the behavior of the airline and its
agent, but stopped blaming the airline itself or the
personhood of the agent. At last, my REBT began to work
again. I have even traveled on that “wrongheaded” airline
again since that time—with, fortunately, much better re-
sults. My ill-fated trip was indeed bad—but I could stand it,
learn from it, and still survive. My high frustration tolerance
was, for a while, rudely interrupted. But not forever.
Depression (Catharine MacLaren)
I moved to New York from Colorado in the summer of
1995. I knew that my active, outdoor lifestyle would have
to change somewhat to accommodate city living, but by
December I was having serious difficulties getting out of
265
We i n r a c h , E l l i s , M a c L a r e n , D i G i u s e p p e , Ve r n o n , Wo l f e , M a l k i n s o n , a n d B a c k x
bed in the morning. A once 7- to 8-hours-a-night sleeper, I
began logging 10 hours a night and still feeling run down
the next day. My energy level was at an all-time low, I
stopped engaging in most social activities, and it was diffi-
cult for me to generate even a spark of excitement about
any aspect of my life. By mid-January I realized that some-
thing was going to have to change.
I read Overcoming Depression (Hauck, 1973) cover to cover
several times. I tried REBT Self-Help forms but had difficulty
identifying core irrational beliefs to dispute. On a practical level,
I started paying closer attention to my exercise and eating
habits. I stepped up my exercise regime and tried to cut “un-
helpful” foods out of my diet. The effects were minimal and
I experienced myself spiraling further down. Hindsight be-
ing 20/20, it comes as no surprise that I finally realized that
there was probably a biological component to my experi-
ence. My mother had been diagnosed several years before
with seasonal affective disorder, and it seemed that my de-
pressive mood had started sometime in November. Colo-
rado being one of the sunniest states in the country, it was
possible that I had unwittingly avoided seasonal downturns
by spending so much time outside throughout the year. My
apartment in New York did not afford me even one window
in my bedroom, so it was pitch black at all times of the day
without a light on. My work schedule was fairly rigorous, so
I spent most of my time indoors or on the subway, and much
less time outside on the weekends than I had in previous
years. I was even doing the majority of my exercise inside.
I read Winter Blues (Rosenthal, 1993) and got down to
work. I replaced all the light bulbs at home and in my of-
fice with full-spectrum bulbs. I obtained a light box and
started waking up 30–50 minutes early to sit in front of it
and read, effectively extending the light in my day. Within
3 weeks I was feeling much better. REBT was helpful in
terms of dealing with the frustration of having to take those
extra steps, but ultimately it was the additional full-
spectrum light that seemed to make the real difference.
Impulsivity (Raymond DiGiuseppe)
One area in which REBT has not helped me (yet) is control-
ling my impulsivity. I tend to take on too many tasks because
I enjoy doing them all and often do not write them down. The
consequences are that I wind up being double-booked at times
and inconveniencing myself and the people with whom I work.
And I become very stressed when I have to reschedule things
or have a huge amount of work piling up. My irrational belief
is that “I shouldn’t have to take the time to fill in the appoint-
ment book. It’s too much of a pain and I can’t stand having to
do it.” I’ve attempted to think to myself, “Just take the time to
write stuff in your appointment book.” But no matter how
many times I have tried to challenge this, it still has not gotten
me to keep an up-to-date appointment book.
Anxiety (Ann Vernon)
One instance when REBT did not work for me was when
my spouse had to have surgery to remove prostate stones.
266 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Despite the doctor’s assurance that there was no reason to
think this was anything more serious, I became extremely
anxious and was certain they would discover he had pros-
tate cancer. My first husband had died from lymph cancer
15 years earlier, and even though I tried to remind myself
that these two situations were not similar and that the doc-
tor had practically guaranteed us that this was not cancer, it
did not stop me from assuming the worst. I attempted to
control the situation by demanding that the surgery be per-
formed immediately because I couldn’t stand the anxiety.
On the day of the surgery I took plenty of paperwork to
the hospital, hoping to distract myself. It did not work. I
kept thinking that if this was cancer, I wouldn’t be able to
stand it. We had a young son, and it would be awful for him
to grow up without his father. I did not think I could live
through watching someone I loved die, because I had al-
ready had to do that as a 1st-year bride. The fact that I had
lived through it never entered my mind. I was also very
angry: it was not fair that this was happening to us.
When the doctor finally came to report that as a matter
of routine they would biopsy the stones and we should
know the results in a few days, I fell apart, saying that “I
couldn’t stand the anxiety of waiting and wondering.” He
was very unempathic, reminding me that there was prob-
ably nothing to worry about, but even if there was cancer,
it would be very treatable. He was probably attempting to
balance my irrationality with rationality, but I did not appre-
ciate it. I burst into tears, saying that he did not understand
what I had already been through with my first husband, and
that if this was cancer it would be awful.
I know the doctor wanted to get rid of me, so he promised
to rush the results. When the results came back negative, my
anxiety quickly dissipated. However, when I was caught up
in the actual experience, I wasn’t able to think rationally. I
had not even considered the fact that I had survived this
before and no doubt could again if in fact it was cancer.
Had the time frame been longer, I might have sought coun-
seling, although I was the only REBT practitioner in the
area. So in addition to my other irrationalities, I was engag-
ing in self-downing: “I should have been able to be more
rational about this. What a hypocrite I am, teaching others
to be rational and not applying it to myself.”
Low Frustration Tolerance (Janet Wolfe)
I was brought up with a strong work ethic and sense of
personal responsibility. These values have helped me be the
person I am, but probably have also helped contribute to
my main REBT “failure”: my tendency to ruminate and
awfulize about people who have been contracted to do work
for me or my organization and either ignore it, screw it up,
or take three times as long to do it as they should have.
Although the majority of the time these situations occur
with outside contractors (repair people, professional con-
sultants, etc.), they can occasionally happen with various
members of my own staff. And no matter how much I try
to tell myself, “They’re just being FFHB’s (fallible, fouled-up

human beings), I find myself enraged, depressed, and rumi-
nating for days on end. Typically, my self-talk is of the same
ilk as that of my clients’ daily expostulations: “I can’t believe
I put in 10 hours of work setting up this project for her, left
her a small task, and she’s taken 2 months to mail it out!”
“How could someone possibly sign a contract, take our hard-
earned money, then so grossly inconvenience us by taking 5
months longer than they said they would?” “It’s awful—I can’t
stand it—this person is an incredibly (stupid) (lazy) (devious)
(incompetent) (crooked) (despicable) human being!”
My friends and colleagues with whom I share my latest
horror story do all the right things—empathize, validate
that the person has indeed screwed up—and then try to
help me see that it is a hassle, not a horror, that humans will
be human, and all those other rational beliefs I try to instill
in my clients when their significant others act unfairly or
incompetently. This may help for a few hours, but left alone
(especially on holiday weekends) laboring at the task that
this other person has already been paid a good deal of money
to do, I usually manage fairly quickly to resume my rumi-
nations—along with considerable anger—not only while I
am working but well into my now quite-diminished (be-
cause of others’ delinquency) leisure time. And although I
am grateful for the many rational shoulders I have to cry
on, the grim fact remains that it is usually me with whom
the buck stops. I hope I can “lick this one” some day, but I
am only guardedly optimistic.
Need for Unconditional Self-Acceptance (Ruth Malkinson)
In Israel, the dominant school of psychotherapy is still psy-
chodynamic, even though REBT, like other cognitive models,
has been empirically proved to be an effective therapy. There-
fore, an invitation to give a workshop on REBT to a team of
professionals who adhere to traditional therapies is always a
challenge. Such was the case when I was invited by a senior
social worker, who practices cognitive therapy himself, in one
of the public mental health clinics, to conduct a one-day work-
shop on REBT for the staff, consisting of psychiatrists, social
workers, psychologists, and occupational therapists. The only
counselors in Israel are school counselors.
I told myself that the workshop must be a success and
that the participants must be convinced of the effective-
ness of REBT. I prepared handouts, video vignettes, role
plays, a list of references, and more. I was experiencing anxi-
ety before meeting the group. On my arrival, I was greeted
by my host, a social worker, who then proceeded to apolo-
gize for not participating in the workshop because of an
unexpected personal commitment (a military parade to
mark the completion of his son’s training). Understandably,
participating in the parade of one’s son is a special event
not to be missed—a point that I failed to accept because I
told myself that he must not leave me alone in the battle-
field to fight for cognitive psychotherapy. I felt deserted.
How could he leave me all by myself in the “lion’s den?” It
should be noted that military service in Israel is compul-
sory for boys and girls reaching the age of 18 years. The
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Rational Emotive Behavior Therapy Successes and Failures
army is embedded in the history of contemporary Israeli
society and is regarded as an important transitional period
for the individual and society.
Feeling bad and hoping for the best was all I could do.
Despite my training in REBT, I was telling myself “I must
be at my best” and “they must like REBT.” Following a short
introduction, I asked participants about their past experi-
ence with REBT and their expectations for the workshop.
They had heard about Ellis and REBT. Their expectations
were essentially neutral—neither high nor low. How awful.
In spite of their lack of enthusiasm, I did not feel terrible
all the time, but just part of the time. Toward the end of the
workshop, the head of the clinic shared a traumatic event
that he had experienced, which was an opportunity for me
to demonstrate the application of REBT under such cir-
cumstances. As a result of the demonstration, the “client”
reported a significant change of his feeling of anxiety to a
level of merely “deep concern.” I felt very proud. My use of
REBT with him had obviously had the desired outcomes.
However, I was profoundly angry, disappointed and frus-
trated when the “client’s” reported change of cognitions and
emotions were interpreted by his colleagues as a “classic pri-
mary death anxiety,” which had nothing to do with my help-
ing him to change his thinking. I felt as though I was a com-
plete failure. I was saying to myself, “his cognitive change
was an excellent ‘proof’ of the effectiveness of REBT and
all of the participants must be convinced. If they were not,
that would make me a complete failure as a teacher.”
CONCLUSIONS
It should come as no surprise that several experienced REBT
experts have both succeeded and failed in applying REBT
to themselves. The implication for practitioners who use
REBT with their clients and perhaps on themselves, as well,
is clear: The successful use of REBT requires considerable
and continuous hard work. REBT is not easy. REBT is not
applicable to every situation. Nor is it realistic to expect
REBT, or any other approach, to work 100% of the time.
That said, it has been used successfully in the personal lives
of those experts who contributed to this article. It could be
argued that in those cases in which the experts failed, it
may have been a function of their fallibility as human be-
ings. For example, those who struggle with LFT probably
get frustrated trying to reduce their LFT. In the final analy-
sis, Paul (1967) got it right. The ultimate question is, of
course, “What treatment, by whom is most effective for this
individual with that specific problem under which set of
circumstances?” (p. 111).
REFERENCES
Ellis, A. (1994). Reason and emotion in psychotherapy (Rev. ed.). New
York: Carol Publishing Group.
Hauck, P. (1973). Overcoming depression. Philadelphia: Westminster Press.
Paul, G. L. (1967). Strategy of outcome research in psychotherapy. Jour-
nal of Consulting Psychology, 31, 109–118.
267
We i n r a c h , E l l i s , M a c L a r e n , D i G i u s e p p e , Ve r n o n , Wo l f e , M a l k i n s o n , a n d B a c k x

Rosenthal, N. (1993). Winter blues: Seasonal affective disorder, what is it
and how to overcome it. New York: Guilford Press.
Walen, S. R., DiGiuseppe, R., & Dryden, W. (1992). A practitioner’s
guide to rational emotive therapy (2nd ed.). New York: Oxford
University Press.
Weinrach, S. G. (1996). Nine experts describe the essence of rational-
emotive therapy while standing on one foot. Journal of Counseling &
Development, 74, 326–331.
Weinrach, S. G., Ellis, A., DiGiuseppe, R., Bernard, M., Dryden, W.,
Kassinove, H., Morris, G. B., Vernon, A., & Wolfe, J. (1995). Rational
Emotive Behavior Therapy after Ellis: Predictions for the future.
Journal of Mental Health Counseling, 17, 413–427.
Weinrach, S. G., Lustig, D., Chan, F., & Thomas, K. R. (1998). Publi-
cation patterns of The Personnel and Guidance Journal/Journal of
Counseling & Development—1978 to 1993. Journal of Counseling &
Development, 76, 427–435.

268 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79

Metaphor and Change in Counseling

William J. Lyddon, Alison L. Clay, and Cheri L. Sparks

Increased general interest among counselors in the way language, narratives, and stories influence clients’ personal and social
realities has drawn particular attention to metaphorical language and its facilitative role in counseling. In this article, the authors
suggest that metaphors and metaphorical knowing may play a significant role in facilitating at least 5 developmental change
processes in counseling: relationship building, accessing and symbolizing client emotions, uncovering and challenging clients’
tacit assumptions, working with client resistance, and introducing new frames of reference.

T
he emergence of the constructivist perspective
in counseling and psychotherapy (Neimeyer &
Mahoney, 1995; Rosen & Kuehlwein, 1996; Sex-
ton & Griffin, 1997) has highlighted a central
role for human language, narrative, and stories in
the creation of personal and social realities (Anderson &
Goolishian, 1988; Efran, Lukens, & Lukens, 1990; Hare-Mustin,
1994; Russell, 1995; Shotter, 1993). It is interesting that one
important consequence of this focus on language has been a
greater appreciation for the metaphorical features of human
knowing and meaning creation (Carlsen, 1996). Indeed, meta-
phors may be found across all domains of human thought,
whether philosophical and scientific or psychological and
personal (Lakoff & Johnson, 1980; Leary, 1990a; Lyddon,
1989). In the counseling context, metaphors may help to
structure and facilitate client–counselor communications and
interactions (Berlin, Olson, Cano, & Engel, 1991). In addi-
tion, transformations in the metaphors clients use to define
their lives and problems may also correspond with signifi-
cant change in counseling (Carlsen, 1996; Goncalves, 1994;
Meichenbaum, 1995). Thus, metaphor is a vehicle both for
communication and for change in counseling (Muran &
DiGiuseppe, 1990).
In this article, we explore the use of metaphor in coun-
seling with a particular focus on the role metaphorical lan-
guage may play in facilitating key developmental change
processes. Toward this end, we first offer a brief historical
and theoretical overview of metaphor.
HISTORICAL AND THEORETICAL BACKGROUND
The study of metaphor may be traced to the Greek phi-
losopher Aristotle, who conceptualized metaphor as “a se-
ries of words in which a comparison is being made between
two or more entities that are literally dissimilar” (Angus &
Rennie, 1988, p. 552). Metaphor, from the Greek metaphera,
which means to transfer or carry across, allows the user to
transfer meaning from one domain to another by “giving to
one thing a name or description that belongs . . . to some-
thing else, on the grounds of some similarity between the
two” (Leary, 1990b, p. 4). Kopp (1971) elaborated stating,
“a metaphor is a way of speaking in which one thing is ex-
pressed in terms of another, whereby this bringing together
throws new light on the character of what is being described”
(p. 17). For example, one person may describe marriage as
a journey, whereas another may view marriage in terms of
a competitive game or even a battle.
It is interesting to note that significant scientific develop-
ments (or paradigm shifts) have not only been referred to
metaphorically as scientific “revolutions” (Kuhn, 1962) but
also have often been accompanied by the use of novel meta-
phors (Leary, 1990a). For example, Gentner and Grudin
(1985) assembled a collection of metaphors used to describe
the human mind from more than 80 years of articles in the
Psychological Review. They found that during the early years of
this journal, spatial and animate-being metaphors dominated,
whereas computer and systems metaphors dominated the later
years. Mahoney (1991) noted that several metaphorical shifts
have occurred in the history of psychobiology, from hydraulic
“humours” to mechanical, electrical, and contemporary
hormonal notions. Sternberg (1990) has drawn similar
conclusions about the history of the study of intelligence. He
has suggested that different theories of intelligence evolved
from different metaphors of intelligence—geographic,
computational, biological, epistemological, anthropological,
sociological, or systems metaphors.
On an individual level, metaphors are believed to aid in
the organization of personal experience (Fox, 1989). Sub-
stantial linguistic evidence supports the metaphorical qual-
ity of human thought processes (Johnson, 1987; Lakoff &
Johnson, 1980), and research on epistemic cognitive style
posits “metaphorism” as a way of knowing apart from either

William J. Lyddon is a professor and director of training, and Alison L. Clay and Cheri L. Sparks are students, all in the Counseling Psychology Doctoral
Program at the University of Southern Mississippi, Hattiesburg. Correspondence regarding this article should be sent to William J. Lyddon, Counseling
Psychology Program, University of Southern Mississippi, Hattiesburg, MS 39406-5025 (e-mail: william.lyddon@usm.edu).

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 269

L y d d o n , C l a y, a n d S p a r k s
rational or empirical knowing (Royce & Powell, 1983). Mills
and Crowley (1986) further suggested that metaphor is the
language of the right brain, a more fluid, imagistic, and
emotional operational system than the left brain (Marlatt
& Fromme, 1987).
Lankton and Lankton (1989) broadly defined metaphor
to include any form of communication or human activity
that “speaks” to the individual. Thus, biblical parables, fairy
tales, myths, stories, rituals, and motion pictures may all be
seen as metaphorical vehicles able to facilitate change, growth,
and understanding. Through use of metaphor, individuals are
able to explore and expand present experience into previ-
ously unrecognized possibilities (Fox, 1989). Fairy tales
stimulate the imagination, while engaging both intellect and
emotion to “see” unexpected solutions for existential
problems (Bettelheim, 1975). Metaphor facilitates under-
standing, in part, of that which defies understanding in
toto—feelings, experiences, moral, and spiritual perspectives
(Lakoff & Johnson, 1980).
Metaphor, as an experiential phenomenon, is part of an
ongoing dialectic in the search for meaning (Angus & Rennie,
1988), a juxtaposing of two seemingly unrelated referents
“producing semantic conceptual anomaly, the symptom of
which is usually emotional tension” (MacCormac, 1985,
p. 5). It is interesting that the emotional tension or disequi-
librium that may accompany significant metaphorical shifts
(and hence, shifts in personal meaning) is viewed from a de-
velopmental constructivist perspective as a natural part of
the change process (Lyddon, 1990; Mahoney, 1991; Rosen,
1996). Thus, as a transformative change agent, metaphor may
be particularly useful in the context of counseling.
METAPHOR AND COUNSELING
According to Mahoney (1991),
Our theories of change—which are our theories of personality, psy-
chopathology, lifespan development, identity, and systems—are most
powerfully captured in the generativities and multiplicities that
are afforded by metaphors. (p. 273)
Indeed, the use of metaphor permeates many facets of
counseling theories as evidenced in the literature. Many
authors have suggested a facilitative role for metaphorical
communication in counseling through stories, anecdotes,
metaphorical objects, and music (e.g., Barker, 1985; Jooste
& Cleaver, 1992; Saari, 1986). Metaphorical strategies have
been implemented with a variety of client populations, from
children to adults, as well as across a wide range of pre-
senting concerns (cf. Adams & Chadbourne, 1982; Bowman,
1992; Cornille & Inger, 1992; Davis & Sandoval, 1978; Huss
& O’Connor, 1995; James & Hazler, 1998; Kopp, 1971;
Lankton & Lankton, 1989; Marlatt & Fromme, 1987;
McClure, 1987; Roberts, 1987; Saari, 1986; Schwartz-
Borden, 1992).
There is much support, therefore, for the notion that use
of metaphor in counseling provides an effective mode of
communication and meaning creation. What is not clear,
270 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
however, are the particular ways in which use of metaphor
functions to facilitate client change. In the following dis-
cussion, we suggest that metaphor is central to at least five
developmental change processes in counseling: (a) relation-
ship building, (b) accessing and symbolizing emotions, (c)
uncovering and challenging clients’ tacit assumptions, (d)
working with client resistance, and (e) introducing new
frames of reference. What follows is a discussion of the role
of metaphor in each of these counseling processes.
Relationship Building
The development of a safe and secure counseling relation-
ship, or working alliance (Bordin, 1976), is believed to be a
significant component of successful counseling process and
outcome (Lambert & Bergin, 1994). Counselor sensitivity
to client-generated metaphors may help to convey under-
standing of the client’s unique way of knowing and at the
same time contribute to the development of a shared lan-
guage and collaborative bond between client and counselor
(Fine, Pollio, & Simpkinson, 1973; Goncalves & Craine,
1990). Although many counselors may generate their own
metaphors, Angus and Rennie (1988) encouraged metaphori-
cal communications in which the counselor and client work
together in the process of “apprehending, articulating, and
elaborating inner association to metaphors” (p. 555). For
example, during her first counseling session a client described
her experience of a romantic relationship as being like play-
ing the childhood game of “Monkey in the Middle.” The
client recounted her role in the relationship as being the
“monkey,” that is, the person jumping for the ball that is
usually thrown just out of her reach. The counselor en-
couraged the client to elaborate on her metaphor by dis-
cussing her perception of the different perspectives from
which the various game participants view the “ball.” She
stated that the player who is throwing the ball behaves so
as to diminish the object, which is casually tossed about
while she (the client), on the other hand, believes the ball
(relationship) to be valuable—something to be treasured
and protected. As a result, the counselor suggested that,
from the client’s perspective, the ball may be compared to
a piece of Waterford crystal, which the client would cher-
ish rather than throw about as the other player does. In this
way, the counselor used the client’s metaphor to
empathically communicate an understanding of the client’s
language and subjective experience. Furthermore, the
counselor’s willingness to elaborate on the client’s meta-
phor invited the client to either agree with the elaboration
or further clarify her feelings in the collaborative context
of the developing counseling relationship.
Accessing and Symbolizing Emotions
Emotional experience and exploration often play a pivotal
role in the change process by organizing the client’s self-
experience and establishing “links between self and environ-
ment” (Greenberg, Rice, & Elliott, 1993, p. 54). Metaphors
may be useful tools for helping clients access, as well as

symbolize, emotions that may have been previously unex-
pressed, unexplored, or even unrecognized. Fox (1989), for
example, asserted that metaphors help to expand the client’s
emotional awareness, increase emotional arousal, and lead to
the expression of a “felt-sense” and “nonliteral experience”
(p. 234). Similarly, Carlsen (1996) noted that use of meta-
phors might facilitate the counseling process by “translating
the intangibles of emotion into some sort of verbal expres-
sion” (p. 350). For example, a female client who has under-
gone a hysterectomy may refer nonemotionally to her scar.
The counselor, in following up, may inquire about the client’s
emotional meaning attached to the physical representation. In
response, the client may speak of an emotional “hole in her
middle” and the pain associated with the loss of part of her
body. Further exploration may focus more specifically not only
on the emotional pain associated with her loss (i.e., grief work)
but also on the new metaphors that may emerge as she begins
to reconstruct her relationship to that which has been lost.
By allowing clients to symbolize the “intangibles of emo-
tion,” metaphors act as vehicles within which clients may
construct new personal meanings of their experiences.
The power of metaphor lies in its ability to touch an affective com-
ponent of the individual’s experience. Emotions are presumed to
make up “feeling memories” which in turn furnish individuals with
a more or less automatic unconscious reaction repertoire to
affectively meaningful stimuli. By providing affectively meaning-
ful stimuli which are incongruous with the individual’s current ways
of seeing him/herself or situation, metaphor may necessitate struc-
tural changes in the person’s personal reality system. (Marlatt &
Fromme, 1987, p. 22)
Thus, metaphors may not only help clients become more
aware of their emotions (“I realize that I feel very frus-
trated playing ‘Monkey in the Middle.’”) but may also help
them to symbolize personal experiences in meaningful ways
(“I was so happy to just be asked to play the game that I
didn’t even stop to realize I wasn’t enjoying it.”).
Uncovering and Challenging Tacit Assumptions
For many clients, becoming more aware of previously tacit
or unconscious beliefs about self and world that influence
current behavior and choices constitutes an important
change process in counseling (Ecker & Hulley, 1996). Meta-
phors may play an important role in facilitating such aware-
ness and expression (e.g., “I now understand that for a long
time I have simply believed that life is a big party, and I’m
not invited” or “I realize that I’ve always felt like an actor in
a tragic play and that I will never get what I want most—
happiness.”). Through metaphor, clients’ tacit beliefs are
unveiled as the unspeakable is uttered, the taken for granted
or the unimagined is expressed and explored (Schnitzer,
1993). Furthermore, by recognizing their tacit assumptions,
clients are in a better position to change or modify them.
As Lenrow (1966) suggested,
Metaphors can highlight the client’s unspoken assumptions about
his capacities for influencing his surroundings effectively in the
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Metaphor and Change in Counseling
future. In this way, metaphor can condense and make vivid the
person’s apparent view of his lot in life. Moreover, metaphor can
highlight the client’s active contribution to his lot and thereby imply
choices open to him for modifying that lot. (p. 147)
The role of the counselor is not to interpret the client’s
metaphors or to directly modify the client’s belief system
but to facilitate, guiding the client’s exploration and mean-
ing in a supportive but challenging manner. For example, as
the client becomes more aware of the tacit beliefs and as-
sumptions that are no longer viable, the counselor may guide
the exploration of alternative metaphors (e.g., “What game
would you rather play than ‘Monkey in the Middle?’” or
“Instead of feeling like a juggler, can you imagine another
way of relating to work, school, and your wife and child?”).
Encouraging a client to explore alternative metaphors is a
way of questioning a point of view while at the same time
maintaining the client’s integrity (Billings, 1991).
Although metaphor is usually thought of in terms of nar-
rative presentation, counselors should also be alert to the
way that clients may reveal metaphors (and thus, tacit as-
sumptions) nonverbally (Modell, 1997). Indeed, as Fox
(1989) asserted, “However clients breathe, laugh, sigh and
move are metaphors” (p. 236). Sims and Whynot (1997)
agreed, adding: “Gesture . . . as well as language may have
metaphorical structure” (p. 342). The client whose hand
gestures with index finger and thumb separated by a space
of approximately 2 inches is indicating an amount of “some-
thing,” the identity of which may be discovered later in
counseling. Likewise, the client who gestures—hands in
front of chest, palms down, and making a downward push-
ing motion—may be expressing nonverbally what she may
not be able to express in words. Careful attention to such
nonverbal metaphorical expressions may provide the coun-
seling relationship with rich content for exploration.
Working With Client Resistance
Helping clients identify and work through their resistance,
or self-protective fear of change, is often an important
aspect of the counseling process (Bugental & Bugental,
1984). As Fox (1989) has pointed out, metaphors often
help clients indirectly access new information about them-
selves, arousing their ambivalence while protecting them
against the potential adverse impact of full discovery. As it
turned out, the client who offered the “Monkey in the
Middle” metaphor for describing a relationship had been
unable to express her feelings regarding the ending of this
relationship. By examining the component parts of the
metaphor, the client was able to recognize and express her
sense of frustration and also to identify the power differen-
tial inherent in the relationship. In addition, the counselor’s
elaboration of the metaphor to examine the nature of the
“ball” allowed the client to gain awareness of the discrepant
valuations of the relationship being made by the “players.”
Thus, metaphorical communication may allow clients to
explore what is “them” by talking about what is “not them.”
In other words, metaphors may function as a safe mediator
271
L y d d o n , C l a y, a n d S p a r k s
between what is and is not currently threatening to the
client (Billings, 1991; Fox, 1989; Jooste & Cleaver, 1992).
Caruth and Ekstein (1966) referred to metaphor as the
client’s alibi, “a conscious allusion which is a way of imply-
ing what he wants to communicate without actually com-
mitting himself, a way of simultaneously keeping and re-
vealing a secret” (p. 38). Indeed, the counselor’s ability to
hear and participate in the creation and elaboration of meta-
phor may provide an alternative vehicle within which to
express and examine painful client feelings and experiences
in a manner that bypasses rigid defenses and facilitates rap-
port (Billings, 1991; Fox, 1989).
Introducing New Frames of Reference
Significant, or second-order, change in counseling often
involves a dramatic shift in a client’s perspective or frame
of reference (Lyddon, 1990). It is not unusual for this type
of change to be organized around a new or novel meta-
phor. As previously noted, metaphors as a device for discovery
may facilitate awareness of aspects of the individual or
relationship that have previously been out of view (Cirillo
& Crider, 1995). When this information becomes available,
however, it sets the stage for the exploration of alternative
perspectives and possibilities. The client who explores the
“Monkey in the Middle” game, for example, may be able
to examine the options she has either to continue playing or
to refuse to play. She may evaluate the discrepancy between
the level of enjoyment afforded to the people on the outside
compared with her position and eventually decide that simply
being accepted as a powerless player in the game is no longer
desirable. Therefore, rather than playing “Monkey in the
Middle,” the client may speak of playing chess (a game in
which both players use strategies of anticipating the other’s
moves) or eventually transform the metaphor from a game
of competition to a game of cooperation and support.
The use of metaphor thus allows clients the opportunity
to create new possibilities and to modify their conceptual
framework within which new solutions may be explored
and adopted (Billings, 1991). Within the counseling rela-
tionship, counselor and client collaborate in examining
these new frames of reference, with the client choosing to
discard those that are no longer useful and adopt those
that are more viable. As a consequence, the client’s sense
of possibility may be expanded, and new meanings for
emotions, thoughts, and behaviors are considered. Davies
(1988), for example, described how the use of the fairy
tale Sleeping Beauty allowed one client to view her eating
disorder from the perspective of being in a deep sleep. As a
result of this reframing, the client explored the possibility
of “waking up” and taking a more active role in counseling
and in her life.
SUMMARY AND CONCLUDING REMARKS
In this article, we have suggested that metaphors not only
play a significant role in the generation of scientific theo-
272 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
ries and understandings but also that metaphorical knowing is
fundamental to the very personal theories and narratives
people construct about their lives, relationships, and per-
sonal problems. In particular, we have suggested that in the
counseling context, counselors’ sensitivity to and use of
client-generated metaphors may play a facilitative role in
at least five developmental change processes important to
successful counseling outcomes. First, a counselor’s ability
to hear and attend to client metaphors not only enlivens
the communication but also facilitates the development of
a collaborative counseling relationship, or working alliance.
From this perspective, empathic understanding may even
be viewed as a process of metaphorical mapping, whereby
the empathizer (counselor) is able to generate a metaphor
that not only results in the communication of verbal un-
derstanding but also expresses an understanding of the
client’s emotional state. A second facilitative role of meta-
phor concerns the ways in which metaphors help clients
access and symbolize emotions. In essence, metaphorical
language offers clients a figurative medium for expressing
emotional understandings that are often difficult (or some-
times impossible) to express in literal terms. Third, in addi-
tion to emotional expressions, metaphors may be useful
vehicles for the identification and clarification of clients’
long held tacit beliefs and assumptions about self and world.
Exploration of these beliefs provides clients with the oppor-
tunity to modify those beliefs that are no longer functional
and to construct alternative, more viable life metaphors
(Goncalves, 1994). A fourth facilitative role concerns the
way metaphorical expressions may help some clients work
more effectively with their ambivalence and fear of change
by allowing them to indirectly express certain feelings that if
dealt with directly would engender strong resistance. In this
way, metaphors may function as a safe bridge between the
part of the client that wants to change and the part that does
not. Fifth, as clients begin to both reevaluate previously held
beliefs and explore alternative, more viable frames of refer-
ence, novel metaphors can often play an important organiz-
ing role for these new developmental possibilities. One cli-
ent described this personal metamorphosis:
I am becoming real . . . I’m taking off this big coat that I have been
lugging around for so long . . . I carried a belief—from the media,
from the culture. Now I am shedding it, getting in touch with the
genuine. (cited in Carlsen, 1996, p. 339)
By using their individual creative styles, counselors have
the opportunity of speaking metaphorically with clients in
very personal and meaningful ways. Whether through dis-
cussion of a current movie or by reference to musical lyrics
or a child’s book such as Williams’s The Velveteen Rabbit,
the environment is “replete with opportunity for therapeutic
suggestion” (Billings, 1991, p. 5). As clients become more
aware of their metaphorical experiences and expressions,
they become empowered to critically examine them.
Cirillo and Crider (1995) encouraged counselors to take
a “fluid view” (p. 518) of metaphorical communication not-

ing that meaning is constructed in the process of dialogical
interaction between speakers. In our view, metaphor may
be seen as an all-terrain vehicle carrying the therapeutic
process off the paved roads of prior meaning structures out
into uncharted territories where new meanings are yet to
be created. Within this mode of travel, client and counselor
may come to concur with Sims and Whynot (1997), who
suggested that “meaning is not a destination. It is a
process” (p. 343).
REFERENCES
Adams, C. H., & Chadbourne, J. (1982). Therapeutic metaphor: An approach
to weight control. The Personnel and Guidance Journal, 60, 510–512.
Anderson, H., & Goolishian, H. (1988). Human systems as linguistic
systems. Family Process, 27, 371–393.
Angus, L. E., & Rennie, D. L. (1988). Therapist participation in meta-
phor generation: Collaborative and noncollaborative styles. Psycho-
therapy, 25, 552–560.
Barker, P. (1985). Using metaphors in psychotherapy. New York: Brunner/Mazel.
Berlin, R. M., Olson, M. E., Cano, C. E., & Engel, S. (1991). Metaphor
and psychotherapy. American Journal of Psychotherapy, 45, 359–367.
Bettelheim, B. (1975). The uses of enchantment. New York: Knopf.
Billings, C. V. (1991). Therapeutic use of metaphor. Issues in Mental Health
Nursing, 12, 1–8.
Bordin, E. S. (1976). The generalizability of the psychoanalytic concept
of the working alliance. Psychotherapy, 16, 252–260.
Bowman, G. (1992). Using therapeutic metaphor in adjustment coun-
seling. Journal of Visual Impairment and Blindness, 86, 440–442.
Bugental, J. F. T., & Bugental, E. K. (1984). A fate worse than death: The
fear of changing. Psychotherapy, 21, 543–549.
Carlsen, M. B. (1996). Metaphor, meaning-making, and metamorphosis.
In H. Rosen & K. T. Kuehlwein (Eds.), Constructing realities: Meaning-
making perspectives for psychotherapists (pp. 337–367). San Francisco:
Jossey-Bass.
Caruth, E., & Ekstein, R. (1966). Interpretation within the metaphor.
American Academy of Child Psychiatry Journal, 5, 35–45.
Cirillo, L., & Crider, C. (1995). Distinctive therapeutic uses of meta-
phor. Psychotherapy, 32, 511–519.
Cornille, T., & Inger, C. (1992). The armor metaphor in marital and fam-
ily therapy. Journal of Family Psychotherapy, 3, 27–42.
Davies, E. (1988). Reframing, metaphors, myths and fairy-tales. Journal
of Family Therapy, 10, 83–92.
Davis, J. M., & Sandoval, J. (1978). Metaphor in group mental health
consultation. Journal of Community Psychology, 6, 374–381.
Ecker, B., & Hulley, L. (1996). Depth-oriented brief therapy. San Fran-
cisco: Jossey-Bass.
Efran, J. S., Lukens, M. D., & Lukens, R. J. (1990). Language, structure,
and change: Frameworks of meaning in psychotherapy. New York:
Norton.
Fine, H. J., Pollio, H. R., & Simpkinson, C. H. (1973). Figurative lan-
guage, metaphor, and psychotherapy. Psychotherapy: Theory, Research
and Practice, 10, 87–91.
Fox, R. (1989). What is meta for? Clinical Social Work Journal, 17, 233–244.
Gentner, D., & Grudin, J. (1985). The evolution of mental metaphors in psy-
chology: A 90-year retrospective. American Psychologist, 40, 181–192.
Goncalves, O. F. (1994). Cognitive narrative psychotherapy: The herme-
neutic construction of alternative meanings. Journal of Cognitive Psy-
chotherapy: An International Quarterly, 8, 105–125.
Goncalves, O. F., & Craine, M. H. (1990) The use of metaphors in cogni-
tive therapy. Journal of Cognitive Psychotherapy: An International Quar-
terly, 4, 135–149.
Greenberg, L. S., Rice, L. N., & Elliott, R. (1993). Facilitating emotional
change: The moment-by-moment process. New York: Guilford.
Hare-Mustin, R. T. (1994). Discourses in the mirrored room: A
postmodern analysis of therapy. Family Process, 33, 19–35.
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Metaphor and Change in Counseling
Huss, E., & O’Connor, T. (1995). The broken balloon: A metaphor for
divorce. Journal of Divorce and Remarriage, 23, 211–223.
James, M. D., & Hazler, R. J. (1998). Using metaphors to soften resis-
tance in chemically dependent clients. Journal of Humanistic Educa-
tion and Development, 36, 122–133.
Johnson, M. (1987). The body in the mind. Chicago: University of
Chicago Press.
Jooste, E. T., & Cleaver, G. (1992). Metaphors and metaphoric objects.
Journal of Phenomenological Psychology, 23, 136–148.
Kopp, S. B. (1971). Guru: Metaphors from a psychotherapist. Palo Alto,
CA: Science and Behavior Books.
Kuhn, T. S. (1962). The structure of scientific revolutions. Chicago: Univer-
sity of Chicago Press.
Lakoff, G., & Johnson, M. (1980). Metaphors we live by. Chicago: Univer-
sity of Chicago Press.
Lambert, M. J., & Bergin, A. E. (1994). The effectiveness of psycho-
therapy. In A. E. Bergin & S. L. Garfield (Eds.), Handbook of psycho-
therapy and behavior change (pp. 143–189). New York: Wiley.
Lankton, C. H., & Lankton, S. R. (1989). Tales of enchantment: Goal-
oriented metaphors for adults and children in therapy. New York: Brunner/
Mazel.
Leary, D. E. (1990a). Metaphors in the history of psychology. Cambridge,
England: Cambridge University Press.
Leary, D. E. (1990b). Psyches muse: The role of metaphor in the history of
psychology. In D. E. Leary (Ed.), Metaphors in the history of psychology
(pp. 1–78). Cambridge, England: Cambridge University Press.
Lenrow, P. B. (1966). Uses of metaphor in facilitating constructive behav-
ior change. Psychotherapy: Theory Research, and Practice, 3, 145–148.
Lyddon, W. J. (1989). Root metaphor theory: A philosophical frame-
work for counseling and psychotherapy. Journal of Counseling and De-
velopment, 67, 442–448.
Lyddon, W. J. (1990). First- and second-order change: Implications for
rationalist and constructivist cognitive therapies. Journal of Counsel-
ing & Development, 69, 122–127.
MacCormac, E. R. (1985). A cognitive theory of metaphor. Cambridge,
England: Bradford Books.
Mahoney, M. J. (1991). Human change processes: The scientific founda-
tions of psychotherapy. New York: Basic Books.
Marlatt, G. A., & Fromme, K. (1987). Metaphors for addiction. Journal
of Drug Issues, 17, 9–28.
McClure, B. A. (1987). Metaphoric illumination in groups. Small Group
Behavior, 18, 179–187.
Meichenbaum, D. (1995). Changing conceptions of cognitive behavior
modification: Retrospect and prospect. In M. J. Mahoney (Ed.), Cog-
nitive and constructive psychotherapies (pp. 20–26). New York: Springer.
Mills, J. C., & Crowley, R. J. (1986). Therapeutic metaphors for children
and the child within. New York: Brunner/Mazel.
Modell, A. H. (1997). The synergy of memory, affects, and metaphor.
Journal of Analytical Psychology, 42, 105–117.
Muran, J. C., & DiGiuseppe, R. A. (1990). Towards a cognitive formulation
of metaphor use in psychotherapy. Clinical Psychology Review, 10, 69–85.
Neimeyer, R. A., & Mahoney, M. J. (Eds.). (1995). Constructivism in psy-
chotherapy. Washington, DC: American Psychological Association.
Roberts, S. (1987). Therapeutic metaphors: A counseling technique. Jour-
nal of the Academy of Rehabilitative Audiology, 20, 61–72.
Rosen, H. (1996). Meaning-making narratives: Foundations for
constructivist and social constructionist psychotherapies. In H. Rosen
& K. T. Kuehlwein (Eds.), Constructing realities: Meaning-making per-
spectives for psychotherapists (pp. 3–51). San Francisco: Jossey-Bass.
Rosen, H., & Kuehlwein, K. T. (Eds.). (1996). Constructing realities: Mean-
ing-making perspectives for psychotherapists. San Francisco: Jossey-Bass.
Royce, J. R., & Powell, A. (1983). Theory of personality and individual
differences: Factors, systems, and processes. Englewood Cliffs, NJ:
Prentice-Hall.
Russell, R. L. (Ed.). (1995). Discourse narrative in psychotherapy. Wash-
ington, DC: American Psychological Association.
Saari, C. (1986). The use of metaphor in therapeutic communication
with young adolescents. Child and Adolescent Social Work, 3, 15–25.
273
L y d d o n , C l a y, a n d S p a r k s

Schnitzer, P. K. (1993). Tales of the absent father: Applying the “story”
metaphor in family therapy. Family Process, 32, 441–458.
Schwartz-Borden, G. (1992). Metaphor: Visual aid in grief work. Omega,
25, 239–248.
Sexton, T. L., & Griffin, B. L. (Eds.). (1997). Constructivist thinking in coun-
seling practice, research, and training. New York: Teachers College Press.
Shotter, J. (1993). Conversational realities: Constructing life through lan-
guage. Thousand Oaks, CA: Sage.
Sims, P. A., & Whynot, C. A. (1997). Hearing metaphor: An approach to
working with family-generated metaphor. Family Process, 36, 341–355.
Sternberg, R. J. (1990). Metaphors of mind: Conceptions of the nature of
intelligence. New York: Cambridge University Press.

274 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79

FOCUS on the Family Caregiver: A Problem-Solving
Training Intervention

Monica F. Kurylo, Timothy R. Elliott, and Richard M. Shewchuk

Problem-solving interventions have documented effectiveness in treating distress among a variety of clientele. The authors dis-
cuss the application of training in social problem solving with family caregivers of persons who have incurred severe physical
disabilities. Specifically, the authors outline training procedures (i.e., Project FOCUS) that enable counselors to assist family
caregivers in developing effective problem-solving skills that may translate to increased wellness for each caregiver, and by
extension, their care recipients.

“I’ve taken all the responsibility for the patient.”
“I want to get over feeling responsible to keep him busy.”
“I have feelings of guilt.”
“It is a challenge to find new ways to do old things, and in dealing
with new everyday solutions.”
—Statements made by caregivers in a focus group convened to
address caregiver issues (cf. Elliott & Shewchuk, 2000).
functioning over time (Kiecolt-Glaser, Dura, Speicher, Trask,
& Glaser, 1991; King, Oka, & Young, 1994; Vitaliano, 1997).
Finally, caregivers may receive little positive reinforcement
for what they do, and they are often forced to neglect their
own needs, leisure pursuits, and personal interests (Quittner,
Opipari, Regoli, Jacobsen, & Eigen, 1992). Simply stated,
the ability of the caregiver to provide support and assis-
tance to their loved one may be reduced if the health and

T
hese statements provide a window to under-
standing the concerns of today’s family caregiver.
The new role of caregiver may be unexpected,
particularly in the case of a sudden onset illness
or injury (e.g., traumatic brain injury or stroke),
and the concomitant duties may at times be overwhelm-
ing. Wellness of the caregiver, as well as the care recipient,
is obviously affected. Research examining caregivers of older
individuals, particularly those with Alzheimer’s disease or
other dementing illnesses, has shown that caregivers dem-
onstrate higher levels of distress than noncaregivers on di-
verse measures of depression, anxiety, well-being, and physi-
cal health (Haley et al., 1995). Family caregivers may be
particularly distressed when their care recipients endure
chronic and uncontrollable bouts of pain and distress
(Miaskowski, Kragness, Dibble, & Wallhagen, 1997). Emo-
tional distress and physical symptoms experienced by
caregivers may be due in part to the subjective appraisal of
caregiving as onerous and unrewarding (Chwalisz, 1996;
Haley, Levine, Brown, & Bartolucci, 1987; Haley et al., 1996).
In addition to emotional distress, caregivers often develop
problems in cardiovascular, neuroendocrine, and immune
well-being of the caregiver erodes over time.
Caregivers of persons (adults and children) with acquired
severe physical disabilities, such as spinal cord injury (SCI),
face many unique circumstances that complicate the
caregiving experience. When a family member acquires an
SCI, for example, the normal pattern of life events in the
family system may be disrupted. A young person with SCI
may be forced to return to the parental home or may be
unable to be “launched” from it (Whiting, Terry, & Strom-
Hendriksen, 1984). Although most states have programs to
promote personal independence after disability, cutbacks
in public funds in many states have limited the availability
of vocational rehabilitation and independent living programs
for many persons with severe physical disabilities (Elliott
& Shewchuk, 1998). Thus, they are compelled to reside
with family members who then provide assistance and per-
sonal care. Furthermore, it is important to note that with
most severe physical disabilities, there is no subtle, devel-
opmental course that may aide family members in plan-
ning for care. Instead, role changes and lifestyle changes
occur abruptly and almost instantaneously with the onset
of the disability. The role changes and lifestyle changes re-

Monica F. Kurylo is a postdoctoral fellow, and Timothy R. Elliott is an associate professor, both in the Department of Physical Medicine and Rehabilitation,
and Richard M. Shewchuk is an associate professor in the Department of Health Services Administration, all at the University of Alabama at Birmingham.
Monica F. Kurylo is the project coordinator and Timothy R. Elliott and Richard M. Shewchuk are codirectors for Project FOCUS. This article was supported
in part by the National Center for Injury Prevention and Control and the Disabilities Prevention Program, National Center for Environmental Health Grant
R49/CCR412718-01 and the National Institute on Disability and Rehabilitation Research Grant #H133B980016A. The contents of this article are solely
the authors’ responsibility and do not necessarily represent the official views of the funding agencies. Correspondence regarding this article should be sent to
Timothy R. Elliott, University of Alabama at Birmingham, Department of Physical Medicine and Rehabilitation, SRC 530 1717 6th Ave S, Birmingham, AL
35249-7330 (e-mail: telliott@uab.edu).

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 275

Kurylo, Elliott, and Shewchuk
quired in these circumstances are necessarily related to long-
term adaptation given that persons with such physical dis-
abilities are likely to live for many years with a disability
that is typically not terminal or life threatening.
The importance of the family caregiver role has been rec-
ognized by professionals interested in initiating interventions
to assist the family caregiver (Myers, 1989). As practitioners
of the wellness paradigm, counselors are uniquely qualified
to provide interventions to family caregivers. For counselors
offering assistance to newly appointed family caregivers, a
focus on wellness provides an excellent backdrop for explora-
tion of interventions focused toward maximizing physical,
psychological, and spiritual health (Myers, 1991, 1992).
According to counseling researchers in the area, wellness
involves the integration of aspects of health (social, mental,
emotional, spiritual, and physical components) and is
maximally achieved when all components are in balance
(Myers, 1991, 1992; Witmer & Sweeney, 1992). Wellness does
not necessitate the absence of illness (as defined by health)
but instead involves enhanced functioning in all areas of life
and throughout the life span (Ginter, 1999; Greenberg, 1985;
Myers, 1991). As such, wellness is a proactive lifestyle in
which a person is responsible for choices of self-care, self-
sufficiency, and empowerment. The goal of counseling, then,
is to promote wellness among clients. Given the lifestyle
change presented by an acute-onset severe physical disabil-
ity, caregivers often experience an imbalance in at least one,
if not several, of the aforementioned areas of health and would
likely benefit from intervention.
FOSTERING CAREGIVER WELLNESS
Interventions stemming from the wellness paradigm to date
have included a focus on providing respite care and social
support. Although some studies have shown benefits of
caregiver support groups and social networks that offer
assistance, respite care, and other buffering functions (e.g.,
Haley et al., 1996; Myers, 1989), evidence indicates that such
support erodes over time (Quittner, Glueckauf, & Jackson,
1990). This suggests that informal efforts to intervene with
family caregivers have not been highly successful or long last-
ing. In addition, it may be difficult to address individual
caregiver needs within group situations or a respite network,
and this may lead the caregiver to feel alone in his or her
concerns. Therefore, individual interventions are necessary.
A cognitive-behavioral intervention approach that has
been proven effective when used in other populations (i.e.,
adults with depression, caregivers of cancer patients) is train-
ing in social problem solving. The social problem-solving
intervention can benefit family caregivers because the tech-
nique addresses each caregiver’s unique needs and demands
within an individual counseling context. Skills learned in
training provided by counselors can augment caregiving,
help decrease caregiver distress, and maximize wellness in
the caregiver.
Particularly during their first year of caregiving, family
caregivers contend with several competing demands, includ-
276 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
ing (a) worry and concern for their loved one’s recovery
and medical care; (b) problems dealing with money, trans-
portation, and making necessary home modifications; (c)
having to take time off work, make changes in their work
schedule, make a job change, or even quit working to ac-
commodate their role of caregiver; and (d) serving as moni-
tor for the individual’s daily programs to maintain good
health (e.g., medication management, exercise, diet, appoint-
ments). This means that while caregivers typically want to
give love and support, they must also take on additional
duties geared toward adequate family care and assume mul-
tiple roles (e.g., financial manager, health care supervisor).
In an effort at reducing distress and maximizing the
caregiver’s (and in turn, care recipient’s) well-being, these
concerns can be addressed through a focused intervention
invested in assisting the caregiver at arriving at solutions to
problems they face in their multiple roles. The purpose of
this article is to describe this problem-solving intervention
for family caregivers (Project FOCUS), and discuss how
counselors can use this intervention with caregivers of in-
dividuals with a severe acquired physical disability.
PROJECT FOCUS
Project FOCUS is a problem-solving training program we
have developed specifically for family caregivers of persons
with spinal cord injury. Family caregivers are encouraged
to build on existing problem-solving tools and apply these
tools to their new situation of providing care for a family
member with a severe physical disability. This commu-
nity-based program currently serves an approximately
equal distribution of African American and Caucasian
individuals. These individuals in our program come pre-
dominantly from lower to lower-middle socioeconomic
status backgrounds. FOCUS is an acronym associated with
the problem-solving process (Facts, Optimism, Creativity,
Understanding, Solve), which assists the caregivers in recall-
ing the problem-solving techniques. The concept within
each step of the process is briefly covered in a handout (see
Appendix) and is explained in greater detail later in this
article. This handout is used as a guide for the caregivers to
follow during training.
Before training in problem-solving skills is initiated, as-
sessment of the caregiver’s problem orientation and prob-
lem-solving skills is performed. Then, card-sorting tasks are
used to determine what specific problems are of concern
to each individual caregiver. After the assessment and card-
sorting tasks are complete, the caregiver is introduced to
the project and problem-solving training. A brief descrip-
tion of the theoretical background of problem solving fol-
lows, along with a detailed explanation of the assessment
and training process.
Theoretical Overview
According to contemporary models of social problem solv-
ing (D’Zurilla & Goldfried, 1971; D’Zurilla & Nezu, 1990,

1999; Nezu & D’Zurilla, 1989), a person’s orientation to his
or her problems determines the manner in which a person
processes information about the self, the environment, and
problematic situations encountered in everyday life. Problem
orientation, or attitude toward problem solving, involves the
ability to (a) ward off negative emotions (e.g., anxiety, de-
pression, and anger) that hamper problem-solving efforts,
(b) promote positive emotions and a sense of competency
that facilitate problem solving, and (c) motivate an individual
toward solving problems (D’Zurilla & Nezu, 1990; D’Zurilla
& Sheedy, 1991; Nezu & D’Zurilla, 1989). Persons with a
positive problem orientation often resolve everyday prob-
lems without much effort, preventing these problems from
exacerbating. Consequently, a sense of competency and
self-efficacy is reinforced as the individual recalls these
successful experiences when more complex problems are
encountered. In contrast, a negative problem orientation is
associated with ineffectual coping, negative emotional
experiences under general and stressful conditions, and
more critical views of the self and personal abilities (Elliott,
Herrick, MacNair, & Harkins, 1994; Elliott, Sherwin,
Harkins, & Marmarosh, 1995), which tend to reinforce a
negative orientation and impair problem-solving efforts
(Nezu, 1987; Nezu & D’Zurilla, 1989). A negative problem
orientation is also associated with more health complaints
(Elliott & Marmarosh, 1994). Problem-solving skills encom-
pass the specific, goal-directed strategies by which individuals
define problems, gather facts about problems, generate alter-
natives, decide on solutions, and implement and monitor
problem-solving strategies, according to the original model
of D’Zurilla and Goldfried (1971).
Effective problem solving has been related to greater use
of instrumental problem-focused coping in times of stress
(MacNair & Elliott, 1992) and an assertive interpersonal
style (Elliott, Godshall, Herrick, Witty, & Spruell, 1991).
Effective problem-solvers might experience fewer health
problems because they may have a more proactive, consci-
entious style of coping. Furthermore, care recipients who
are with caregivers who possess many impulsive and careless
problem-solving tendencies have more difficulty accepting
their condition and are at risk for developing preventable
health complications over time (Elliott, Shewchuk, &
Richards, 1999). In addition, when caregivers of persons
with recent-onset SCI have a negative orientation toward
problem solving, they have more problems with depres-
sion, anxiety, and ill health throughout the first year of
caregiving, regardless of the actual demands of caregiving
or level of physical impairment of the care recipient (Elliott,
Shewchuk, & Richards, in press).
The effectiveness of problem-solving training has been
documented in treatment of clinical depression among
adults (Nezu & Perri, 1989) and elderly individuals (Arean
et al., 1993), suicidal ideation (Lerner & Clum, 1990), dis-
tress among persons with chronic illness (Roberts et al.,
1995), and in improving self-management skills of under-
graduates (Richards & Perri, 1978). This model has also
been successful in alleviating distress among family
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
FOCUS on the Family Caregiver
caregivers of individuals with cancer (Houts, Nezu, Nezu,
& Bucher, 1996).
Measuring Social Problem Solving
Before beginning the intervention, we first determine each
caregiver’s problem orientation and specific problem-solving
skills using the Social Problem Solving Inventory–Revised
(SPSI–R; D’Zurilla, Nezu, & Maydeu-Olivares, 1995). A
caregiver rates the extent to which each of 52 statements is
true of him or her on a 5-point Likert scale (0 = not at all
true of me, to 4 = extremely true of me). The measure has five
scales: (a) Positive Problem Orientation (e.g., “Whenever I
have a problem, I believe that it can be solved”); (b) Nega-
tive Problem Orientation (e.g., “When my first efforts to solve
a problem fail, I get discouraged and depressed”); (c) Ratio-
nal Problem-Solving Skills (e.g., “Before I try to solve a prob-
lem, I set a specific goal so that I know exactly what I want
to accomplish”); (d) Impulsivity/Carelessness (e.g., “When I
have a decision to make, I do not take the time to consider
the pros and cons of each option”); and (e) Avoidance (e.g.,
“I wait to see if a problem will resolve itself first before try-
ing to solve it myself”) (D’Zurilla & Chang, 1995). Scores on
all five scales are then examined as a profile by which the
clinician can understand the caregiver’s attitude toward prob-
lem solving and tendencies when problem solving.
Card Sort Procedures
After measuring each caregiver’s problem-solving orienta-
tion and skills, two unique card-sorting tasks are used to
determine which areas may be problems for the caregiver
and his or her loved one. The card-sorting (or Q-sort) task
used in our intervention is based on 24 problems identified
by a focus group of 7 caregivers (4 Caucasian women, 2
African American women, and 1 African American man;
specific information and procedures concerning development
of these cards may be obtained from the second author or
the following publications: Elliott & Shewchuk, 1999, 2000).
A single problem is listed on a card for a total of 24 cards.
Each caregiver in our project is asked to examine the deck
of 24 cards and place all 24 cards into different piles based
on their perceived similarities. The caregiver is instructed
to separate them into as many or as few piles as desired
and to move cards into different piles as necessary until
satisfied with how the cards are arranged. The caregiver is
then asked to give each group a name so that the counselor
can understand the unique theme of each group and what
separates each group from the others. This technique en-
courages the caregiver to think about how problems might
overlap or share similar characteristics. Understanding the
similarities between problems may be helpful for generat-
ing possible solutions that may have been used successfully
in other similar situations.
A second method of card sorting is also used in which
each caregiver is asked to rank order on a continuum from
1 (most problematic) to 7 (least problematic) the 24 cards as
they apply to the caregiver’s unique situation. All 24 cards
277
Kurylo, Elliott, and Shewchuk
are used, and the caregiver is instructed to place a certain
number of cards at each level of the continuum (2 cards
under #1 and #7, 3 cards under #2 and #6, 4 cards under #3
and #5, and the remaining 6 cards under #4). This forced-
choice format is used to determine which item(s) is (are)
the most problematic and may need to be explored further
during the problem-solving training. This method of card
sorting has many advantages. It requires a new caregiver
to consider problems that may be outside immediate
awareness or may help the caregiver to label a particular
experience as a problem. This technique also allows the
caregiver to talk with the counselor about problems that
he or she may have been previously unwilling or unpre-
pared to discuss (Brown, 1986; Elliott & Shewchuk, 2000).
The counselor may use this approach to help the caregiver
prioritize problems that need immediate attention from
those that are less urgent. This technique also allows the
caregiver to consider problems that may arise in the future
and plan how they wish to prevent these problems from
occurring or lessen the problems’ potential impact. The
caregiver can then plan for potential consequences of and
obstacles to his or her goals.
Problem-Solving Training
Problem-solving training for the caregiver includes five
major components (adapted from the original problem-solving
model proposed by D’Zurilla & Goldfried, 1971): (a) facts/
problem definition, (b) optimism/orientation, (c) creativity/
generation of alternatives, (d) understanding/decision making,
and (e) solve/implementation and verification. This training
protocol also parallels the protocol for problem-solving
training for depression outlined in Nezu, Nezu, and Perri
(1989). It is important to note that the five components
should be viewed as a continuous and interlinking process
rather than five stages to be followed in serial order. For
example, the counselor may decide to begin with the optimism
component of the model when working with a caregiver
who is less motivated (as determined through the SPSI–R
assessment device) and continuously address the optimism
component throughout the training process.
THE FIVE COMPONENTS OF FOCUS
Facts/Problem Definition
This component involves description of a problem that was
identified as such in the card-sorting procedures. The caregiver
is asked to articulate the specific problem and then break it
down into manageable parts. The caregiver is instructed to
seek all available facts about the problem in an effort to
answer who, what, when, where, why, and how of the situ-
ation. The caregiver is also instructed to ask experts for
information regarding the problem situation if necessary. For
example, if the problem situation concerns a constellation of
physical symptoms, the caregiver would be encouraged to
contact the physician’s office. When describing the problem,
the caregiver is encouraged to give the facts in unambiguous,
278 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
concrete terms, separate facts from assumptions, and dif-
ferentiate relevant from irrelevant information.
Within this phase of training, the caregiver is assisted in ar-
ticulating a specific, attainable goal in an effort to overcome
the identified problem. The counselor encourages realistic goals
and discourages “perfect” goals. The caregiver is assisted in
determining potential obstacles to meeting the established
goal—solutions to these obstacles will also be generated in the
creativity phase. If the problem situation is complex (i.e., in-
volves several problem areas that are subsumed under one
large problem situation, such as several smaller interpersonal
conflicts that have an effect on one’s relationship with an-
other person), the caregiver is encouraged by the counselor to
focus on one aspect of the situation at a time because smaller,
more discrete goals are typically easier to solve.
Optimism/Orientation
The primary goal in this segment in the problem-solving
training paradigm is to assist the caregiver in developing a
sense of optimism regarding his or her abilities to problem
solve. This includes instilling a belief that one is sufficiently
skilled to solve the problem and instilling a sense of motivation
to engage in the problem-solving process while simulta-
neously regulating emotional experiences to maintain a sense
of confidence. Often, when the counselor recognizes the im-
portance and breadth of the caregiver’s role, the caregiver
begins to feel a sense of pride in accomplishments to date,
which motivates continuation in this training process. The
caregiver is asked to provide an example of recent per-
sonal successes in problem solving and to consider the mo-
tivational characteristics of a personal role model who is
an effective problem solver. Furthermore, the caregiver
can feel more optimistic regarding problem solving with
the recognition that (a) problems are a common (“nor-
mal”) part of everyone’s life, (b) many other caregivers
have similar problems, (c) problems can be predicted and
prevented, and (d) there have been personal problem-solving
successes in his or her past. The idea that optimism must
be accompanied with realism is stressed. In other words,
the caregiver should have realistic expectations about the
time and effort that will be necessary to identify and use
chosen strategies, and the caregiver should not expect to
“move mountains.”
In addition to the aforementioned, the individual is assisted
in recognizing a problem when it occurs and thereby recog-
nizing when problem solving is applicable. One manner
in which this can be accomplished is through recognition
of emotional cues (e.g., anger, sadness, frustration, irritability)
that a problem exists. We call these cues “red flags” or signs
that the person should be problem solving. The caregiver is
encouraged not to be impulsive when solving problems and
not to avoid solving problems because both of these processes
usually complicate the initial problem. Specifically, we dis-
cuss that impulsive attempts often result in careless mistakes.
The caregiver is encouraged to take the time necessary to
analyze situations, to tolerate discomfort when solutions are

not immediate, and to withhold judgment until the problem-
solving process is complete.
After discussing the aspects of a positive problem orienta-
tion at length with the caregiver, the counselor leads the
caregiver in an exercise designed to increase motivation for
solving problems. One exercise we use is a reversed role-play
scenario. Specifically, the counselor adopts a maladaptive prob-
lem orientation, and the caregiver is asked to successfully
counter-argue the attitudes presented by the counselor (e.g.,
“I don’t have any problems at all,” “Only someone who is
experiencing the exact same problem can be helpful to me;
no one else can understand,” “I should avoid making decisions
no matter what the consequences”). By counter-arguing the
attitudes, the caregiver can develop skills to use to counter-
argue personal attitudes that may decrease motivation and
impede problem solving at a later time.
When working with individuals who impulsively solve prob-
lems (as assessed with the SPSI–R), a different technique is
used. The caregiver is instructed to recognize the occurrence
of a problem situation and label the situation as a problem.
This can help deter the tendency to react automatically. First,
the caregiver is asked to generate a list of areas in life in which
problems occur (caregiving, relationships, job, children and
education, etc.). Then the caregiver creates a second list that
includes specific current problems as well as predicted areas
of vulnerability unique to the caregiver.
For those individuals who are typically unassertive when
problems arise, we use a method to help the caregiver to
recognize “red flags” as cues to the existence of problems.
In this exercise, the caregiver and counselor generate a list
of feelings and explore how the feelings serve as cues for
the presence of a problem. The caregiver is encouraged to
stop and think when these emotions are detected and to
avoid reacting automatically.
Once the caregiver is familiar with the problem-solving
process, the person is asked to provide a brief written de-
scription of the aforementioned recent problem situation
that was articulated in the Facts/Problem Definition seg-
ment. This exercise emphasizes how the first two segments
of the model intersect. Then the caregiver is instructed to
indicate how he or she initially reacted to the situation (both
emotionally and behaviorally) and describe how those re-
actions compare with a positive problem orientation. This
illustrates to the caregiver how negative emotions can be
used as a cue that a problem exists. It also provides an op-
portunity for the caregiver to understand what aspects of a
positive problem orientation will be most useful in main-
taining a positive outlook through the problem-solving pro-
cess, given the available facts of the problem situation.
Creativity/Generation of Alternatives
This portion of the problem-solving training involves actively
brainstorming multiple solutions to the identified problem.
Because a caregiver may believe that there is only one correct
answer or may tend to solve the problem with the first idea
that comes to mind, training in this area is critical. First, the
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
FOCUS on the Family Caregiver
caregiver is instructed to think of as many solutions to the
problem as possible and write each one down on a worksheet.
When the caregiver reaches an impasse, the counselor sug-
gests solutions. It is crucial that judgments are withheld at this
point in the process, which means that “silly” or “impossible”
solutions count equally as potential solutions. The caregiver is
reinforced for deferring judgment and for giving as many al-
ternative strategies as possible. The caregiver is encouraged to
be as specific and concrete as possible in generating strategies.
The following techniques are offered to increase the number
of items generated: (a) combining different solutions together
to make another solution, (b) changing a solution slightly to
create more solutions (adding or subtracting from it), (c) imag-
ining oneself and his or her role model(s) in the actual
situation(s) and considering what each would do and say in
the situation, and (d) recalling potential solutions to other simi-
lar problems that were evaluated during the card-sorting pro-
cedure. Although there is no required number of solutions,
the caregiver is instructed that having more options will
increase optimism about his or her ability to solve problems.
Understanding/Decision Making
After having generated several options, it is now time for
the caregiver to integrate all the information gathered to
this point and consider which solutions to implement. Be-
fore deciding on a solution, the caregiver is encouraged to
consider the potential outcomes of the chosen solutions
and weigh the costs and benefits of each. This is done by
rating each alternative as it relates to feasibility, achieving
goals, perceived fit between the option and the problem at
hand, and coping with obstacles. This step is the essence of
the “stop and think” paradigm, in which the problem solver
weighs the pros and cons of each solution alternative be-
fore proceeding with implementation. Finally, the caregiver
rates the likelihood that he or she will implement each so-
lution alternative (0 = not at all likely, 1 = somewhat likely, 2
= very likely). Recall of the facts identified earlier in the
process may be beneficial in processing information during
this segment. Of course, maintaining a positive problem
orientation is also crucial during this phase because the
caregiver will be most likely to implement those solutions
about which he or she feels most optimistic.
Solve/Implementation and Verification
The final phase in the problem-solving training process is
the act of solving the problem and then systematically review-
ing the outcome to determine how the solution worked and
the degree to which the actual outcome approximates the
expected one. Because the counselor is not typically present
during this stage, the caregiver is requested to provide feed-
back at the next meeting with the counselor. The caregiver is
asked to evaluate how effective his or her solution was in
solving the problem and to identify the actual effects on each
individual involved in the problem situation and solution. This
self-monitoring component is crucial to promote under-
standing about what made the chosen solution effective
279
Kurylo, Elliott, and Shewchuk
or ineffective and how to implement similar or alternative
solutions in future problem situations. This component builds
self-esteem, self-efficacy, and self-empowerment.
FINAL POINTS OF CONSIDERATION
This five-phase problem-solving training process requires
approximately 2 to 3 hours to complete during an initial
meeting with a caregiver. Manuals are available that de-
scribe the use of this intervention with other groups (e.g.,
terminal cancer patients and their family members) and
provide informative examples (D’Zurilla & Nezu, 1999;
Nezu, Nezu, Friedman, Faddis, & Houts, 1998; Nezu, Nezu,
& Perri, 1989). In the Project FOCUS model, our training
with caregivers also includes (a) regular telephone contacts
with the caregivers emphasizing and reinforcing effective
problem solving each month during the year; (b) regular
mailings of informative and instructive materials over the
year to the caregiver; and (c) follow-up evaluations and
“booster sessions” at 6 months and 12 months to reassess
problem-solving orientation and skills, and level of distress
(i.e., depressed mood, anxiety), and to retrain and reinitiate
the caregiver as necessary. These components reinforce use
of effective problem-solving principles throughout the year.
The caregiving career seems to follow a dynamic course
over the first year after an injury to a family member
(Shewchuk, Richards, & Elliott, 1998). Therefore, we dis-
tribute problem-solving training over the duration of a year
rather than over the course of several weeks, as suggested
by other social problem-solving researchers (Nezu & Perri,
1989). In addition, our approach considers the immediate
and specific problems reported by each individual caregiver
and matches the intervention with the problems unique to
each caregiver (Elliott & Shewchuk, 2000). This can pro-
mote self-empowerment among caregivers and foster a sense
of pride and accomplishment.
We believe that family caregivers of persons who receive
training in problem solving will have greater well-being, will
be more likely to use rational problem-solving skills, and will
report less distress. By extension, persons with severe physi-
cal disabilities whose caregivers are trained in problem solv-
ing may likely have fewer secondary complications in the
first year following the injury and thus require less medical
attention. We intend to promote health, well-being, and in-
dependence among caregivers and their family members.
CONCLUSION
Counselors are uniquely qualified to assist caregivers in
maximizing their health and well-being through individu-
alized cognitive-behavioral intervention provided by the
Project FOCUS model. The model illustrates that caregivers
can benefit from problem-solving training that enhances
their ability to regulate emotional experiences when en-
countering problems and provides them with the cogni-
tive-behavioral skills for effective problem solving. In using
this model, counselors consider the immediate and specific
280 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
problems of a caregiver. The caregiver is empowered with
self-care skills and the ability to handle situations that arise
in an effective fashion; the person seeks more information
when needed and takes a proactive role in all aspects of the
caregiver experience. These skills may promote indepen-
dence from formal health care systems and empower
caregivers and their family members to achieve a greater
level of wellness and self-sufficiency.
REFERENCES
Arean, P. A., Perri, M., Nezu, A., Schein, R., Christopher, F., & Joseph, T.
(1993). Comparative effectiveness of social problem-solving therapy
as treatments for depression in older adults. Journal of Consulting and
Clinical Psychology, 61, 1003–1010.
Brown, S. R. (1986). Q-technique and method: Principles and procedures. In W.
D. Berry & M. S. Lewis-Beck (Eds.), New tools for social scientists: Advances and
applications in research methods (pp. 57–76). Beverly Hills, CA: Sage.
Chwalisz, K. (1996). The perceived stress model of caregiver burden:
Evidence from the spouses of persons with brain injuries. Rehabilitation
Psychology, 41, 91–114.
D’Zurilla, T. J., & Chang, E. C. (1995). The relations between social prob-
lem solving and coping. Cognitive Therapy and Research, 19, 547–562.
D’Zurilla, T. J., & Goldfried, M. (1971). Problem solving and behavior
modification. Journal of Abnormal Psychology, 78, 104–126.
D’Zurilla, T. J., & Nezu, A. M. (1990). Development and preliminary evalu-
ation of the Social Problem Solving Inventory. Psychological Assessment, 2,
156–163.
D’Zurilla, T. J., & Nezu, A. M. (1999). Problem-solving therapy: A social
competence approach to clinical intervention (2nd ed.). New York: Springer.
D’Zurilla, T. J., Nezu, A. M., & Maydeu-Olivares, A. (1995). Manual for
the Social Problem Solving Inventory–Revised. Unpublished manuscript,
State University of New York at Stony Brook.
D’Zurilla, T. J., & Sheedy, C. (1991). Relation between social problem-
solving ability and subsequent level of psychological stress in college
students. Journal of Personality and Social Psychology, 61, 841–846.
Elliott, T., Godshall, F., Herrick, S., Witty, T., & Spruell, M. (1991). Prob-
lem-solving appraisal and psychological adjustment following spinal cord
injury. Cognitive Therapy and Research, 15, 387–389.
Elliott, T., Herrick, S., MacNair, R., & Harkins, S. (1994). Personality cor-
relates of self-appraised problem-solving ability: Problem orientation
and trait affectivity. Journal of Personality Assessment, 63, 489–505.
Elliott, T., & Marmarosh, C. (1994). Problem-solving appraisal, health
complaints, and health-related expectancies. Journal of Counseling &
Development, 72, 531–537.
Elliott, T., Sherwin, E., Harkins, S., & Marmarosh, C. (1995). Self-appraised
problem solving ability, affective states, and psychological distress. Jour-
nal of Counseling Psychology, 42, 105–115.
Elliott, T., & Shewchuk, R. (1998). Recognizing the family caregiver: Inte-
gral and formal members of the rehabilitation process. Journal of Voca-
tional Rehabilitation, 10, 123–132.
Elliott, T., & Shewchuk, R. (1999). Using focus groups to understand the
experience of family caregivers and care recipients. Manuscript submitted
for publication.
Elliott, T., & Shewchuk, R. (2000). Problem-solving therapy for family
caregivers of persons with severe physical disabilities. In C. Radnitz
(Ed.), Cognitive-behavioral interventions for persons with disabilities
(pp. 309–327). New York: Aronson.
Elliott, T., Shewchuk, R., & Richards, J. S. (1999). Caregiver social problem-
solving abilities and family member adjustment to recent-onset physical
disability. Rehabilitation Psychology, 44, 104–123.
Elliott, T., Shewchuk, R., & Richards, J. S. (in press). Family caregiver social
problem-solving abilities and adjustment during the initial year of the
caregiving role. Journal of Counseling Psychology.
Ginter, E. J. (1999). David K. Brooks’ contribution to the developmen-
tally based life-skills approach. Journal of Mental Health Counseling,
21, 191–202.

Greenberg, J. S. (1985). Health and wellness: A conceptual differentiation.
Journal of School Health, 55, 403–406.
Haley, W. E., Levine, E., Brown, S. L., & Bartolucci, A. (1987). Stress, ap-
praisal, coping and social support as predictors of adaptational outcome
among dementia caregivers. Psychology and Aging, 2, 323–330.
Haley, W. E., Waff, D. L., Coleton, M. I., Ford, G. R., West, C. A. C., Collins,
R. P., & Isobe, T. L. (1996). Appraisal, coping, and social support as mediators
of well-being in Black and White family caregivers of patients with Alzheimer’s
disease. Journal of Consulting and Clinical Psychology, 64, 121–129.
Haley, W. E., West, C. A. C., Wadley, V. G., Ford, G. F., White, F. A., Barrett, J.
J., Harrell, L. E., & Waff, D. L. (1995). Psychological, social, and health
impact of caregiving: A comparison of Black and White dementia family
caregivers and non-caregivers. Psychology and Aging, 10, 540–552.
Houts, P. S., Nezu, A. M., Nezu, C. M., & Bucher, J. A. (1996). The prepared
family caregiver: A problem-solving approach to family caregiver educa-
tion. Patient Education and Counseling, 27, 63–73.
Kiecolt-Glaser, J. K., Dura, J. R., Speicher, C. E., Trask, O. J., & Glaser, R.
(1991). Longitudinal changes in immunity and health. Psychosomatic Medi-
cine, 53, 345–362.
King, A. C., Oka, R. K., & Young, D. R. (1994). Ambulatory blood pressure
and heart rate responses to the stress of work and caregiving in older women.
Journals of Gerontology, 49, 239–245.
Lerner, M. S., & Clum, G. A. (1990). Treatment of suicidal ideators: A prob-
lem-solving approach. Behavior Therapy, 21, 403–411.
MacNair, R. R., & Elliott, T. (1992). Self-perceived problem-solving ability,
stress appraisal, and coping over time. Journal of Research in Personality, 26,
150–164.
Miaskowski, C., Kragness, L., Dibble, S., & Wallhagen, M. (1997). Differences
in mood states, health status, and caregiver strain between family caregivers
of oncology outpatients with and without cancer-related pain. Journal of
Pain and Symptom Management, 13, 138–147.
Myers, J. E. (1989). Adult children and aging parents. Dubuque, IA: Kendall/Hunt.
Myers, J. E. (1991). Wellness as the paradigm for counseling and development:
The possible future. Counselor Education and Supervision, 30, 183–193.
Myers, J. E. (1992). Wellness, prevention, development: The cornerstone of
the profession. Journal of Counseling & Development, 71, 136–139.
Nezu, A. M. (1987). A problem solving formulation for depression: A litera-
ture review and proposal of a pluralistic model. Clinical Psychology Review,
7, 121–144.
APPENDIX
Project FOCUS
Facts: Get the facts! First, define the problem. When you
have determined what the specific problem is, your goal
becomes solving that problem. In order to determine the
necessary steps leading to your goal (i.e., solving the prob-
lem), you will need to decide what information (facts) you
already have to assist you in solving the problem, and de-
cide what else you need to solve the problem. You may
need to ask an expert (e.g., nurse, physical therapist, men-
tal health practitioner, physician) in order to get the neces-
sary information (facts). Utilize the experts. We are here to
assist you. No question is a dumb question.
Optimism: Be optimistic, but realistic. You will feel better
about yourself and your ability to problem-solve when you
are hopeful and positive. Think of a role model who handles
problems effectively (either someone you know personally—
a friend or relative—or know about—president, hero). How
would they feel in this situation? Think: “I CAN do this!”
Creativity: Be creative in your ideas! What other ways
can you solve the problem that you have not thought of
before? Brainstorm! Think of as many ways as possible that
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
FOCUS on the Family Caregiver
Nezu, A. M., & D’Zurilla, T. (1989). Social problem solving and negative
affect. In L. P. Kendall & D. Watson (Eds.), Anxiety and depression: Distinc-
tive and overlapping features (pp. 285–315). San Diego, CA: Academic Press.
Nezu, A. M., Nezu, C. M., Friedman, S. H., Faddis, S., & Houts, P. S. (1998).
Helping cancer patients cope: A problem-solving approach. Washington, DC:
American Psychological Association.
Nezu, A. M., Nezu, C. M., & Perri, M. G. (1989). Problem-solving therapy for
depression: Theory, research, and clinical guidelines. New York: Wiley.
Nezu, A. M., & Perri, M. G. (1989). Social problem-solving therapy for uni-
polar depression: An initial dismantling investigation. Journal of Consulting
and Clinical Psychology, 57, 408–413.
Quittner, A. L., Glueckauf, R. L., & Jackson, D. (1990). Chronic parenting
stress: Moderating versus mediating effects of social support. Journal of
Personality and Social Psychology, 59, 1266–1278.
Quittner, A. L., Opipari, L., Regoli, M., Jacobsen, J., & Eigen, H. (1992). The
impact of caregiving and role strain on family life: Comparisons between
mothers of children with cystic fibrosis and matched controls. Rehabilita-
tion Psychology, 37, 275–290.
Richards, C. S., & Perri, M. G. (1978). Do self-control treatments last? An evalu-
ation of behavioral problem solving and faded counselor contact as treat-
ment maintenance strategies. Journal of Counseling Psychology, 25, 376–383.
Roberts, J., Brown, G. B., Streiner, D., Gafni, A., Pallister, R., Hoxby, H.,
Drummond-Young, M., LeGris, J., & Meichenbaum, D. (1995). Problem solv-
ing counseling or phone-call support for outpatients with chronic illness:
Effective for whom? Canadian Journal of Nursing Research, 27, 111–137.
Shewchuk, R. M., Richards, J. S., & Elliott, T. (1998). Dynamic processes in
health outcomes among caregivers of patients with spinal cord injuries.
Health Psychology, 17, 125–129.
Vitaliano, P. P. (Ed.). (1997). Physiological and physical concomitants of
caregiving: Introduction to special issue. Annals of Behavioral Medicine, 19,
75–77.
Whiting, R. A., Terry, L., & Strom-Hendriksen, H. (1984). From home
to college, from college to home: An interactional approach to treat-
ing the symptomatic disabled college student. In J. C. Hansen (Ed.),
Families with handicapped members (pp. 30–43). Rockville, MD:
Aspen.
Witmer, J. M., & Sweeney, T. J. (1992). A holistic model for wellness and
prevention over the life span. Journal of Counseling & Development,
71, 140–148.
you can solve the problem (both old and new ways). You
will be determining later which ideas may work best, but
for now, think of as many as you can.
Understanding: Do you understand what the problem is?
Understanding the problem, the facts about the problem,
and the possible solutions to the problem will assist you in
solving the problem. It is also important to know and
understand the possible consequences (both positive and
negative) for each of the ideas of how to solve the problem
that you thought of previously. Finally, it will also be
important to have an understanding of the difficulties, or
potential obstacles, involved in reaching your goal.
Solve: Solving the problem is the final step. Given the
preceding information (facts, ideas of how to solve the prob-
lem, and understanding), you can develop a plan of action
(e.g., steps), which will lead to your goal, as you outlined
above. Then solve the problem and see the outcome. Does
the outcome match your expectations, your goals? If not,
go back to your list of ideas and try another idea. Try not to
be impulsive. Stop and Think.
281
Where Has Our Theory Gone? Learning Theory
and Intentional Intervention

Sterling Gerber

Reflecting on the principle that every effective counseling intervention works to the extent that it incorporates sound learning
principles, this article emphasizes intentional selection of intervention strategies with reference to 4 learning models: association,
reinforcement, cognition-perceptual, and cognition-rational/linguistic. Three contexts—developmental, social, and spiritual—are
described that require modification of intervention strategy, and a case example is given.

C
ounselor educators and practitioners who were
educated in the 60s will remember the tri-
chotomy of choices relative to philosophical
and theoretical foundations. There existed the
“directive counseling” school of thought,
whose major proponent was E. G. Williamson (1939, 1950).
This position rested firmly in the guidance tradition of
defining the client as one who was deficient or limited
regarding some knowledge, skill, or wisdom in which the
counselor was proficient. The task of the counselor was to
ascertain that deficiency and to prescribe or direct infor-
mation or procedures to rectify the problem.
The “nondirective counseling” school was strongly in
vogue, due largely to the work of Carl Rogers (1942, 1951,
1961). Built on the premise that clients are basically ca-
pable and socialized and that the source of their self-defeat
lies largely in perceptual distortion or “defensiveness,” the
procedure of choice was to create an atmosphere of uncon-
ditional positive regard. As the client would disclose infor-
mation of a potentially risky nature and be met with a
nonjudgmental response, he or she was helped to broaden
the perceptual frame of reference and attain a more realis-
tic, nondefensive view of reality. Such expansion of per-
ception freed the client to do those things that were neces-
sary and acceptable to the client and generally were already
existent within his or her repertoire.
The third option was called “eclectic counseling” and was
put forth by Frederic Thorne (1950). It grew out of the
observation that rigid adherence to any theoretical position
precluded the flexibility to fit the broad range of client
personalities and problems that were encountered by most
practitioners. Counselor education curricula typically had
several components including developmental theories, per-
sonality theories, learning theories, counseling theories, and
assessment procedures (which rested heavily on psychometric
theory). The number and complexity of theories provided a
rich array of positions to work within. To the neophyte,
selecting a preferred position from among them was nearly
impossible. As a result, many counselors resolved their pro-
fessional identity crisis by adopting the eclectic label and
assumed that they could extract methods and procedures from
any source and integrate them into a workable practice.
Early in my professional formation, I received two impera-
tives that made it difficult to follow the zeitgeist and become
eclectic. One was the Gestalt maxim that “meaning comes
from context” and, as an extension of that, the power of any
counseling method (now named intervention procedure or
strategy) comes from its theoretical context. The second was
the directive given by an influential professor, “Do not be-
come eclectic. Believe in something!” If eclecticism involved
taking various methods out of their theoretical context, and
indeed this is what Thorne prescribed, then they existed in a
diluted, weak, or powerless form. If the power rests in the
theoretical context, then it is conceivable that two methods
taken from opposing contexts would cancel each other.
The apparent dilemma was to choose from a restricted
number of theory-based counseling approaches that seemed
to be limited in their scope of application or to become
eclectic, which was “forbidden.” This forced a third option,
one that was not readily apparent. It is possible to use vari-
ous intervention strategies in their appropriate context—
theory pure—by selecting from among extant theory bases
the particular one that shows the closest match to client
circumstance and style. It will have a higher probability of
successful resolution of client concerns than one that does
not fit and higher than an amalgamation of methods that is
applied out of context. The awareness of strength in match-
ing intervention to specific client characteristics is not new.
Over the past several years a trend has developed, that of
integrative metaframeworks (see for example Mueller,

Sterling Gerber is a professor in the Counseling, Educational, and Developmental Psychology Department at Eastern Washington University, Cheney. Corre-
spondence regarding this article should be sent to Sterling Gerber, Department of Counseling, Educational, and Developmental Psychology, Martin Hall 135,
Eastern Washington University, Cheney, WA 99004-2437 (e-mail: Sterling.Gerber@mail.ewu.edu).

282 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79

 Learning Theory and Intentional Intervention

Dupuy, & Hutchins, 1995; Preston, 1998). What is new with

this model is its reliance on learning theory as the basis for Class Family Context
making decisions regarding intervention strategies.
One Stage
An immediate cry arises, “It is highly impractical if not Association Selection of
impossible to become proficient in all the theory-based basic
approaches. There are just too many of them. If I select a Reinforcement intervention
Two Stage
few or several, what are the chances that those I select will Cognition–
strategy
match my client population?” according to
Perceptual which family
There is an answer to that concern. Although it requires client circum-
that counselors learn more than one theory approach (thus Cognition– stance and
Rational/
avoiding the limitations of theory-pure discipleship), it per- Linguistic
style fit.
mits avoidance of the theory-poor practice of eclecticism. Many
professional counselors, accustomed to labeling themselves as
eclectic, bristle at the implication of “less than okay” proce-
dures, yet their uses of intervention strategies have evolved, Modifications to the
from experience or intent, to incorporate those that tend to application of the Developmental
be complementary. The intent is not to criticize what may be strategy result from
aspects of client Social
a workable frame or to quibble over the meaning of the term contexts.
eclectic. Rather it is to explicate what is seen as a cognitive Spiritual
structure that will enhance intentionality and effectiveness.
In addition to the premise that the power of an interven-
tion comes from its theoretical context, it is necessary to FIGURE 1
recognize that all client change comes as the result of learn-
ing. Another way of phrasing this principle is that all coun- Selecting Intervention Strategy to Match Client
seling interventions work because they exist in the context Circumstance and Style
of a learning theory. Given this observation, the number of
positions needed to deal with all clients is reduced from all
counseling, personality, development and learning theories The second and confounding contemporary force is the
to “just” all learning theories. Even so, this task is more than diminution of emphasis on theory, particularly learning
daunting. Although possible, it still remains impractical. theory, in counselor preparation programs. Referring to an
early draft on revision of the Council for Accreditation of
Counseling and Related Educational Programs (CACREP;
A LEARNING THEORY BASE FOR INTERVENTION 1998) standards, there were eight curricular areas: Human
A careful study of learning theories and of counseling in- Growth and Development, Social and Cultural Foundations,
terventions interpreted from their learning theory under- Helping Relationships, Groups, Career and Lifestyle De-
pinnings results in a blending of many seemingly diverse velopment, Appraisal, Research and Program Evaluation,
positions into two classes, four families, and three contexts and Professional Orientation. The two areas that might be
(see Figure 1). The two classes are one-stage and two-stage related to theory, Human Growth and Development and
models. The four families are association, reinforcement, cog- Helping Relationships, did not specify learning theory.
nition-perceptual, and cognition-rational/linguistic. The con-
Human Growth and Development—studies that provide an under-
texts are developmental, social, and spiritual. It is possible standing of the nature and needs of individuals at all developmental
for a counselor to learn at least one approach within each levels. Studies in this area include, but are not limited to the follow-
of these four families, with awareness of modifications to ing: a. theories of human development across the life-span; b. major
fit the contexts, and integrate them into a repertoire rich theories of personality development; c. human behavior including
an understanding of developmental crises, disability, psychopa-
enough to match essentially all of his or her clients.
thology, and cultural factors as they affect both normal and abnormal
Two contemporary forces in counselor education create a behavior.
major inconsistency and challenge. The first is a societal press Helping Relationships—studies that provide an understanding
for accountability that has spawned an emphasis on inten- of counseling and consultation processes. Studies in this area in-
tionality in the selection and carrying out of intervention clude, but are not limited to the following: a. helping skills, and
counseling and consultation theories including coverage of relevant
strategies (Gerber, 1999). It is only through knowing, in research and factors considered in applications; b. counselor or con-
advance, the unique aspects of the client concerns, patterns sultant characteristics and behaviors that influence helping processes
of self-defeating behavior, and preferred patterns of reacting including gender and ethnicity differences, verbal and nonverbal
(client circumstance and style) that an appropriate interven- behaviors and personal characteristics, orientations, and skills; and
tion can be selected. Matching the client circumstance and c. client or consultee characteristics and behaviors that influence
helping processes including gender and ethnicity differences, ver-
style to a particular family of learning-theory-based interven- bal and nonverbal behaviors and personal characteristics, traits,
tion permits rational and intentional selection of methods capabilities, life circumstances, and developmental level. (CACREP,
that have a high probability of working. 1998, p. 4)

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 283

Gerber
Subsequent iterations of the accreditation standards restored
some reference to learning theory. That it was ever in question
is instructive. Another factor in this equation is the experi-
ence of professors in finding textual materials for teaching
that subject. Many, if not most texts, communicate a narrow-
ing focus to one of research and academic interest (see
Hergenhahn & Olson, 1993). The problem of translating theory
to practice is aggravated by teaching materials and program
guidelines that treat learning theory as an esoteric, if necessary
at all, component of counselor training.
Frequently the profession is in the “schizoid” position of
telling its trainees to be more accountable and sending them
out without the cognitive and perceptual preparation to
permit them to do the job correctly (see Gladding, 1996;
Lawless, Ginter, & Kelly, 1999). One mark of a professional
is the ability to explicate the rationale for his or her proce-
dures (Ginter, 1999). Those who prepare professionals re-
ally ought to equip them for the task. The remainder of
this article provides an introduction to the classes, families,
and contexts of learning-based, intentional intervention. A
more complete treatment can be found in Enhancing Coun-
selor Intervention Strategies: An Integrational Approach
(Gerber, 1999).
One problem with learning theory approaches in the past
is that they were based on a definition of learning that was
restrictive or inclusive of only a small portion of human
learning dynamics. Obviously it is difficult to explain cog-
nitive or perceptual change in associational or reinforce-
ment terms and vice versa. Of necessity, the definition of
learning must be broader or integrational in nature. For the
purposes of this treatise, learning is considered to be a con-
struct, meaning that it is not accessible to direct empirical
assessment and, hence, is implied or inferred from observed
change. This involves changes in behavior, cognition, or both
in ways of thinking (process) and in results of thinking
(products)—perception, or affect, including feelings, atti-
tudes, and values. The definition excludes changes due to
drugs, fatigue, illness, instinct, and maturational changes
in readiness. Learning occurs following or as a result of
experience, direct or vicarious; thinking; or insight. It is
relatively permanent, usually enhanced by practice, and is
purposeful for self-maintenance, survival, or prediction or
anticipation of future conditions. Often, there are physi-
ological correlates: Learning either causes or results from
changes in the central nervous system.
Classes of Learning Models
The symbols SàR are often used to indicate a cause and
effect relationship between stimulus and response. Some
theories frame the human as a biological machine that “runs”
according to which “buttons are pushed.” To elicit a certain
behavior, you must press a specific button.
Sometimes the person is included in the formula:
SàOàR. This recognizes organismic qualities or charac-
teristics that might affect the response. Because the inner
workings of a person are closed to direct observation, they
284 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
are inferred from systematic observation of differences in SàR
trials. It can be said that S causes R, or that S cues the person
to give the predictable R. This occurs as a unitary process, a
single stage, triggered by the stimulus. Another model says
that the probability of a particular response is enhanced or
diminished by conditions in the person, SàOàR. This also is
a single-stage model. The button is pushed and the machine
responds. Even though the response might be one of lowered
intensity because of “dirty bearings” or “frayed connections,” it
still is a specific response to a particular stimulus.
The probability of a specific stimulus eliciting a particu-
lar response can sometimes be affected by what follows
the response. This formula is SàRàS. The change in stimu-
lus condition, subsequent to and consequent on the re-
sponse, acts to increase the probability of recurrence of the
response on another presentation of the original stimulus.
This also is a single-stage dynamic because the relationship
between S and R is a one-step, highly predictable one. Dif-
ferences across observations of different people in similar
stimulus circumstances and in the same individual from
time to time are explained by organismic changes that in-
hibit or enhance normal response or by changes in conse-
quential stimulus conditions that affect the probability of
responding, an external locus of control.
Another approach for looking at the dynamics across
an SàOàR event is to recognize that the person (O) is
an active agent. This usually requires allowing for the
person’s ability to perceive, think, judge, and decide. This
is a two-stage process. The first stage is the occurrence
of a stimulus and its perceptual impact on the person
(SàO). What does the person perceive and how clear or
accurate is that perception? This explains variations in
factors such as attention, sensory receptor differences,
and perceptual set across individuals and in the same in-
dividual over time. In effect, the same physical event can
be perceived at a variance from its physical science prop-
erties. The second stage is the decision and action pro-
cess (OàR). The person takes an action based not on
the stimulus but, rather, on the decision he or she has
made. Stage one is stimulus and perception. Stage two is
decision and action. Differences in individuals’ responses
across similar stimulus circumstances and in the same
individual to repetitions of a similar stimulus situation
result from decision, an internal locus of control.
Families of Learning Models
The two classes give rise to four families, two of which are
one-stage models and two are two-stage models. For refer-
ence in this context and as fairly common descriptors in the
area of learning theory, these are referred to as association,
reinforcement, cognition-perceptual, and cognition-rational/
linguistic. Each is described briefly, including an abbreviated
set of directions for application.
Association. Association theories, such as those of Pavlov
and of Guthrie (Sahakian, 1976), have the following com-
mon dynamics:

• The stimulus event occurs prior to or at the same time as the
response and either elicits or cues the response.
• A connection or bond is established between the stimulus and
response (afferent-efferent neural events) or between two stimuli
and a reflexive response (afferent-afferent-efferent neural events).
• Contiguity—the simultaneous or overlapped, or closely proximate
occurrence of the stimulus and response—is the single and pow-
erful condition for the bond or association of the two to be made.
• Teaching or conditioning is a process of arranging stimulus and
response events to occur contiguously to one another. (Gerber,
1999, p. 44; see Endnote)
Two applied approaches that are based in association
theory are those of Wolpe (1969) and of Bandler and
Grinder (1975; Neuro-Linguistic Programming). The inter-
vention will vary depending on which of the two basic as-
sociation approaches best fits the client situation.
Generally, it is best to use a Pavlovian-based approach if
the response to be modified or to be brought under differ-
ential stimulus control is reflexive. For a mature client, a
vicarious strategy may be appropriate (à la Wolpe). Some
reflexive retraining can be accomplished by use of a
Guthrian “threshold” method. If the response is not reflex-
ive, a Guthrian-based approach is preferred. There are two
applications of Pavlovian-based interventions (see Appen-
dix A): (a) bringing an existing reflex under stimulus con-
trol and (b) overriding an undesirable conditioned response
by substituting a more desirable one to an existing stimu-
lus. The intitial determination in Guthrian-based interven-
tions (see Appendix B) is whether (a) a new behavior is to
be conditioned to a specific cue or (b) an existing behavior
is to be replaced in a prescribed context.
One caveat of an integrational approach is that although
it may be possible to explain learning and to manage change
from more than one family of models, the preferred inter-
vention is the one that is most parsimonious in application.
It creates the highest probability for change at the least
expense of time and resources.
Reinforcement. Reinforcement theories, like that of Skinner
(Sahakian, 1976), have the following common dynamics:
• Learning is a process of the contiguous occurrence of three re-
lated or contingent events: the stimulus, the response, and the
consequent stimulus or reinforcer.
• Specific stimuli that are intended to become controls for spe-
cific responses must be discriminated by the learner from among
other similar but not identical stimuli.
• The desired response or target behavior must be caused to occur
contiguously to the discriminated stimulus; this may be accom-
plished by selective reinforcement or by the conditioning of suc-
cessive approximations of the target response.
• A consequent stimulus is any event subsequent to the response
that has the effect of increasing the probability of the discrimi-
nated stimulus calling out the target response.
• The consequent stimulus must be caused to occur contiguously
to the target response.
• Teaching or conditioning is a process of managing the contiguous
relationship of the three contingent events. (Gerber, 1999, p. 45)
Although the intricacies of a reinforcement theory strat-
egy can be complex, most can be reduced for analysis and
management to six steps:
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Learning Theory and Intentional Intervention
1. Identify the target behavior in terms of discrete units
(countable instances of each or time intervals that
include their occurrence)
2. Decide if the target behavior is to occur in a restricted
context that includes signals (discriminated stimuli)
for the target response
3. Establish a baseline (the number of discrete occur-
rences within a specified time)
4. Identify reinforcers (generally small objects or to-
kens that can be managed easily and delivered readily
on occurrence of the target response)
5. Run the learning trials (repetitions of the Sd à R à
Sreinfpattern)
6. Perform a criterion test (repetition of the procedure
used to establish the baseline).
Cognition-perceptual. Perception theories, such as
Wertheimer (1959), Köhler and Koffka (as cited in Sahakian,
1976), Tolman (1949), and Bandura (1969, 1974), have the
following common dynamics:
• What a person perceives varies with the properties of physical
stimulation.
• The perceptual variations or distortions follow patterns or “rules.”
• Perception and understanding are molar events that cannot be
reduced without altering their meaning.
• Learning is a process of varying perceptual frames until one is
accomplished that corresponds to a universal structure.
• Perceptual learning is not incremental; it happens in an all-or-
none fashion and is experienced as a revelation.
• Past perceptual units or Gestalts are stored in memory and be-
come data for future decisions.
• Teaching or conditioning is a process of managing experiences,
real or vicarious, until the learner accomplishes insight or per-
ceptual “homeostasis.” (Gerber, 1999, p. 46)
Applied models based in perceptual learning theory include those
of Bruner (1966), Seligman (1975), Perls (1969), Watzlawick
(1978), and Rogers (1951). Maslow (1962) also conforms with
these theorists, as do the objects-relations theorists.
Although it may be an oversimplification, applications of cog-
nitive-perceptual learning can be placed into just two categories:
framing and reframing. When a client has no prior experience to
draw on in a given circumstance (perceptual deficit), the creation
of a frame is called for. If the client has a self-defeating frame
from prior experience (perceptual surplus), that frame needs to
be loosened and a more functional frame created.
Basic to a perceptual intervention is an awareness of the
client’s circumstance and style. The client’s behavior will flow
from his or her unique perceptual awareness, hence it is im-
portant to discover that frame of reference. Through the use
of active listening, several questions must be answered:
• What is the client doing or not doing that is self-defeating?
• What sense do you make of the client’s behavior in the context
of his or her situation?
• Is there a perceptual deficit? Is the client trying to do something
for which he or she has no previous experience, either direct or
vicarious?
• Does the client’s behavior make sense from the client’s distorted or
incorrect perception? Is the client working from a misleading frame?
285
Gerber
• From a perception of the client’s style, what kind(s) of expe-
rience, direct or vicarious, would be effective in creating or
helping the client to replace a perceptual frame? (Gerber, 1999,
p. 112)
After answering the questions, the counselor selects a strat-
egy that will fit the client’s circumstance and style. In in-
stances in which the client is using an ineffective perceptual
frame, it is necessary to loosen that frame before introducing
a preferred perception. This is done through techniques such
as role reversal, splits, and figure-ground shifts (Gerber, 1999).
Cognition-Rational/Linguistic. Festinger (1957), Försterling
(1986), and Rokeach (1973) are primary theorists, whereas
Ausubel (1963), Ellis (1958, 1962), Beck, Rush, Shaw, and
Emery (1979), and Berne (1961) have intervention strategies
based in cognition-rational/linguistic learning theory. The works
of all of these have the following dynamics in common:
• Experience becomes coded and remembered; humans rely to a
great extent on language codes.
• Language permits labeling of referents and of experience; it also
permits progressively higher levels of abstraction resulting in
organization or structure of knowledge.
• People can modify or combine structures to arrive at new and
creative knowledge.
• Reasoning skills permit analysis of past and present experiences
and prediction of future outcomes.
• Thought precedes action; most behavior results from cognitive
analysis of data and from decision.
• Teaching or conditioning is a process of providing (a) standard-
ized labels and/or descriptions for experiences and referents; (b)
organizational structures of abstractions to enhance memory re-
trieval, reasoning, and classification of new experiences; and (c)
supplanting of inefficient or self-defeating structures with new
ones. (Gerber, 1999, p. 47)
Because both cognition-rational/linguistic and cognition-
perceptual approaches are two-stage models that involve
mental functions, they have similar patterns of application
dynamics. Deciding on an appropriate cognition-rational/
linguistic intervention will include the following steps. First,
in answering the following questions:
• What is the client doing or not doing that is self-
defeating?
• What is the experiential context of the undesirable
response?
• What explanation does the client give for his or her
response?
This last question is misleading in its simplicity. In most
cases, the client will not verbalize clearly or accurately the
underlying rationale. It has to be deduced by the counselor
from an analysis of the context and the response. The coun-
selor may ask himself or herself, “What rationale logically
precedes the response in this context?”
Is the rationale ill-defined, whimsical, or nonsensical? If
so, treat the problem as a cognitive deficit and provide a
cognitive structure that will direct self-enhancing behavior.
Nonexistent and weak structures are easily replaced by
strong, contextually appropriate and logical structures.
286 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Is the rationale loose, weak, or strong and well-defended? If
weak, treat the problem as a cognitive deficit. If strong, use a
cognitive surplus approach. Work to loosen the structure by
using cognitive dissonance principles. Introduce dissonant
information (doubt) thereby producing an “I’m not so sure”
state that will be receptive to a new structure—particularly if
the new structure eliminates the dissonance.
The second part of cognitive restructuring requires in-
troduction of a competing structure, the use of which rules
out the initial structure (mutually exclusive where possible),
and that will work for the client in the particular context.
The pattern of cognitive self-defeat tends to be general-
ized across many client responses. Some approaches to
cognition-rational/linguistic intervention include a teach-
ing phase to help the client learn and use the intervention
process in his or her own behalf. Instruction and practice in
identifying defeating self-statements or cognitive distor-
tions, confronting the illogicality, and substituting better
structures are provided.
Conditions of cognition-rational/linguistic deficit seem
unusually easy to remedy. They involve a relatively simple
process of teaching the client what is missing. It may be
that the difficulty comes in recognizing that a deficit occurs.
Because the intervention is so simple, in many situations it is
used at the outset. If it works, then the deficit has been
filled. If it does not work, then the problem is more assur-
edly a cognition-rational/linguistic surplus or an affective
or behavioral one.
CONTEXTS FOR APPLICATION OF LEARNING MODELS
It is relatively easy to identify cognitive deficits or surpluses
or behavioral deficits or surpluses and apply integrally sound
strategies to resolve identified problems. There are, how-
ever, confounding variables that make the process more
complex. People learn their way through life, including both
successful coping responses and self-defeating or inappro-
priate ones. Intervention techniques are teaching techniques
geared to produce self-enhancing learned responses to
replace problematic ones.
So far, this article has described four major types of learn-
ing families, each with different dynamics and each particu-
larly useful for producing specific types of response. Because
people are different one from another in many ways, it is
necessary to consider characteristics of individual clients that
require modifications of intervention strategies. These char-
acteristics may block client receptiveness to a given approach,
or they may facilitate acceptance of another approach. They
can be categorized for discussion; however, it is important
to recognize that each client can be better served when
operating from an understanding of his or her unique cir-
cumstance and style and his or her context and response to
that context. Three categories of individual differences—
developmental, social, and spiritual dynamics—are considered
as contextual issues, things that need to be taken into account
when personalizing an intervention strategy for each particu-
lar client. Contextual issues are client variables that mandate

modification of learning-based intervention strategies. They
further define the circumstance and style of the client.
Developmental Context
As people develop through the various ages and stages of life,
they meet challenges or crises (Gilligan, 1982; Ginter, 1999;
Havighurst, 1949; Levinson, 1978; Sheehy, 1976). Preventive
mental health is concerned with anticipating and teaching for
such transitions. Counseling provides assistance for people who
have difficulty moving through these challenges.
Developmental norms are useful for identifying the prob-
lem dynamics encountered by people at various stages of
progress. They are helpful, too, in clarifying anomalies of
development that may typify a given client. It is helpful to
classify the kinds of abilities available to people of different
ages. Piaget (1952) has shown that there are qualitative dif-
ferences in the cognitive abilities of children as they grow
from very young to at least adolescence. Such data support
my contention that the intervention of choice for children
up to 8 years of age (plus or minus 2 years) and of people
with corresponding mental capacity is either some form of
operant conditioning or perceptual framing.
In addition to the obvious variation in intervention strat-
egy dictated by a client who is at the concrete operations
stage in contrast to the formal operations stage of Piaget
(Ivey & Ivey, 1998; Phillips, 1969), there are similar indica-
tors in other models. Consider the shift in motivation and
attentional focus between clients identified in Erickson’s (1980)
identity and intimacy stages. The kind of self-statements
invoked in a cognitive restructuring paradigm would be
likely to differ along the egocentric versus interpersonally
expansive lines. Likewise, clients identified in the Perry
Scheme (Moore, 1992), a developmental model incorpo-
rating both cognitive and value dimensions, as operating in
a dualistic mind-set, are markedly different from those work-
ing in contextual relativism. The extent to which they can
be expected to engage in self-directed experiences or to
which they rely on an external source of knowledge and
validation will dictate differences in application of inter-
ventions in each of the families of learning theory.
Within any of these developmental models, there are
adjustments required in the use of in vivo and vicarious or
fantasy generated experience and in determining whether
a client needs step-by-step modeling and practice of an iden-
tified experience or if he or she can carry out the prescribed
activity after receiving simple instructions.
Social Context
Systems theorists contend that it is counterproductive to
isolate the “identified client” for treatment and then intro-
duce that client back into an unchanged context (Bateson,
1958; Minuchin, 1974; Minuchin & Fishman, 1981; Ruesch
& Bateson, 1968). The system is the client. From the empha-
sis on multiculturalism (Gazda, Ginter, & Horne, 2001; Sue
& Sue, 1999) comes the awareness that group morés, ethics,
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Learning Theory and Intentional Intervention
customs, and traditions create limits on the accessibility of
client or system to intervention. They also indicate the types
of counselor-initiated activities that are acceptable. That
social context is a critical variable is attested to by Sue, Ivey,
and Pedersen (1996):
A person’s identity is formed and continually influenced by his or
her context. Working effectively with clients requires an under-
standing of how the individual is embedded in the family, which in
turn requires an understanding of how the family is affected by its
place in a pluralistic culture. . . . Virtually all clients who come for
help have cultural issues underlying their concerns. (pp. 15–16)
Consideration of the social context is a critical compo-
nent to effective counseling. This is supported with studies
of group influence, including peer group dominance for
adolescents, family dynamics (Ansbacher & Ansbacher,
1956), groupthink (Janis, 1967), and the Abilene paradox
(Harvey, 1974). The family is seen by Adler (Ansbacher &
Ansbacher, 1956) to be the primary and most influential
group for comparison and for establishment of lifestyle.
Other groupings are of major importance to the identity of
clients. They include gender (Baber, 1992; Beall & Sternberg,
1993; Gilligan, 1982; Minton, 1992; Notman & Nadelson,
1991), racial, cultural (Aponte, Rivers, & Wohl, 1995;
Axelson, 1993; Crane, 1997; Dunn & Griggs, 1995; Wehrly,
1995), regional (Good & Good, 1986), and socioeconomic
classifications. Some knowledge of major identity charac-
teristics common to such groups is helpful in working with
a particular client.
In the past, both psychologists and other people assumed that indi-
viduals’ behavior and achievements reflected their stable, internal
personality traits. In contrast, current research shows that behavior
and achievement, and even our personality traits, are heavily influ-
enced by other’s beliefs, perceptions, expectations, and treatment,
as well as by our knowledge of expectations for members of our
sex. . . . . Both men’s and women’s behavior, achievements, and
personality traits are far more quickly responsive to social situ-
ational forces than was previously supposed. (Beall & Sternberg,
1993, pp. 33–34)
Some readily apparent examples that illustrate the impor-
tance of social context sensitivity are clients presenting a “not-
okay” self-image (Harris, 1967) from early family socializa-
tion, the contrast between self-views and worldviews of fe-
male cohorts of age 50 and age 30, and the difference in
cultural structure evidenced by a Western focus on individual
responsibility and Eastern or Native American emphases. It
must be remembered that reliance on stereotypical charac-
teristics may blur what is uniquely important about the cli-
ent. It is necessary to attend to the client’s disclosure and
definition of what his or her affiliations mean to him or her.
Spiritual Context
Another aspect of client context is spirituality. This may or
may not include religion, although a large proportion of
clients either belong to a church or identify with a group
for whom spirituality is of high interest.
287
Gerber
Consideration of two major emphases seems warranted: (a) The ex-
tent to which clients define themselves relative to religious prin-
ciples, indoctrination, and practices as evidenced in self-descriptions
such as, “I am a devout Christian” or “I am not a very good Methodist”
or “I am an avid Buddhist” or “My parents/priest would be disappointed
in me;” and (b) The reliance on a relationship with a higher power,
informally as a support or more formally through prayer or medi-
tation. (Gerber, 1999, p. 183)
Membership in a church does not necessarily accompany a
high degree of religiosity or of spirituality. Sometimes member-
ship may apply more to a social context than a spiritual one.
The large portion of the world population projected to be church
affiliated is sufficient cause to recognize and to legitimize religious
or spiritual factors either as determinants or as modifiers of human
thought, attitude, emotion, and behavior. Even people who profess
anti-religious or anti-spiritual sentiment are defined, at least in part,
by the religious context they oppose. (Gerber, 1999, pp. 183–184)
If clients do not exist in a spiritual or religious vacuum
and if their experiences and manifest behavior are influ-
enced by spiritual or religious factors, then it is necessary
that therapists allow for and account for such factors. From
a counseling standpoint, less so from the scientist’s view-
point in which it is important to verify the existence of
spiritual entities and causal relationships between religious
practices and results, it is important to recognize the power
for direction within the client’s phenomenal experience and
from the client’s cognitive structure. Whether or not they
have been empirically isolated and their veracity and power
demonstrated, spiritual beliefs have significance for clients
(Bergin, 1991; Fukuyama & Sevig, 1997; Ingersoll, 1997;
Paloutzian, 1996; Richards & Potts, 1995).
Although many topics may have intense spiritual meaning
for a particular client, there are some that often have reli-
gious or spiritual implications. These include suicide, extra-
marital sex, criminality, homosexuality, poor self-image,
questions of self-worth, coping with death, rebellion against
and feelings of disrespect for parents, and family break-up.
These may be apparent in clients who claim a degree of spiri-
tuality without religious affiliation. The conceptualizations
may be ill defined and vague (religion puts words on them,
with the risk of distorting and being dogmatic). Because the
client is the resident expert on himself or herself, it is not
necessary to have an exhaustive knowledge of his or her
professed religious affiliation or spiritual philosophy. Such
knowledge provides a stereotypical and dogmatic structure
of his or her abstract spiritual nature. Of more importance is
the meaning and experience of the client’s spiritual context
from his or her phenomenal viewpoint.
CASE EXAMPLE
The intent of this article was to give a relatively brief over-
view of a complex process for personalizing counseling. For
that reason, the segments of the model were described in
simple, relatively uncomplicated ways. Examples of each
aspect of the model and of various integrations of model
dynamics would be helpful in gaining a complete under-
288 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
standing. For more complete explanations and for numer-
ous case examples, the reader is referred to Gerber (1999).
One case, paraphrased from the reference text, will serve
as an example of client circumstance and style and of two
optional intervention strategies, each based on learning
theory and each being internally consistent.
The client is a 10-year-old girl, diagnosed as having a learn-
ing disability. She attends a combination of special education
and regular classes. Teachers report that she is disruptive,
nonattentive, and noncompliant. Her personal history indi-
cates that she was born to a single parent with whom
she spent the first 8 years. Her mother has a history of
substance abuse and had been reported on numerous
occasions to Child Protective Services for alleged child
neglect. At age 9 she began a series of placements in foster
homes, ending with assignment to her maternal grandpar-
ents who now have permanent custody. She has a history
of malnutrition, although at present her health is normal.
In classroom observations, her behavior was generally age
appropriate. She typically hesitated on introduction to new
activities then became fully invested. Her class disruptions
bordered on being “cute,” never so bad as to cause major
problems and yet too obvious to ignore. Variations in her
attention were random. Noncompliance occurred mostly
when she was challenged with misbehavior or when experi-
encing difficulty with a task. The special education teacher
used standard operant procedures for most tasks. The client
was inconsistent in her responses, on some occasions jump-
ing right into active involvement with moderate success
and on other occasions showing either daydreaming or off-
task activity. One peculiar response of note was her insist-
ing to sit close to her regular classroom teacher during story
time—sitting at the feet of the teacher and often leaning
against the teacher’s leg. Another observational vignette
showed her not being aware of which problems she was to
answer in class during oral math review yet, when oriented
to a specific item, giving marginally correct answers.
An interview with the grandparents disclosed the follow-
ing concerning her transition into their home. At first, she
ate more than necessary, almost as though she did not trust
that there would be more. She lied about insignificant things
such as what she was watching on television. The grandpar-
ents had received direction in setting ground rules, chart-
ing responses, and contracting for privileges. This approach
had some moderating effect but was not working consis-
tently. The grandmother was quite concerned about the cli-
ent. The grandfather tended to say, “She’ll be okay. Just give
her time.” He especially enjoyed having her follow him
around the farm as he did his evening chores.
In an interview with the client, most responses and ac-
tions were typical for her age. She answered questions halt-
ingly and watched the interviewer intently as she answered.
Rephrases of the same question elicited varying responses.
When given an invitation to freely describe a school expe-
rience, she started readily then hesitated, changing her story
frequently, again watching the interviewer intently as though
for guidance or affirmation of her story.

One interpretation of this behavior just described is that
the client has been conditioned by an unpredictable reality to
manifest externally cued survival responses. At the same time,
she did not accomplish a repertoire of socially acceptable
responses. Because of the developmental level of the student
(i.e., 6–10 years of age) and taking the diagnosis of learning
disability into account, I suggest it is likely that she will
respond best to an operant intervention strategy. Inconsistencies
in her responses at school and home are due to three problems
in application of operant procedures: (a) inconsistency on
the part of teachers and grandparents, (b) absence of system
adjustments to maintain progress in target response, and
(c) lack of working programs for important target responses
such as staying on task, appropriately responding to adults,
increased production between reinforcements, monitoring
or “staying with” class assignments, and being ready to respond
at her turn. The intervention strategy that follows from this
involves identifying one or two unwanted responses in each
setting—regular classroom, special education classroom, and
grandparents’ home—and building a behavior management
program for each. As the client consistently shows desired
target response, additional wanted and not present responses
will be programmed for reinforcement.
Here is an alternate set of assumptions and a second
intervention strategy that logically follows. The client has
some major developmental deficiencies due to a lifetime
of near-noncontingency relationships between activity and suc-
cess. She has been striving to earn acceptance in a world of
changing rules. She was neither valued nor accepted as a per-
son, no matter what her condition and no matter what her
response. From the cognition-perceptual family of learning
theory, it can be asserted that the client is in a deficit state, not
having had the experiences that provide a basis for normal,
socialized responses. Her inconsistency in attention, variability
in responses—both acceptable and unacceptable—to the
same situation, seeking of personal closeness and physical
contact in the nonthreatening story context, and her fitting
in as a companion in farm chores all indicate a need for
personal validation.
One strategy that fits this frame requires the creation of
periods of time and activities for which quantity or quality
of performance does not matter. The critical factor is for the
client to receive adult attention and acceptance, no matter
what her actions have been. One example is already in place—
her accompanying her grandfather on his chores during which
time he focuses on her as an active participant with no stan-
dards for acceptable performance. With the development of
a feeling of security and worth in these settings, the despera-
tion to be okay in other situations will subside and she will
naturally pay attention more consistently and perform more
often within the limits of acceptability. From a Maslovian
perspective, once she has learned that she is okay, fulfilled
her love and belonging needs, and accomplished personal
validation, she will be able to function at the higher level of
contingency based, self-esteem activities.
The two strategies just outlined operate on mutually ex-
clusive principles. One uses strict adherence to contingency
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Learning Theory and Intentional Intervention
relationships, and the other succeeds by avoiding contin-
gency dynamics. Which is better? The one that works is. At
this juncture, it is too early to judge. The selection of strat-
egy is one of the professional decisions inherent in the func-
tion of counselor. Having a rationale and a theory base that
permits efficient management of therapeutic dynamics per-
mits the counselor to begin with the strategy chosen and to
evaluate its immediate results. Obviously if it does not work
as expected, selecting an alternative strategy is indicated.
This particular case example has obvious and strong impli-
cations for tailoring the intervention in response to develop-
mental and social contexts. It does not, however, include refer-
ence to the spiritual context. If there existed strong religious
opposition from those involved to the use of extrinsic rein-
forcers or if there were moralistic intrusions regarding the
rightness or wrongness of single parenting, or conceptions of
the client being “bad” or “evil,” perhaps some variations in ap-
proach would be necessary. There is nothing inherent in this
approach that requires searching for obscurities in the client’s
circumstance. It works from a phenomenological base of an
expanding counseling-mediated perception shared by the
client and counselor (Stewart, 1983).
REPRISE AND CONCLUSION
The profession of counseling (i.e., training programs) has
moved away from its theoretical foundations (Weinrach,
1991). A rudderless ship is an appropriate metaphor. It might
be supposed that the movement away from learning theory
within the profession was purposeful and probably resulted
from the previous assessment of learning theory as being
ineffectual. A functional solution is possible.
A careful study of learning theories and of counseling
interventions interpreted from their learning theory under-
pinnings results in a blending of many seemingly diverse
positions into two classes, four families, and three contexts.
The two classes are one-stage and two-stage models. The
four families are association, reinforcement, cognition-
perceptual, and cognition-rational/linguistic. The contexts
are developmental, social, and spiritual. It is possible for a
counselor to learn at least one approach within each of these
four families, with awareness of modifications to fit the
contexts, and integrate them into a repertoire rich enough
to match virtually all of his or her clients.
Whether or not a new model is particularly appealing to
counselor educators and practitioners depends on their re-
ceptiveness and the model’s salience. This particular model
represents an important set of dynamics, a means to the
end of truly being intentional and accountable in the prac-
tice of the counseling profession.
REFERENCES
Ansbacher, H., & Ansbacher, R. (1956). The individual psychology of Alfred
Adler. New York: Basic Books.
Aponte, J., Rivers, R., & Wohl, J. (1995). Psychological interventions and
cultural diversity. Boston: Allyn & Bacon.
289
Gerber
Ausubel, D. (1963). The psychology of meaningful verbal learning: An
introduction to school learning. New York: Grune & Stratton.
Axelson, J. (1993). Counseling and development in a multicultural soci-
ety (2nd ed.). Pacific Grove, CA: Brooks/Cole.
Baber, A. (1992). Naked at gendergap: A man’s view of the war between
the sexes. New York: Carol Publishing Group.
Bandler, R., & Grinder, J. (1975). The structure of magic. Palo Alto, CA:
Science and Behavioral Books.
Bandura, A. (1969). Principles of behavior modification. New York: Holt,
Rinehart, and Winston.
Bandura, A. (1974). Behavior theory and the models of man. American
Psychologist, 29, 859–869.
Bateson, G. (1958). Naven, a survey of the problems suggested by a com-
posite picture of a New Guinea tribe drawn from three points of view.
Stanford, CA: Stanford University Press.
Beall, A., & Sternberg, R. (1993). The psychology of gender. New York:
Guilford Press.
Beck, A., Rush, A., Shaw, B., & Emery, G. (1979). Cognitive therapy of
depression. New York: Guilford Press.
Bergin, A. (1991). Values and religious issues in psychotherapy and
mental health. American Psychologist, 46, 397–403.
Berne, E. (1961). Transactional analysis in psychotherapy. New York:
Grove Press.
Bruner, J. (1966). Toward a theory of instruction. New York: Norton.
Council for Accreditation of Counseling and Related Educational Pro-
grams. (1998). CACREP standards 2000, second draft. Unpublished
manuscript.
Crane, R. (1997, February 10). Counseling multicultural clients. Ad-
vance: For Speech-Language Pathologists & Audiologists, 7(6), 7.
Dunn, R., & Griggs, S. (1995). Multiculturalism and learning style: Teach-
ing and counseling adolescents. Westport, CT: Praeger.
Ellis, A. (1958). Rational psychotherapy. Journal of General Psychology,
59, 35–49.
Ellis, A. (1962). Reason and emotion in psychotherapy. New York: Lyle
Stuart and Citadel Press.
Erickson, M. (1980). The collected papers of Milton H. Erickson on hyp-
nosis (E. Rossi, Ed.). New York: Irvington.
Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA:
Stanford University Press.
Försterling, F. (1986). Attributional conceptions in clinical psychol-
ogy. American Psychologist, 41, 275–285.
Fukuyama, M., & Sevig, T. (1997). Spiritual issues in counseling: A
new course. Counselor Education and Supervision, 36, 224–232.
Gazda, G., Ginter, E., & Horne, A. (2001). Group counseling and group
psychotherapy: Theory and application. Boston: Allyn & Bacon.
Gerber, S. (1999). Enhancing counselor intervention strategies: An
integrational viewpoint. Philadelphia: Accelerated Development/
Taylor & Francis.
Gilligan, C. (1982). In a different voice: Psychological theory and women’s
development. Cambridge, MA: Harvard University Press.
Ginter, E. (1999). David K. Brooks’ contribution to the developmen-
tally based life-skills approach. Journal of Mental Health Counseling,
21, 191–202.
Gladding, S. (1996). Counseling: A comprehensive profession (3rd ed.).
Englewood Cliffs, NJ: Merrill.
Good, B., & Good, M. (1986). The cultural context of diagnosis and
therapy: A view from medical anthropology. In M. Miranda & H. Kitano
(Eds.), Mental health research and practice in minority communities:
Development of culturally sensitive training programs (DHHS Publica-
tion No. ADM 86-1466, pp. 1–8). Rockville, MD: National Institute
of Mental Health.
Harris, T. (1967). I’m OK—you’re OK: A practical guide to transactional
analysis. New York: Harper & Row.
Harvey, J. (1974). The Abilene paradox: The management of agreement.
Organizational Dynamics, 3, 63–80.
Havighurst, R. (1949). Developmental tasks and education. Chicago: Uni-
versity of Chicago Press.
Hergenhahn, B., & Olson, M. (1993). An introduction to theories of learn-
ing. Englewood Cliffs, NJ: Prentice-Hall.
290 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Ingersoll, E. (1997). Teaching a course on counseling and spirituality.
Counselor Education and Supervision, 36, 224–232.
Ivey, A., & Ivey, M. (1998). Reframing DSM-IV: Positive strategies from
developmental counseling and therapy. Journal of Counseling & De-
velopment, 76, 334–350.
Janis, I. (1967). Victims of groupthink: A psychological study of foreign
policy decisions and fiascos. Boston: Houghton-Mifflin.
Lawless, L., Ginter, E., & Kelly, K. (1999). Managed care: What mental
health counselors need to know. Journal of Mental Health Counseling,
21, 50–65.
Levinson, D. (with Darrow, C., Klein, M., & McKee, B.). (1978). The
seasons of a man’s life. New York: Knopf.
Maslow, A. (1962). Toward a psychology of being. Princeton, MA: Van
Nostrand.
Minton, H. (Ed.). (1992). Gay and lesbian studies. New York: Haworth.
Minuchin, S. (1974). Families and family therapy. Cambridge, MA:
Harvard University Press.
Minuchin, S., & Fishman, H. (1981). Families therapy techniques. Cam-
bridge, MA: Harvard University Press.
Moore, W. (1992). Student and faculty epistemology in the college class-
room: The Perry scheme of intellectual and ethical development. Un-
published manuscript.
Mueller, R., Dupuy, P., & Hutchins, D. (1995). A review of the TFA
counseling system: From theory construction to application. Journal
of Counseling & Development, 72, 573–577.
Notman, M., & Nadelson, C. (Eds.). (1991). Women and men: New perspec-
tives on gender differences. Washington, DC: American Psychiatric Press.
Paloutzian, R. (1996). Invitation to the psychology of religion. Boston:
Allyn & Bacon.
Perls, F. (1969). Gestalt therapy verbatim. Moab, UT: Real People Press.
Phillips, J. (1969). The origin of intellect: Piaget’s theory. San Francisco:
Freeman.
Piaget, J. (1952). The origins of intelligence in children. New York: Inter-
national Universities Press.
Preston, J. (1998). Integrative brief therapy. San Luis Obispo, CA:
Impact.
Richards, P., & Potts, R. (1995). Using spiritual interventions in psycho-
therapy: Practices, successes, failures, and ethical concerns for Mor-
mon psychotherapists. Professional Psychology Research and Practice,
26, 163–170.
Rogers, C. (1942). Counseling and psychotherapy. Boston: Houghton Mifflin.
Rogers, C. (1951). Client-centered therapy. Boston: Houghton Mifflin.
Rogers, C. (1961). On becoming a person. Boston: Houghton Mifflin.
Rokeach, M. (1973). The nature of human values. New York: Free Press.
Ruesch, J., & Bateson, G. (1968). Communication: The social matrix of
psychiatry. New York: Norton.
Sahakian, W. (1976). Introduction to the psychology of learning. Chicago:
Rand McNally.
Seligman, M. (1975). Helplessness: On depression, development, and
death. San Francisco: Freeman.
Sheehy, G. (1976). Passages: Predictable crises of adult life. New York:
Bantam Books.
Stewart, J. (1983). Interpretive listening: An alternative to empathy.
Communication Education, 32, 379–391.
Sue, D., Ivey, A., & Pedersen, P. (1996). A theory of multicultural coun-
seling and therapy. Pacific Grove, CA: Brooks/Cole.
Sue, D., & Sue, D. (1999). Counseling the culturally different: Theory and
practice (3rd ed.). New York: Wiley.
Thorne, F. C. (1950). Principles of personality counseling: An eclectic view-
point. Brandon, VT: Journal of Clinical Psychology.
Tolman, E. (1949). Purposive behavior in animals and men. Berkeley:
University of California Press.
Watzlawick, P. (1978). The language of change: Elements of therapeutic
communication. Palo Alto, CA: Basic Books.
Wehrly, B. (1995). Pathways to multicultural counseling competence: A
developmental journey. Pacific Grove, CA: Brooks/Cole.
Weinrach, S. (1991). Selecting a counseling theory while scratching
your head: A rational-emotive therapist’s personal journey. Journal of
Mental Health Counseling, 13, 367–378.

Wertheimer, M. (1959). Productive thinking, enlarged edition. New York:
Harper.
Williamson, E. G. (1939). How to counsel students: A manual of techniques for
clinical counselors. New York: McGraw-Hill.
Williamson, E. G. (1950). Counseling adolescents. New York: McGraw-Hill.
Wolpe, J. (1969). The practice of behavior therapy. New York: Pergammon
Press.
APPENDIX A
Two Applications of Pavlovian-Based Interventions
a. existing reflex—new stimulus.
1. Identify the US [unconditioned stimulus] for the
existing reflex (UR).
2. Select a substitute stimulus (CS) to which the re-
flex is to be conditioned.
3. Pair the US and the CS contiguously over many tri-
als until presentation of the CS alone elicits the re-
flex that is now the CR.
b. existing stimulus—undesirable conditioned response.
1. Identify the US for that undesired reflex (UR1) that
has become a conditioned response (CR1) elicited
by the existing stimulus (CS1).
2. Determine a desired reflexive response (UR2) that
is mutually exclusive of the undesired reflex (UR1).
3. Find the stimulus (US2) that elicits the desired
reflex (UR2).
4. Pair US2 with several previously neutral stimuli
(CS2) so as to gain stimulus control over UR2.
5. Present CS2 to elicit CR2 at the same time as a
diluted form of CS1, resulting in only the elicitation
of CS2.
6. Gradually increase the intensity of CS1 while CS2
is eliciting CR2 until CS1 becomes an eliciting stimu-
lus only for the desired reflex, now CR2. (Gerber,
1999, p. 66)
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Learning Theory and Intentional Intervention
Endnote. The following copyright information applies to this
and subsequent quotations for this work. From Enhancing
Counselor Intervention Strategies: An Integrational Viewpoint,
by S. Gerber, 1999. Copyright 1999 by Taylor & Francis, Inc.,
http://www.routledge-ny.com. Reprinted with permission.
APPENDIX B
Intitial Determination in Guthrian-Based Interventions
a. new response.
1. Identify or create obvious contextual elements that
consistently occur and can be used as cues.
2. Cause the desired response to occur in the presence
of the cues, making certain that the cues are viable
and are attended to by the learner.
3. Repeat step two numerous times so that the various
elements of the complex cues can be associated with
the desired response.
4. Change the situation so that no subsequent response
is associated with the same cues.
b. replace response.
1. Determine a response that is to replace the undesir-
able response. Note that non-response, e.g., remain-
ing quiet and still, is an option as a replacement
response for inappropriate activity.
2. Identify the situation that cues the undesirable
response.
3. In the presence of those cues, cause the substitute
response to occur. This can be done by direct ma-
nipulation or by using the fatigue or incompatible
response approaches to habit change.
4. Change the situation so that no subsequent re-
sponse is associated with the same cues. (Gerber,
1999, p. 66)
291
 RESEARCH
Clinical Judgment in Case Conceptualization and
Treatment Planning Across Mental Health Disciplines

Janet E. Falvey

This study compared clinical judgment on case conceptualization and treatment planning (TP) tasks among psychologists,
mental health counselors, and social workers using Clinical Treatment Planning Simulations depicting anxiety and affective
disorders. Orientation and work setting did not significantly influence judgment; experience exerted a “modest” curvilinear
effect. Mental health counselors scored higher than social workers and similar to psychologists on both tasks. The best predic-
tors of TP scores for the anxiety disorder were interventions addressing stress management and family dynamics; strong
predictors for the affective disorder included focus on family dynamics and the therapeutic relationship. Implications of findings
for training are discussed.

O
ver the past decade, the managed care in-
dustry has become the new gatekeeper of
mental health practice (Lawless, Ginter, &
Kelly, 1999). Access to service delivery has
increasingly shifted from state licensing
boards to managed care provider panels, where profes-
sionals competing for contracts submit credentials such
as education, licenses, work setting, experience, and client
utilization patterns (Shueman, 1997; Stromberg, Ratcliff,
& Scheutze, 1997). However, empirical links between these
practitioner variables and treatment outcomes are equivo-
cal at best (Dawes, 1994; Garb, 1998). The use of such
criteria as benchmarks of clinical competency thus remains
poorly validated. The impetus for these professions to
specify factors that enhance clinical practice has therefore
intensified, fueled by health policy issues and consumer
protection interests.
In support of common clinical opinion, some research
suggests that demographics such as profession, orientation,
degree, work setting, and experience inform diagnosis, case
conceptualization, and treatment planning (Bishop & Richards,
1984; Falvey, 1992b; Garb, 1989; Gil-Ali & Newman, 1984;
Plous & Zimbardo, 1986). For example, Lambert and
Wertheimer (1988) found that diagnostic accuracy in
reviewing clinical case histories improved significantly with
relevant education and experience among graduate students
and paraprofessionals. Kopta, Newman, McGovern, and
Sandrock (1986) used two 10-page case vignettes to com-
pare conceptualization, interventions, and treatment costs
among psychologists and doctoral-level interns with vary-
ing theoretical orientations, finding significantly distinct
patterns of conceptualization and intervention across
orientations. Houts (1984) also found significant differences
across orientation for client responsiveness and treatment
prognosis among doctoral-level trainees. Butcher and Scofield
(1984) used a standardized simulation and process-tracing
approach to assess judgment in case review and treatment
planning among 15 practitioners with 2–5 years postdegree
experience, finding evidence that either advanced education
or increased experience enhanced judgment on these tasks.
Falvey and Hebert (1992) found significant differences in treat-
ment planning among a sample of 137 clinical mental health
counselors and 62 graduate trainees reviewing standardized
case simulations.
Other findings, however, conclude that clinician demo-
graphics do not seem to significantly influence treatment
decisions (Dawes, 1989; Faust et al., 1998; Rock, Bransford,
Maisto, & Morey, 1987; Turner & Kofoed, 1984; Worthington
& Atkinson, 1993) or that they do not usefully distinguish
how clinicians think and process client data (Spengler &
Strohmer, 1994; Strohmer & Spengler, 1993). For example,
Clavelle and Turner (1980) compared paraprofessionals,
social workers, and psychologists on judgments regarding
the need for hospitalization and medication using two simu-
lated intake interviews, finding no greater consensus in de-
cision making among professionals than paraprofessionals.

Janet E. Falvey is an associate professor of counseling in the Department of Education at the University of New Hampshire and a practicing psychologist.
Correspondence regarding this article should be sent to Janet E. Falvey, Morrill Hall Room 207, University of New Hampshire, Durham, NH 03824.

292 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79

 C l i n i c a l J u d g m e n t i n C a s e C o n c e p t u a l i z a t i o n a n d Tr e a t m e n t P l a n n i n g
O’Donohue, Fisher, Plaud, and Curtis (1990) used a struc-
tured interview and retrospective self-report method to
review three recently terminated cases of each of 25 clini-
cians from a community mental health center to assess their
assessment methods, treatment goals and methods, finding
the lack of any systematic decision process for these ac-
tions in over 90% of the 75 cases reviewed. Turner and
Kofoed sampled five disciplines and three work settings
among 75 mental health practitioners reading an inpatient
case report and selecting treatment options, finding that
work setting had no impact and only social workers showed
indication of any “set” related to treatment planning.
Research to date thus remains inconclusive regarding
the impact of therapist demographics on common clinical
judgment tasks. Equally disturbing are conflicted findings
regarding whether experienced or even expert clinicians
outperform trainees or lay persons on diagnostic, assess-
ment, and treatment planning tasks (Berman & Norton,
1985; Christensen & Jacobson, 1994; Dawes, 1994; Falvey
& Hebert, 1992; Garb, 1989; Kleinmuntz, 1990; Mahoney,
1988; Shaw & Dobson, 1988). Given that such demograph-
ics often dictate access to professional practice, credentialing,
and reimbursement, it seems critical to extend this research.
Case conceptualization and treatment planning are uni-
versal clinical judgment tasks that lend themselves to such
efforts. From intake to termination, clinicians must gather
and analyze case information, formulate hypotheses, and
implement treatment decisions. Given the quantity and
ambiguity of information presented, this is a daunting task.
Complex clinical data will not “speak for itself”; it must be
interpreted (Kleinmuntz, 1990; O’Donohue et al., 1990;
Turk & Salovey, 1988). Providing the framework for subse-
quent interventions, conceptualization and treatment plan-
ning go beyond simple diagnostic matching to consider the
symptoms, context, and history of a problem; the demand
characteristics of the therapeutic environment; client goals
and motivation; and anticipated process dynamics in treat-
ment (Makover, 1996; Mordock, 1994; O’Donohue et al.,
1990; Tillett, 1996). Moreover, these processes are linked.
To understand how clinicians make treatment decisions, one
needs to understand how they formulate case histories
(Garb, 1998). Efforts to identify cognitive strategies and
biases that characterize these judgment tasks have yielded
intriguing findings to date.
First, limited human information-processing capacity affects
how practitioners collect, synthesize, and recall information
in complex judgment tasks (Faust, 1986; Garb, 1998;
Kleinmuntz, 1990). Cognitive shortcuts (heuristics) and
judgment biases are common. A considerable body of research
concurs that mental health clinicians (including experts)
selectively screen client data, generate few alternative hypoth-
eses early in the assessment process, retain initial diagnoses
even in the presence of subsequent refuting evidence, and
tend to select treatments that confirm (rather than test) their
initial judgments (Ambady & Rosenthal, 1992; Elstein,
Shulman, & Sprafka, 1978; Garb, 1998; Hogarth, 1987;
Kleinmuntz, 1990; Schwartz & Griffin, 1986; Snyder &
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Thomsen, 1988; Turk & Salovey, 1988; Tutin, 1993). Judgment
is frequently influenced by heuristics such as availability
(ease of recall of relevant information), representativeness
(similarity to other cases or theoretical models), illusory
correlation (judgment based on personal belief rather than
standard criteria), and overconfidence (increasing confidence,
but not judgment accuracy, with additional case informa-
tion) (Faust, 1986; Garb, 1998; Turk & Salovey, 1988).
Cognitive complexity, the ability of an individual to inte-
grate large amounts of multidimensional information, clearly
influences clinical decision making (Spengler & Strohmer,
1994). However, there are few empirical studies of the cog-
nitive processes of clinicians engaged in treatment planning
(Garb, 1998). One related line of inquiry has examined
patterns of case review: the sources, quantity, and sequence
of clinical data gathering. Findings suggest that the order in
which case material is accessed and how much of it is
reviewed may affect treatment decisions (Butcher & Scofield,
1984; Butcher, Scofield, & Baker, 1985; Chapman, Bergus, &
Elstein, 1996; Ellis, Robbins, Schult, Ladony, & Banker, 1990;
Elstein et al., 1978; Friedlander & Stockman, 1983). Experts
tend to review less but more relevant data than do less experi-
enced clinicians in the assessment process. There is also some
evidence that clinical judgment may be case specific; that is,
performance on one case may not predict performance on
another case (Elstein et al., 1978; Falvey & Hebert, 1992;
Luborsky et al., 1986; O’Donohue et al., 1990).
Given lack of empirical consensus on the influence of
professional variables on clinical judgment to date, the cur-
rent study replicates and extends previous research by ex-
ploring the impact of practitioner demographics, case re-
view processes, and clinical focus on case conceptualization
and treatment planning across two common mental health
disorders. Participants from three professions responded to
standardized case simulations representing anxiety and af-
fective disorders. Performance was evaluated on two dimen-
sions: case conceptualization (CC), the ability to identify
important signs, symptoms, and dynamics of a client; and
treatment planning (TP), the ability to synthesize this
information into a viable intervention strategy. These skills
are widely recognized as critical case management compe-
tencies for all mental health providers (Garb, 1998; Makover,
1996; Mordock, 1994; O’Donohue et al., 1990; Strupp &
Butler, 1990; Tillett, 1996; Waddington, 1997).
This study was designed to explore four research questions
regarding performance on these standardized simulations:
(a) Is clinical judgment consistent or case specific across
two distinct disorders? (b) What is the impact of therapist
demographics (profession, work setting, clinical experience,
and orientation) on CC and TP scores? (c) What factors
in the case review process affect performance on these
judgment tasks? and (d) What areas of clinical focus
enhance performance on these judgment tasks? The study
also assessed content validity of the Clinical Treatment
Planning Simulations (CTPS; Falvey, 1992b) for assess-
ing clinical judgment among master’s- and doctoral-level
therapists.
293
Falvey

METHOD notes). Information in these reports was generated from

criteria identified and ranked by national panels of inter-
Participants disciplinary experts on affective disorders (N = 56) and
anxiety disorders (N = 50). These experts included research-
One hundred sixty-eight participants included 49 psycholo-
ers and practitioners from six professions, averaging 15 years
gists and 59 social workers from a tristate New England
of clinical, and 7 years of research experience with these
region and 60 clinical mental health counselors from East
disorders, and documenting an average of four refereed pub-
Coast states. Demographics were obtained from the one-
lications and seven national/international presentations ad-
page biodata form submitted by participants. All profes-
dressing these disorders (Falvey, 1992b). The CTPS cases pro-
sionals were state licensed or certified, reporting a mean of
vide standard stimuli for assessing clinical judgment in case
13.7 years of clinical experience (SD = 4.1, range 2–46
conceptualization, diagnostic, and treatment planning tasks.
years). The sample comprised 86 women and 82 men, with
Interrater coding reliability for these cases has been reported
a mean age of 45.4 (SD = 9.4) and a range of 29 to 74 years.
as high (.88), and the simulations have demonstrated con-
Ethnic representation was 155 Caucasian (92%), 9 African
tent and discriminant validity across experts, experienced
American (5%), and 4 Hispanic (2%) participants. Primary
practitioners, and trainees (Falvey & Hebert, 1992).
clinical orientations were reported as psychodynamic (n =
Two scores are derived from written responses to each
67, 40%), cognitive-behavioral (n = 52, 31%), family sys-
CTPS simulation. A Case Conceptualization (CC) score
tems (n = 28, 17%), and humanistic (n = 21, 13%). Al-
consists of the following information:
though most cited various work settings during their ca-
reers, nearly half (n = 79, 47%) of these professionals were
1. Initial assessment sources—specific interviews, reports
currently working in independent practice, whereas 50
and referrals reviewed or requested to gain familiarity
(30%) worked in agencies, 19 (11%) in inpatient settings,
with the case (e.g., intake interview, medical records)
and 20 (12%) in academic settings. In self-reporting areas
2. Relevant history and status—symptoms, client presen-
of clinical expertise, 70 clinicians (42%) reported expertise
tation, and aspects of current or developmental his-
in treating anxiety disorders, and 106 (63%) cited expertise
tory identified as salient in the written case summary
in treating affective disorders, the two diagnoses represented
in this research. All participants reported membership in
The Treatment Planning (TP) score includes the follow-
the primary professional organization of their discipline (e.g.,
ing information:
American Psychological Association [APA]; American
Counseling Association [ACA]; National Association of
1. Clinical considerations—recommendations regarding
Social Workers [NASW]), and most listed additional pro-
format of therapy, its expected frequency and dura-
fessional affiliations.
tion, prognosis, and anticipated process dynamics (e.g.,
Chi-square analyses were computed to examine poten-
resistance, motivation)
tial distinctions across the three groups of professionals
2. Treatment interventions—clinical or medical interven-
participating in this study. There were no significant differ-
tions, or both, and referrals identified as relevant in
ences in years of experience across groups. Psychologists
the treatment protocol
differed significantly in work setting from both master’s-
3. Treatment goals—changes in target areas of intrapsy-
level groups (χ2 = .002, p < .01); they more frequently re-
chic and interpersonal functioning specified in the
ported independent practice (61%) rather than agency
treatment protocol
(10%) employment. By contrast, more mental health coun-
selors and clinical social workers were employed in agency
Written responses are coded and computer scored using weights
settings (33% and 43%, respectively). Significant differences
assigned by the experts who developed CC and TP criteria for
in two primary clinical orientations were also distinguished
each CTPS case. Further information on coding and scoring
across professions (χ2 = .0009, p < .001). Twenty-four psy-
procedures is described in Falvey and Hebert (1992).
chologists (49%) reported a cognitive-behavioral orienta-
To assess content validity of the CTPS cases, all participants
tion, whereas 32 clinical social workers (54%) reported a
were asked to complete a 10-item CTPS Evaluation Form at
psychodynamic orientation. Among clinical mental health
the conclusion of each of the two cases. A Likert scale (1 = not
counselors, 22 (37%) reported each of these orientations.
at all, 2 = slightly, 3 = moderately, 4 = significantly, 5 = extremely)
was used to rate the format (clarity of instructions, clarity of
Instruments
material presentation, task adequate to reflect their clinical
Two CTPSs (Falvey, 1992b) consist of standardized case experience); fidelity (degree of realism of case file, compre-
simulations representing anxiety and affective disorders. hensiveness of case as an initial evaluation, reflects actual
Each CTPS case includes eight sealed reports containing intake procedures); and utility (usefulness for teaching/
information commonly found in mental health files (i.e., training programs, for self-assessment by clinicians, in com-
emergency services contact, request for services, intake petency assessment for credentialing purposes) of each
evaluation, previous treatment records, medical history, CTPS case. Participants were also asked to estimate the
couples’ interview, current medical report, and progress degree of difficulty and time spent completing the case.

294 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79

 C l i n i c a l J u d g m e n t i n C a s e C o n c e p t u a l i z a t i o n a n d Tr e a t m e n t P l a n n i n g
Procedure
An informational mailing was sent to all current APA and
NASW members in a tristate New England region and to
all certified clinical mental health counselors (CCMHCs)
in Eastern coastal states, describing the project as an inter-
disciplinary study of clinical decision making and inviting
their participation if currently a mental health provider.
No additional incentive or reimbursement for participa-
tion was offered. A total of 358 professionals expressed
interest by returning an enclosed biodata form on which
they summarized graduate training, licenses/certifications,
years of clinical experience, areas of expertise, clinical work
setting(s), orientation(s), and professional affiliation(s). On
receipt of the biodata forms, these practitioners were assigned
a three-digit ID number and mailed the following research
materials: two CTPS case simulations representing adults
with an anxiety and an affective disorder, respectively;
response forms with instructions for reviewing each CTPS
case file and writing a case conceptualization and treat-
ment plan; a 10-item CTPS Evaluation Form for each case,
and a stamped manila envelope for returning all materials
to the researcher on completion.
From among this pool of practicing psychologists, social
workers, and mental health counselors, 168 professionals
(47%) subsequently completed and returned the research
materials. Given the large number of nonparticipants (n =
190) from among those initially interested, chi-square analy-
ses were computed across available demographic variables
(age, sex, experience, expertise) of the two groups. One
significant finding emerged, distinguishing years of experi-
ence (χ2 = 8.91, p < .01). Nonparticipating clinicians aver-
aged less clinical experience (M = 10.3 years) than did par-
ticipants in this study (M = 13.7 years).
In reviewing the CTPS cases, participants were con-
fronted with two judgment tasks: analysis—abstracting
relevant client data into a case conceptualization; and
synthesis—assembling this information into a viable treat-
ment plan (Smith, 1983). Instructions were printed on
the response forms in the sequence in which they were
followed. First, participants recorded which reports they
chose to review in each CTPS case file and the order in
which they were chosen. When they determined that they
had enough information about the client, they were instructed
to write a case conceptualization (CC) including salient
client demographics, symptoms, clinical presentation, psy-
chosocial history and their clinical and diagnostic impres-
sions. The second task was to write a treatment plan for
that client using whatever format they chose. Third, partici-
pants were requested to complete the CTPS Evaluation
Form included with each case and return all materials to
the researcher.
Thirteen participants (8%) returned only one CTPS
case; they were included in all analyses except for the
assessment of judgment consistency. Each participant’s
ID number was recorded on all returned materials to as-
sure anonymity of responses.
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
RESULTS
Coding and Scoring
Two trained raters independently coded all CTPS responses
(N = 323) for the three professions represented in this study.
Raters were both licensed psychologists and counselor edu-
cators; one also held national counselor certification. One
rater had 26 years experience as an educator/researcher and
13 years of clinical experience. The other rater had 12 years
experience as an educator/researcher and 16 years of clini-
cal experience. Both raters were trained in content analysis
and had participated in pilot testing the coding keys. The
coding key for each disorder consists of a one-page listing
of all CC and TP categories of information identified by
the expert panel. Raters marked all categories identified by
a respondent in their essay response to the CC and TP tasks.
Raters had no information regarding individual respondents
during the coding process.
Interrater reliability for these raters using the point-by-
point statistic (Kelly, 1977) ranged from .85 to .89 across
cases, with overall reliability at .87. A computer program
scored all coded responses, deriving a CC and TP score for
each case based on expert rankings of the relevance/impor-
tance of each possible category. Table 1 presents the range
and mean scores for respondents completing each disorder.
CTPS Content Validity
CTPS format (clarity of instructions and case materials)
was rated highly by psychologists across both cases (Ms =
4.1), and only slightly less so for social workers (Ms = 4.0)
and mental health counselors (M = 3.9; M = 3.8). Fidelity
of the cases in representing actual clinical judgments tasks
was rated similarly, with means between 3.9 and 4.2 across
professions for both disorders. Utility, the usefulness of these
simulations for specific purposes, yielded slightly more di-
verse ratings. For teaching and training purposes respon-
dents rated the CTPS cases as quite useful, with mean rat-
ings between 4.1 and 4.4 across professions and cases. Their
potential as professional credentialing tools was considered
good; mental health counselors rated them highest (Ms =
4.0), and social worker and psychologist ratings for this pur-
pose averaged between 3.7 and 3.9.
Although time-consuming, completing these materials
was not particularly difficult for experienced clinicians. Chi-
square analyses revealed no significant differences for time
on task or degree of difficulty across the disciplines. Psy-
chologists and social workers found the CTPS tasks to be
somewhat easier (ratings between 2.3 and 2.6) than did
mental health counselors (Ms = 2.9). Psychologists took
less time to complete each case (averaging 66–72 minutes)
than did social workers or counselors (averaging 78–90 min-
utes), with all professionals requiring more time to complete
the anxiety disorder than the affective disorder. The CTPS
thus seems to sample task domains that are familiar and rou-
tine for both master’s- and doctoral-level practitioners.
295
Falvey

TABLE 1

Means and Standard Deviations of Case Conceptualization (CC) and Treatment Planning (TP) Scores
Across Profession, Experience, Work Setting, and Clinical Orientation

Anxiety Disordera Affective Disorderb

CC TP CC TP
Variable n M SD M SD M SD M SD

Profession
Psychologist 49 1898.7 509.4 1889.2 537.9 2390.3 390.9 1527.0 530.9
Social worker 59 1813.9 485.3 1753.6 483.3 2339.1 508.8 1263.4 392.0
Mental health counselor 60 2110.3 397.4 1907.5 595.8 2514.4 555.1 1749.3 672.7
Experience (years)
Less than 5 24 1980.0 428.3 1875.8 450.3 2422.7 507.8 1520.7 500.0
5–15 90 2048.3 436.4 1919.4 552.5 2432.7 505.4 1583.6 632.6
More than 15 54 1760.4 515.0 1721.0 549.6 2373.2 486.0 1415.6 523.8
Work setting
Inpatient 19 1997.6 472.0 1881.9 588.5 2499.7 304.8 1617.6 558.4
Agency/CMHC 50 1979.6 517.4 1833.5 599.1 2488.5 572.4 1539.3 591.4
Private practice 79 1921.7 473.4 1840.2 506.1 2361.2 485.3 1503.1 548.4
School/university 20 1887.6 421.6 1880.8 523.6 2386.1 502.1 1445.4 721.4
Clinical orientation
Psychodynamic 67 1992.5 472.2 1779.6 490.1 2460.2 475.6 1379.4 497.7
Cognitive-behavioral 52 1977.2 417.5 1915.9 601.4 2465.2 433.3 1708.0 683.5
Humanistic 21 1899.5 506.3 1923.4 365.7 2421.1 550.1 1438.6 546.1
Family systems 28 1802.6 562.9 1827.2 653.0 2227.0 596.1 1560.9 507.5

Note. CMHC = community mental health center.

a
Possible range of scores: CC = 0–3128; TP = 0–5923. bPossible range of scores: CC = 0–3604; TP = 0–5115.

Judgment Consistency
Some research has concluded that judgment is case spe-
cific; that is, performance on one case does not predict per-
formance on another case (Elstein et al., 1978; O’Donohue
et al., 1990). This assertion infers that clinical decisions re-
flect knowledge about specific clients or disorders rather
than representing characteristic thinking processes of prac-
titioners (Hogarth, 1987). The first research question, there-
fore, addressed whether participant performance on these
judgment tasks was consistent within and across cases or
whether their decision outcomes were largely independent
of one another.
Pearson product–moment correlations were computed for
CC and TP scores within and across the two CTPS cases.
Findings revealed a modest relationship between CC and
TP within cases (anxiety disorder r = .37, affective disorder
r = .34) and a moderately strong relationship between CC
scores (r = .64) and TP scores (r = .51) across cases. All
correlations were significant at the .001 level. Although
there was considerable group variability on these tasks, per-
formance outcomes were thus relatively consistent for in-
dividual participants.
Demographic Variables
The second research question addressed the impact of thera-
pist demographics on clinical judgment. To examine the
influence of profession (psychologist, social worker, men-
tal health counselor), clinical experience (< 5 years, 5–15
years, > 15 years), work setting (inpatient, agency, indepen-
dent practice, or academic), and orientation (psychody-
namic, cognitive-behavioral, family systems, or humanistic)
on performance, analyses of variance (ANOVAs) were com-
puted across CC and TP scores for each case. Alpha was set
at .01 to control for four multiple comparisons. Table 1
presents score means and standard deviations across demo-
graphic variables of interest in this study.
ANOVA revealed significant CC effects on the anxiety dis-
order for profession, F(2, 116) = 3.99, p < .01, and experience,
F(2, 116) = 5.94, p < .01, with no interactions. Mental health
counselors scored significantly higher (M = 2110.3) than so-
cial workers (M = 1813.9), and all professionals in the study
who reported between 5 and 15 years of experience scored
significantly higher (M = 2048.3) than did those with more
than 15 years in the field (M = 1760.4). CC scores on the
affective disorder yielded no main effect or interactions.
Across TP scores, ANOVA revealed no main effects or
interactions for the anxiety disorder. For the affective dis-
order, a main effect was found for profession, F(2, 111) =
7.66, p < .001, with no interactions. Mental health counse-
lors again scored significantly higher (M = 1749.3) than
did social workers (M = 1263.4).
Case Review Process
The third research question examined behavioral indices
to determine which factors in participants’ review of these

296 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79

 C l i n i c a l J u d g m e n t i n C a s e C o n c e p t u a l i z a t i o n a n d Tr e a t m e n t P l a n n i n g

cases were the strongest predictors of performance. Multi-
variate regression analyses were computed across case re-
view variables to assess their contribution to CC and TP
scores, with the alpha coefficient set at .01 to control for
experiment-wise error rate across four multiple tests. The
regression models accounted for a small but significant pro-
portion of the variance in CC and TP scores for both cases.
Table 2 summarizes these analyses.
Using CC as the criterion, predictor variables included
the number of reports reviewed, use of formal (Diagnos-
tic and Statistical Manual of Mental Disorders, DSM) diag-
nosis, diagnostic accuracy, reported task difficulty, reported
expertise with this disorder, and time on task. For the
anxiety disorder, significant predictors, R2 = .12, F(6, 146) =
3.19, p < .01, included the use of a DSM diagnosis (p < .01)
and longer time on task (p < .01). Significant predictors for
the affective disorder, R2 = .23, F(6, 139) = 6.87, p < .0001,
included reviewing more reports (p < .0001) and longer time
on task (p < .001).
Using TP scores as the criterion, predictor variables included
time on task, degree of difficulty, treatment format (struc-
tured vs. unstructured), brief treatment approach (specified
as 6 months or less), use of DSM diagnosis, and reported
expertise with this disorder. For the anxiety disorder, signifi-
cant predictors, R2 = .17, F(6, 146) = 4.83, p < .001, included
longer time on task (p < .001), DSM diagnosis (p < .01), and
use of a structured treatment format (p < .01). For the affec-
tive disorder, significant predictors, R2 = .22, F(6, 139) = 6.49,
p < .0001, included longer time on task (p < .0001) and a
structured treatment format (p < .0001).
Clinical Focus
The final research question examined areas of clinical focus for
their impact on CC and TP scores. Which areas of attention in
their case reviews and what treatment recommendations were
the strongest predictors of performance among these profes-
sionals? In order to develop concise models to predict outcome
scores, intercorrelations were first computed across predictor
and criterion variables. The CC score was found to be signifi-
cantly related to all predictor variables across both cases. For
the anxiety disorder, intercorrelations among the independent
variables were found between client presentation and medical
history (r = .24, p < .01) and substance abuse history (r = .25,
p < .01), and between substance abuse and medical histories
(r = –.26, p < .01). For the affective disorder, intercorrelations
among independent variables were found between current
symptoms and psychosocial history (r = 1.00) and between
suicide risk and current support system (r = –.22, p < .01).
Across both cases, TP scores were also found to be sig-
nificantly related to each predictor variable. For the anxi-
ety disorder, 4 among 11 treatment considerations and in-
terventions were found to be highly intercorrelated with
several other variables and were thus eliminated from the
model. For the affective disorder, 5 among 13 treatment
considerations and interventions were similarly highly
intercorrelated and eliminated from that model. Multivari-
ate stepwise regression analyses were computed for the re-
maining clinical focus variables related to CC and TP for
each case, with alpha level set at .01 to control for experi-
ment-wise error rate across the four multiple tests. Table 3
presents results of these analyses.
Using CC as the criterion, six predictor variables entered
into the regression equation for the anxiety disorder accounted
for 59% of the variance in case conceptualization scores.
Client presentation accounted for nearly one third of the
variance in Step 1 of the model (adj R2 = .31, F = 74.4, p <
.0001). Medical history, added in Step 2, was also a strong
predictor, accounting for 15% of CC score variance.

TABLE 2

Multiple Regression With Case Review Variables as Predictors of Case Conceptualization (CC) and
Treatment Planning (TP) Scores

Predictor Variables
No. Reports Formal Diagnostic Degree of Area of Time on Brief TX Treatment
Reviewed Diagnosis Accuracy Difficulty Expertise Task Approach Format
Criterion
Variable B t B t B t B t B t B t B t B t F R2

Anxiety
disorder
CC .11 1.38 .21 2.70* .01 .08 .11 1.34 –.05 –0.59 .17 2.12* 3.19* .12
TP .19 2.37* .04 0.44 .09 1.22 .29 3.66** .10 1.30 .15 1.85* 4.83** .17
Affective
disorder
CC .30 3.97*** .12 1.62 .03 .45 .09 1.20 –.03 –0.45 .26 3.44** 6.87*** .23
TP .10 1.34 .11 1.41 –.01 –0.13 .30 4.00*** .11 1.41 .30 3.99*** 6.49*** .22

Note. TX = treatment.
*p < .01. **p < .001. ***p < .0001.

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 297

Falvey

TABLE 3

Multiple Regression Models With Clinical Focus Variables as Predictors of Case Conceptualization and
Treatment Planning Scores

Predictor Variable B SE B β t p F R2 Adj R2

Case conceptualization
Anxiety disorder 40.4* .61 .59
Substance abuse history 185.9 51.1 .19 3.6 .0004
Intake interview 881.7 220.7 .20 4.0 .0001
Medical history 306.3 50.8 .32 6.0 .0000
Current symptoms 906.0 308.7 .15 2.9 .0038
Psychological testing 201.6 61.9 .17 3.3 .0014
Client presentation 410.1 51.3 .43 8.0 .0000
Affective disorder 55.1* .72 .71
Substance abuse history 343.8 51.3 .30 6.7 .0000
Current symptoms 927.2 276.0 .15 3.4 .0010
Prior treatment records 411.8 72.0 .26 5.7 .0000
Medical history 269.5 44.2 .27 6.1 .0000
Intake interview 800.7 273.2 .13 2.9 .0039
Suicide assessment 466.1 47.6 .44 9.8 .0000
Current support system 291.7 46.1 .29 6.3 .0000
Treatment plan
Anxiety disorder 32.6* .59 .58
Cognitive restructuring 291.4 63.6 .25 4.6 .0000
Psychoeducation 187.9 60.9 .17 3.1 .0024
Explore client history 180.7 57.1 .17 3.2 .0019
Realign family relationships 402.2 58.5 .36 6.9 .0000
Build rapport 191.6 56.9 .17 3.4 .0010
Stress management 375.9 59.7 .35 6.3 .0000
Socialization 289.1 77.3 .19 3.7 .0003
Affective disorder 42.1* .69 .68
Cognitive restructuring 211.8 58.5 .17 3.6 .0004
Coping skills training 278.4 73.7 .18 3.8 .0002
Psychoeducation 288.4 60.5 .23 4.8 .0000
Suicide precautions 262.3 70.9 .17 3.7 .0003
Realign family relationships 494.2 53.9 .42 9.2 .0000
Medication consultation 153.9 58.4 .12 2.6 .0093
Building rapport 411.9 53.2 .35 7.7 .0000
Socialization 442.9 75.8 .27 5.8 .0000

*p < .0001.

For the affective disorder, seven predictor variables entered
into the regression equation accounted for 71% of CC score
variance. Suicide assessment accounted for one third of the
variance in Step 1 (adj R2 = .33, F = 77.9, p < .0001). It was
followed by substance abuse history, which increased adj R2
by 14% in Step 2. Medical history contributed an additional
8% of the variance in Step 3 of this model.
Using TP as the criterion, seven predictor variables en-
tered into the regression model for the anxiety disorder
accounted for 58% of treatment planning score variance.
By far, the best single predictor (adj R2 = .26, F = 58.2, p <
.0001) was stress management. It was followed by realign-
ing marital and family relationships, which increased adj R2
by 13% in Step 2 of the equation.
Eight predictor variables entered into the regression model
for the affective disorder accounted for 68% (adj R2) of TP score
variance. The best single predictor (adj R2 = .24, F = 51.9, p <
.0001) was realigning marital and family relationships, whereas
building therapeutic rapport contributed 15% of score variance
in Step 2. Socialization activities (9%) and cognitive restructur-
ing (8%) were also relatively strong TP score predictors.
DISCUSSION
The rapidly shifting landscape of mental health service pro-
vision demands that clinical training programs, credentialing
bodies, and continuing education venues identify the spe-
cific skills and training associated with competent service
delivery. Although public discourse about the qualifications
of various provider groups to conduct psychotherapy
abounds in the professional journals, it is largely guild driven
rather than empirically grounded. Until a body of research
establishes which variables are associated with best clinical
practices and develops methodologies for their assessment,
provider screening criteria will remain largely at the discre-
tion of the managed care industry.

298 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79

 C l i n i c a l J u d g m e n t i n C a s e C o n c e p t u a l i z a t i o n a n d Tr e a t m e n t P l a n n i n g
Assessment of practitioner competencies is admittedly
quite complex. Defining relevant practice domains, devel-
oping assessment tools, and establishing valid competency
criteria require considerable empirical rigor. Several decades
of clinical judgment research have been marked by contro-
versy and criticism over the lack of consensus regarding
how judgment and decision making can be learned, taught,
and measured (Garb, 1998). The large dispersion of scores
in the current study reaffirms that even within cohorts of
profession, experience, work setting, and orientation, men-
tal health providers are a diverse group. Clinical judgment
is a highly individual cognitive process, even among those
one would assume to be similarly trained or credentialed.
Notwithstanding these limitations, this study demon-
strates a multidisciplinary approach to evaluating two widely
recognized clinical judgment tasks common to all mental
health practitioners. A self-selecting bias favored participa-
tion by more experienced clinicians over less experienced
ones in the recruitment process, and this demands caution
in interpreting these results. Given previous research that
suggests that clinicians gain confidence (although not nec-
essarily competence) with experience (Dawes, 1989; Garb,
1989), one may speculate that less experienced clinicians
may have felt more intimidated about having their skills
compared with professionals from other disciplines on a
largely unstructured judgment task.
Both a limitation and a strength come from the experi-
enced sample participating in this study. The majority of
research on clinical judgment in mental health has focused
on trainees or new professionals, whereas this study attracted
predominantly “mid-career” professionals. With 86% of par-
ticipants reporting more than 5 years of clinical experience
in predominantly outpatient settings (including 32% report-
ing more than 15 years experience), the following findings
are most applicable to that population.
Addressing the Criterion Issue
Given the centrality of case conceptualization and treat-
ment planning in most psychotherapy models, efforts to
define current best practices seem worthwhile (Nathan,
1998). However, to do so requires some valid way of as-
sessing these clinical judgment skills. Against what criteria
should individual performance be measured?
As in the recent development of clinical practice guide-
lines, it is advisable to use independent subject-matter ex-
perts with both clinical and research expertise to develop
criterion-referenced standards of performance (Baker,
Lichtenberg, & Moye, 1998; Edmunds, 1996; Friedman,
Prywes, & Benbassat, 1989; Loveland, 1985). CTPS simu-
lations in this study relied on large multidisciplinary expert
panels to identify and weight all case information as well as
all case conceptualization and treatment planning criteria.
These experts represented all three professions sampled
in this study and three of the four clinical orientations
(the humanistic orientation was cited as influential but
not primary among experts). Weighting criterion scores
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
by averaging these experienced clinician ratings has been
cited as a viable method to establish reliability and validity
of the CTPS cases (Garb, 1998).
Client symptoms in each case met all DSM criteria for the
target diagnosis. The largely unstructured response format
minimized cueing effects and simulated real-life decision
contexts in which clinicians must generate options under
conditions of uncertainty. This increased the fidelity of the
instrument, although it required complex coding and scor-
ing procedures. High validity ratings by practitioners in this
and other studies support the viability of CTPS simulations
as a strategy for evaluating clinical judgment on these tasks.
Implications for Training and Practice
Four analyses were conducted to delineate how participants
approached these tasks, how they performed in relation to
one another, and clinical variables that contributed to their
CC and TP scores. How findings can be linked to strategies
for improving clinical training and practice is the focus of
the following discussion.
The finding that clinical judgment was consistent among
these clinicians is encouraging. Individual CC and TP scores
were modestly yet significantly correlated within and across
both cases. Given that less than half of these clinicians re-
ported expertise with anxiety disorders and fewer than two
thirds reported expertise with affective disorders, this finding
challenges earlier conclusions that judgment simply reflects
knowledge about specific disorders. It instead lends support
to other empirical evidence of internally consistent reasoning
styles among experienced clinicians (Falvey, 1992a; Turner &
Kofoed, 1984) and suggests that there may be distinct cogni-
tive patterns that characterized their decision processes.
The significant difference found across professions on CC
scores for the anxiety but not the affective disorder may be
explained by the lower level of expertise with anxiety dis-
orders (42%) that this sample reported. Conversely, the
higher percentage of expertise with affective disorders
(63%) may account for less variability among those scores.
That is, variability in scores could be expected to decrease
as overall expertise increases. Does expertise also enhance
performance? Falvey and Hebert (1992) reported signifi-
cantly higher TP scores on CTPS cases among experts as
compared with experienced clinicians completing these
cases, lending some support to this speculation. However,
other research remains equivocal on whether experience
improves performance (Garb, 1998; Hogarth, 1987).
Although several trends across disciplines were suggested
by the data, it is not clear from this research what heuris-
tics were involved in the case review process and how they
may have influenced decision making. A follow-up study,
using a process-tracing approach in which another sample
of psychologists, social workers, and mental health coun-
selors “think aloud” while reviewing and responding to a
CTPS case, is examining this line of inquiry further (Falvey
& Bray, 2001). To the extent that specific cognitive styles
and their associated decision outcomes can be identified,
299
Falvey
preprofessional supervision and in-service training may be
targeted to correct problematic heuristics and enhance
clinical judgment.
The second analysis examined the impact of practitioner
demographics on performance. Among the master’s-level
professionals included in this study (N = 119), mental health
counselors scored higher than clinical social workers on all
judgment tasks. This supports other findings that profes-
sion does influence clinical judgment (Clavelle & Turner,
1980; Falvey, 1992b; Turner & Kofoed, 1984) and suggests
that certified mental health counselors are at least compa-
rable with their clinical social work counterparts on these
case management tasks.
It is interesting that CC and TP scores for mental health
counselors were also slightly higher than those of psycholo-
gists across both cases, although not significantly. Given that
most CCMHCs receive their graduate training from psy-
chologist faculty members, they could be expected to ap-
proach decision-making tasks similarly. It is also possible
that CCMHCs were highly motivated on these tasks due
to recent efforts of that group to achieve parity with other
mental health professionals. However, participants were
unaware of specific comparison groups used in the study,
thus partially controlling for rivalry and expectancy effects.
These findings suggest that training and experience at the
master’s level is sufficient for completing the clinical judg-
ment tasks of case conceptualization and treatment plan-
ning. Clinical mental health counselors performed as well
as clinical psychologists across both tasks and both cases,
although counselors found the tasks slightly more difficult
than did psychologists. Previous studies concluded that
master’s-level practitioners spend the majority of their pro-
fessional time in direct service delivery (Robiner, Arbisi, &
Edwall, 1994; Trent, 1993; Watkins, Campbell, & McGregor,
1989) and thus could be expected to have competency in
the case management skills assessed in this study. For com-
mon outpatient disorders, then, client assessment and treat-
ment planning seems well within the scope of master’s-
level clinicians. The degree to which these skills are related
to subsequent successful treatment outcomes is a recom-
mended future research direction.
Using the CTPS cases as standard stimuli, results of this
study also suggest that psychology-based training is more
beneficial than social work training on these clinical judg-
ment tasks. Psychologists and mental health counselors
scored higher than social workers on both tasks across both
cases. This was particularly evident in treatment planning
in which social workers considered fewer interventions and
goals, relied less on cognitive-behavioral strategies that are
empirically supported for these disorders, and achieved con-
sistently lower scores than the other practitioners. Given
the lack of interaction between profession and experience,
work setting, or orientation, and considering that these were
experienced and credentialed professionals with similar con-
tinuing education requirements, this finding seems robust.
Social work training may thus benefit from an increased fo-
cus on goal-setting and cognitive-behavioral interventions.
300 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Work setting and clinical orientation did not distinguish
performance on either CTPS case. Setting was also nonsig-
nificant in another interdisciplinary study of clinical deci-
sion making (Turner & Kofoed, 1984) and thus seems to
have little impact on these judgment tasks. The finding for
orientation was unexpected and failed to support previous
evidence that this variable does influence judgment (Beutler,
Crago, & Arizmendi, 1986; Bishop & Richards, 1984; Gil-
Ali & Newman, 1984; Houts, 1984; Kopta et al., 1986).
Some research has suggested that the impact of orientation
may diminish over time, affecting the performance of train-
ees and entry-level professionals more than experienced
clinicians (Berven, 1985; Falvey & Hebert, 1992). Although
this study found no interaction between orientation and
experience, it primarily sampled experienced professionals
(e.g., 86% reporting over 5 years in the field).
Several other factors may have contributed to this finding.
In the current study, two primary orientations (psychody-
namic and cognitive-behavioral) accounted for more than
two thirds of the participants in each profession. This may
reflect a self-selection bias among those choosing to partici-
pate because humanistic and family systems clinicians are
less likely to rely on the formal clinical assessment proce-
dures that these tasks represent (Bellack & Hersen, 1990). In
addition, in noting their primary orientation, participants had
the option of ranking other orientations that significantly
influenced their work. Most did; in fact, only 23 (14%) listed
just one orientation, and well over 50% of the participants
from each profession listed three or more orientations as in-
fluential. This finding supports evidence of multiple orienta-
tions among psychologists (Zook & Walton, 1989) and sug-
gests that pragmatic eclecticism rather than theoretical alle-
giance was the mode among these practitioners.
From the aforementioned evidence, one could speculate
that orientation has more impact on clinicians’ decision
making early in their careers (as assessed in the earlier re-
search) when a limited knowledge base favors theoretical
allegiance. With increased experience, practitioners tend to
become more pragmatic and theoretically eclectic, thus the
impact of orientation is probably muted to a large degree.
The finding that experience significantly affected CC
scores on the anxiety disorder suggests that experience
does have some influence on judgment. Further inspec-
tion of group means suggests that this is a “modest” cur-
vilinear relationship. Previous research concluding that
clinicians do not seem to learn from experience (Dawes,
1989; Garb, 1998) may be clarified by this finding. In
this study, midcareer professionals having 5 to 15 years
of experience (n = 90) performed better than either
newly trained (n = 24) or more senior practitioners (n =
54) on CC and TP scores. It seems that perhaps we do
learn from our clinical experience, up to a point. With
experience comes the tendency to rely more heavily on
cognitive heuristics, commonly experienced as clinical wis-
dom (Brenner & Howard, 1976; Clavelle & Turner, 1980;
Turk & Salovey, 1988). One might hypothesize that this
“wisdom” provides so many templates that it interferes with
 C l i n i c a l J u d g m e n t i n C a s e C o n c e p t u a l i z a t i o n a n d Tr e a t m e n t P l a n n i n g
the ability to consider unique aspects of a case. Alternately,
it is possible that practitioners trained prior to the 1980s
did not receive extensive training in case conceptualization
and treatment planning. However, managed care utiliza-
tion reviews and professional continuing education require-
ments over the past decade would be expected to update
the skills of those professionals.
The third set of analyses addressed aspects of the case
review process that influenced case conceptualization and
treatment planning scores. Taking time to review more of the
available reports, reliance on DSM diagnostic taxonomy, and
use of structured treatment plans predicted better performance
than did other variables. These findings suggest that there are
diminishing returns in emphasizing efficiency over thorough-
ness on these tasks and support the value of formal documen-
tation systems typically required by third parties.
Spending more time on these tasks predicted higher CC
and TP scores. Half of the participants (49%) reported taking
1 hour or less to complete each case. Scores improved as time
increased, with professionals taking 2 hours or longer averag-
ing the highest CC and TP scores for each case. The number
of reports reviewed also predicted CC scores for the affective
disorder in which 33 participants (21%) who reviewed fewer
than half of the eight possible reports averaged CC scores
more than 300 points lower than those who reviewed more
case material. This challenges previous findings that adding
incremental data during case review increased confidence but
not judgment accuracy (Garb, 1998; Hogarth, 1987).
Identification of a formal (DSM) diagnosis was also associ-
ated with higher scores across both cases and significantly
higher CC scores for the anxiety disorder. Most participants
did record DSM diagnoses: 80% of psychologists, 75% of so-
cial workers, and 58% of mental health counselors. It is curi-
ous that the accuracy of diagnoses was not a significant pre-
dictor across either disorder. A probable explanation for this is
“fine tuning”; that is, incorrect or unclear diagnoses often tar-
geted signal symptoms of the disorders (e.g., anxious, depressed
mood) but did not adhere to formal taxonomy. Despite limi-
tations of the DSM (Strupp & Butler, 1990), this classification
system does seem to enhance case conceptualization.
It was notable that considerably fewer mental health
counselors used DSM diagnoses and structured treatment
plans than did the other groups. This helps to explain the
high variability in counselor scores on these tasks. Mental
health counselors recorded more extreme scores at both
ends of the range and thus seem to constitute a more di-
verse group than did the other professions. In a relatively
recent nationwide survey of CCMHCs, less than two thirds
of respondents (64%) believed that they were adequately
trained in use of the DSM (Mead, Hohenshil, & Singh,
1997). This study unfortunately did not assess which
CCMHCs were trained in CACREP-approved programs
(that require competency in the DSM diagnostic system)
versus other training programs; that may account for the
wide variability in scores. Clearly, this finding supports
the inclusion of course work on formal diagnostic systems
in graduate counseling curricula.
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
An important finding regarding patterns of case review
among these clinicians was that the use of a structured format
for treatment planning resulted in significantly higher TP scores
across both disorders. Treatment format was considered struc-
tured if it included a problem list, a sequence of recommended
interventions, and targeted goals of these interventions. This
finding supports the utility of focused treatment plans as com-
pared with general treatment discussions. More than half of
the psychologists and social workers submitted structured
treatment plans for both cases, whereas less than one third
of CCMHCs used structured formats for each case. This also
seems to be an area of needed training for mental health
counselors because it contributed to the wide standard de-
viation among TP scores for that group.
The final set of analyses in this study examined variables
of clinical focus for their impact on CC and TP scores. Case
conceptualization was clearly enhanced by focusing on the
client’s current clinical presentation, which accounted for
nearly one third of score variance. Medical history for both
cases and substance abuse history for the affective case were
also important predictors. Formal psychological testing did
not emerge as an important predictor. This finding supports
other evidence that current symptoms and client status are
critical variables to consider in intake assessments (Bellack
& Hersen, 1990; Greenblatt & Kleinmuntz, 1984).
Realigning marital and family relationships emerged as
an important predictor of treatment planning scores across
both cases. Although only 17% of these participants reported
a family systems orientation, those who were most effec-
tive on the treatment planning task clearly were thinking
systemically in developing an intervention plan. This em-
phasis was reiterated in the affective disorder, in which at-
tention to improving interpersonal relationships (through
therapeutic rapport, in the family, and in the client’s social
network) emerged as significantly stronger predictors than
either cognitive-behavioral or insight-oriented strategies in
treatment planning. Intrapsychic interventions thus seem
less prominent than an interpersonal focus in successfully
conceptualizing treatment for these disorders.
Implications for Clinical Assessment
Practitioners sampled in this study concurred that the CTPS
cases portray realistic and comprehensive outpatient files
similar to what they encounter in their clinical work envi-
ronments. They found the cases to be particularly suited
for preservice and in-service training, as well as having po-
tential for use in professional credentialing. These findings
support the viability of the CTPS for competency-based
assessment in the allied health professions (Falvey & Hebert,
1992; Loveland, 1985; McLeod, 1992). Standardized simu-
lations have been widely used for training and credentialing
in medical specialties (Friedman et al., 1989; Krahn &
Blanchaer, 1986; Melnick, 1990; Schwartz & Griffin, 1986)
and have demonstrated good potential for use in counsel-
ing and psychology (Butcher et al., 1985; Hile, Campbell,
& Ghobary, 1994; Smith, 1983; Turk & Salovey, 1988). The
301
Falvey
current study demonstrates the potential of such simula-
tions to discern decision processes and outcomes.
When used for teaching purposes, the ability of trainees
to receive immediate feedback on their performance as
compared with a criterion group provides a valuable learn-
ing tool (Friedman et al., 1989). The CTPS coding system
is formatted such that CC and TP criteria identified by the
experts can be viewed, along with weightings for each cri-
teria. Trainees can see which aspects of a case they con-
curred with the experts on and where they may have missed
important aspects of case conceptualization or treatment
planning. Training for a particular disorder could focus on
(for example) identifying target goals, specifying the length
of treatment, selecting adjunct services or referrals, or in-
cluding specific relational and behavioral interventions as
needed based on expert feedback.
The use of CTPS simulations for credentialing is some-
what less promising. The simulations were rated by practi-
tioners in this study as offering a high fidelity representa-
tion of their actual clinical practice. Their usefulness for
credentialing purposes was also rated highly across all
groups. However, the large standard deviations found across
most variables of interest reveal that there is considerable
individual variation in performance on these tasks. This may
limit their utility for standard competency assessments such
as licensure or certification.
Caveats and Conclusions
This investigation suggests that there are many similarities
and some important distinctions across professionals in re-
sponding to universal case management tasks. It also dem-
onstrates a methodology for evaluating clinical decision pro-
cesses and outcomes. However, although a more realistic
assessment of clinical skill than traditional written exams,
simulations do not replicate face-to-face interactions be-
tween practitioner and client (Frederikson, 1986). It is thus
not clear that how one responds to a written simulation
will accurately reflect their behavior with actual clients. The
use of case simulations may therefore best be incorporated
into a testing battery that includes theoretical knowledge
(assessed with traditional written exams), clinical knowl-
edge (evaluated through performance on standardized simu-
lations), and applied skill (demonstrated through work
samples of actual psychotherapy practice).
Major findings of this research suggest that mental health
professionals are consistent in their performance on case
conceptualization and treatment planning tasks across diag-
noses, that professional training and experience seem to have
some impact on these outcomes, and that it is possible to
identify specific aspects of case review and treatment plan-
ning that enhance performance on these judgment tasks.
Perhaps most encouraging about this methodology is its
potential to provide trainees and professionals with direct
feedback about their performance.
Although all participants in the study were practicing
clinicians, their weekly client contact hours were not as-
302 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
sessed. This may have influenced findings because those
who conduct case conceptualizations and prepare treatment
plans regularly could be expected to differ from those who
only infrequently do so. Future research may benefit from
evaluating this variable in conjunction with performance.
In this era of managed care, standardized simulations such
as the CTPS can become very useful in providing empirical
support for professional accountability. To the extent that we
can identify and assess those practitioner variables associated
with successful performance on common clinical judgment
tasks, the mental health disciplines will be well positioned to
regain clinical authority over our professional practice.
REFERENCES
Ambady, N., & Rosenthal, R. (1992). Thin slices of expressive behavior as
predictors of interpersonal consequences. Psychological Bulletin, 111,
256–274.
Baker, R. B., Lichtenberg, P. A., & Moye, J. (1998). A practice guideline for
assessment of competency and capacity of the older adult. Professional
Psychology: Research and Practice, 29, 149–154.
Bellack, A. S., & Hersen, M. (1990). Handbook of comparative treatments for
adult disorders. New York: Wiley.
Berman, J. S., & Norton, N. C. (1985). Does professional training make a
therapist more effective? Psychological Bulletin, 98, 401–407.
Berven, N. L. (1985). Reliability and validity of standardized case manage-
ment simulations. Journal of Counseling Psychology, 32, 397–409.
Beutler, L. E., Crago, M., & Arizmendi, T. G. (1986). Research on thera-
pist variables in psychotherapy. In L. S. Garfield & A. E. Bergin (Eds.),
Handbook of psychotherapy and behavior change (pp. 257–310). New
York: Wiley.
Bishop, J., & Richards, T. (1984). Counselor theoretical orientation as related
to intake judgments. Journal of Counseling Psychology, 31, 398–401.
Brenner, D., & Howard, K. (1976). Clinical judgment as a function of experi-
ence and information. Journal of Clinical Psychology, 32, 721–728.
Butcher, J. E., & Scofield, M. (1984). The use of a standardized simulation and
process tracing for studying clinical problem-solving competence. Counse-
lor Education and Supervision, 24, 70–84.
Butcher, J. E., Scofield, M., & Baker, S. (1985). Clinical judgment in
mental health treatment planning: An empirical investigation. Ameri-
can Mental Health Counselors Association Journal, 7, 116–126.
Chapman, G. B., Bergus, G. R., & Elstein, A. S. (1996). Order of information
affects clinical judgment. Journal of Behavioral Decision Making, 9, 201–211.
Christensen, A., & Jacobson, N. S. (1994). Who (or what) can do psycho-
therapy: The status and challenge of nonprofessional therapies. Psychologi-
cal Science, 5, 8–14.
Clavelle, P. R., & Turner, A. D. (1980). Clinical decision-making among profes-
sionals and paraprofessionals. Journal of Clinical Psychology, 36, 833–838.
Dawes, R. M. (1989). Experience and validity of clinical judgment: The
illusory correlation. Behavioral Sciences and the Law, 7, 457–467.
Dawes, R. M. (1994). House of cards: Psychology and psychotherapy built
on myth. New York: Free Press.
Edmunds, M. (1996). Clinical practice guidelines: Opportunities and
implications. Annals of Behavioral Medicine, 18, 126–132.
Ellis, M. V., Robbins, E. S., Schult, D., Ladony, N., & Banker, J. (1990).
Anchoring errors in clinical judgments: Type I error, adjustment, or
mitigation? Journal of Counseling Psychology, 37, 343–351.
Elstein, A. S., Shulman, L. S., & Sprafka, S. A. (1978). Medical problem
solving: An analysis of clinical reasoning. Cambridge, MA: Harvard
University Press.
Falvey, J. E. (1992a). Science or serendipity? What we know about
clinical judgment in mental health. Journal of Mental Health Coun-
seling, 14, 458–470.
Falvey, J. E. (1992b). From intake to intervention: Interdisciplinary per-
spectives on MH treatment planning. Journal of Mental Health Counsel-
ing, 14, 471–489.
 C l i n i c a l J u d g m e n t i n C a s e C o n c e p t u a l i z a t i o n a n d Tr e a t m e n t P l a n n i n g
Falvey, J. E., & Bray, T. (2001). How therapists think while making clinical judg-
ments: A multidisciplinary process-tracing study. Unpublished manuscript.
Falvey, J. E., & Hebert, D. J. (1992). Psychometric study of the Clinical
Treatment Planning Simulations (CTPS) for assessing clinical judgment.
Journal of Mental Health Counseling, 14, 490–507.
Faust, D. (1986). Research on human judgment and its application to clini-
cal practice. Professional Psychology: Research and Practice, 17, 420–430.
Faust, D., Guilmette, T. J., Hart, K., Arkes, H. R., Fishburne, F. J., & Davey, L.
(1998). Neuropsychologists’ training, experience, and judgment accu-
racy. Archives of Clinical Neuropsychology, 3, 145–163.
Frederikson, N. (1986). Construct validity and construct similarity: Meth-
ods for use in test development and test validation. Multivariate Behav-
ioral Research, 21, 3–28
Friedlander, M., & Stockman, S. (1983). Anchoring and publicity effects in
clinical judgment. Journal of Clinical Psychology, 39, 637–643.
Friedman, M., Prywes, M., & Benbassat, J. (1989). Variability in doctors’
problem-solving as measured by open-ended written patient simulations.
Medical Education, 23, 270–275.
Garb, H. N. (1989). Clinical judgment, clinical training, and professional
experience. Psychological Bulletin, 105, 387–396.
Garb, H. N. (1998). Studying the clinician: Judgment research and psychologi-
cal assessment. Washington, DC: American Psychological Association.
Gil-Ali, D., & Newman, F. L. (1984). Patient information and its use by thera-
pists differing in theoretical orientation. Psychotherapy, 21, 479–482.
Greenblatt, R. L., & Kleinmuntz, B. (1984). Statistical modeling of clinical
intake decisions. Journal of Clinical Psychology, 40, 1528–1537.
Hile, M. G., Campbell, D. M., & Ghobary, B. (1994). Automation for clini-
cians in the field: The validity of a performance support system. Behavior
Research Methods, Instruments & Computers, 26, 205–208.
Hogarth, R. (1987). Judgment and choice (2nd ed.). New York: Wiley.
Houts, A. (1984). Effects of clinician theoretical orientation and patient
explanatory bias on initial clinical judgments. Professional Psychology:
Research and Practice, 15, 284–293.
Kelly, M. B. (1977). A review of the observational data-collection and reli-
ability procedures reported in the Journal of Applied Behavior Analysis.
Journal of Applied Behavior Analysis, 10, 97–101.
Kleinmuntz, B. (1990). Why we still use our heads instead of formulas:
Toward an integrative approach. Psychological Bulletin, 107, 296–310.
Kopta, S. M., Newman, F. L., McGovern, M., & Sandrock, D. (1986). Psy-
chotherapeutic orientations: A comparison of conceptualizations, inter-
ventions, and treatment plan costs. Journal of Consulting and Clinical Psy-
chology, 54, 369–374.
Krahn, C. G., & Blanchaer, M. C. (1986). Using an advanced organizer to
improve knowledge application by medical students in computer-based
clinical simulations. Journal of Computer-Based Instruction, 13, 71–74.
Lambert, L. E., & Wertheimer, M. (1988). Is diagnostic ability related to
relevant training and experience? Professional Psychology: Research and
Practice, 19, 50–52.
Lawless, L. L., Ginter, E. J., & Kelly, K. R. (1999). Managed care: What
mental health counselors need to know. Journal of Mental Health Coun-
seling, 21, 50–65.
Loveland, E. H. (1985). Measurement of competence: Some definitional
issues. Professional Practice of Psychology, 6, 129–143.
Luborsky, L., Crits-Christoph, P., McLellan, A. T., Woody, G., Piper, W.,
Liberman, B., Imber, S., & Pilkonis, P. (1986). Do therapists vary much in
their success? American Journal of Orthopsychiatry, 56, 501–502.
Mahoney, M. J. (1988). Rationalism and constructivism in clinical judg-
ment. In D. Turk & P. Salovey (Eds.), Reasoning, inference, & judgment
in clinical psychology (pp. 155–181). New York: Free Press.
Makover, R. B. (1996). Treatment planning for psychotherapists. Washington,
DC: American Psychiatric Press.
McLeod, J. (1992). What do we know about how best to assess counsellor
competence? Counselling Psychology Quarterly, 5, 359–372.
Mead, M. A., Hohenshil, T. H., & Singh, K. (1997). How the DSM system is
used by clinical counselors: A national study. Journal of Mental Health
Counseling, 19, 383–401.
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Melnick, D. E. (1990). Computer-based clinical simulation: State of the art.
Evaluation & the Health Professions, 13, 104–120.
Mordock, J. B. (1994). Treatment planning in counseling. In J. L. Ronch,
W. Van Ornum, & N. C. Stilwell (Eds.), The counseling sourcebook: A
practical reference on contemporary issues (pp. 227–233). New York: Cross-
road Publishing.
Nathan, P. E. (1998). Practice guidelines: Not yet ideal. American Psycholo-
gist, 53, 290–299.
O’Donohue, W., Fisher, J. E., Plaud, J. J., & Curtis, S. D. (1990). Treatment
decisions: Their nature and their justification. Psychotherapy, 3, 421–427.
Plous, S., & Zimbardo, P. (1986). Attributional biases among clinicians: A
comparison of psychoanalysts and behavior therapists. Journal of Con-
sulting and Clinical Psychology, 54, 568–570.
Robiner, W. N., Arbisi, P. A., & Edwall, G. E. (1994). The basis for the
doctoral degree for psychology licensure. Clinical Psychology Review,
14, 227–254.
Rock, D. L., Bransford, J. D., Maisto, S. A., & Morey, L. (1987). The study of
clinical judgment: An ecological approach. Clinical Psychology Review, 7,
645–661.
Schwartz, S., & Griffin, T. (1986). Medical thinking: The psychology of medi-
cal judgment and decision making. New York: Springer-Verlag.
Shaw, B. F., & Dobson, K. S. (1988). Competency judgments in the training
and evaluation of psychotherapists. Journal of Consulting and Clinical
Psychology, 56, 666–672.
Shueman, S. A. (1997). Confronting health care realities: A reply to Sank
(1997). Professional Psychology: Research and Practice, 28, 555–558.
Smith, I. L. (1983). Use of written simulations in credentialing programs.
Professional Practice of Psychology, 4, 21–50.
Snyder, M., & Thomsen, C. J. (1988). Interactions between therapists and
clients: Hypothesis testing and behavioral confirmation. In D. Turk & P.
Salovey (Eds.), Reasoning, inference, and judgment in clinical psychology
(pp. 124–152). New York: Free Press.
Spengler, P. M., & Strohmer, D. C. (1994). Clinical judgment biases: The
moderating roles of counselor cognitive complexity and counselor client
preferences. Journal of Counseling Psychology, 41, 8–17.
Strohmer, D. C., & Spengler, P. M. (1993). Studying mental health counse-
lor clinical judgment: A response to Falvey and colleagues. Journal of
Mental Health Counseling, 15, 465–474.
Stromberg, C., Ratcliff, R., & Scheutze, J. (1997). Surviving the
minefield of managed care contracting. The Psychologist’s Legal Up-
date (#9). Washington, DC: National Register of Health Service
Providers in Psychology.
Strupp, H. H., & Butler, S. F. (1990). Psychotherapy. In A. S. Bellack & M.
Hersen (Eds.), Handbook of comparative treatments for adult disorders (pp.
3–16). New York: Wiley.
Tillett, R. (1996). Psychotherapy assessment and treatment selection. Brit-
ish Journal of Psychiatry, 168, 10–15.
Trent, J. T. (1993). Issues and concerns in master’s-level training and em-
ployment. Journal of Clinical Psychology, 49, 586–592.
Turk, D. C., & Salovey, P. (Eds.). (1988). Reasoning, inference, and judgment
in clinical psychology. New York: Free Press.
Turner, J., & Kofoed, L. (1984). Decision-making, professional discipline,
and program affiliation: Selection of an inpatient treatment alternative.
Journal of Clinical Psychology, 40, 858–865.
Tutin, J. (1993). The persistence of initial beliefs in clinical judgment. Jour-
nal of Social and Clinical Psychology, 12, 319–335.
Waddington, L. (1997). Clinical judgment and case formulation. British
Journal of Clinical Psychology, 36, 309–311.
Watkins, C. E., Campbell, V. L., & McGregor, P. (1989). APA-affiliated
master’s-level counselors. Counseling Psychologist, 17, 289–300.
Worthington, R. L., & Atkinson, D. R. (1993). Counselors’ responsibil-
ity and etiology attributions, theoretical orientations, and counseling
strategies. Journal of Counseling Psychology, 40, 295–302.
Zook, A., & Walton, J. M. (1989). Theoretical orientations and work
settings of clinical and counseling psychologists: A current perspec-
tive. Professional Psychology: Research and Practice, 20, 23–31.
303
Counselors’ Attribution of Responsibility, Etiology, and
Counseling Strategy

Jerry L. Kernes and J. Jeffries McWhirter

This study surveyed 167 counselors working at university counseling centers on their etiology and responsibility attributions and
models of helping. Participants responded to vignettes describing either a male or female client experiencing symptoms of either
an identity or adjustment problem. Counselors endorsed all of P. Brickman et al.’s (1982) models of helping for both problem types.
Predictions concerning etiology attributions were partially supported. Counselors selected attributions logically consistent with
an internal cause for the identity problem. However, counselors did not make external attributions for the adjustment problem. No
significant results were observed for the influence of client sex.

C
ounselors constantly seek to understand the
behavior of their clients. In the process of at-
tempting to understand their clients, counse-
lors observe clients’ behaviors and the context
in which those behaviors are imbedded and
make inferences concerning what could have produced such
behavior. These inferences may indict dispositional characteris-
tics of the client, environmental forces, or a combination of both
(Heider, 1958). Diagnostic decisions, symptom recognition, and
predictions concerning treatment response and outcome can
be influenced by counselors’ explanations for the cause of
clients’ presenting problems (Lopez & Wolkenstein, 1990).
The role attributions play in counselor decision making
has not been sufficiently explored (Murdock & Freemont,
1989). The majority of clinical attributional research has
focused on questions of self-perception (see Harvey &
Galvin, 1984, for review). Few studies of clinical attribution
have focused on person perception (Lopez & Wolkenstein,
1990). Studies that have been conducted in this area have
produced mixed results. For example, Strohmer, Haase,
Biggs, and Keller (1982) examined counselors’ attributions
along three dimensions: the degree to which the client
could control difficulties (controllability), whether the
problem was long term or temporary (stability), and the
degree to which the problem pervaded the client’s life
(globality). Counselors were then asked to predict the likeli-
hood of the client’s progress in counseling. The researchers
concluded that attributions might not significantly influ-
ence counselor decision making. In a later study, counselors’
attributions of controllability in diagnosis of affective
disorders were found to play a role in diagnostic decisions
(Strohmer, Biggs, Keller, & Thibodeau, 1984). Specifically,
the researchers found that attributions significantly affected
diagnostic discriminations between bipolar and unipolar
disorders.
Additional evidence for the influence of counselors’ attri-
butions on counseling decisions is provided by Batson (1975)
and Murdock and Freemont (1989). Batson compared the
attributions and treatment recommendations of professional
and nonprofessional helpers in a simulated referral agency.
In general, participants in the study tended to match treat-
ment referrals to attributions. Clients whose problems
were attributed to personal factors (mental disorder) were
given referrals to agencies concerned with personal change
(mental hospital, residential treatment center), whereas
clients whose problems were viewed as stemming from
situational factors (unemployment, lack of job skills) were
given referrals to social change agencies (state employment
agency, social services). Murdock and Fremont, likewise, had
counselors rate clients along four attributional dimensions
(locus, stability, globality, and controllability). Counselors
were then asked to make decisions on treatment urgency,
duration of the problem, and on ideal treatment modality.
Attributions regarding the stability of the cause and dura-
tion of the problem best predicted treatment decisions
(Murdock & Freemont, 1989).
If attributions play a significant role in counselors’ decision-
making processes, issues of responsibility are one important
set of attributions to consider. People tend to hold others
more responsible for their situations if they perceive those

Jerry L. Kernes is a doctoral candidate, and J. Jeffries McWhirter is a professor, both in the Counseling Psychology Program at Arizona State University,
Tempe. This article is based on the master’s thesis completed by Jerry L. Kernes under the direction of J. Jeffries McWhirter. The research was funded, in part,
by a grant from the Research Support Program at Associated Students of Arizona State University, Graduate College and Vice President for Research. An
earlier version of this article was presented in August 1998 at the 106th Annual Convention of the American Psychological Association in San Francisco. The
authors thank Richard T. Kinnier and Barbara Kerr for their comments and assistance in conducting this study. Correspondence regarding this article should
be sent to Jerry L. Kernes, c/o J. Jeffries McWhirter, Counseling Psychology Program, Division of Psychology in Education, Arizona State University, Box
870611, Tempe, AZ 85287-0611 (e-mail: Jerry.Kernes@asu.edu).

304 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79

 C o u n s e l o r s ’ A t t r i b u t i o n o f R e s p o n s i b i l i t y, E t i o l o g y, a n d C o u n s e l i n g S t r a t e g y
individuals as having been in control of their previous behav-
iors (see Weiner, 1993). In addition, people are less likely to
want to help others whom they perceive to be responsible for
their situations (Schmidt & Weiner, 1988). According to
Feinberg (1970), the question of moral responsibility in-
volves two separate issues—blame and control. Blame is
attributed to individuals if they are seen as responsible
for creating their problems. Control is attributed to people
if they are held responsible for changing or influencing
events in their lives. These dimensions undoubtedly influ-
ence counselors’ clinical decisions. Specifically, counselors’
attributions concerning clients’ responsibility for causing
and solving their problems may affect treatment decisions
and helping behavior.
Brickman et al. (1982) provided a theoretical framework
for classifying models of helping and coping according to
attributions of responsibility for the cause and solution of a
client’s problems. Their analysis yielded four fundamentally
different orientations to the world said to exist in the minds
of helpers, aggressors, and recipients of help or aggression.
Brickman and his colleagues contended that people might
be unaware of the assumptions they hold regarding respon-
sibility for the causes and solutions of problems but that
“they cannot, as social actors, avoid making such
assumptions”(p. 370). Karuza, Zevon, Rabinowitz, and
Brickman (1982) pointed out that in a clinical setting the
assumptions implicit in these orientations might color the
diagnosis, treatment choice, intervention strategy, and be-
havior of helpers. Brickman et al.’s model of helping and
coping has been applied to a number of settings and clini-
cal concerns including cancer patients’ beliefs about their
cancer, prevention of suicidal behavior, interventions with
the elderly, drug addiction, congruence in beliefs between
client and counselor, alcoholism, and cross-cultural counsel-
ing (see Avants, Margolin, & Singer, 1993; Jack & Williams,
1991; Karuza, Zevon, Gleason, Karuza, & Nash, 1990; Morojele
& Stephenson, 1992; Tracey, 1988; West & Power, 1995; Young
& Marks, 1986, respectively).
FOUR ORIENTATIONS OF THE MODEL
Medical Model
In this model, clients are not held responsible for either
the cause of their problem or its solution. Clients are seen
as suffering from illness they did not cause and must accept
the interventions of trained experts in order to improve.
Helpers are to use their training to identify the client’s prob-
lems and provide the necessary services for change. According
to Brickman et al. (1982), the advantage of this model is
that it allows clients to seek and accept help without being
blamed for their weakness. However, the model also promotes
a sense of client dependency.
Moral Model
This model can be seen as the converse of the medical model
and is best typified by self-help movements. In this model,
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
clients are seen as responsible for both causing and solving
their problems. Clients view themselves, or are viewed by oth-
ers, as creating their problems through a deficit in moral char-
acter (i.e., laziness, stubbornness, lack of willpower) and are
expected to strive to overcome problems. Helpers are essen-
tially viewed as motivators or coaches. The advantage of this
model is that clients are recognized as being totally respon-
sible for their lives and are likely to be more motivated to
produce changes. The drawback of this model is that it can be
taken to the extreme to believe that victims are responsible
for their own victimization (Brickman et al., 1982).
Compensatory Model
This model holds clients responsible only for solving their
problems but not for causing them. Clients are seen as suf-
fering from the failure of their social environments to meet
their needs. The relationship between client and helper is a
partnership, with the helper assuming a subordinate role.
Helpers are expected to be teachers and provide educa-
tion, skill building, and opportunities. Examples of this
model include community action programs and rehabilita-
tive therapy. Brickman et al. (1982) have claimed that the
advantage of this model is that it actively involves clients in
finding solutions to their problems while discounting the
clients’ past failures. The potential drawback of this model,
however, is that it may make clients feel undue pressure at
having to continually solve problems they did not create
and may foster a negative view of the world.
Enlightenment Model
This model holds clients responsible for causing their prob-
lems but not for solving them. In this model, clients are seen
as guilty individuals whose lives are out of control. The aim
of this model is to provide clients with enlightenment about
the nature of their problems and the difficulty inherent in
solving them. Clients are expected to submit to the will of
the authority figure or helper who knows the course of ac-
tion to follow. The advantage of this model is that it provides
clients with a sense of relief that their problems are beyond
their control and a sense of shared community and suffering.
A disadvantage of this model is that it may lead clients to
structuring their entire life around the source of authority.
In addition to the role played by attributions of responsi-
bility, counselors’ etiology attributions also seem important
in their influence on counselor decision making and help-
ing behavior. According to Dumont (1993), the formation
of an etiology assessment “inevitably implicates a particu-
lar vision of the development of a disorder and the treat-
ment of choice for remediating it” (p. 197). Most of the
research in this area has focused on larger assessments of
causality. For example, Weiner (1979) identified three di-
mensions along which attributional judgments may vary:
locus of cause (internal vs. external), the pervasiveness of
the cause (global vs. specific), and the degree to which the
cause is seen as changing over time (stable vs. unstable).
Likewise, Hansen (1980) distinguished between three gen-
305
Kernes and McWhirter
eral types of explanations observers can make for a behavior:
They may attribute the behavior to dispositions of the person
(traits, personality, etc.); they may attribute the behavior to
characteristics of the stimulus with which the person is inter-
acting; or they may attribute the behavior to the circumstances
in which the behavior is taking place. Finally, research has iden-
tified a number of expectancies and biases (e.g., dispositional,
representative, availability) implicated in determining where
a client’s problem lies (see Batson, O’Quin, & Pych, 1982;
Morrow & Deidan, 1992). It is surprising that very little re-
search has examined specific etiology attributions. Most of
the research that has been conducted on specific etiology as-
sumptions has looked only at clients’ assumptions about the
nature of their problems.
Townsend (1975) studied cross-cultural beliefs of mental
health clients and found that American mental health clients
tended to believe their disorders were due to environmental
factors, whereas German mental health clients tended to be-
lieve both that mental disorders are reactions to psychological
events and are curable and that mental disorders are inherited
and cannot be cured. Foulks, Persons, and Merkel (1986) at-
tempted to link beliefs about causes of psychological prob-
lems to therapeutic process. They developed a 47-item inven-
tory of psychiatric patients beliefs about their own illnesses.
They found that patients who endorsed medical-model causes
for their illness and rejected nonmedical-model causes made
more visits to psychiatric clinics than patients who endorsed
a nonmedical model of causation and rejected a medical
model. Similarly, Pistrang and Barker (1992) developed an
inventory of clients’ beliefs about the causes and treatments
of their own psychological problems. Beliefs about cause and
treatment were strongly associated (e.g., clients who saw their
problems as resulting from psychodynamic issues such as
early childhood problems also endorsed a psychodynamic
type of treatment).
Still, little is known about the specific etiology beliefs and
treatment recommendations of mental health professionals.
Atkinson, Worthington, Dana, and Good (1991) investigated
clients’ and counselors’ beliefs about the etiology of psycho-
logical problems and preferences for counseling orientations.
They used an empirical procedure to produce a list of spe-
cific beliefs about the causes of psychological problems (ir-
rational concerns, career or academic difficulties, physical ill-
ness, trauma or pain, lack of social skills, genetics, and bad
luck). Results indicated that the majority of both clients and
counselors ranked irrational concerns as the cause of psy-
chological problems. The major focus of the study, however,
was on the etiology attributions made by clients, therefore
providing little data on the attributions of counselors. In ad-
dition, the study examined etiology attributions of psycho-
logical problems in general rather than focusing on attribu-
tions made for specific disorders.
Worthington and Atkinson (1993) applied the Brickman
et al. (1982) model of responsibility attribution to the at-
tributions made by counseling center counselors of clinical
vignettes describing a student with an adjustment disorder
or an identity disorder. Among their findings was that coun-
306 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
selors varied their helping orientation according to the type
of disorder presented. Specifically, counselors subscribed
to a compensatory model of helping for the adjustment disor-
der and were equally divided between the compensatory and
the moral models of helping for the identity disorder. Coun-
selors did not endorse the medical or the enlightenment mod-
els for either disorder type. Worthington and Atkinson also
examined the etiology attributions the counseling center coun-
selors made for these disorders. In general, counselors selected
etiology attributions logically consistent with the assumption
of an internal cause for an identity disorder and an external
cause for an adjustment disorder. Specifically, counselors ranked
lack of self-understanding as the primary cause of the iden-
tity disorder (86%) and specific trauma as the primary cause
of the adjustment disorder (55%). Moreover, specific trauma
was never endorsed as a primary etiology attribution for
the identity disorder, whereas lack of self-understanding
received only minimal endorsement as the primary cause
of adjustment disorder.
One limitation Worthington and Atkinson (1993) ac-
knowledged in their investigation was the examination of
counseling center counselors’ attributions for only female
clients. They suggested that attributions of responsibility
and etiology and recommended counseling strategy might
vary as a function of client sex, and they called for addi-
tional research investigating counselors’ attributions of male
clients. A long line of research on gender bias and stereo-
types indicates that gender may play an important part in
attributions made by mental health professionals. Becker
and Lamb (1994); Broverman, Broverman, Clarkson,
Rosenkrantz, and Vogel (1970); O’Malley and Richardson
(1985); Robertson and Fitzgerald (1990); and Swenson and
Ragucci (1984) all spoke of differential evaluations made
of male and female clients. Much of this research indicated
that women are perceived in more negative terms than are
men and that psychotherapy may function to reinforce
traditional roles for women (see Abramowitz, Abramowitz,
Jackson, & Gomes, 1973; American Psychological Associa-
tion Task Force, 1975). Other research suggested that
counseling and psychotherapy might also function to
reinforce traditional roles for men (see Costrich, Feinstein,
Kidder, Marcek, & Pascale, 1975; Fitzgerald & Cherpas, 1985;
Robertson & Fitzgerald, 1990). Still other research reported
that mental health professionals might not hold differential
views of men and women (see Phillips & Gilroy, 1985; Poole
& Tapley, 1988; Smith, 1980).
The present study seeks to build on the findings of
Worthington and Atkinson (1993) and Brickman et al.
(1982) regarding counselors’ attributions of responsibility
and etiology and to extend these findings by adding client
sex as an independent variable. The purpose of the present
study is to investigate the relationship between disorder
type and client sex on counselors’ ratings of client respon-
sibility, problem etiology, and recommended counseling
strategy. On the basis of the findings of Worthington and
Atkinson, we predicted that the majority of counselors
would endorse a compensatory model of helping for an ad-
 C o u n s e l o r s ’ A t t r i b u t i o n o f R e s p o n s i b i l i t y, E t i o l o g y, a n d C o u n s e l i n g S t r a t e g y
justment disorder and would be “equally” divided in their
endorsement of a moral model and a compensatory model
of helping for an identity disorder. We also predicted that
counselors would select etiology attributions logically con-
sistent with an internal cause for identity problems and an
external cause for adjustment problems.
In addition to these replication hypotheses, based on re-
search on dispositional biases in counseling, we predicted
that counselors would attribute greater internal causality,
greater stability, and greater controllability to the symptoms
of clients described as exhibiting an identity problem than
to clients described as exhibiting an adjustment problem.
Likewise, we predicted that counselors would report being
more likely to engage in interventions aimed at changing the
client’s environment or activities for clients exhibiting an
adjustment problem and would be more likely to engage in
interventions aimed at changing the client’s thoughts or feel-
ings for clients exhibiting an identity problem.
METHOD
Participants
Participants were 167 counselors (85 women and 82 men)
from university counseling centers in the United States and
Canada. Participants ranged in age from 24 to 65 years, with
a mean age of 46.2 years and standard deviation of 9.15
years. The participants held advanced degrees in the fol-
lowing specialties: counseling psychology (81), clinical psy-
chology (41), counselor education (24), social work (11),
and other (10). The participants held the following posi-
tions: center director (90), senior staff (52), assistant direc-
tor/training director/program coordinator (14), staff (5), and
other (6). Years of professional experience beyond highest
degree ranged from 1 to 36, with a mean of 12.5 years and
standard deviation of 9.58 years.
The following theoretical orientations were represented among
the participants: psychodynamic (34), cognitive-behavioral
(37), humanistic-existential (18), eclectic (69), family systems
(4), and other (5). The ethnic (self-reported) representation of
the participants was Native American (3), African American
(5), European American/White (154), Asian American (1),
Chicano/Latino/Hispanic (1), multiethnic (1), and other (2, who
did not indicate ethnicity).
Instruments
Counselor Preferences and Practices Survey. A revised form
of the Counselor Preferences and Practices Survey (CPPS;
Worthington & Atkinson, 1993) was used in the present
study to assess attributions of responsibility and etiology
and to examine recommended counseling strategies. The
CPPS consists of four parts. The first part requests basic
demographic information (age, sex, ethnicity) of the par-
ticipants. The second part requests information regarding
years of professional experience, type of degree, and theo-
retical orientation. In the present study, theoretical orien-
tation was collapsed into six categories (psychodynamic,
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
cognitive-behavioral, humanistic-existential, family systems,
eclectic, and other) because of Worthington and Atkinson’s
finding that few respondents subscribed to some of the theo-
ries listed on the CPPS.
The third and fourth parts of the CPPS consist of two
vignettes taken from the DSM-III Case Book (Spitzer,
Skodol, Gibbon, & Williams, 1981) and a series of ratings
that addressed client responsibility for causing and resolv-
ing problems, counselor etiology beliefs, and recommended
counseling strategy. One vignette describes a female client
exhibiting symptoms of an adjustment problem, and the
other describes a female client exhibiting symptoms of
an identity problem. Worthington and Atkinson (1993)
developed the vignettes to test Brickman et al.’s (1982)
assumption that counselors would vary their model of
helping based on the client’s responsibility for causing and
resolving the problem and on Furman and Ahola’s (1989)
assumption that counselors would vary their attributions
of etiology based on the type of problem presented. Two
additional vignettes were created in the present study by
altering the sex description of the clients in the original
vignettes to include descriptions of male clients with adjust-
ment and identity problems.
After reading the vignettes, participants rated a client’s
responsibility for causing and resolving their problems on
separate 6-point scales that ranged from 1 (not at all) to 6
(completely). Participants also rank ordered the top 3 causes
of the client’s problems from a list of 12 causes of psycho-
logical problems (genetics, unresolved feelings, specific
trauma, social isolation, stress, lack of self-understanding,
dysfunctional family, sick society, maladaptive learning, bio-
logical imbalances, physical illness, irrational thinking) taken
from existing etiology models (see Atkinson et al., 1991;
Daws, 1967; Foulks et al., 1986; Kedric, 1985; Maloney,
1985; Robertson & Fitzgerald, 1990). Finally, participants
were asked to indicate their likelihood of engaging in coun-
seling strategies aimed at changing the client’s environment,
changing the client’s interactions, and changing the client’s
thoughts and feelings. Participants were asked to rate their
likelihood of engaging in each strategy on separate 6-point
scales that ranged from 1 (extremely unlikely) to 6 (extremely
likely). No reliability or validity data is available for the CPPS.
Helping-Coping Attribution Scale. The therapist version of
the Helping-Coping Attribution Scale (HCAS; Tracey, 1988)
was used to measure counselors’ helping and coping attribu-
tions toward vignette clients. Tracey developed the HCAS to
measure the blame and control dimensions of Brickman et
al.’s (1982) model of responsibility attribution. The HCAS
contains four items. Two of the items measure attribution of
the cause of the client’s current difficulty (e.g., “this client is
responsible for his or her current problem” and “this client’s
problem is more a result of the situation he or she is in rather
than his or her own inability to cope”). The remaining two
items were designed to measure attributions regarding the
client’s responsibility for resolving the problem (e.g., “solving
the client’s problem is more the client’s responsibility than
mine” and “this client would not be able to change without
307
Kernes and McWhirter
my or others’ aid”). Counselors rated each of the four items
using a 7-point Likert-type scale that ranged from 1 (very
strongly disagree) to 7 (very strongly agree). The items on the
two scales were summed to produce two scores (blame and
control). High scores on the Blame scale indicated the client
was viewed as responsible for the cause of the problem, and
high scores on the Control scale indicated the client was
viewed as responsible for the solution to the problem.
Tracey (1988) reported 1-week test–retest reliability for
the Blame and Control subscales of the therapist version of
the HCAS as .71 and .68, respectively. Internal consistency
of the subscales was estimated using the Spearman-Brown
Prophecy formula to be .45 for the Blame subscale and .50
for the Control subscale. Tracey also reported that the Blame
and Control subscales were relatively independent of each
other with correlations between scales at –.15.
Causal Dimension Scale. A revised version of Russell’s
(1982) Causal Dimension Scale (CDS) was used to assess
counselors’ attributions of the locus of causality, stability, and
controllability of clients’ problems. The CDS was originally
designed to assess how an individual perceived the causes
they have stated for an event. The CDS contains nine items
with three questions each measuring the subscales of Locus
of Causality, Stability, and Controllability. Responses were
measured on 9-point semantic differential scales. A total score
for each subscale was arrived at by summing the responses
for each scale (Locus of Causality items = 1, 5, 7; Stability
items = 3, 6, 8; Controllability items = 2, 4, 9). High scores
on these scales indicated a cause was perceived as internal,
stable, and controllable, whereas low scores indicated a cause
was perceived as external, unstable, and uncontrollable.
Russell investigated the validity of the individual semantic
differential scales by subjecting each item to separate analy-
ses of variance. For each item, the largest main effect was
observed for the dimension the item was intended to mea-
sure. The subscales were found to be only moderately re-
lated to one another, with correlations ranging from .19 to
.28. Russell found the three scales to be internally consis-
tent. Alpha coefficients for the Locus of Causality, Stability,
and Controllability subscales were .87, .84, and .73, respec-
tively. Similarly, Vallerand and Richer (1988) reported alpha
coefficients for the Locus of Causality, Stability, and Con-
trollability scales at .80, .73, and .50. Abraham (1985) tested
the usefulness of the CDS as a tool for measuring attribu-
tions made about the mental health of others. Alpha coeffi-
cients for the Locus of Causality, Stability, and Controllabil-
ity subscales were .68, .90, and .88. The CDS was modified
for use in the present study by changing the wording of the
items from self-attributional statements (e.g., “Is the cause
something that reflects an aspect of yourself?”) to
attributional statements made of clients (e.g., “Is the cause
something that reflects an aspect of the client?”).
Procedure
A packet consisting of a cover letter, the CPPS, the HCAS,
the CDS, and a self-addressed stamped envelope was mailed
308 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
randomly to each counseling center director listed in
Gallagher’s (1995) National Survey of Counseling Center
Directors in the United States and Canada. In the cover let-
ter, directors were asked for their assistance in gathering
data by passing the materials on to an experienced staff
member. Specifically, they were informed that the condi-
tions of the study necessitated responses from experienced
mental health professionals with ongoing caseloads of cli-
ents. Directors were invited to participate themselves if they
believed they met the criteria.
Participants read packets consisting of one of four versions
of the CPPS (male identity problem vignette with male ad-
justment problem vignette, male identity problem vignette
with female adjustment problem vignette, female identity
problem vignette with female adjustment problem vignette,
or female adjustment problem vignette with male identity
problem vignette). Vignettes were counterbalanced accord-
ing to problem type and client sex to prevent any potential
order effects. Three hundred thirty packets were mailed in
the initial mailing. Six packets from this initial mailing were
returned as nondeliverable. A follow-up letter was mailed 3
weeks after the initial mailing. A second follow-up letter was
mailed 3 weeks after the first reminder. This procedure re-
sulted in the return of 177 questionnaires, which represented
a 53.6% return rate. Ten of these questionnaires were re-
jected due to incomplete or missing data.
RESULTS
Endorsement of Model
Ratings on the attributions of responsibility for causing and
solving the problems described in the vignettes were summed
across the two measures assessing responsibility (CPPS and
HCAS) and recorded as low (ratings of 1–10) or high (rat-
ings of 11–20). Each model in Brickman et al.’s (1982)
conceptualization posits a different degree of responsibil-
ity for causing and solving the problem at hand. Briefly, the
medical model does not hold clients responsible for either
causing or solving their problems. The moral model holds
clients responsible for both the cause and the solution of
their problems. The compensatory model holds clients re-
sponsible only for solving their problems, whereas the en-
lightenment model holds them responsible only for caus-
ing their problems. Cross-tabulations were developed to
assess these models and the results presented in Table 1.
Table 1 indicates that participants subscribed to all four
of the models for both problem types presented. For the
adjustment problem, counselors were relatively evenly di-
vided in their endorsements of a compensatory model
(48.80%) and a moral model (44.64%). Counselors also
endorsed medical and enlightenment models (4.16% and
2.38%, respectively). For the identity problem, counselors
generally favored a moral model (51.20%) over a compen-
satory model (37.95%). Again, small percentages of coun-
selors also endorsed medical and enlightenment models
(4.81% and 6.02%, respectively).
 C o u n s e l o r s ’ A t t r i b u t i o n o f R e s p o n s i b i l i t y, E t i o l o g y, a n d C o u n s e l i n g S t r a t e g y

TABLE 1

Counselors’ Endorsement of Model by Problem Type and Client Sex

Problem Type Client Sex

Brickman et al.’s (1982) Model Type Adjustment % Identity % Male % Female %

Medical 7 4.16 8 4.81 5 2.90 10 6.17

Moral 75 44.64 85 51.20 76 44.18 84 51.85
Compensatory 82 48.80 63 37.95 83 48.25 62 38.27
Enlightenment 4 2.38 10 6.02 8 4.65 6 3.70
Note. Percentages reflect rounding error.

To determine if counselors altered their model of help-
ing to match the disorder presented, a 2 (disorder type) by
4 (helping model) cross-tabulation was submitted to a chi-
square analysis. The chi-square value was nonsignificant,
χ2(3, 334) = 5.741, p = .125, thus indicating that there was
no association between model endorsement and type of
problem presented.
Table 1 also indicates that counselors endorsed all four
models of helping for both male and female clients. For male
clients, counselors were relatively evenly divided in their
endorsements of a compensatory model and moral model of
helping (48.25% and 44.18%, respectively). A small percent-
age of counselors also endorsed enlightenment and medical
models of helping (4.65 and 2.90%, respectively). For female
clients, counselors generally favored a moral model of helping
(51.85%) to a compensatory model (38.27%). A small per-
centage of counselors also endorsed medical and enlighten-
ment models (6.17% and 3.70%, respectively).
Etiology Attributions
The prediction that counselors would select etiology attri-
butions logically consistent with an internal cause for an
identity problem and an external cause for an adjustment
problem was partially supported. As can be seen in Table 2,
etiology attributions varied across the two disorders. For the
identity disorder, the majority of counselors (81.32%) clearly
selected etiology attributions consistent with an internal cause
(lack of self-understanding). The internal attributions of ir-
rational thinking and unresolved feelings accounted for 5.42%
and 4.21% of etiology attributions, respectively. Thus, nearly
91% of counselors’ attributions reflected a belief in an inter-
nal cause for identity problems.
For the adjustment problem, counselors selected more
internal than external attributions. The internal attributions
of unresolved feelings, lack of self-understanding, and irra-
tional thinking accounted for nearly 55% of counselors re-
sponses (38.09%, 8.92%, and 7.73%, respectively). The ex-
ternal attribution of specific trauma accounted for only
33.92% of counselors’ attributions.
A 2 (problem type) by 12 (etiology attribution) was sub-
jected to chi-square analysis to determine if counselors
altered their attributions based on problem type presented.
Chi-square results were significant, χ2(9, 334) = 199.171,
p = .005. Thus, counselors’ attributions were associated with
problem type presented. A 2 (client sex) by 12 (etiology
attribution) cross-tabulation was performed to determine
whether counselors made differential etiology attributions
based on client sex. As can be seen in Table 2, counselors
selected similar etiology attributions for both male and

TABLE 2
Counselors’ Etiology Attributions by Problem Type and Client Sex

Problem Type Client Sex

Etiology Attribution Adjustment % Identity % Male % Female %

Genetics 1 0.59 1 0.60 1 0.58 1 0.61

Sick society 0 0.00
Social isolation 2 1.19
Dysfunctional family 0 0.00
Biological imbalance 2 1.19
Lack of self-understanding 15 8.92
Unresolved feelings 64 38.09
Stress 8 4.76
Specific trauma 57 33.92
Physical illness 0 0.00
Maladaptive learning 6 3.57
Irrational thinking 13 7.73
Note. Percentages reflect rounding error.
0 0.00 0 0.00 0 0.00
1 0.60 3 1.74 0 0.00
1 0.60 1 0.58 0 0.00
2 1.20 1 0.58 3 1.85
135 81.32 74 43.02 76 46.90
7 4.21 39 22.67 32 19.75
6 3.61 8 4.65 6 3.70
1 0.60 28 16.27 30 18.51
0 0.00 0 0.00 0 0.00
3 1.80 5 2.90 4 2.46
9 5.42 12 6.97 10 6.17

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 309

Kernes and McWhirter
female clients. Specifically, counselors selected lack of self-
understanding as the primary etiology attribution for both
men and women (43.02% and 46.90%, respectively). A large
number of counselors also selected the internal attribution
of unresolved feelings for both men and women (22.67%
and 19.75%, respectively). Specific trauma was identified
as the most important external etiology attribution for both
men and women (16.27% and 18.51%, respectively).
Ratings on the Causal Dimension Scale
A 2 × 2 multivariate analysis of variance (MANOVA) (disor-
der type by client sex) was conducted for the three ratings of
the Causal Dimension Scale (locus of causality of the client’s
problem, stability of the client’s problem, and controllability
of the client’s problem).
There was a significant main effect of type of problem
(Λ = .852 F = 19.067, p = .001, η = .15). A follow-up one-
way analysis of variance (ANOVA) indicated significant
differences between groups for ratings of locus of causality,
F(1, 332) = 47.319, p = .001. In general, clients described
as experiencing an identity problem were rated as having
greater internal locus of causality than those clients described
as experiencing an adjustment problem. A follow-up one-
way ANOVA also indicated a significant difference between
groups for ratings of problem stability, F(1, 332) = 10.140,
p = .002. Counselors ascribed more permanence to the
symptoms of clients described as having an adjustment
problem than to clients described as having an identity
problem. A follow-up one-way ANOVA also revealed
significant differences between groups for ratings of the con-
trollability of the problem, F(1, 332) = 30.845, p = .001.
Counselors ascribed more control over their problems to
clients with an identity problem than to clients with an
adjustment problem. Means and standard deviations are
presented in Table 3.
Ratings on Specific Interventions
A 2 × 2 MANOVA (disorder type by client sex) was
conducted for ratings of the three specific interventions of
restructuring the client’s environment, changing the client’s
interactions, and modifying the client’s thoughts or feelings.
There was a significant main effect of type of problem
TABLE 3
Counselors’ Ratings on the Causal
Dimension Scale
Causal Dimension Scale
Locus of Causality Stability Controllability
Variable M SD M SD M SD
Adjustment 16.68 4.69 11.32 4.77 11.64 4.41
Identity 19.95 3.95 9.84 3.62 14.21 4.05
310 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
(Λ = .976 F = 2.74, p = .043, η = .02). A follow-up one-way
ANOVA indicated a significant difference between groups for
restructuring the client’s environment, F(1, 332) = 7.620,
p = .006. In general, counselors endorsed this intervention as
more appropriate for clients with an adjustment problem (M =
3.48) than for clients with an identity problem (M = 3.09).
DISCUSSION
Predictions concerning counselors’ endorsement of theo-
retical models were not supported. No significant relation-
ship was detected between endorsement of theoretical
model and disorder type. Counselors endorsed all of
Brickman et al.’s (1982) models of helping for both disor-
der types. These results are in contrast to Worthington and
Atkinson’s (1993) findings that counseling center counse-
lors endorsed only the compensatory and moral models of
helping across disorder types. The number of counselors
endorsing medical and enlightenment models of helping in
the present study was quite small, however. For the adjust-
ment disorder, 4.16% of counselors endorsed a medical model
and 2.38% endorsed an enlightenment model. For the iden-
tity disorder, 4.81% of counselors endorsed a medical model
and 6.02% endorsed an enlightenment model. A more chal-
lenging finding is that counselors “equally” endorsed a com-
pensatory and moral model for helping clients with an
adjustment disorder and favored a moral model for helping
clients with an identity disorder. These findings contrast with
Worthington and Atkinson’s finding that counseling center
counselors favored a compensatory model for helping cli-
ents with an adjustment disorder and were equally divided
in their endorsement of moral and compensatory models for
helping clients with an identity disorder.
Although no significant relationship was observed be-
tween endorsement of theoretical model and disorder type,
clearly the majority of college counseling center counse-
lors adhere to either a moral or compensatory model of
helping (91.3%). These findings, although not as encourag-
ing as those of Worthington and Atkinson (1993), should
be viewed favorably if one accepts Brickman et al.’s (1982)
assumption that “models in which people are held respon-
sible for solutions (the compensatory and moral models)
are more likely to increase people’s competence than mod-
els in which they are not held responsible for solutions (the
medical and enlightenment models)” (p. 375).
Although counselors in the present study generally favor
the moral and compensatory models of helping, they do
not seem to distinguish between these models in their
conceptualization of client problems. Both the moral and
compensatory models view clients as responsible for find-
ing solutions to their problems. The issue of responsibility
for causing those problems differentiates these models, with
the moral model holding clients responsible for causing their
problems and the compensatory model avoiding discussion
of causal responsibility. It is possible that the scenario prob-
lems were too innocuous to enlist any significant differences
between endorsement of these two model types. Moreover,
 C o u n s e l o r s ’ A t t r i b u t i o n o f R e s p o n s i b i l i t y, E t i o l o g y, a n d C o u n s e l i n g S t r a t e g y
if these problems were too innocuous, causal responsibility
might have been distorted. Further research should examine
whether counselors endorse different model types when the
problems experienced by clients are more severe.
Predictions concerning etiology attributions were partially
supported. Counselors clearly selected attributions logically
consistent with an internal cause for the identity problem.
Specifically, lack of self-understanding accounted for 81.32%
of the responses. This finding is consistent with Worthington
and Atkinson (1993). Attributions were less supportive of
predictions for the adjustment problem. In fact, counselors
more frequently endorsed internal etiology attributions than
external attributions. Although specific trauma did account
for a large percentage of counselors’ attributions (33.92%),
this is less than Worthington and Atkinson’s finding of 55% of
counselors endorsing specific trauma.
As seen in Table 2, lack of self-understanding, the most
frequently endorsed etiology attribution for the identity
problem, received only minimal endorsement for the ad-
justment problem (8.92%). Likewise, specific trauma and
unresolved feelings, the most frequently endorsed etiol-
ogy attributions for the adjustment problem, received only
minimal endorsement for the identity problem (.60% and
4.21%, respectively). These results indicate that counse-
lors do make differential etiology attributions based on
disorder type presented.
Predictions concerning the three dimensions of the Causal
Dimension Scale (causality, stability, controllability) were
generally supported. As predicted, the symptoms of clients
with an identity problem were rated as having greater in-
ternal locus of causality than were symptoms of clients with
an adjustment problem. Similarly, the prediction that coun-
selors would attribute greater controllability to the symp-
toms of clients with an identity problem was also supported.
These findings suggest that counselors view identity problems
as more dispositional in nature than adjustment problems.
However, the prediction that identity problems would be rated
as more stable than adjustment problems was not supported.
In fact, counselors ascribed more permanence to the symp-
toms of adjustment problem clients. This finding is curious.
Perhaps it is possible that counselors paid attention to the
larger context of clients’ symptoms (i.e., reaction to a mother’s
death for the adjustment problem and a general sense of not
knowing who one is for the identity problem). Erikson’s (1980)
personality theory suggests that identity confusion, although
pervasive, lasts only for a circumscribed period from adoles-
cence through early adulthood. In this sense, a mother’s death
is a permanent event, whereas identity confusion is tempo-
rary. This explanation is tenuous. Future research might ex-
plore how variations in descriptions of disorder types might
affect ratings of symptom permanence. Further research would
be more informative if it allowed counselors to provide a ra-
tionale for their responses. For example, counselors could be
asked to provide commentary at the end of the vignettes de-
scribing their working hypotheses.
Whether these findings reflect a “dispositional bias” or
are merely good clinical judgment remains an open ques-
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
tion. Certainly Diagnostic and Statistical Manual of Mental
Disorders, fourth edition (DSM-IV; American Psychiatric
Association, 1994) criteria for the diagnosis of adjustment
and identity problems indicates different loci of client re-
sponsibility in contributing to their problems. There is a
question as to whether it is biased for clinicians to assume
that clients presenting with different problems are differ-
entially responsible for the cause of those problems. Lopez
(1989) has suggested there is a conceptual difference be-
tween a clinician who errs in adjusting diagnostic norms
and one who errs because of prejudicial attitudes. It is pos-
sible that counselors may make an error by not considering
individual differences in shaping their interventions and sim-
ply evaluating client needs based on categorization into such
gross membership as “adjustment problem” or “identity
problem.” However, this error differs from one made be-
cause of the influence of prejudicial attitudes based on cli-
ent background.
A similar question arises regarding responsibility for over-
coming one’s problems. If a client’s difficulties lie with the
fact that there is something “about the client” (i.e., person-
ality traits or characteristics), counselors are less likely to
hold that client responsible for overcoming their difficul-
ties than they are clients whose problems stem from a re-
action to an event. That is, clinicians may view dispositional
characteristics as less amenable to change. Again, the prob-
lems presented in the current study were rather innocuous.
It is quite possible that a different pattern of responsibility
attribution would emerge if the scenarios had presented
more severe forms of identity and adjustment problems.
Certainly, one can logically argue that counselors should
hold those who are contributing to the cause of their prob-
lems as more responsible for the resolution of those prob-
lems than they would clients whose problems are largely
reactions to significant events. The answers to those ques-
tions lie outside the scope of the present study.
Predictions for specific interventions were supported for
changing the client’s environment. Counselors viewed mak-
ing changes in a client’s environment as more appropriate for
clients with an adjustment problem than for clients with an
identity problem. These results are generally consistent with
Royce and Muehlke’s (1991) findings that internal attribu-
tions of clients’ problems were associated with treatment aimed
at changing the person, whereas stable attributions were linked
to strategies aimed at changing behaviors and systems. It is
unlikely that given such limited information as case vignettes,
counselors were able to form a complete treatment plan.
However, the fact that any specific interventions were en-
dorsed indicates how quickly such a plan can be set in mo-
tion. Meehl (1960) suggested that clinicians formulate a
client’s problems in a few sessions and that these formula-
tions remain largely unchanged over the course of therapy.
Client sex seemed to play an insignificant role in the
present study. No significant results were observed for the
influence of client sex in counselors’ endorsement of theo-
retical models or in etiology attributions. Likewise, client
sex failed to influence ratings on the CDS and ratings
311
Kernes and McWhirter
concerning specific interventions. This seems in opposition
to claims made for differential attributions based on client
sex (see Phillips & Gilroy, 1985, for an overview). One plau-
sible explanation for this finding is that counselors may be
trained to pay particular attention to a client’s sex when
making clinical decisions in order to avoid any potential “gen-
der biasing.” In an “innocuous” situation such as presented
here, client sex is not seen as a salient attribute in making
diagnostic and treatment decisions. At the same time, how-
ever, counselors are inextricably linked to their own gender
role socialization and hence may make decisions based on
such. Although a client’s sex may not play much of a role in
clinical decisions, counselors cannot escape entirely the in-
fluence their own sex may have in such decisions. Further
research should examine the interplay of client sex, counse-
lor sex and presenting issue when a client’s sex might be
germane (pregnancy, impotence, etc.). In addition, future
research should more precisely investigate the potential role
differences in counselor training (i.e., specialty areas) might
play in counselors’ attributions. Further research must also
expand the restrictive nature of the definition of bias and of
the biased clinician (Lopez, 1989). For example, research
must investigate not only instances of “overpathologizing” of
women but also explore the existence of “minimizing bias”
in the evaluation of feminine gender role symptomatology.
In addition to some of the limitations already discussed,
the present study is limited by threats to internal and exter-
nal validity common to analogue studies. Although reliabil-
ity and validity data are available for some of the instruments
used in the present study, the lack of reliability and validity
data for the CPPS and the marginal reliability of the HCAS
limits the conclusions that can be drawn. However, the CPPS
does use some of the suggestions Worell and Robinson (1994)
make for improving analogue research methods, and the
CPPS and HCAS represent the only measures specifically
designed to test Brickman et al.’s model of responsibility
attribution. The development of more reliable and valid mea-
sures is clearly needed. Results of this study cannot be
directly generalized to college counseling centers. Although
this study did employ university counseling center counse-
lors as participants, judgments made of fictional “clients” may
only remotely parallel diagnostic and treatment decisions
made with actual student clients. Logically, additional research
needs to include counselor evaluations of actual clients.
REFERENCES
Abraham, I. L. (1985). Causal attributions of depression: Reliability of
the Causal Dimension Scale in research on clinical inference. Psycho-
logical Reports, 56, 415–418.
Abramowitz, S. I., Abramowitz, C., Jackson, C., & Gomes, B. (1973).
The politics of clinical judgment. What nonliberal examiners infer
about women who do not stifle themselves. Journal of Consulting and
Clinical Psychology, 41, 385–391.
American Psychiatric Association. (1994). Diagnostic and statistical
manual of mental disorders (4th ed.). Washington, DC: Author.
American Psychological Association Task Force. (1975). Report of the
task force on sex bias and sex-role stereotyping in psychotherapeutic
practice. American Psychologist, 30, 1169–1175.
312 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Atkinson, D. R., Worthington, R. L., Dana, D. M., & Good, G. E. (1991).
Etiology beliefs, preferences for counseling orientations, and counsel-
ing effectiveness. Journal of Counseling Psychology, 38, 258–264.
Avants, S. K., Margolin, A., & Singer, J. (1993). Psychological interven-
tions and research in the oncology setting. An integrative framework.
Psychotherapy, 30, 1–10.
Batson, C. D. (1975). Attribution as a mediator of bias in helping. Jour-
nal of Personality and Social Psychology, 72, 455–466.
Batson, C. D., O’Quin, K., & Pych, V. (1982). An attribution theory analy-
sis of trained helpers’ inferences about clients’ needs. In T. A. Wills
(Ed.), Basic processes in helping relationships (pp. 59–80). New York:
Academic Press.
Becker, D., & Lamb, S. (1994). Sex bias in the diagnosis of borderline
personality disorder and posttraumatic stress disorder. Professional
Psychology: Research and Practice, 25, 55–61.
Brickman, P., Rabinowitz, V. C., Karuza, J., Jr., Coates, D., Cohn, E., Kid-
der, L. (1982). Models of helping and coping. American Psychologist, 37,
368–384.
Broverman, I. K., Broverman, D. M., Clarkson, F. E., Rosenkrantz, P., &
Vogel, S. R. (1970). Sex-role stereotypes and clinical judgments of
mental health. Journal of Consulting and Clinical Psychology, 34, 1–7.
Costrich, N., Feinstein, J., Kidder, L., Marcek, J., & Pascale, L. (1975).
When stereotypes hurt: Three studies of penalties of sex-role rever-
sals. Journal of Experimental Social Psychology, 11, 520–530.
Daws, P. (1967). The aetiology of mental disorder. Bulletin of the British
Psychological Society, 20, 45–47.
Dumont, F. (1993). Inferential heuristics in clinical problem formula-
tion: Selective review of their strengths and weaknesses. Professional
Psychology: Research and Practice, 24, 196–204.
Erikson, E. H. (1980). Identity and the life cycle. New York: Norton.
Feinberg, J. (1970). Doing and deserving: Essays in the theory of responsibility.
Princeton, NJ: Princeton University Press.
Fitzgerald, L. F., & Cherpas, C. C. (1985). On the reciprocal relationship
between gender and occupation. Rethinking the assumptions concerning
masculine career development. Journal of Vocational Behavior, 27,
109–122.
Foulks, E. F., Persons, J. B., & Merkel, R. L. (1986). The effect of patients’
beliefs about their illnesses on compliance in psychotherapy. American
Journal of Psychiatry, 143, 340–344.
Furman, B., & Ahola, T. (1989). Adverse effects of psychotherapeutic
beliefs: An application of attribution theory to the critical study of
psychotherapy. Family Systems Medicine, 7, 183–195.
Gallagher, R. (1995). National survey of counseling center directors [mono-
graph]. Alexandria, VA: International Association of Counseling Services.
Hansen, R. D. (1980). Commonsense attribution. Journal of Personality
and Social Psychology, 39, 996–1009.
Harvey, J. H., & Galvin, K. S. (1984). Clinical implications of attribution
theory and research. Clinical Psychology Review, 4, 15–33.
Heider, F. (1958). The psychology of interpersonal relations. New York: Wiley.
Jack, R., & Williams, J. (1991). Attribution and intervention in self-
poisoning. British Journal of Medical Psychology, 64, 345–358.
Karuza, J., Jr., Zevon, M. A., Gleason, T. A., Karuza, C. M., & Nash, L.
(1990). Models of helping and coping, responsibility attributions, and
well-being in community elderly and their helpers. Psychology and
Aging, 5, 194–208.
Karuza, J., Jr., Zevon, M. A., Rabinowitz, V. C., & Brickman, P. (1982).
Attribution of responsibility by helpers and recipients. In T. A. Wills
(Ed.), Basic processes in helping relationships (pp. 107–129). New York:
Academic Press.
Kedric, H. C. (1985). Attitudes toward the etiology of psychopathology
in members of a fundamentalist Christian church. Dissertation Ab-
stracts International, 47, 1703B. (University Microfilms No. 86-14914)
Lopez, S. R. (1989). Patient variable biases in clinical judgment: Con-
ceptual overview and methodological considerations. Psychological
Bulletin, 106, 184–203.
Lopez, S. R., & Wolkenstein, B. H. (1990). Attributions, person perception, and
clinical issues. In S. Graham & V. S. Folkes (Eds.), Attribution theory: Applica-
tions to achievement, mental health, and interpersonal conflict (pp. 103–121).
Hillsdale, NJ: Erlbaum.
 C o u n s e l o r s ’ A t t r i b u t i o n o f R e s p o n s i b i l i t y, E t i o l o g y, a n d C o u n s e l i n g S t r a t e g y
Maloney, H. N. (1985). Theories of psychopathogy [Monograph]. Pasadena,
CA: Fuller Theological Seminary.
Meehl, P. (1960). The cognitive activity of the clinician. American Psy-
chologist, 15, 19–27.
Morojele, N., & Stephenson, G. (1992). The Minnesota model in the
treatment of addictions: A social psychological assessment of changes
in beliefs and attributions. Journal of Community and Applied Social
Psychology, 2, 25–41.
Morrow, K. A., & Deidan, C. T. (1992). Bias in the counseling process:
How to recognize and avoid it. Journal of Counseling & Development,
70, 571–577.
Murdock, N. L., & Freemont, S. K. (1989). Attributional influences in coun-
selor decision making. Journal of Counseling Psychology, 36, 417–422.
O’Malley, K. M., & Richardson, S. (1985). Sex bias in counseling: Have
things changed? Journal of Counseling and Development, 63, 294–299.
Phillips, R. D., & Gilroy, F. D. (1985). Sex-role stereotypes and clinical
judgments of mental health: The Broverman’s findings reexamined.
Sex Roles, 12, 179–193.
Pistrang, N., & Barker, C. (1992). Client’s beliefs about psychological
problems. Counseling Psychology Quarterly, 5, 325–335.
Poole, F. D., & Tapley, A. (1988). Sex roles, social roles, and clinical judg-
ments of mental health: The Broverman’s findings reexamined. Sex
Roles, 12, 265–272.
Robertson, J., & Fitzgerald, L. F. (1990). The (mis)treatment of men:
Effects of client gender role and life style on diagnosis and attribution
of pathology. Journal of Counseling Psychology, 37, 3–9.
Royce, W. S., & Muehlke, C. V. (1991). Therapists’ causal attributions of
clients’ problems and selection of intervention strategies. Psychologi-
cal Reports, 68, 379–386.
Russell, D. (1982). The Causal Dimension Scale: A measure of how in-
dividuals perceive causes. Journal of Personality and Social Psychology,
42, 1137–1145.
Schmidt, G., & Weiner, B. (1988). An attribution-affect-action theory of
behavior: Replications of judgments of help-giving. Personality and
Social Psychology Bulletin, 14, 610–621.
Smith, M. L., (1980). Sex bias in counseling and psychotherapy. Psycho-
logical Bulletin, 87, 392–407.
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Spitzer, R. L., Skodol, A. E., Gibbon, M., & Williams, J. B. W.
(1981). DSM-III casebook. Washington, DC: American Psychiatric
Association.
Strohmer, D. C., Biggs, D. A., Keller, K. E., & Thibodeau, J. R. (1984).
Clinical judgment and affective disorders. Journal of Counseling Psy-
chology, 31, 99–103.
Strohmer, D. C., Haase, R. F., Biggs, D. A., & Keller, K. E. (1982). Process
models of counselor judgment. Journal of Counseling Psychology, 29,
597–606.
Swenson, E. V., & Ragucci, R. (1984). Effects of sex-role stereotypes and
androgynous alternatives on mental health judgments of psychothera-
pists. Psychological Reports, 54, 475–481.
Townsend, J. M. (1975). Cultural conceptions, mental disorders and so-
cial roles: A comparison of Germany and America. American Socio-
logical Review, 40, 739–752.
Tracey, T. J. (1988). Relationship of responsibility attribution congru-
ence to psychotherapy outcome. Journal of Social and Clinical Psy-
chology, 7, 131–146.
Vallerand, R. J., & Richer, F. (1988). On the use of the causal dimension
scale in a field setting: A test with confirmatory factor analysis in suc-
cess and failure situations. Journal of Personality and Social Psychology,
54, 704–712.
Weiner, B. (1979). A theory of motivation for some classroom experi-
ences. Journal of Educational Psychology, 71, 3–25.
Weiner, B. (1993). On sin versus sickness: A theory of perceived respon-
sibility and social motivation. American Psychologist, 48, 957–965.
West, R., & Power, D. (1995). Alcoholic’s beliefs about the responsibil-
ity for, and recovery from, their condition. Drug and Alcohol Review,
14, 55–62.
Worell, J., & Robinson, D. (1994). Reinventing analogue methods for
research with women. Psychology of Women Quarterly, 18, 463–476.
Worthington, R. L., & Atkinson, D. R. (1993). Counselors’ responsibil-
ity and etiology attributions, theoretical orientations, and counseling
strategies. Journal of Counseling Psychology, 40, 295–302.
Young, R. A., & Marks, S. E. (1986). Understanding attributional pro-
cesses in cross-cultural counseling. International Journal for the Ad-
vancement of Counseling, 9, 319–330.
313
Risk Factors for Suicidality Among Clients With Schizophrenia

Robert C. Schwartz and Benjamin N. Cohen

Among clients with schizophrenia, suicidality is associated with extreme personal distress, an increased number of inpatient
hospitalizations, increased health care expenditures, and early mortality. This study attempted to identify risk factors for current
suicidality in clients diagnosed with schizophrenia (N = 223). Results indicated that severity of depressive symptoms most strongly
correlated with degree of suicidality. Younger age and recent traumatic stress each significantly predicted suicidality independent
of depressive symptoms. Stepwise regression procedures showed that the combination of depression, younger age, and trau-
matic stress might provide a general prediction model for suicidality among clients diagnosed with schizophrenia. Counseling
implications of these findings are outlined.

S
chizophrenia is a common mental disorder with
international prevalence rates of approximately
1%. In the United States, prevalence rates range
from .6% to 1.9% (World Health Organization,
1995). Symptoms of schizophrenia are usually
chronic and pervasive, leading to over $16 billion in annual
health care expenditures (American Psychiatric Association
[APA], 1997). Costly hospitalizations and early mortality
are also common results of schizophrenic symptoms. One
of the most common reasons for premature death among
clients with schizophrenia is suicide (Black & Fischer, 1992).
Estimates of suicidal ideation (i.e., thoughts of killing oneself)
range from 60% to 80%, and approximately 20%–42% of these
clients actively attempt suicide (Drake, Gates, Whitaker, &
Cotton, 1985). Studies also indicate that approximately
10%–15% of clients with schizophrenia complete suicide
(Black, Warrack, & Winokur, 1985; Landmark, Cernovsky, &
Mersky, 1987; Miles, 1977; Roy, Mazonson, & Pickar, 1984).
For example, Caldwell and Gottesman (1990) found that
50% of clients with schizophrenia attempt suicide, and up
to 13% complete the act. Given the significance of this prob-
lem, it is important to identify risk factors associated with
suicidality among clients with schizophrenia.
Prior research suggests that certain psychotic symptoms
may correlate with suicidality. In a retrospective autopsy
study, Heila, Isometsa, Henriksson, and Heikkinen (1997)
reported that depression and severe psychosis were related
to suicides. These findings support prior research indicating
that depressive symptoms (e.g., extreme hopelessness) cor-
relate with suicides among clients with schizophrenia (S.
Cohen, Lavelle, Rich, & Bromet, 1994; Drake & Cotton, 1986;
Milch, 1990; Roy, 1982). Regarding psychotic symptoms,
several authors (S. Cohen et al., 1994; Gallagher, Dinan,
Sheehy, & Baker, 1995; K. J. Kaplan & Harrow, 1996,
1999) reported that increased hallucinations and delu-
sions were indicative of suicidal risk. Therefore, severe
psychotic symptoms may increase depression, leading to
a heightened risk of suicide (Fenton, McGlashan, Victor,
& Blyler, 1997; K. J. Kaplan & Harrow, 1999).
Rossau and Mortensen (1997) asserted that men are
significantly more likely to complete suicide than are women,
indicating that demographic characteristics may correlate with
suicidal risk. Other investigators have proposed that no clear
relationship exists between suicidal behavior and sex (Amador
et al., 1996; S. Cohen et al., 1994). As J. Cohen, Test, and Brown
(1990) explained, age may be more highly correlated with
suicidality than sex. There may be an inverse relationship be-
tween age and suicidal risk (Breier & Astrachan, 1984; Langley
& Bayatti, 1984). That is, younger clients tend to become more
suicidal more often. However, S. Cohen et al. (1994) and Amador
et al. (1996) failed to link age with suicidal behavior. Therefore,
further research is necessary to explore the relationship between
demographic characteristics and suicidality among clients with
schizophrenia.
It has also been hypothesized that severe life stressors
may elicit suicidality among clients with schizophrenia.
However, only one study found to date has attempted to
correlate life stressors with suicidality in clients with schizo-
phrenia. In this study, Weiden and Roy (1992) asserted that
a combination of psychological stressors and poor overall
functioning may lead to suicides. Perhaps severe life stres-
sors combined with poor global functioning and inadequate
coping mechanisms result in suicidal behaviors (Weiden &
Roy, 1992).
As Heila et al. (1997) explained, many of the studies cited
earlier are hampered by methodological limitations. These

Robert C. Schwartz is an assistant professor and coordinator of the Addiction Counseling Certification Program in the Department of Counseling and Special
Education at The University of Akron, Akron, Ohio. Benjamin N. Cohen is a doctoral student in the Department of Counseling, Educational Psychology, and
Special Education at Michigan State University, East Lansing. Correspondence regarding this article should be sent to Robert C. Schwartz, Department of
Counseling and Special Education, The University of Akron, 127 Carroll Hall, Akron, OH 44325-5007 (e-mail: rcs@uakron.edu).

314 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79

 Risk Factors for Suicidality Among Clients With Schizophrenia
include small sample size, the use of nonstandardized as-
sessment instruments, and retrospective (e.g., autopsy)
investigations that did not evaluate current suicidality.
In addition, we noted that the assessment of important
variables (e.g., medical/physical problems, family relationship
problems, and work/school problems) were not included in
prior research designs. It is crucial to discriminate immediate
risk factors if eventual suicidal behaviors are to be forestalled.
If immediate risk factors for suicidality are identified, clini-
cians may be better equipped to preempt suicidal gestures
(Sullivan, Young, & Morgenstern, 1997). This study attempted
to expand on prior research by investigating the relationship
between current suicidality and demographic characteristics,
psychological and emotional symptoms, life stressors, and
overall functioning in a large sample of clients with schizo-
phrenia. We hypothesized that a combination of current
symptoms, life stressors, and demographic characteristics
would predict suicidal risk. Specifically, we hypothesized that
as life stressors and symptoms of illness intensified, suicidality
would become more severe among clients with schizophrenia
who had certain demographic characteristics.
METHOD
Participants
Participants for this study were selected from a crisis stabili-
zation unit (CSU) at a large, 12-county community agency
in Florida. The CSU is an emergency mental health
facility designed to assess symptoms (including suicidal
risk), diagnostic criteria, social problems, global functioning,
and treatment needs of clientele presenting with emotional
crises. All clientele participate in a comprehensive evalua-
tion conducted by trained mental health personnel. Then,
standardized instruments are completed, and a diagnosis (if
any) is given. Archival data on all adult clients diagnosed
with schizophrenia (any subtype) over a continuous 6-month
period were included in this study (N = 223). Participants
were not discriminated based on income, age (if over 18
years), severity of symptoms, physical disabilities, current
or prior treatment history, or living status (i.e., homeless,
supported housing, or independent living environments).
The only inclusion criterion for this study was a confirmed
diagnosis of schizophrenia. Although clients gave written
informed consent for participation in clinical evaluations,
consent for participation in this study was not obtained
because archival data were used from medical records. The
identity of participants was protected at all stages of data
collection and analysis.
Participants were between the ages of 18 and 79 years
(M = 38.8, SD = 10.1); 135 (61%) were women and 88
(39%) were men. On average they evidenced a long-standing
treatment history (M years of treatment = 13.2, SD = 8.8).
Regarding race, 117 (52%) were Caucasian and 106 (48%)
were African American. Participants were generally from a
lower socioeconomic class; average monthly income was
$460 (SD = $302). In general, participants were unem-
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
ployed. Average days worked during the previous month
were 1.9 (SD = 6.0). These characteristics are representa-
tive of clients with schizophrenia and thus reflect the
broader population (APA, 1997).
All clients in this study met the diagnostic criteria for
schizophrenia according to the Diagnostic and Statistical
Manual of Mental Disorders, fourth edition (DSM-IV; APA,
1994). Diagnoses were based on the Structured Clinical
Interview for DSM-IV (First, Spitzer, Gibbon, & Williams,
1995). Clients with psychotic disorders other than schizo-
phrenia (e.g., schizoaffective disorder or delusional disor-
der) were excluded from participation in this study because
the symptoms and course of other psychotic disorders dif-
fer significantly from those of schizophrenia (APA, 1994).
Instruments
The Global Assessment of Functioning (GAF) Scale (APA,
1994) is a 100-point instrument used to rate clients’ cur-
rent overall functioning based on a standard global mental
health–illness continuum. The GAF is a widely used in-
strument with good validity and reliability that matches
clients’ self-reported functioning (Jones, Thornicroft, Coffey,
& Dunn, 1995; Piersma & Boes, 1995). To obtain interrater
reliability estimates for this study, 20% (45) of evaluations
were conducted simultaneously by two clinicians. Pearson
product–moment correlations of GAF scores showed an
interrater agreement of r = .91.
The Functional Assessment Rating Scale (FARS; Ward &
Dow, 1994) is a 17-item instrument used to assess emotional
symptoms and psychosocial impairments in clients with
severe mental illnesses, especially those with psychotic
disorders. Each item is scored independently after a struc-
tured clinical interview. Items are scored according to how
severe a certain symptom or area of impairment is for that
client. A standardized 9-point rating system is used for all
items (1 = absent, 3 = mild, 5 = moderate, 7 = severe, 9 =
extreme). In this study, the FARS was used to assess the
following symptoms and psychosocial impairments: severity
of suicidality, depression, cognitive impairments, disturbed
thought processes, recent traumatic stress, medical/physical
problems, work/school problems, self-care deficits, and family
relationship problems. Severity of suicidality served as the
dependent variable, and all other items were used as indepen-
dent variables. All variables were specifically operationalized
in behavioral terms, and all clinicians were trained to use the
FARS before clinical evaluations commenced.
According to Ward and Dow (1994), the FARS has good
interrater reliability, good stability reliability, and good con-
struct validity. In a more recent test of the instrument’s
psychometric properties, Schwartz (1999) reported mean
interrater reliability correlations of r = .88, mean stability
reliability correlations of r = .86, mean concurrent reliabil-
ity correlations of r = .89, and good construct validity. To
test interrater agreement on FARS items used in this study,
20% (45) of evaluations were conducted simultaneously
by two clinicians. Pearson product–moment correlations
315
Schwartz and Cohen

revealed the following statistics on interrater agreement: mentation included demographic information, a medical his-
traumatic stress (r = .95), depression (r = .93), suicidality tory, prior treatment history, a social history, and a family
(r = .92), cognitive impairments (r = .92), disturbed thought history. Clients then participated in two standardized clini-
processes (r = .90), medical/physical problems (r = .89), cal interviews, the Structured Clinical Interview for the Func-
self-care deficits (r = .88), family relationship problems tional Assessment Rating Scale (SCI-FARS; Ward et al., 1995)
(r = .80), and work/school problems (r = .61). and the Structured Clinical Interview for DSM-IV (First et
al., 1995). Total interview duration averaged approximately
FARS Items 90 minutes. After each individual interview, clinicians
diagnosed the client (if appropriate) according to DSM-IV
Degree of suicidality was the dependent variable in this
criteria. All clinical interviews were conducted by master’s-
study. Suicidality was defined as the severity of reported
or doctoral-level clinicians specializing in the assessment of
and observed dangerousness to self. Severity of suicidality
mental disorders, and all diagnoses were confirmed by a board-
was operationalized as the degree to which the client mani-
fested suicidal ideation, intent to harm oneself, lethality of a certified psychiatrist. Then, each client was rated using the
current plan (if any), and opportunity or means of complet- GAF scale (APA, 1994) and the FARS (Ward & Dow, 1994).
All raters were blind to the protocol and existence of this
ing that plan. Clinicians specifically assessed the frequency
study at the time of evaluations. The University of Florida
and nature of suicidal thoughts, whether and to what degree
Institutional Review board, the Idaho State University Hu-
the client wanted and intended to commit suicide, whether
man Subjects Committee, and the Meridian Behavioral
they had a current plan (and if so how organized and lethal
the plan was), and how achievable the plan would be given Healthcare Risk Management Department approved the
their current psychological symptoms and their physical re- research design and methodology for this study.
sources. Only current suicidality was evaluated because past
suicidal ideation or gestures could confound results. RESULTS
The independent variables in this study were separated into
First, frequency distributions were obtained for all FARS
symptoms of illness and social problems or impairments.
items and for the GAF. Then, a multiple regression omni-
Regarding symptoms of illness, depression was measured
bus test was conducted to determine whether the combi-
according to degree of reported and observed hopelessness,
nation of symptoms, social problems, global functioning,
helplessness, depressed mood, and current neurovegitative
and demographic characteristics significantly correlated with
symptoms (e.g., disturbed sleep or appetite patterns). Cogni-
ratings of suicidality. Eleven independent variables were
tive impairments included cognitive disorganization, dis-
included in the omnibus test: 3 symptom variables (sever-
orientation, memory problems, poor abstract thought, and
ity of depression, cognitive impairments, and disturbed
poor attention or concentration. These impairments were mea-
thought processes), 5 social variables (traumatic stress,
sured using a mental status examination. Disturbed thought
medical/physical problems, work/school problems, self-care
processes referred to psychotic symptoms, such as delusions,
deficits, and poor family relationship), 1 measure of global
hallucinations, paranoia, and grandiosity. Medical/physical
functioning (the GAF score), and 2 demographic variables
problems included any current impairing physical condition,
(age and gender).
such as injuries or illnesses. Medical/physical problems were
Follow-up multiple regression tests were used to evalu-
rated according to self-report or clinician observation.
ate the predictive significance of each independent vari-
In terms of social problems or impairments, traumatic
able while controlling for the influence of other indepen-
stress was operationalized as upsetting memories, night ter-
dent variables. That is, specific tests were used as follow-up
rors, hypervigilance, and flashbacks related to a traumatic
measures to determine which independent variable(s) sig-
event (e.g., physical or sexual abuse). Work/school prob-
nificantly contributed to the overall prediction model. Fi-
lems were operationalized as the inability to cope with the
nally, a stepwise regression procedure was conducted to sta-
demands of work/school, poor performance on the job or
tistically determine the most effective overall prediction
with school grades, being terminated from work or expelled
model regarding suicidality. An alpha level of p < .01 was
from school, or inability to maintain consistent employ-
used for all statistical tests. A power analysis revealed that,
ment or registration status. Self-care deficits were defined
given a predicted moderate effect size and an alpha level of
as a reduced ability to independently perform basic adult
p < .01, a power of at least .80 could be expected using the
daily living skills, such as personal hygiene, cooking, clean-
sample size obtained in this study (J. Cohen, 1992; J. Cohen
ing, or money management. Family relationship problems
& Cohen, 1983).
referred to arguments at home, conflicted relationships with
Results indicated that, on average, participants manifested
parents or siblings, lack of emotional or social support, or
a mild degree of suicidality (M = 1.8, SD = 1.7), although a
lack of desired contact with family members.
full range of suicidality was observed in the sample popula-
tion (Range = 1–9). On average, clients also exhibited se-
Procedure
vere impairments in overall functioning as measured by the
On unsolicited presentation at the CSU, the psychosocial GAF scale (M = 42.5, SD = 15.3). The most severe symp-
history of each participant was documented. This docu- toms and psychosocial problems in this sample included

316 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79

 Risk Factors for Suicidality Among Clients With Schizophrenia

disturbed thought processes (M = 4.7, SD = 2.5), cognitive TABLE 2

impairments (M = 4.4, SD = 2.1), and work/school prob-
lems (M = 4.3, SD = 2.7), respectively. These results are Results of Follow-Up Regression Tests for
consistent with a population of clients with schizophrenia Predictor Variables
presenting for psychological evaluation during a period of
crisis (APA, 1997). Table 1 displays a frequency distribu- Standardized
tion for all FARS items assessed in this study. FARS Item Beta Weights t value p value
Results showed that the omnibus regression model was sta-
Depression .39 7.40 *p = .001
tistically significant (F = 10.8, df = 11,211, p < .0001). For the Age –.04 –4.40 *p = .001
omnibus multiple regression model, the proportion of shared Traumatic stress .17 3.00 *p = .002
variance equaled .44, suggesting that 44% of the variance in Gender .41 2.00 p= .04
ratings of suicidality were accounted for by the 11 independent Disturbed thoughts .07 1.40 p= .17
Medical/physical
variables included in the model. Specific follow-up multiple re- problems .08 1.20 p = .23
gression tests showed that three variables were significantly and Self-care deficits .06 1.10 p = .24
independently associated with ratings of suicidality. These Work/school problems –.04 –1.10 p = .26
Cognitive impairments .05 0.84 p = .40
tests indicated that, when the influence of all other vari- GAF Scale score –.01 –0.81 p = .41
ables were controlled, severity of depression (t = 7.4, p = .001), Poor family relationship .01 0.29 p = .77
age (t = –4.4, p = < .001), and traumatic stress (t = 3.0, p = .002)
were significantly related to ratings on suicidality. Table 2 Note. See Table 1 Note. GAF = Global Assessment of Functioning.
*p < .01.
shows results of specific regression tests, including standardized
beta weights (parameter estimates) for all variables.
Finally, a stepwise regression procedure was used to obtain spent approximately one third of their life in treatment.
the most effective overall prediction model for suicidality. They were from a low socioeconomic group (based on
Results suggest that depressive symptoms alone accounted monthly income) and were generally unemployed. Most
for 28% of the variance in ratings of suicidality. When age was participants lived independently, with family members, or
added to the prediction model as a second variable, the two- in supported housing units in the community. Thus, we
variable combination accounted for 33% of the variance in believe that empirical results should adequately general-
ratings of suicidality. After traumatic stress was entered into ize to a population of clients with schizophrenia who are
the prediction equation, the three-variable model accounted receiving treatment for their illness.
for 38% of the variance in ratings of suicidality. The addition Our research hypothesis was supported by results obtained
of other independent variables added only minimally (and in- from multiple regression analyses, expanding the knowledge
significantly) to the prediction model. Therefore, although base regarding risk factors for suicidality among clients with
depressive symptoms were found to most strongly associate schizophrenia. First, severity of depression was found to
with ratings of suicidality, the most significant combination of significantly positively correlate with degree of suicidality.
risk factors was the three-variable model including depressive As depressive symptoms (e.g., hopelessness, helplessness,
symptoms, age, and recent traumatic stress. neurovegitative symptoms) increased, suicidal ideation and
intent also increased. Other investigators (S. Cohen et al.,
DISCUSSION 1994; Drake & Cotton, 1986; Heila et al., 1997; Milch, 1990;
Rossau & Mortensen, 1997; Roy, 1982; Taiminen, 1994) have
In general, the participants selected for this study manifested found this correlation as well. Moreover, it is well documented
characteristics common to clients with schizophrenia (APA, in clinical training manuals that depression is a dysphoric
1997). Participants were middle-aged and, on average, had mental state that may induce suicidal ideation or culminate
in suicide if not properly treated (H. I. Kaplan & Sadock,
TABLE 1 1995). Thus, the combination of severe depression and
schizophrenia may constitute a psychiatric emergency (Roy,
Frequency Distribution for FARS Ratings 1995). For example, King (1994) recently found that over a
7-year period 84% of persons who completed suicide had
FARS Item M SD symptoms of schizophrenia or severe depression, or both,
Disturbed thoughts 4.7 2.5 immediately preceding the suicidal act. Therefore, assessing and
Work/school problems 4.3 2.7 monitoring depressive symptoms is extremely important for
Poor family relationship 3.1 2.0 professionals serving clients with schizophrenia. We propose
Traumatic stress 2.2 1.8
Suicidality 1.8 1.7 that future research should attempt to ascertain which
Cognitive impairments 4.4 2.1 specific aspects of depression lead to increased suicidality in
Self-care deficits 4.0 2.4 this client population. If specific aspects of the depressive
Depression 2.6 1.9
Medical/physical problems 1.9 1.5
syndrome can be identified and linked to suicidal potential,
then clinicians can be better prepared to intervene with
Note. FARS = Functional Assessment Rating Scale. clients manifesting these symptoms.

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 317

Schwartz and Cohen
Two other variables—age and traumatic stress—were
each independently correlated with ratings of suicidality
in this sample. Age was negatively correlated with degree of
suicidality; that is, younger clients tended to evidence more
lethality than older clients. This finding has also been
previously reported in the literature (APA, 1997; Breier
& Astrachan, 1984; J. Cohen et al., 1990; Langley &
Bayatti, 1984), lending additional support to the hypoth-
esis that younger clients with schizophrenia should be
monitored more conscientiously for suicidality, especially
when depressed (Gupta, Black, Arndt, Hubbard, &
Andreasen, 1998). However, research has yet to deter-
mine why younger clients with schizophrenia tend to
manifest increased suicidality. Perhaps they are more
profoundly affected by psychosocial stressors due to un-
derdeveloped coping mechanisms. Or possibly they dis-
play more severe psychotic symptoms related to the early
age of onset of this disorder, resulting in more severe au-
ditory command hallucinations or persecutory delusions.
A third hypothesis is that younger clients with schizo-
phrenia may become more depressed and discouraged
about their future due to the long-term and pervasive
nature of this disorder. As Schwartz and Petersen (1999)
explained, increased insight may lead to heightened
suicidality particularly among younger clients. Future re-
search should attempt to ascertain the primary factor(s)
resulting in increased suicidality in younger clients with
schizophrenia. The early age of onset of schizophrenia
coupled with an increased risk of suicide among younger
clients necessitates heightened preventive interventions
focused on this client population.
Recent traumatic stress was shown to positively corre-
late with suicidality in this sample of clients with schizo-
phrenia. As severity of traumatic stress increased, suicidal
ideation and intent also increased. To our knowledge, the
impact of recent traumatic stress on suicidality has not been
studied previously with this population of clients. The afore-
mentioned finding is intuitively sensible, however, because
one’s intrapsychic perceptions and coping abilities in rela-
tion to extreme stressors certainly affects one’s overall out-
look of life and death (Roy, 1995). If a client who is schizo-
phrenic confronts a traumatic event, the intensity of that
event, as well as the degree to which the client is able to
manage the resulting affect, may have a direct impact on
danger to self. Degree of suicidality may also be a function
of what type of traumatic event occurred, the duration of
traumatic stress, and whether or not therapeutic interven-
tions were received. These specific factors were not included
in the data collection for this study. Clearly, further investi-
gation concerning the relationship between traumatic stress
and suicidality is warranted because appropriate clinical
interventions may preempt suicidal behaviors if the critical
aspects of the traumatic experience were known. Until ad-
ditional empirical research is conducted, we suggest that
practitioners conscientiously monitor (and possibly increase
therapeutic interventions with) clients with schizophrenia
reporting recent traumatic stress.
318 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
COUNSELING IMPLICATIONS
Many of the findings reported in this study have direct clini-
cal implications. Most important, practitioners should be
aware that a combination of severe depressive symptoms
and recent traumatic stress among younger clients with
schizophrenia may indicate a substantially increased risk of
suicidality. Severity of depression and traumatic stress can
be directly and reliably assessed by counselors (e.g.,
Zimmerman, 1994), and we strongly advise that clinicians
evaluate these areas regardless of whether or not they are
presented by clients as primary treatment concerns. Treat-
ment for depression and traumatic stress are both widely
available and therapeutically effective if used in a timely
manner. Therefore, if counselors conscientiously monitor
symptoms associated with an increased risk of suicidality
among clients with schizophrenia, preventive interventions
can be used more effectively. In addition, it is important
for counselor training to include suicide assessment, in
general, as well as the impact on the counselor both emo-
tionally and professionally of a client’s suicide (Foster &
McAdams, 1999).
Should a mental health professional confront a client
with schizophrenia who manifests suicidality, standard
crisis management procedures exist that serve as general
guidelines for appropriate interventions (see Gilliland &
James, 1997). Practitioners should first assess the client’s
lethality (ideation, intent, plan, and means). Then, symptoms
specifically associated with suicidal thoughts (e.g., depres-
sion and traumatic stress) should be evaluated. Next, the
extent and availability of social support networks (e.g., friends,
family members, case managers) should be ascertained.
Finally, treatment options should be discussed and clinical
interventions should be initiated immediately if necessary.
Thus, it is important to investigate biological, psychologi-
cal, and social aspects of suicidality with clients who have
schizophrenia (Harkavy-Friedman & Nelson, 1997). Treat-
ment options may range from more intensive outpatient
counseling to the addition of new psychotropic medica-
tions to inpatient hospitalization, depending on the client’s
potential danger to self (APA, 1997). The optimum treat-
ment strategy should use the least restrictive environment
necessary to provide increased personal safety and improved
psychological functioning.
It should be noted, however, that this study did not inves-
tigate past suicidal behaviors or suicides. Thus, results may
be generalizable only to those clients reporting current
suicidality. In addition to the fact that current suicidality
does not always lead to suicidal gestures, there were several
other limitations to this study. Even though most indepen-
dent variables (e.g., medical/physical problems) were not
found to associate with suicidality in this study or in prior
research, psychosocial variables not included in this inves-
tigation may ultimately demonstrate an impact on
suicidality. For example, substance abuse has been shown
to correlate with increased suicidality in patients with
schizophrenia (Heila et al., 1997; Roy, 1995). Finally, the
 Risk Factors for Suicidality Among Clients With Schizophrenia
independent variables that were assessed in this study were
global areas of functioning. As stated previously, we did not
evaluate which specific subcomponents of each variable
(e.g., depression and traumatic stress) accounted for our
statistical results. Given the complex nature of suicidality
among clients with schizophrenia, we recommend that in-
vestigators and clinicians use a biopsychosocial approach
when conducting research or implementing interventions.
We agree with Harkavy-Friedman and Nelson (1997) that
ongoing assessment and clinical interventions must be used
with clients with schizophrenia to reduce the likelihood of
suicidality. In addition, further empirical research in this
complex area is clearly warranted.
REFERENCES
Amador, X. F., Friedman, J. H., Kasapis, C., Yale, S. A., Flaum, M., & Gorman,
J. M. (1996). Suicidal behavior in schizophrenia and its relationship to
awareness of illness. American Journal of Psychiatry, 153, 1185–1188.
American Psychiatric Association. (1994). Diagnostic and statistical
manual of mental disorders (4th ed.). Washington, DC: Author.
American Psychiatric Association. (1997). Practice guidelines for the
treatment of patients with schizophrenia. American Journal of Psy-
chiatry, 154(Suppl.), 1–63.
Black, D. W., & Fischer, R. (1992). Mortality in DSM-III-R schizophre-
nia. Schizophrenia Research, 7, 109–116.
Black, D. W., Warrack, G., & Winokur, G. (1985). The Iowa record-linkage
study, I–III. Archives of General Psychiatry, 42, 71–88.
Breier, A., & Astrachan, B. (1984). Characterization of schizophrenic patients
who commit suicide. American Journal of Psychiatry, 141, 206–209.
Caldwell, C. B., & Gottesman, I. I. (1990). Schizophrenics kill themselves too:
A review of risk factors for suicide. Schizophrenia Bulletin, 16, 571–589.
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159.
Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation analy-
sis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.
Cohen, J., Test, M. A., & Brown, R. L. (1990). Suicide and schizophrenia:
Data from a prospective community treatment study. American Journal
of Psychiatry, 147, 602–607.
Cohen, S., Lavelle, J., Rich, C. L., & Bromet, E. (1994). Rates and correlates
of suicide attempts in first-admission psychotic patients. Acta Psychiatrica
Scandinavica, 90, 167–171.
Drake, R. E., & Cotton, P. G. (1986). Depression, hopelessness and suicide
in chronic schizophrenia. British Journal of Psychiatry, 148, 554–559.
Drake, R. E., Gates, C., Whitaker, A., & Cotton, P. G. (1985). Suicide among
schizophrenics: A review. Comprehensive Psychiatry, 26, 90–100.
Fenton, W. S., McGlashan, T. H., Victor, B. J., & Blyler, C. R. (1997).
Symptoms, subtype, and suicidality in patients with schizophrenia
spectrum disorders. American Journal of Psychiatry, 154, 199–204.
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. (1995). Struc-
tured Clinical Interview for Axis I DSM-IV disorders. New York: State
Psychiatric Institute, Biometrics Research Department.
Foster, V. A., & McAdams, C. R. (1999). The impact of client suicide in
counselor training: Implications for counselor education and supervi-
sion. Counselor Education and Supervision, 39, 22–33.
Gallagher, A. G., Dinan, T. G., Sheehy, N. P., & Baker, L. J. V. (1995).
Chronic auditory hallucinations and suicide risk factors in schizophre-
nia. Irish Journal of Psychology, 16, 346–355.
Gilliland, B. E., & James, R. K. (1997). Crisis intervention strategies (3rd
ed.). Pacific Grove, CA: Brooks/Cole.
Gupta, S., Black, D. W., Arndt, S., Hubbard, W. C., & Andreasen, N. C.
(1998). Factors associated with suicide attempts among patients with
schizophrenia. Psychiatric Services, 49, 1353–1355.
Harkavy-Friedman, J. M., & Nelson, E. A. (1997). Assessment and interven-
tion for the suicidal patient with schizophrenia. Psychiatric Quarterly, 68,
361–375.
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Heila, H., Isometsa, E. T., Henriksson, M. M., & Heikkinen, M. E.
(1997). Suicide in schizophrenia: A nationwide psychological au-
topsy study on age and sex-specific clinical characteristics of 92
suicide victims with schizophrenia. American Journal of Psychiatry,
154, 1235–1242.
Jones, S. H., Thornicroft, G., Coffey, M., & Dunn, G. (1995). A brief men-
tal health outcome scale: Reliability and validity of the Global Assess-
ment of Functioning. British Journal of Psychiatry, 166, 654–659.
Kaplan, H. I., & Sadock, B. J. (1995). Comprehensive textbook of psychia-
try (6th ed.). Baltimore: Williams & Wilkins.
Kaplan, K. J., & Harrow, M. (1996). Positive and negative symptoms as
risk factors for late suicidal activity in schizophrenics versus
depressives. Suicide and Life-Threatening Behavior, 26, 105–121.
Kaplan, K. J., & Harrow, M. (1999). Psychosis and functioning as risk
factors for later suicidal activity among schizophrenia and
schizoaffective patients: A disease-based interactive model. Suicide
and Life-Threatening Behavior, 29, 10–24.
King, E. (1994). Suicide in the mentally ill: An epidemiological sample and
implications for clinicians. British Journal of Psychiatry, 165, 658–663.
Landmark, J., Cernovsky, Z. Z., & Mersky, H. (1987). Correlates of sui-
cide attempts and ideation in schizophrenia. British Journal of Psy-
chiatry, 151, 18–20.
Langley, G., & Bayatti, N. (1984). Suicide in Exe Vale Hospital. British
Journal of Psychiatry, 145, 463–467.
Milch, W. E. (1990). The change of symptomatology in hospitalized sui-
cidal patients. Crisis, 11, 44–51.
Miles, C. P. (1977). Conditions predisposing to suicide: A review. Jour-
nal of Nervous and Mental Disease, 164, 231–246.
Piersma, H. L., & Boes, J. L. (1995). Agreement between patient self-
report and clinician rating: Concurrence between the BSI and the
GAF among psychiatric inpatients. Journal of Clinical Psychology, 51,
153–157.
Rossau, C. D., & Mortensen, P. B. (1997). Risk factors for suicide in
patients with schizophrenia: Nested case-control study. British Jour-
nal of Psychiatry, 171, 355–359.
Roy, A. (1982). Suicide in chronic schizophrenia. British Journal of Psy-
chiatry, 141, 171–177.
Roy, A. (1995). Psychiatric emergencies. In H. I. Kaplan & B. J. Sadock
(Eds.), Comprehensive textbook of psychiatry (6th ed., pp. 1739–1752).
Baltimore: Williams & Wilkins.
Roy, A., Mazonson, A., & Pickar, D. (1984). Attempted suicide in chronic
schizophrenia. British Journal of Psychiatry, 144, 303–306.
Schwartz, R. C. (1999). Reliability and validity of the Functional As-
sessment Rating Scale. Psychological Reports, 84, 389–391.
Schwartz, R. C., & Petersen, S. (1999). The relationship between insight
and suicidality among patients with schizophrenia. Journal of Nervous
and Mental Disease, 187, 376–378.
Sullivan, G., Young, A. S., & Morgenstern, H. (1997). Behaviors as risk
factors for rehospitalizations: Implications for predicting and prevent-
ing admissions among the seriously mentally ill. Social Psychiatry and
Psychiatric Epidemiology, 32, 185–190.
Taiminen, T. (1994). Depression among schizophrenic patients: A selec-
tive review. Psychiatrica Fennica, 25, 185–194.
Ward, J. C., & Dow, M. G. (1994). The Functional Assessment Rating
Scale (Florida Mental Health Institute). Tampa: University of South
Florida, Department of Community Mental Health.
Ward, J. C., Dow, M. G., Saunders, T., Penner, K., Halls, S., & Sachs-
Ericsson, N. (1995). Structured Clinical Interview for the Functional
Assessment Rating Scale (Florida Mental Health Institute). Tampa:
University of South Florida, Department of Community Mental
Health.
Weiden, P., & Roy, A. (1992). General versus specific risk factors for
suicide in schizophrenia. In D. Jacob (Ed.), Suicide and clinical prac-
tice (pp. 75–100). Washington, DC: American Psychiatric Press.
World Health Organization. (1995). Schizophrenia: Epidemiology of men-
tal disorders and psychosocial problems. Geneva, Switzerland: Author.
Zimmerman, M. (1994). Interview guide for evaluating DSM-IV psychiat-
ric disorders and the mental status examination. Philadelphia: Psych
Press Products.
319
Helping Seventh Graders Be Safe and Successful:
A Statewide Study of the Impact of Comprehensive
Guidance and Counseling Programs

Richard T. Lapan, Norman C. Gysbers, and Gregory F. Petroski

The researchers examined the relationships between statewide implementation of comprehensive guidance and counseling
programs and indicators of safety and success for seventh graders. Hierarchical linear modeling was used to analyze data from
22,601 seventh graders attending 184 Missouri schools and 4,868 middle school teachers. After researchers controlled for differ-
ences between schools due to socioeconomic status and enrollment size, students attending middle schools with more fully
implemented comprehensive programs reported (a) feeling safer attending their schools, (b) having better relationships with their
teachers, (c) believing that their education was more relevant and important to their futures, (d) being more satisfied with the
quality of education available to them in their schools, (e) having fewer problems related to the physical and interpersonal milieu
in their schools, and (f) earning higher grades.

S
chools and school counselors in the United States
today face a daunting array of challenges gener-
ated from profound social, cultural, and economic
changes that occurred in the last half of the twen-
tieth century and that will continue into the
twenty-first century. The unprecedented number of vio-
lent acts in public schools helped motivate the Congress of
the United States to consider legislation that would pro-
vide schools with money to hire more elementary school
counselors. Keeping students safe has become a primary
focus for everyone involved in America’s schools. School
counselors have an important role in promoting and main-
taining student safety.
At a point in American history when schools are searching
for effective ways to respond to issues of violence, drugs,
dropouts, teenage pregnancy, adolescent suicide and depres-
sion, the quality of life (QOL) available to students in schools
has become a critical educational issue. In the rehabilitation
counseling literature, enhanced QOL has been linked to
expanded opportunities individuals have to promote
personal growth, fulfillment, and self-esteem (Kosciulek,
1999; Pain, Dunn, Anderson, Darrah, & Kratochvil, 1998).
QOL has been defined as an overall assessment, based on
both a subjective and objective evaluation, of one’s emo-
tional, physical, social, and material well-being (Felce & Perry,
1996). The subjective evaluation of students’ QOL centers
on their perceived personal satisfaction with the quality of
education available in their schools. The objective evalua-
tion of students’ QOL focuses on their personal assessment
of the physical and the interpersonal contexts that shape
learning opportunities and social interactions within their
schools. To help students satisfy their individual needs, in-
crease the likelihood that they will experience some con-
trol over their school environments, be successful academi-
cally, and have real opportunities to make choices in their
current and future learning, schools and school counselors
must find ways to improve the QOL available to students.
In addition, public schools are charged with the task of
promoting the development of future generations of citizens
who can constructively participate in a democratic society
that is competing in a constantly changing, increasingly
technologically sophisticated global marketplace. Federal
legislation (e.g., the School-to-Work Opportunities Act
of 1994) and reports (e.g., Secretary’s Commission on
Achieving Necessary Skills, 1991) have reflected wide-
spread concern regarding the preparation needed today
by young people to attain real economic opportunity. Full
and equitable membership in such an economic future
requires sophisticated and advanced academic preparation,
particularly in the areas of mathematics, science, and tech-
nology. The stakes are high, both for individuals and the
nation as a whole.

Richard T. Lapan is an associate professor, Norman C. Gysbers is a full professor, and Gregory F. Petroski is a doctoral student, all in the Department of
Educational and Counseling Psychology at the University of Missouri at Columbia. This research was supported in part by a grant from the Missouri
Department of Elementary and Secondary Education. Special thanks are extended to Missouri’s School-to-Careers State Management Team. In particular, the
authors thank Doug Sutton, Pam Spires, Robert Ruble, Donna Schulte, and Michelle Corcoran for their continued support, encouragement, and feedback.
Correspondence regarding this article should be sent to Richard T. Lapan, 16 Hill Hall, University of Missouri at Columbia, Columbia, MO 65211 (e-mail:
LapanR@missouri.edu).

320 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79


Finding sustainable and systematic ways to motivate stu-
dents to achieve academic success continues to be a major
task facing school practitioners, educational researchers, and
policy makers. In the educational psychology literature,
research is accumulating that highlights the important role
of the relationships between teachers and students in mo-
tivating academic achievement and social competencies for
elementary and middle school students (Wentzel, 1999).
Teacher–student relationships that are characterized by
warmth, an absence of conflict, and open communication
promote better school adjustments. Students benefiting from
such relationships with teachers display a stronger sense
of community in the school, more competent social be-
haviors, and improved academic performance (Schaps,
Battistich, & Solomon, 1997). In the counseling literature,
academic gains have been found when curriculum activities
forge relevant, interesting connections to possible educational
and career futures for students (e.g., Baker & Taylor, 1998;
Evans & Burck, 1992; Hoyt, 1998; Lapan, Gysbers, Hughey,
& Arni, 1993). Although they found significant gains across
the K–12 years, Evans and Burck (1992) argued that infu-
sion of career education guidance activities into the
academic curriculum led to even greater gains for pre–
high school students.
SCHOOL COUNSELOR IMPACT ON STUDENT DEVELOPMENT
There is a growing body of empirical research highlighting
the positive impact school counselors have on overall student
development K–12, including academic development
(e.g., Borders & Drury, 1992; Gerler, 1992; Macdonald &
Sink, 1999; Paisley & Borders, 1995; Whiston & Sexton, 1998).
Evidence for this has emerged from research conducted both
inside the field of school counseling and from sources out-
side of the discipline. For example, Lapan, Gysbers, and Sun
(1997) found in a statewide sample that high school stu-
dents who attended schools where more fully implemented
comprehensive guidance and counseling programs were in
place reported more positive school experiences. Across 236
small, medium, and large size high schools, surveys of a total
of 22,964 students found positive relationships between
guidance program implementation and (a) higher self-
reported grades, (b) student perceptions that their high
school education was better at preparing them for the
future, and (c) student perceptions that their schools were
making more career and college information available. In
addition, positive associations were found between guid-
ance program implementation and indicators of school climate
that assessed QOL (e.g., student feelings of belonging and
safety in school, perceptions that classes were less often
disrupted, and that peers were better behaved). Similarly,
Nelson and Gardner (1998) found that students in Utah
high schools with more fully implemented comprehen-
sive guidance and counseling programs rated their overall
education as better, took more advanced mathematics and
science courses, and had higher scores on the ACT College
Entrance Examination.
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Helping Seventh Graders Be Safe and Successful
In a survey of Missouri school counselors K–12, Gysbers,
Lapan, and Blair (1999) found that school counselors who
were more fully implementing their state’s comprehensive
guidance and counseling program model reported substantial
increases over time in the delivery of the kinds of guidance
and counseling services suggested in recent national legis-
lative initiatives. These activities included (a) having school
counselors conduct more structured group and classroom ac-
tivities; (b) spending more time with students; and (c) being
more extensively used by teachers, administrators, and
parents as resources and consultants. School counselors reported
(a) becoming more active in the school and local community,
(b) spending less time in their offices and more time deliv-
ering developmental counseling services, and (c) spending
less time on clerical and administrative tasks.
Support for the positive and preferred roles that school
counselors should, and often do, perform is also to be found
in studies conducted independent of research on comprehen-
sive guidance and counseling programs. For example, Schneider
and Stevenson (1999) published a comprehensive study of
the dramatic rise in educational and occupational ambitions
for adolescents across the United States. Relying on the
extensive national data set provided by the Alfred P. Sloan
Study of Youth and Social Development, these authors
reported that high schools that did not encourage and
emphasize student planning for the future were more likely
to have students with misaligned ambitions. Misaligned
ambitions were defined as a significant mismatch between
the amount of education an adolescent expects to attain
and the education required for those occupations the adoles-
cent wants to enter.
In conducting in-depth analyses of high schools that more
effectively help adolescents develop aligned ambitions,
Schneider and Stevenson (1999) pointed to the important
role of school counselors. The most effective high school in
their study had a school counselor to student ratio of 225 to
1. School counselors were expected to work closely with
both students and their parents to (a) help develop and
implement educational and career plans, (b) focus on stu-
dent behavior, and (c) deal with personal problems that
students might come to experience during their high school
years. Schneider and Stevenson concluded that this emphasis
on planfulness by school counselors greatly assisted stu-
dents to develop a sense of personal agency and encour-
aged students to take responsibility for both their educa-
tion and decisions they made about their future.
Blustein, Phillips, Jobin-Davis, Finkelberg, and Roarke
(1997) conducted a theory-building investigation of fac-
tors that promote more successful school-to-work transi-
tions for individuals who enter the workforce directly after
leaving high school. Individuals who had more positive ex-
periences with high school counselors expressed much
greater levels of satisfaction with their current jobs than
individuals who had more negative experiences with coun-
selors. Blustein et al. reported that more satisfied workers
indicated that their school counselors had been proactive
in reaching out to them and had provided both emotional
321
Lapan, Gysbers, and Petroski
and instrumental support for them in high school. In con-
trast, individuals who were more dissatisfied with their jobs
thought that school counselors had too many students to
take care of and did not provide them with the kind of indi-
vidualized attention that could have been most helpful.
MISSOURI COMPREHENSIVE GUIDANCE PROGRAM (MCGP)
In 1971, the University of Missouri at Columbia was awarded
a U.S. Department of Education grant to assist all states with
the development and implementation of career guidance,
counseling, and placement programs in their schools. On the
basis of the model developed in this project, Missouri school
counselors, state guidance officials from the Missouri
Department of Elementary and Secondary Education (1999),
the leaders of the Missouri School Counselor Association,
and university faculty who train school counselors in
Missouri have worked together to develop and implement
the MCGP (see Gysbers et al., 1999, as well as Lapan et al.,
1997, for more extensive reviews of this history). The MCGP
has three structural elements that integrate (a) program
content, (b) an organizational plan and delivery system for
guidance and counseling services, and (c) the political and
human resources needed to successfully carry out the pro-
gram. School counselor time is differentially apportioned across
the K–12 years to implement the four program components
of the MCGP: (a) Guidance Curriculum, (b) Individual
Planning, (c) Responsive Services, and (d) System Support
(Gysbers, Starr, & Magnuson, 1998).
Starting in the 1984–85 school year and ending during
the 1997–98 school year, school counselors and adminis-
trators from 441 of the 525 school districts across Missouri
received extensive training in the development and imple-
mentation of the MCGP in their home school districts. The
MCGP has become the state mandated training model for
school counselors. The extent of its implementation is one
of the criteria on which all school districts are evaluated in
Missouri’s mandatory school district accreditation process,
that is, the Missouri School Improvement Program (MSIP).
Because of this statewide attempt to implement a com-
mon program model, Missouri provides a unique oppor-
tunity to study the effects of comprehensive guidance and
counseling programs on student development. In addition
to Missouri, this training model has been adopted, in whole
or in part, in many states and school districts across the
country. Results from statewide studies in Missouri have
clear relevance and applicability for school counselors, state
officials, policy makers, and university faculty in other
states who are engaged in the difficult process of devel-
oping and implementing comprehensive guidance and
counseling programs.
PURPOSE OF THE STUDY
The purpose of the present study was to examine, on a
statewide basis, the impact of more fully implemented com-
prehensive guidance and counseling programs on student
322 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
(a) perceptions of safety in school, (b) satisfaction with their
education, (c) grades, (d) perceptions of their relationships
with teachers, and (e) perceptions of the importance and
relevance of education to their future. Comprehensive guid-
ance and counseling programs have been conceptualized and
designed to positively affect each of these factors. This study
examined whether or not such relationships exist for seventh-
grade students because the seventh grade is a critical time in
individual development and in the transition process from
late elementary to middle and high school years.
Missouri’s database, generated from the statewide school
district accreditation process, was used to explore relation-
ships between teacher ratings of the extent to which the
MCGP was being implemented in their schools and the
student factors described above. One limitation of some of
the earlier research on the impact of comprehensive guid-
ance and counseling programs was a reliance on school coun-
selor self-report data (Lapan et al., 1997). It was predicted
in the present study that findings from earlier research with
high school students that used counselor self-report data
would be replicated at the seventh-grade level, using teach-
ers as observers of the level of implementation of recom-
mended MCGP guidance activities in their schools.
RESEARCH QUESTIONS
The following research questions were examined in this
study:
1. Are student indicators of safety and success posi-
tively related to student characteristics (i.e., sex,
minority status, objective assessment of the quality
of life in one’s school, and amount of time spent on
homework)?
2. After controlling for differences between schools due
to enrollment size and socioeconomic level, are more
fully implemented comprehensive guidance and
counseling programs related to each of the five indi-
cators of student safety and success?
3. Do school enrollment size, socioeconomic level, and
level of implementation of the school counseling pro-
gram significantly affect the regression slopes be-
tween student characteristics and student indicators
of safety and success?
METHOD
Sample
MSIP seventh-grade student and middle school teacher survey
data collected from 1992 to 1996 were analyzed in this study.
For each of these years, MSIP collected stratified random samples
representative of all schools in Missouri. The data set analyzed
in the present study included 22,601 seventh graders and 4,868
teachers from 184 schools. Approximately 50% of the
students were girls (n = 11,344) and 16% of the total sample
were minority students (n = 3,682). Approximately 69%

of these middle school teachers were women (94% were
Caucasian American and 5% were African American). Ap-
proximately 52% of these teachers had bachelor’s degrees,
with 46% having master’s degrees.
The sample was representative of the diversity of schools
in Missouri. The state classified the locations of the 184
schools in the following ways: approximately 12% were des-
ignated as large cities, 8% as midsize cities, 32% as suburban
areas, 27% as towns, and 22% as rural locales. The number of
schools sampled each year was as follows: 33 in 1992, 33 in
1993, 41 in 1994, 42 in 1995, and 35 in 1996.
Missouri’s School Improvement Program (MSIP)
The MSIP (Missouri Department of Elementary and Sec-
ondary Education, 1999) is an extensive and mandated
statewide school district evaluation and accreditation pro-
cess. Each year an MSIP external accreditation team evalu-
ates a statewide representative sample of school districts,
assuring that all districts in the state are assessed on a 5-year
cycle. Quality standards and indicators for all educational
programs, including guidance, have been developed and
implemented statewide. Before an MSIP accreditation team
visit, school personnel, school board members, parents, and
students complete surveys. Of particular interest for the
present study, classroom teachers are asked several behav-
iorally specific questions about the extent to which the
counseling activities of the MCGP are being implemented
in their schools. Teachers are in a key observer role because
they have direct knowledge of the extent and kinds of coun-
selor involvement in their classrooms and with their stu-
dents. Questions answered by teachers were taken directly
from the four program components of the MCGP.
All students in a school district that is being evaluated by
MSIP are expected to complete a questionnaire assessing
critical aspects of their school experiences. The seventh-
grade survey covers a wide range of factors critical to student
perceptions of safety and success in school. Data gathered from
MSIP student questionnaires have become a valuable source
of insight into student attitudes and behaviors across Missouri
(e.g., see Sun, Hobbs, & Elder, 1994). In the present study,
principal investigators identified survey items that assessed
critical aspects of student safety and school success. All
selected items have clear face validity with the issues of
student safety and success central to the present study. In the
following sections, all items analyzed in the present study are
listed with their corresponding student-level or school-level
variables.
Student-Level Variables
Sex and minority status. A dichotomous variable was created
to distinguish between girls and boys (girls = 0, boys = 1). A
dichotomous variable was created to distinguish between
majority Caucasian American students and minority students
(i.e., African Americans, Hispanic Americans, and Native
Americans; majority = 0, minority = 1). The percentage of
Asian American students in the sample was small and was
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Helping Seventh Graders Be Safe and Successful
not included in the analyses. The present study focused on
student attitudes of those minority groups at greater risk for
school failure and lower academic achievement.
Safety and relationships with teachers. Students were asked
to rate on a 5-point scale (1 = strongly disagree to 5 = strongly
agree) the item “I feel safe at school” (M = 3.42, SD = 1.21).
On the same 5-point scale, students were asked to rate the
following four items referring to their relationships with
teachers: (a) “There is good communication between teach-
ers and students,” (b) “If a student has a problem there are
teachers who will listen and help,” (c) “In my school every-
one is given a chance to succeed in class,” and (d) “Students
and teachers often work together on projects in class.” A prin-
cipal component factor analysis was performed on these four
items. Only one factor was extracted because it was the only
factor to have an eigenvalue greater than 1.0. Loadings of
the four items on this factor ranged from a minimum of .68
to a maximum of .78. The estimate of internal consistency
across these four items was adequate for research purposes
(coefficient alpha = .74). Therefore, a summed scale score
across all four items was calculated for each student and used
in all subsequent analyses (M = 10.13, SD = 3.23).
Relevance of education to one’s future. Two items (using the
same 5-point Likert scale described earlier) were combined
to measure this variable: (a) “The kind of education I’m get-
ting here will help me later on” and (b) “What I study in
school seems important to me.” Summative scores across both
items were calculated and used in all subsequent analyses to
measure the perceived importance of one’s education and
its connection to one’s future (M = 7.82, SD = 1.60).
Subjective estimate of QOL. One item asked students
to make a global assessment of the quality of education
available to them in their school. A 5-point scale was
used to measure these student perceptions (1 = poor to 5 =
excellent; M = 3.77, SD = .93).
Objective estimate of QOL. Four items asked students to
rate the environmental conditions and interpersonal milieu
in their schools on a 5-point scale (1 = strongly disagree to
5 = strongly agree). These items were (a) “It is often too hot
or too cold in this building,” (b) “There is often so much
noise that it is hard to study,” (c) “Our classes are often
interrupted,” and (d) “At least half the time in school is
wasted.” A principal component factor analysis was per-
formed on these four items. Only one factor was extracted
because it was the only factor to have an eigenvalue greater
than 1.0. Loadings of the four items on this factor ranged
from a minimum of .60 to a maximum of .74. The estimate
of internal consistency across these four items was adequate
for research purposes (coefficient alpha = .59). Therefore,
summed scale scores were calculated for each student and
used in all subsequent analyses (M = 8.09, SD = 3.16).
Grades. Students were asked to pick one of eight categories
that best described their grades in school. An 8-point scale
was created to assign values to student self-reported grades.
These values were as follows: mostly A = 8, half A and B = 7,
mostly B = 6, half B and C = 5, mostly C = 4, half C and
D = 3, mostly D = 2, and below D = 1 (M = 6.09, SD = 1.77).
323
Lapan, Gysbers, and Petroski

Homework. One item asked students to rate how much time was adequate for research purposes (coefficient alpha = .95).
they usually spent on homework each day. A 6-point scale was A summed scale score across these seven items was calcu-
constructed to assign values to each of six possible student lated for each teacher (M = 26.32, SD = 3.55). To compute a
responses. These values ranged from “I don’t usually have it school-level variable assessing the extent to which the guid-
assigned” = 1 to “More than 2 hours” = 2 (M = 2.63, SD = 1.16). ance program was being implemented in that school, each
individual teacher’s summed score was then averaged across
School-Level Variables all middle school teachers in that school who responded to
the MSIP survey. The mean number of middle school teach-
Enrollment. MSIP provides information on the actual num- ers completing the survey in each school was 25 (SD = 15).
ber of students attending each school that was sampled.
In the present study, mean student enrollment was 552 Data Analysis
seventh-grade students (SD = 262) per school. The mini-
mum number of students was 97 and the maximum was The three research questions addressed in this study were
1,245. The total number of students attending each middle tested using hierarchical linear modeling (HLM). Bryk and
school was used as a covariate to control for effects due to Raudenbush (1992) developed HLM to address the nesting
school enrollment size. problem in which research participants are simultaneously
Socioeconomic status (SES) level. MSIP also provides data embedded within several social layers. For example, students
on the percentage of students in each school who receive come from various family backgrounds and participate in
free and reduced lunch. In the present study, the percent- classrooms and school districts that are very different from
age of students receiving free and reduced lunch varied be- each other. As illustrated in Figure 1, HLM uses multiple
tween a minimum of approximately 4% to a maximum of linear regression to predict outcomes for group members
almost 98%. The percentage of students receiving free and that are a function of both the characteristics of the group to
reduced lunch in each school was used as a covariate to which they belong and the characteristics of the individuals
control for effects due to varying SES levels between schools. themselves (Arnold, 1992). Starting at the individual
By using school enrollment size and SES in this way, the
present study was able to statistically control for two of the
most important factors that discriminate between urban,
Characteristics of schools Between-school models:
suburban, and rural schools. By first removing the effects Covariates: Student attitudes are due
of enrollment and SES on student attitudes and school ex- Enrollment size to factors that vary
periences, we were able to study the possible impact of Socioeconomic level between schools.
Independent Predictor: (Research Question 2)
more fully implemented guidance programs on student
Teacher ratings of the
safety and success independent of these potentially con- extent to which the 4
founding factors. program components of ➤
Guidance. MSIP requires classroom teachers to rate seven the MCGP are being
implemented in their
items assessing the extent to which counseling activities school.
Student indicators of safety
and success
outlined in the four program components of the MCGP Safety
have been carried out in their classrooms and schools. These Relationship with teachers
Between-school models:
items are (a) “The guidance program provides classroom Regression slopes Relevance of education to
activities for students on a regular basis,” (b) “Counselors between student one’s future
Satisfaction (Subj. QOL)
work with the students to assist them with their personal characteristics and
Grades
student attitudes are
concerns,” (c) “Counselors assist students with their educa- influenced by factors that
tional and career plans,” (d) “Counselors consult with staff vary between schools.
➤

and parents concerning students’ personal and academic (Research Question 3)

progress,” (e) “Counseling is provided for individuals and
small groups,” (f) “Students are referred to other professional Characteristics of
students Within-student models:
resources in the school and community as appropriate,” and
➤

Sex Student attitudes are due

(g) “Counselors inform staff and community about the aims Minority status to characteristics of
and purposes of the guidance program.” Obj. QOL students that vary within
Time spent on schools.
A 5-point scale was provided to teachers to rate the extent (Research Question 1)
homework
to which each of these activities was actually being carried
out in their schools (1 = strongly disagree/never to 5 = strongly
agree/always). A principal component factor analysis was FIGURE 1
performed on these seven items. Only one factor was Relating Student Outcomes to Between-School
extracted because it was the only factor to have an eigen- Differences in the Implementation of Comprehensive
value greater than 1.0. Loadings of the seven items on this School Counseling Programs
factor ranged from a minimum of .69 to a maximum of .80.
The estimate of internal consistency across these seven items Note. See Table 1 Note.

324 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79

 Helping Seventh Graders Be Safe and Successful

student level, HLM calculates regression equations for out-
comes that are a function of characteristics of students
within each school. For each of the 184 schools participat-
ing in the present study, separate regression equations were
calculated for the five indicators of student safety and suc-
cess (i.e., perceived safety, relationships with teachers, rel-
evance of education to one’s future, satisfaction with edu-
cation available to one, and self-reported grades).
Each of these indicators was predicted to be a function
of four student characteristic variables (i.e., sex, minority
status, objective assessment of quality of life in one’s school
building, and time spent on homework). These equations
produce an intercept value and beta weight for each of
the five indicators of student success and safety for every
school in this study. The intercept is the average value for
each of the four indicators in every school. The beta
weights estimate the association between student indica-
tors and student characteristics in each school. HLM tests
these relationships for statistical significance and reports
the results as within-school unconditional models (Research
Question 1). Unconditional models summarize student-
level regression results across all schools. They are defined
as unconditional models because the potential effects of
between-school predictors on these values have not been
taken into account.
Second, HLM uses between-school factors to predict the
variability in these intercepts and beta weights across
schools. This second set of regression equations attempts to
explain these distributions of scores as a function of factors
that vary between schools. Three school-level predictors
(i.e., school enrollment size, socioeconomic level, and level
of implementation of the school counseling program) were
hypothesized to significantly affect each of the five indica-
tors of student success and safety (Research Question 2).
In addition, each of the school-level predictors was hypoth-
esized to have a significant influence on the relationships
(i.e., beta weights) between student characteristics and stu-
dent indicators of safety and success (Research Question
3). HLM computes coefficients and standard t statistics to
test each relationship for statistical significance (Bryk &
Raudenbush, 1992).
Before performing the HLM analyses, scatter plots of the
dependent variables in relation to the ordinal independent
variables were examined. No evidence was found to suggest
curvilinear relationships between variables. In addition, plots
of residuals did not indicate the existence of polynomial
relationships. Only statistically significant HLM findings are
reported in the Results section. Additional data are avail-
able from the principal author on request.
RESULTS
Correlations Between Student Characteristics and
Indicators of Safety and Success
Correlations between student characteristics and student
indicators of safety and success are reported in Table 1.
Minimal gender and minority differences were found. How-
ever, seventh-grade girls reported earning higher grades than
boys did, and nonminority students indicated that they were
earning higher grades than minority students. For all stu-
dents, a better relationship with teachers was strongly re-
lated to (a) feeling safer in school, (b) being more satisfied
with their education and feeling that it was more relevant
and important to their future, (c) earning higher grades and
doing more homework, and (d) attending a school where
there were fewer problems with the interpersonal and physi-
cal environment. Believing that one was safer in school was
related to (a) earning higher grades, (b) believing that one’s
education was more relevant to one’s future, and (c) being
more satisfied with both the subjective and the objective
components assessing the QOL available in one’s school. The
objective assessment of the QOL in one’s school was related
to all student-level variables except sex and minority status.

TABLE 1

Intercorrelations Among Student-Level Variables

Variable 1 2 3 4 5 6 7 8 9

1. Safety — .46 .38 .38 –.33 .19 –.09 –.08 .09

2. Teachers — .56 .51 –.33 .14 –.03 –.05 .12
3. Relevance — .49 –.27 .22 .03 –.09 .18
4. Subj. QOL — –.32 .24 –.05 –.05 .12
5. Obj. QOL — –.15 .07 .04 –.09
6. Grades — –.18 –.19 .09
7. Minority status — –.01 –.01
8. Sex — –.11
9. Homework —

Note. For students: N = 22,601. “Teachers” refers to students’ perceived relationship with teachers. “Relevance” refers to students’ perceptions
of the relevance and importance of their education to their future. “Subj. QOL” refers to students’ subjective evaluation of their level of satisfaction
with the quality of education available to them in their schools. “Obj. QOL” refers to students’ objective evaluation of the physical and interpersonal
milieu in their schools.

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 325

Lapan, Gysbers, and Petroski

Correlations Between School-Level Variables TABLE 3

Correlations between school-level variables are reported in
Table 2. These coefficients were calculated between mean Final Estimation of Significant Hierarchical Linear
school-level values for variables used in the present study Modeling Fixed Effects
(n = 184). More complete implementation of a comprehen-
sive guidance and counseling program was consistently and Fixed Effect Coefficient SE T ratio
positively associated with school-level indicators of student
safety and success. As zero order correlations, these coeffi- Dependent variable: Safety
cients suggest that level of school counseling program imple-
School-level predictor
mentation shared between 3% and 8% of its variance with SES –0.011 .001 –7.63***
these school-level indicators of student safety and success. Enrollment –0.018 .005 –3.62***
Lower school-level SES ratings were associated with Guidance 0.023 .007 3.23**
Student-level predictor
school-level means that indicated students (a) did not feel Obj. QOL slope intercept –0.130 .016 –8.44***
as safe in their schools, (b) were less satisfied with both the SES × Obj. QOL slope 0.001 .000 2.74*
subjective and the objective components of the QOL avail-
able to them in their schools, and (c) were earning lower Dependent variable: Relationship with teachers
grades. However, correlations between school-level means School-level predictor
suggested that students from poorer schools had equally Guidance 0.053 .020 2.61*
good relationships with teachers and believed that their Student-level predictor
education was just as relevant and important to their future as Homework slope intercept 0.202 .018 11.20***
Obj. QOL slope intercept –0.358 .050 –7.22***
did students from wealthier schools. Correlations between SES × Obj. QOL slope 0.002 .001 3.51**
school-level means also suggested that students from larger
schools were earning higher grades and feeling more satisfied Dependent variable: Satisfaction with the quality of education
with the education available to them than were students from in one’s school
smaller schools. School-level predictor
SES –0.005 .001 –4.03**
Enrollment 0.012 .004 2.76**
Research Question 1 Guidance 0.031 .006 5.14***
Statistically significant relationships between student char- Student-level predictor
Homework slope intercept 0.053 .006 8.55***
acteristics and indicators of student safety and success are Obj. QOL slope intercept –0.093 .013 –7.28***
reported in Table 3. Students who believed that their learn- SES × Obj. QOL slope 0.001 .000 3.23*
ing environment was not as encumbered by disruptions in
their interpersonal and physical milieu were more likely to Dependent variable: Relevance and importance of education
(a) feel safer (T = –8.44, p < .001), (b) have better relation- to one’s future
ships with teachers (T = –7.22, p < .001), (c) be more satis-
fied with their education (T = –7.28, p < .001), (d) see School-level predictor
Guidance 0.022 .008 2.63*
their education as being more relevant and important to Student-level predictor
their future (T = –5.87, p < .001), and (e) earn higher grades Homework slope intercept 0.183 .012 14.83***
(T = –8.00, p < .001). Students who reported doing more Obj. QOL slope intercept –0.138 .024 –5.87***
SES × Obj. QOL slope 0.001 .000 3.01*
homework indicated that they (a) had better relationships
Dependent variable: Grades
TABLE 2 School-level predictor
SES –0.011 .002 –6.12***
Correlations Between School-Level Variables Enrollment 0.014 .005 2.46*
Guidance 0.027 .008 3.27**
Student-level predictor
Variable Guidance Enrollment SES Sex slope intercept –0.450 .051 –8.90***
SES × Sex slope –0.006 .001 –4.11***
1. Safety .17* –.08 –.50** Minority slope intercept –1.040 .102 –10.23***
2. Teachers .18* .05 –.13 SES × Minority slope 0.010 .002 4.00***
3. Relevance .19* .10 .02 Homework slope intercept 0.077 .015 5.15***
4. Subj. QOL .29** .28** –.37** Obj. QOL slope intercept –0.132 .016 –8.00***
5. Obj. QOL –.18* .11 .32** SES × Obj. QOL slope 0.001 .000 4.05***
6. Grades .13 .28** –.62** Enrollment × Obj. QOL
slope 0.004 .001 4.06***
Note. See Table 1 Note. For schools, N = 184. “Guidance” refers to the
extent of implementation of the guidance program. “Enrollment” re-
fers to the number of students in each school. “SES” (socioeconomic Note . See Tables 1 & 2 Notes . Only statistically significant
status) refers to the socioeconomic level of each school. coefficients are reported in the table.
*p < .05. **p < .001. *p < .05. **p < .01. ***p < .001.

326 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79


with teachers (T = 11.20, p < .001), (b) were more satisfied
with their education and believed that their education
was more relevant and important to their future (T = 8.55,
p < .001), and (c) were earning higher grades (T = 5.15, p <
.001). Girls reported earning higher grades than boys did
(T = –8.90, p < .001), and minority students indicated they
were earning lower grades than majority students were
(T = –10.23, p < .001).
Research Question 2
Results from between-school HLM analyses testing relation-
ships between school-level predictors and student indicators
of safety and success are also reported in Table 3. After con-
trolling for between-school differences in SES and enrollment
size, more fully implemented school counseling programs sig-
nificantly predicted (a) student perceptions of being safer in
their schools (T = 3.23, p < .01), (b) better relationships
between students and teachers (T = 2.61, p < .05), (c) greater
satisfaction of students with the education they were receiv-
ing in their schools (T = 5.14, p < .001), (d) perceptions that
one’s education was more relevant and important to one’s
future (T = 2.63, p < .05), and (e) earning higher grades (T =
3.27, p < .01). Students from lower SES schools believed that
they (a) were not as safe in their schools (T = –7.63, p < .001),
(b) were not as satisfied with the quality of education avail-
able to them in their schools (T = –4.03, p < .01), and (c) were
earning lower grades (T = –6.12, p < .001). Students from
larger schools believed that they (a) were not as safe in their
schools (T = –3.62, p < .001), (b) were more satisfied with
the quality of education available to them in their schools
(T = 2.76, p < .01), and (c) were earning higher grades
(T = 2.46, p < .05). These findings mostly corroborate results
found in the zero order correlations between variables that
were reported in Table 2. However, the HLM procedure
found a significant relationship between school counseling
program implementation and student self-reported grades
in which the zero order correlation between these variables
was nonsignificant.
Research Question 3
Results from between-school HLM analyses testing whether
or not the between-school variables significantly affected
regression slopes between student characteristics and stu-
dent indicators of safety and success are also reported in
Table 3. Guidance program implementation did not sig-
nificantly influence any of these slope relationships. How-
ever, students in poorer schools were more negative in their
ratings of the objective QOL available to them in their
schools. These lower QOL ratings were associated with
student responses that indicated that these students were
(a) not feeling as safe in school (T = 2.74, p < .05), (b) hav-
ing poorer relationships with teachers (T = 3.51, p < .01),
(c) less satisfied (T = 3.23, p < .05), (d) understanding their
education to be less relevant to their future (T = 3.01, p <
.05), and (e) earning lower grades (T = 4.05, p < .001).
Responses from students attending larger schools suggested
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Helping Seventh Graders Be Safe and Successful
that there was a relationship between more disruptions in
school and lower grades (T = 4.06, p < .001).
DISCUSSION
Many writers, both within the United States and from the
international guidance community, have consistently high-
lighted the need for rigorous program evaluation research
studies that would inform school counseling program reform
and improvement efforts (e.g., Borders & Drury, 1992; Gerler,
1992; Gysbers, Hughey, Starr, & Lapan, 1992; Killeen,
Sammons, & Watts, 1999; Lapan & Kosciulek, 2001; Oomen,
Athanassoula-Reppa, Barnes, Petri, & Hautvast, 1997).
Findings from the present study have both replicated and
extended a growing body of evaluation research on guidance
program effectiveness and the work of school counselors
across the K–12 years (e.g., Blustein et al., 1997; Gerler,
1992; Gysbers et al., 1999; Lapan et al., 1993; Lapan et
al., 1997; R. S. Lee, 1993; Schneider & Stevenson, 1999;
St. Clair, 1989). These studies have consistently shown
that when school counselors are more fully engaged in
implementing preferred work tasks outlined in frameworks
for comprehensive guidance and counseling programs,
counselors move out of marginalized positions and into
roles that more effectively promote essential educational
and career objectives for students (Gysbers & Henderson,
2000; Watkins, 1994).
The present study provided additional evidence that the
implementation of comprehensive school counseling pro-
grams was consistently associated with important indica-
tors of student safety and success. School counselors who
were more fully engaged in providing students with a unique
network of emotional and instrumental support services were
more likely to exert a more positive impact on students than
counselors who did not implement such activities. This range
of school counseling activities included (a) spending more
time in classrooms, (b) assisting students with personal prob-
lems as well as educational and career plans, (c) consulting
with parents and school personnel, (d) providing individual
and group counseling services, (e) referring students as
needed, (f) and communicating to others both within the
school and local community about the goals and aims of the
guidance program. For both girls and boys, seventh graders
in schools where these guidance activities were more avail-
able reported having (a) better relationships with teachers,
(b) higher grades, (c) a belief that their education was more
important to them and relevant to their future, (d) and more
enhanced subjective and objective perceptions of the QOL
available to them in their schools.
Improving Student–Teacher Relationships
Middle school students’ motivation to pursue both social
goals (e.g., to behave appropriately in school) and task-related
goals (e.g., to learn and achieve academically) has been
strongly linked to the kinds of relationships that they
perceive they have with their teachers (Wentzel, 1999). Stu-
327
Lapan, Gysbers, and Petroski
dents who perceive their teachers to be caring and support-
ive, respectful, fair, holding positive expectations for them,
and having more open communication make greater aca-
demic gains and use more prosocial behaviors in classrooms
(Brophy, 1983; Eccles & Midgley, 1989; Feldlaufer, Midgley,
& Eccles, 1998; Murdock, 1999; Raudenbush, 1984;
Weinstein, 1989; Wentzel, 1997, 1998, 1999). Student per-
ceptions of a lack of equity and fairness in their relationships
with teachers have been found to be a major correlate of
dropping out of school (Fine, 1986; Wehlage & Rutter, 1986).
Findings from the present study corroborate the central im-
portance of student perceptions of their relationships with
teachers in promoting student safety and success.
Seventh graders who reported better relationships with
teachers were much more likely to indicate having more
positive experiences in school. These students felt safer in
school, reported earning higher grades, and believed that
their education was more important to them and relevant
to their future. Students strongly connected more positive
relationships with teachers to their subjective and objec-
tive evaluation of the QOL available to them in their school.
That is, they felt more satisfied with their education and
experienced their learning environment to be less encum-
bered by negative distractions (e.g., students misbehaving
and disrupting classes).
Students who attended schools where a more fully imple-
mented MCGP was in place consistently indicated having
better relationships with teachers. When trying to understand
the connection between more effective student–teacher
relationships and more positive achievement-related and
prosocial behaviors, Wentzel (1999) speculated that middle
school students may experience supportive relationships with
teachers as effective parenting practices that promote devel-
opmentally appropriate levels of autonomy, competence,
and control (Baumrind, 1971, 1991). Developmentally
appropriate levels of autonomy, competence, and control are
enhanced for students (a) where interactions with teachers
are respectful and democratic; (b) where clear performance
expectations are set based on a recognition of individual
differences; (c) and where communication is open, construc-
tive, and nurturing (Wentzel, 1999). These results suggest
that school counselors who are more fully implementing com-
prehensive guidance and counseling programs are more likely
to effectively engage larger numbers of students in activities
that would help to promote these kinds of relationships.
Offsetting SES
Seventh graders in economically poorer schools believed
that school was important and relevant to their future and
reported having nurturing and fair relationships with teach-
ers as frequently as did students from more financially
advantaged schools. However, these girls and boys, both
minority and majority students, did not feel as satisfied with
the QOL available in their schools. They indicated that more
time was wasted, more classes interrupted, too much noise
occurred that made it harder for them to study, and that
328 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
their school building was often too hot or too cold. These
students reported earning lower grades, doing less home-
work, and not feeling as safe in their schools.
Our results showed that, to some degree, a more fully
implemented comprehensive school counseling program
was of benefit to seventh graders regardless of the socio-
economic level of their school. More fully implemented
guidance and counseling programs were related to greater
student feelings of both safety and success in poorer as well
as in wealthier schools.
Limitations
Two potential limitations to this study need to be ad-
dressed. First, the present study relied on student and
teacher self-report data. More objective measures of stu-
dent achievement and behavior in school as well as trained
raters to assess the extent of counseling activities would
add to the results. However, previous researchers have
found strong connections between student ratings of these
kinds of items and more objective measures of student
performance and behavior in school (e.g., Wentzel, 1999).
Although teachers are not neutral observers, they do pro-
vide a critically important perspective on the delivery of
school counseling services.
In the present study, it is important to note that teachers
rated the occurrence of school counseling activities for
which they had direct behavioral evidence. They were not
asked to rate the quality of school counseling services, only
whether or not specific activities were occurring in their
classrooms and schools. Teachers are important partners in
comprehensive guidance and counseling programs and on
rating scales, such as the one used in the present study, a
valuable independent source of information on school coun-
seling program implementation.
Second, although the tests for school counseling pro-
gram implementation were statistically significant, the
effect sizes were small. However, it is important to note
that these relationships were found across 184 schools,
more than 22,000 seventh graders, and 4,868 middle school
teachers, and we used a measurement strategy that did
not exploit common method variance that could inflate
such correlations. Validity coefficients across a population,
as opposed to a sample, would be expected to be smaller
than ratings that are more similar to reliability estimates.
In the present study, teachers rated the extent to which
certain counseling activities were occurring in their schools
and classrooms. Students rated the quality of their school
experiences, not the counseling program specifically.
Teacher and student ratings were then aggregated within
each school and analyzed. These procedures made it very
difficult to find statistical significance and exploit chance
variations due to similar measures used and data collec-
tion procedures used. Teachers and students completed
their different forms at different times, under different
test administration procedures, with no preconceptions
that such an investigation would take place.

Implications
Given the daunting array of challenges that face our nation
and our schools, policy makers need to clearly understand that
school counselors will be better able to help solve current
problems that plague America’s schools if they are employed
not merely to fill a position but to fully implement a compre-
hensive guidance and counseling program in their schools
(Gysbers & Henderson, 2000). Comprehensive guidance and
counseling programs have provided school counselors K–12
with the organizational structure to focus efforts, organize work
schedules, and allocate time necessary for implementing
proactive school counseling activities and services that
promote critical aspects of student development. Continued
implementation of comprehensive programs through the work
of school counselors can deliver these critical services to the
entire student body. Unlike results found in earlier studies
(e.g., V. E. Lee & Ekstrom, 1987), implementation of compre-
hensive programs will provide equitable access to guidance
and counseling services for students from rural areas, economi-
cally disadvantaged homes, and minority families.
Results from the present study certainly do not suggest
that school counselors are the only, or even the primary,
means by which the United States can improve public
school education. However, results do suggest that school
counselors’ implementation of comprehensive guidance and
counseling programs can be a consistently positive piece of
the puzzle that needs to be assembled if we are to make a
genuine effort to help all of our nation’s children achieve
academically and attain their desired future. Full implemen-
tation of such school counseling programs is a sustainable,
cost-effective national strategy for assisting all students to
feel safer and be more successful in school.
REFERENCES
Arnold, C. L. (1992). An introduction to hierarchical linear models. Mea-
surement and Evaluation in Counseling and Development, 25, 58–90.
Baker, S. B., & Taylor, J. G. (1998). Effects of career education interven-
tions: A meta-analysis. The Career Development Quarterly, 46, 376–385.
Baumrind, D. (1971). Current patterns of parental authority. Develop-
mental Psychology Monograph, 4(1, Pt. 2).
Baumrind, D. (1991). Effective parenting during the early adolescent
transition. In P. A. Cowan & M. Hetherington (Eds.), Family transi-
tions (pp. 111–164). Hillsdale, NJ: Erlbaum.
Blustein, D. L., Phillips, S. D., Jobin-Davis, K., Finkelberg, S. L., & Roarke,
A. E. (1997). A theory-building investigation of the school-to-work
transition. The Counseling Psychologist, 25, 364–402.
Borders, L. D., & Drury, S. M. (1992). Comprehensive school counseling
programs: A review for policymakers and practitioners. Journal of
Counseling & Development, 70, 487–498.
Brophy, J. (1983). Research on the self-fulfilling prophecy and teacher
expectations. Journal of Educational Psychology, 75, 631–660.
Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models:
Applications and data analysis methods. Newbury Park, CA: Sage.
Eccles, J., & Midgley, C. (1989). Stage-environment fit: Developmentally
appropriate classrooms for young adolescents. In C. Ames & R. Ames
(Eds.), Research on motivation in education (Vol. 3, pp. 139–186). San
Diego, CA: Academic Press.
Evans, J. H., & Burck, H. D. (1992). The effects of career education
interventions on academic achievement: A meta-analysis. Journal of
Counseling & Development, 71, 63–68.
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Helping Seventh Graders Be Safe and Successful
Felce, D., & Perry, J. (1996). Exploring current conceptions of quality of
life: A model for people with and without disabilities. In R. Renwick,
I. Brown, & M. Nagler (Eds.), Quality of life in health promotion and
rehabilitation (pp. 51–62). Thousand Oaks, CA: Sage.
Feldlaufer, H., Midgley, C., & Eccles, J. S. (1998). Student, teacher, and
observer perceptions of the classroom before and after the transition
to junior high school. Journal of Early Adolescence, 8, 133–156.
Fine, M. (1986). Why urban adolescents drop into and out of public
high school. Teachers College Record, 87, 393–409.
Gerler, E. R. (1992). What we know about school counseling: A reaction to
Borders and Drury. Journal of Counseling & Development, 70, 499–501.
Gysbers, N. C., & Henderson, P. (2000). Developing and managing your
school guidance program (3rd ed.). Alexandria, VA: American Counsel-
ing Association.
Gysbers, N. C., Hughey, K., Starr, M., & Lapan, R. T. (1992). Improving
school guidance programs: A framework for program, personnel, and
results evaluation. Journal of Counseling & Development, 70, 565–570.
Gysbers, N. C., Lapan, R. T., & Blair, M. (1999). Closing in on statewide
implementation of a comprehensive guidance program model. Profes-
sional School Counseling, 2, 357–366.
Gysbers, N. C., Starr, M. F., & Magnuson, C. S. (1998). Missouri compre-
hensive guidance: A model for program development and implementation
(Rev. ed.). Jefferson City, MO: Missouri Department of Elementary
and Secondary Education.
Hoyt, K. B. (1998). Career education and the transition from schooling-
to-employment. Youth Policy, 15 & 16, 10–20.
Killeen, J., Sammons, P., & Watts, A. G. (1999). The effects of careers educa-
tion and guidance on attainment and associated behavior. Cambridge,
England: The National Institute for Careers Education and Counselling.
Kosciulek, J. F. (1999). The consumer-directed theory of empowerment.
Rehabilitation Counseling Bulletin, 42, 196–213.
Lapan, R. T., Gysbers, N. C., Hughey, K., & Arni, T. J. (1993). Evaluating
a guidance and language arts unit for high school juniors. Journal of
Counseling & Development, 71, 444–452.
Lapan, R. T., Gysbers, N. C., & Sun, Y. (1997). The impact of more fully
implemented guidance programs on the school experiences of high
school students: A statewide evaluation study. Journal of Counseling &
Development, 75, 292–302.
Lapan, R. T., & Kosciulek, J. F. (2001). Toward a community career system
program evaluation framework. Journal of Counseling & Development,
79, 3–15.
Lee, R. S. (1993). Effects of classroom guidance on student achieve-
ment. Elementary School Guidance and Counseling, 27, 163–171.
Lee, V. E., & Ekstrom, R. B. (1987). Student access to guidance counseling
in high school. American Educational Research Journal, 24, 287–310.
Macdonald, G., & Sink, C. A. (1999). A qualitative developmental analysis
of comprehensive guidance programmes in schools in the United States.
British Journal of Guidance and Counselling, 27, 415–430.
Missouri Department of Elementary and Secondary Education. (1999).
Missouri School Improvement Program: Procedures handbook. Jefferson City,
MO: Author.
Murdock, T. B. (1999). The social context of risk: Status and motivation
predictors of alienation in middle school. Journal of Educational Psy-
chology, 91, 62–75.
Nelson, D. E., & Gardner, J. L. (1998). An evaluation of the comprehensive
guidance program in Utah public schools. Salt Lake City, UT: The Utah
State Office of Education.
Oomen, A., Athanassoula-Reppa, A., Barnes, A., Petri, M., & Hautvast,
D. (1997). A comparative study of careers education and guidance in
six schools in Greece, the Netherlands and the United Kingdom. Edu-
cation for Careers in the Union. APS, Utrecht, The Netherlands.
Pain, K., Dunn, M., Anderson, G., Darrah, J., & Kratochvil, M. (1998).
Quality of life: What does it mean in rehabilitation? Journal of Rehabili-
tation, 64(2), 5–11.
Paisley, P. O., & Borders, L. D. (1995). School counseling: An evolving
specialty. Journal of Counseling & Development, 74, 150–153.
Raudenbush, S. (1984). Magnitude of teacher expectancy effects of pupil IQ as
a function of the credibility of the expectancy induction: A synthesis of the
findings from 18 experiments. Journal of Educational Psychology, 76, 85–97.
329
Lapan, Gysbers, and Petroski
Schaps, E., Battistich, V., & Solomon, D. (1997). School as a caring com-
munity: A key to character education. In A. Molnar (Ed.), Ninety-
sixth yearbook of the National Society for the Study of Education (pp.
127–139). Chicago: University of Chicago Press.
Schneider, B., & Stevenson, D. (1999). The ambitious generation: America’s
teenagers motivated but directionless. New Haven, CT: Yale University Press.
School-To-Work Opportunities Act, Pub. L. No. 103-289. (1994).
Secretary’s Commission on Achieving Necessary Skills. (1991). What
work requires of schools: A SCANS report for America 2000. Washing-
ton, DC: U.S. Department of Labor.
St. Clair, K. L. (1989). Middle school counseling research: A resource for
school counselors. Elementary School Guidance and Counseling, 23,
219–226.
Sun, Y., Hobbs, D., & Elder, W. (1994). Multi-level analyses of television
viewing among high school students: A contrast between non-
metropolitan, rural and other communities. Journal of Research in Rural
Education, 10, 97–107.
Watkins, C. (1994). Whole-school guidance? British Journal of Guidance
and Counselling, 22, 143–150.
330 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Wehlage, G. G., & Rutter, R. A. (1986). Dropping out: How much do
schools contribute to the problem? Teachers College Record, 87,
374–392.
Weinstein, R. (1989). Perceptions of classroom process and student
motivation: Children’s views of self-fulfilling prophecies. In C.
Ames & R. Ames (Eds.), Research on motivation in education (Vol.
3, pp. 187–221). San Diego, CA: Academic Press.
Wentzel, K. R. (1997). Student motivation in middle school: The role of
perceived pedagogical caring. Journal of Educational Psychology, 89,
411–419.
Wentzel, K. R. (1998). Social support and adjustment in middle school:
The role of parents, teachers, and peers. Journal of Educational
Psychology, 90, 202–209.
Wentzel, K. R. (1999). Social-motivational processes and interpersonal
relationships: Implications for understanding motivation at school.
Journal of Educational Psychology, 91, 76–97.
Whiston, S. C., & Sexton, T. L. (1998). A review of school counseling
outcome research: Implications for practice. Journal of Counseling &
Development, 76, 412–426.
High School Students’ Career-Related
Decision-Making Difficulties

Itamar Gati and Noa Saka

The present study examined the construct of career-related decision-making difficulties among 1,843 Israeli adolescents: choos-
ing a high school (9th grade), choosing high school elective courses (10th grade), and deciding on a military job preference (11th
grade). Three versions of the Career Decision-Making Difficulties Questionnaire (CDDQ) were constructed to match the 3 deci-
sion situations. The structures of the 10 difficulty categories of the revised CDDQ were found similar to that proposed by I. Gati,
M. Krausz, and S. H. Osipow (1996). Boys reported higher difficulties than girls in external conflicts and dysfunctional beliefs.
Research and counseling implications are discussed.

P
eople typically make their first career-related de-
cisions during adolescence. Such decisions may
have lifelong consequences for the individual’s
vocational future, psychological well-being,
health, and social acceptance (Mann, Harmoni,
& Power, 1989). Adolescent students often have to make
decisions concerning their choice of high school and their
high school elective courses. These decisions affect the stu-
dents’ educational and vocational opportunities. In addi-
tion, most Israeli high school students, both boys and girls,
must make another career-related decision—the type of
army service they prefer. In all three situations, the stu-
dents are given the option to choose an alternative. How-
ever, the final decision, in most cases, is in the hands of the
involved officials, who also take into account factors over
which the students have no influence. Finally, parents and
additional significant others explicitly or implicitly affect
the adolescent’s decision.
Although some of the adolescents who are required to
make these early career decisions do so relatively easily,
many others face difficulties before or during the actual
process of decision making. These difficulties may lead them
to attempt to transfer the responsibility for making the
decision to someone else or to delay or even avoid making
a decision. This may ultimately lead to a less than optimal
decision. In addition, the way students handle these deci-
sions may have an effect on the way they will deal with
future career decisions. Finally, the stress involved in the
process may affect various aspects of the adolescents’ daily
life. For adolescents, who are trying to clarify and consoli-
date their world, the skill of decision making is central in
dealing with various situations (Scott, Reppucci, & Woolard,
1995). Indeed, Taveira, Silva, Rodriguez, and Maia (1998)
found that adolescents reported that fairly high levels of
stress were associated with career exploration and decision-
making activities in general. On the other hand, a certain
degree of decision-specific affective distress among adoles-
cents can also be adaptive because it increases their moti-
vation to seek help and thus decreases the chances for poor
or ill-informed decisions (Larson & Majors, 1998).
To help students in making career decisions, school coun-
selors have to locate the difficulties the adolescents face
and provide them with guidance on how to overcome, or
at least minimize, these difficulties. The goal of this study
was to characterize and categorize the different types of
career-related decision-making difficulties faced by Israeli
adolescents, relying on a general theoretical model for ana-
lyzing such difficulties recently proposed by Gati, Krausz,
and Osipow (1996).
ADOLESCENTS AS CAREER DECISION MAKERS
Research has shown that educational issues and future vo-
cational decisions are of great importance to adolescents
(Bibby & Postersky, 1985; Collins & Harper, 1974; Eme,
Maisiak, & Goodale, 1979; Nicholson & Antill, 1981; Rutter,
1980; Violato & Holden, 1988). Friedman (1991) investi-
gated types of decisions with which Israeli adolescents (9th
and 11th graders) are concerned and found that problems
involving educational issues concern adolescents most (43%
of the problems revolved around studies and career). Among
the educational concerns, the most serious ones were choos-

Itamar Gati is a professor of psychology and education, and Noa Saka is a graduate student, both at the Hebrew University in Jerusalem. The authors thank
Michal Malka, Adina Orenstein, Dafna Sharav, and Miri Zarihan, for their contribution in various phases of the research, and Itay Asher, Liat Barkai-
Goodman, Naomi Goldblum, Tali Ever-Hadani, Liat Kibari, Elchanan I. Meir, Leah Israelevich, Gal Ram, Talia Sagiv-HaCohen, and Laura Naftali for
their comments on an earlier version of this article. This research was supported by a grant from the chief scientist of the Department of Education of Israel and
carried out at the NCJW Research Institute for Innovation in Education at the Hebrew University of Jerusalem. Correspondence regarding this article should
be sent to Itamar Gati, Department of Psychology, The Hebrew University, Jerusalem 91905, Israel (e-mail: msgati@mscc.huji.ac.il).

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 331

Gati and Saka
ing high school elective courses (46%) and choosing a high
school (26%).
The significant developmental process that takes place
during adolescence results in improved cognitive abilities,
which enhance decision-making capabilities. Piaget and
Inhelder (1969), for example, described such improvements
in terms of the transition from concrete cognitive processes
to abstract ones and from subjective to objective thinking.
They also described such processes as acquiring confidence
in handling complex environments and relying more on facts
and exploration. Lewis (1981) found a positive relation-
ship between adolescents’ age and their decision-making
capabilities. This included increased awareness of the risks
and implications involved in making a decision, seeking more
advice from adults or peers, and awareness of the implica-
tions of receiving advice from someone with vested inter-
ests. Other studies have found cognitive changes, such as
improvements in memory, organization of cognitive pro-
cesses, and the ability to better process information and
apply knowledge (Keating & Bobbitt, 1978; Sternberg &
Nigro, 1980; Sternberg & Rifkin, 1979). Consequently, ado-
lescents develop the ability to think of the universe of pos-
sibilities, to frame the alternatives, and to think in abstract
terms (Friedman & Mann, 1993; Ormond, Luszcz, Mann,
& Beswick, 1991; Steinberg, 1985).
Thus, as the need to make significant decisions arises dur-
ing adolescence, the ability to make them develops as well.
Moreover, research has shown that it is possible to improve
adolescents’ decision-making skills (e.g., Mann, Harmoni,
Power, Beswick, & Ormond, 1988; Owens, 1983; Silverman
& Wells, 1987). As alluded to earlier, the present study is
aimed at improving career counselors’ understanding of the
various difficulties faced by adolescents in making their
educational career decisions. This should eventually con-
tribute to developing further interventions aimed at facili-
tating adolescents’ career decision-making processes by re-
ducing potential difficulties.
CAREER INDECISION
The difficulties related to making career and educational deci-
sions have been associated in the past with the notion “career
indecision.” Previous research has devoted much attention to
categorizing the types of problems related to indecision,
focusing on several theoretical views (Campbell, & Cellini,
1981; Crites, 1978; Fouad, 1994; Osipow & Fitzgerald,
1996; Rounds & Tinsley, 1984; Super, 1953). Much of this
discussion focused on categorizing the types of difficulties
related to indecision and on formulating a typology of indi-
viduals according to the problems and difficulties they face.
The empirical research has focused on the development of
measures for examining individual differences in career inde-
cision (see reviews by Gati et al., 1996; Osipow, 1999; Slaney,
1988; Spokane & Jacob, 1996). Some of these measures
emerged from the daily practice of career counseling and
they typically provide a global rather than a refined
assessment of the individual’s difficulties.
332 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
TAXONOMY OF CAREER DECISION-MAKING DIFFICULTIES
The taxonomy of difficulties in career decision making
developed by Gati et al. (1996) was used as a theoretical
basis for the present research. This taxonomy is based on
decision theory, which has come to play an important role
in understanding the processes involved in career decision
making (e.g., Brown, 1990; Gati, 1986; Gati & Asher, 2001;
Jepsen & Dilley, 1974; Katz, 1966; Mitchell & Krumboltz,
1984; Osipow, 1987; Osipow & Fitzgerald, 1996; Phillips,
1994; Pitz & Harren, 1980; Walsh & Osipow, 1988). The
difficulties in the taxonomy were based on deviations from
the model of an “ideal career decision maker.” The ideal career
decision maker is a person who is aware of the need to make
a career decision, is willing to make such a decision, and is
capable of making the decision “right” (i.e., a decision that is
based on an appropriate process and is compatible with the
individual’s goals and resources). Any deviation from the
model of the ideal career decision maker is regarded as a
potential difficulty that may affect the individual’s decision-
making process in one of two possible ways: (a) by preventing
the individual from making a career decision or (b) by leading
to a less than optimal career decision.
The proposed taxonomy (Gati et al., 1996) includes three
major difficulty categories, which are further divided into
10 specific categories of difficulty. The first major category,
Lack of Readiness, includes three categories of difficulties
that may arise before the beginning of the career decision-
making process: (a) lack of motivation to engage in the ca-
reer decision-making process; (b) general indecisiveness con-
cerning all types of decisions; and (c) dysfunctional beliefs,
including irrational expectations (Nevo, 1987) concerning
the career decision-making process (e.g., “I believe there is
only one ideal career for me”).
The two other major difficulty categories, Lack of Infor-
mation and Inconsistent Information, include categories of
difficulties that may arise during the actual career decision-
making process. Lack of Information includes four catego-
ries of difficulties: (a) lack of knowledge about the steps
involved in the process, (b) lack of information about the
self, (c) lack of information about the various alternatives
(i.e., occupations), and (d) lack of information about the ways
of obtaining additional information. The major category
Inconsistent Information includes three categories of diffi-
culty: (a) unreliable information, that is, difficulties related to
unreliable or contradictory information (e.g., above average
high school grades, but a low SAT score); (b) internal
conflicts, that is, conflicts within the individual, such as con-
tradictory preferences or difficulties concerning the need
to compromise; and (c) external conflicts, that is, conflicts
involving the influence of significant others. The structure
of the three major categories and the 10 categories of diffi-
culties is summarized in Figure 1.
Further distinctions within each category were based on
both theoretical considerations and expected practical sig-
nificance. For example, within the category of lack of in-
formation about the self, a distinction was made between lack
 High School Students’ Career-Related Decision-Making Difficulties
CAREER DECISION-MAKING DIFFICULTIES
Prior to the Process
Lack of Lack of
Readiness Information
due to about
Lack of Indeci- Dysfunc- The Self
Motivation siveness tional Career
Beliefs Decision-
Making
Process
FIGURE 1
A Taxonomy of Career Decision-Making Difficulties
of information regarding the individual’s preferences (“What
do I want?”) and capabilities (“What can I do?”). (For a more
detailed description of the taxonomy, see Gati et al., 1996.)
To empirically test the proposed theoretical taxonomy,
Gati et al. (1996) constructed the Career Decision-Making
Difficulties Questionnaire (CDDQ), in which each of the
difficulties in the 10 categories was represented by a state-
ment (e.g., “I usually try to avoid commitment”). Studying
both American and Israeli samples of young adults, Gati et
al. (1996) found that the empirical structure of the three
major categories and the 10 specific categories was highly
similar to the theoretical structure presented in Figure 1.
Osipow and Gati (1998) examined the construct and concur-
rent validity of the CDDQ. They analyzed the responses of
403 college students from the United States and examined
the empirical relations of the CDDQ with two measures
associated with difficulties in career decision making—the
Career Decision Scale (CDS; Osipow, Carney, & Barak,
1976; Osipow & Winer, 1996), which is a widely used in-
decision questionnaire (Betz, 1992; Slaney, 1988), and the
Career Decision-Making Self-Efficacy Scale (CDMSE;
Taylor & Betz, 1983), which includes five distinct, theo-
retically defined scales, corresponding to the five career-choice
competencies suggested by Crites (1978). As hypothesized,
Osipow and Gati found that the CDDQ was positively cor-
related with the CDS and negatively correlated with the
CDMSE. In addition, the scores on the CDDQ indicated
that decided students had significantly fewer difficulties than
undecided ones. Lancaster, Rudolph, Perkins, and Patten’s
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
During the Process
Inconsistent
Information
due to
Occupa- Ways of Unreli- Internal External
tions (i.e., Obtaining able Conflicts Conflicts
alterna- Informa- Informa-
tives) tion tion
(1999) research supported the construct validity of the
CDDQ and reported a large difference between decided and
undecided groups in the total CDDQ score.
THE GOAL OF THE PRESENT RESEARCH
In the present study, we adapted the taxonomy of difficul-
ties proposed by Gati et al. (1996) to match the difficulties
faced by Israeli adolescent students in three grades and de-
cision situations: choosing a high school in the 9th grade
(which is the last year of junior high school in Israel), choos-
ing high school elective courses in the 10th grade, and de-
ciding on a military job preference in the 11th grade. We
report three studies corresponding to these three decision
contexts. For each decision context, we first revised and
adapted the original taxonomy (Gati et al., 1996) and then
constructed a questionnaire on the basis of the adapted tax-
onomy. Using these questionnaires, we empirically exam-
ined the structure of the difficulties and compared it with
the original theoretical taxonomy. Specifically, in each of
the three decision situations we examined whether the
decision-making difficulties are adequately represented by
the 10 categories proposed by Gati et al. (1996) and whether
the empirical relations among the categories resemble the
theoretical taxonomy (i.e., whether the 10 categories are
indeed clustered into the three major categories—Lack of
Readiness, Lack of Information, and Inconsistent Informa-
tion). In addition, we examined gender differences, differ-
ences between “decided” and “undecided” students, and dif-
333
Gati and Saka
ferences among the various grades in the relative salience
of difficulties involving conflicts with significant others.
STUDY 1
Toward the end of junior high school (i.e., the end of the 9th
grade), most students in Israel can apply to the high school
they prefer. In certain cases it is possible to make the transi-
tion from junior high to high school within the same school;
in other cases, the student must transfer to another school.
This school transition allows students to choose a high school
specializing in areas of particular interest to them, such as
art, music, science, or technology. Usually the students have
several alternatives to select from, and, in the larger cities,
the number of alternatives can reach more than a dozen. It
should be noted, however, that the students’ grades, behav-
ior, and occasionally also aptitude test scores affect their
chances of acceptance to their preferred high school.
Method
Participants. Five hundred and seventy-nine 9th-grade students
(297 girls, 266 boys, 16 did not report sex) from 22 classes (in
6 junior high schools) in Israel participated in the study. In
this sample, 90% of the students were native Israelis.
Questionnaire. We revised the 44 statements comprising
the original CDDQ (Gati et al., 1996) and adapted them
to the context of choosing a high school. As part of this
adaptation, we decided to combine into one some of the
statements that earlier were found to be highly correlated,
and we deleted a few other irrelevant statements. These
changes were based on the results of item analyses of the
original CDDQ, using data collected prior to the current
study (see Gati et al., 1996; Gati, Osipow, Krausz, & Saka,
2000; Osipow & Gati, 1998). In addition, 20 school counse-
lors and 100 ninth-grade students were asked to list difficul-
ties involved in the choice of high school, and we examined
whether all the difficulties listed by the school counselors
and the students, themselves, could be represented by the
statements in the revised questionnaire. After several revi-
sions, 35 statements were included in the final version of the
questionnaire.
The first page of the anonymous questionnaire included
general background information such as sex, name of cur-
rent school, grade, number of children in the family, position
in birth order, birthplace, and year of immigration to Israel.
Following this were two questions regarding whether the
student had some alternative (i.e., a specific high school) in
mind and what was his or her degree of confidence in that
alternative (on a 5-point scale, 1 = not confident at all, 5 =
very confident). Next, the students were asked to report to
what degree they believed their choice would affect their
vocational future and to what degree they believed their own
preferences would determine which high school they would
finally attend (both on a 5-point scale).
The subsequent four pages included 35 statements, each
corresponding to a particular difficulty (e.g., “It is usually
334 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
difficult for me to make a decision”). The participants were
asked to rate the degree to which each statement described
them on a 7-point scale (1 = does not describe me at all to 7 =
describes me well). At the end of the questionnaire, the students
were asked to rate the overall severity of their difficulties in
making the decision and to list any additional difficulties.
Procedure. The students responded to the questionnaires
during a class. The questionnaires were handed out to the
students by one of the research assistants in the presence of
a teacher or by the homeroom teacher, who was given the
relevant instructions and explanation regarding the
questionnaire’s purposes and importance. The question-
naires were distributed and the students’ questions, if any,
were answered. No time limitation was given; the time taken
to fill in the questionnaire varied from 20 to 30 minutes.
Analyses. First, we computed the following scores for each
participant: (a) the score of each of the 10 scales represent-
ing the 10 difficulty categories (defined as the mean of the
items included in each scale), (b) the score of the three
major categories (defined as the mean of the respective
scales), and (c) an overall difficulty score (the mean of the
10 scale scores). Next, we computed Cronbach’s alpha re-
liability for each of the 10 scales, the three major difficulty
categories, and the overall difficulty score. Then, using the
Pearson product–moment correlation, we computed the
intercorrelations among all the scales. Finally, to examine
the empirical structure of the 10 questionnaire scales, we
used a cluster analysis computer program—ADDTREE
(Sattath & Tversky, 1977). ADDTREE represents the
intercorrelations among the scales in the form of an addi-
tive or “path length” tree, in which the variables are divided
into clusters according to the size of the correlations be-
tween them, with highly correlated variables being located
in the same cluster.
Results
Table 1 presents the means, standard deviations, and
Cronbach’s alpha reliabilities of the 10 scale scores, the three
major categories, and the overall score. As in previous stud-
ies (Gati et al., 1996; Gati et al., 2000; Osipow & Gati,
1998), Cronbach’s alpha reliabilities of the 10 scale scores
varied, ranging from .40 for the scale of Dysfunctional Be-
liefs to .82 for External Conflicts; the median scale reliabil-
ity was .68. Among the three major categories, Lack of
Readiness had the lowest reliability (.62); the reliability of
the two other major categories was much higher (.88 for
Lack of Information, and .87 for Inconsistent Information).
The reliability of the questionnaire as a whole was .91.
The mean overall difficulty score of students who had
already considered an alternative was significantly lower
than that of students who had not yet considered one, t(570)
= 2.70, p < .01. Although the mean severity of difficulties
reported by the students who had already considered an
alternative (3.58) was, as expected, lower than that of those
who had not (3.68), the difference was relatively small and
not statistically significant, t(560) = 0.53, p > .05.
 High School Students’ Career-Related Decision-Making Difficulties

TABLE 1 As can be seen in Figure 2, the empirical structure is simi-

lar, although not identical, to that of the theoretical model
Means, Standard Deviations, Reliabilities of the underlying the questionnaire (see Figure 1). As expected,
Scales of the CDDQ, and Their Correlation With the major category, Lack of Readiness, emerged separately,
Severity in Study 1 (N = 579) suggesting that there is a distinction between difficulties
Correlation arising prior to the career decision-making process and those
Number With arising during the actual process. Although the third major
Scale/Category of Items M SD α Severity category, Inconsistent Information, also emerged as ex-
pected, the second category, Lack of Information, seemed
Lack of Readiness
Lack of Motivation 3 2.80 1.43 .59 .14* less distinct. First, in contrast to the hypothesized struc-
Indecisiveness 4 3.92 1.20 .54 .26* ture, the scale Lack of Information About the Process ap-
Dysfunctional peared independently instead of being located under Lack
Beliefs 3 3.50 1.34 .40 .08
Lack of Information
of Information. This deviation may be attributed to the fact
About the that lack of knowledge about how to engage in a decision-
Process 3 3.09 1.53 .78 .44* making process can arise prior to the actual engagement in
About the Self 4 2.81 1.31 .73 .43* such a process as well as during the process. Second, the
About the
Alternatives 3 2.74 1.56 .60 .34* scale Lack of Information About the Self emerged sepa-
About Ways of rately from the other two scales of Lack of Information.
Obtaining This deviation reflects the relatively high correlations of
Additional
Information 2 2.31 1.48 .65 .33*
this scale with difficulties involving Unreliable Information
Inconsistant and Internal Conflicts (.57 and .58). The two scales of Lack
Information of Information, which represent objective information (Lack
Unreliable of Information About the Alternatives and Lack of Infor-
Information 3 2.99 1.59 .72 .44*
Internal Conflicts 6 3.21 1.41 .76 .50* mation About Ways of Obtaining Additional Information),
External Conflicts 4 2.63 1.52 .82 .42* were grouped together as expected. In addition, Internal
Lack of Readiness 10 3.41 0.92 .62 .22* and External Conflicts were included in the same subcluster,
Lack of Information 12 2.74 1.20 .88 .47* as expected.
Inconsistent
Information 13 2.94 1.27 .87 .53* Next, we computed the correlations of the overall subjec-
Total CDDQ 35 3.00 0.97 .91 .51* tive severity reported by the students with each of the 10
difficulty scales. These correlations are presented on the right-
Note. CDDQ = Career Decision-Making Difficulties Questionnaire.
*p < .001. hand side of Table 1. All correlations were positive and,
except for the scale of Dysfunctional Beliefs, statistically
significant (p < .001). The correlations were higher for the
We computed the intercorrelations among the 10 scale Lack of Information and Inconsistent Information major cat-
scores. The clustering structure derived from this egories, and lower for difficulties related to Lack of Readiness.
intercorrelation matrix (by ADDTREE) is presented in Figure
2. The distance between any pair of scales in this clustering STUDY 2
structure is represented by the sum of horizontal segments
on the shortest path connecting them. The variance linearly In the 10th grade, Israeli high school students have to decide
accounted for by the distances in the clustering structure is how many credit points (3–5) they prefer to take in each of
.96, indicating that the clustering structure in the figure ad- their obligatory courses (e.g., mathematics, English, Hebrew
equately summarizes the empirical relations among the scales. literature), which elective courses to take as advanced-level
courses (e.g., sciences: physics, chemistry, etc.; humanities:
history, languages etc.; social sciences: psychology, econom-
Lack of Motivation ics, etc.) and at what levels. The elective courses selected and
Indecisiveness the levels in both types of courses (reflected in the number
Dysfunctional Beliefs
of credit points) may affect future educational possibilities;
Lack of Information About the Process
for example, to study engineering, students must take at least
Lack of Information About Alternatives
one advanced science course in high school.
Lack of Information About Ways of Obtaining Information
Lack of Information About the Self
Method
Unreliable Information
Internal Conflicts Participants. Seven hundred and eighty-seven 10th-grade
External Conflicts students (465 girls, 300 boys, 22 did not report sex) in 23
FIGURE 2 classes (in 8 high schools) in Israel participated in the study.
In this sample, 87% of the students were native Israelis.
The Structure of the 10 Difficulty Scales in Procedure. We adapted the CDDQ to the context of choos-
Grade 9 (N = 579) ing high school elective courses using the same procedure as

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 335

Gati and Saka

in Study 1. The content of the items was revised to match not yet considered one, t(778) = 4.94, p < .001. In addition,
the specific decision; the revised and adapted questionnaire the mean severity rating of the former group (3.71) was
consisted of 36 questions. The data collection procedure and lower than that of the latter (5.03), t(758) = 5.46, p < .001.
analyses were similar to those of Study 1. The clustering structure derived from the intercorrelation
matrix (by ADDTREE) is presented in Figure 3. The vari-
Results ance linearly accounted for by the distances in the clustering
structure is .96. As can be seen in Figure 3, the empirical
Table 2 presents the means, standard deviations, and structure is highly similar to that obtained for the 9th-grade
Cronbach’s alpha reliabilities of the 10 scale scores of the students and similar but not identical to the theoretical model
questionnaire, its three major categories, and its overall score. underlying the questionnaire (see Figure 1). As expected,
As can be seen in Table 2, Cronbach’s alpha reliabilities of two of the major categories—Lack of Readiness and Incon-
the 10 scale scores show considerable variation. As in Study sistent Information—emerged, whereas the four scales of
1, the lowest reliability (.34) was observed for the scale of Lack of Information were less cohesive. First, Lack of Infor-
Dysfunctional Beliefs; the other scales had moderate to high mation About the Process appeared independently, as in the
reliabilities ranging from .57 for the scale of Lack of Moti- 9th-grade sample. Second, Lack of Information About Self
vation to .84 for External Conflicts (the median scale reli- appeared independently but close to the clusters of Lack of
ability was .65). Among the three major categories, Lack of Information About Alternatives and Lack of Information
Readiness had the lowest reliability (.58); the reliabilities About Ways of Obtaining Additional Information.
of the two other major categories were much higher (.88 Next, we computed the correlations between the subjective
for Lack of Information, and .86 for Inconsistent Informa- severity reported by the students and each of the 10 difficulty
tion). The reliability of the questionnaire as a whole was scales. These correlations are presented on the right-hand
.91, the same as that in Study 1. Again, the mean difficulty column of Table 2. The pattern of correlations was similar to
score of students who had already considered an alterna- that in Study 1, with a low correlation for Lack of Readiness.
tive was significantly lower than that of students who had
STUDY 3
TABLE 2
At the end of high school, by the age of 18, the majority of
Means, Standard Deviations, Reliabilities of the Israeli youth are drafted for mandatory military service (typi-
Scales of the CDDQ, and Their Correlation With cally, 3 years for young men and 2 years for young women).
Severity in Study 2 (N = 787) Although the actual military job and its location are deter-
mined by the army, according to the individual’s specific
Correlation skills and capabilities and the army’s needs, the adolescent
Number With can contribute to the army’s decision regarding his or her
Scale/Category of Items M SD α Severity
military job. During the second half of the 11th grade, stu-
Lack of Readiness dents are usually asked to indicate their military job prefer-
Lack of Motivation 3 3.18 1.42 .57 .24*
Indecisiveness 4 3.85 1.28 .59 .34*
ence. This is often done by ranking the options on a given
Dysfunctional list of possible military jobs, which is compiled by the army
Beliefs 3 3.36 1.20 .34 –.02 on the basis of medical, psychometric, and other additional
Lack of Information information about the individual. Many schools have “open
About the
Process 3 3.27 1.51 .72 .48* days,” on which army representatives come to the school
About the Self 5 3.06 1.40 .76 .53* and describe various military jobs to the students.
About the
Alternatives 3 3.06 1.46 .63 .47*
About Ways of Lack of Motivation
Obtaining
Dysfunctional Beliefs
Additional
Information 2 2.28 1.39 .66 .30* Indecisiveness
Inconsistant Lack of Information About the Process
Information
Lack of Information About the Self
Unreliable
Information 3 3.19 1.55 .65 .44* Lack of Information About Alternatives
Internal Conflicts 6 3.08 1.26 .73 .56* Lack of Information About Ways of Obtaining Information
External Conflicts 4 2.43 1.50 .84 .36* Unreliable Information
Lack of Readiness 10 3.46 0.86 .58 .29*
Internal Conflicts
Lack of Information 13 2.92 1.18 .88 .55*
Inconsistent External Conflicts
Information 13 2.90 1.23 .86 .52*
Total CDDQ 36 3.08 0.94 .91 .56* FIGURE 3
Note. See Table 1 Note. The Structure of the 10 Difficulty Scales in
*p < .001. Grade 10 (N = 787)

336 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79

 High School Students’ Career-Related Decision-Making Difficulties

Deliberating about one’s future military job can sometimes TABLE 3

be very stressful because this job will be one’s occupation
for 2 or 3 years and can affect one’s future civilian occupa- Means, Standard Deviations, Reliabilities of the
tion. The decision process may involve many difficulties, Scales of the CDDQ, and Their Correlation With
which can be conceptualized by the categories in the tax- Severity in Study 3 (N = 477)
onomy. First, information difficulties are often prominent in Correlation
this process because many military jobs are classified. Sec- Number With
ond, external conflicts may arise when the individual’s pref- Scale/Category of Items M SD α Severity
erences are not congruent with those of his or her family Lack of Readiness
(e.g., the individual may want to be in a combat unit, but the Lack of Motivation 3 2.89 1.26 .47 .02
family may express concerns about this preference). Another Indecisiveness 4 3.37 1.25 .60 .26*
Dysfunctional
difficulty may arise from the fact that it is the army that Beliefs 3 3.82 1.45 .64 –.15
ultimately determines the individual’s military occupation Lack of Information
(which may be incompatible with his or her preferences). About the
This may decrease the individual’s motivation to engage in Process 3 3.47 1.67 .76 .66*
About the Self 4 2.82 1.32 .66 .58*
the decision-making process in the first place. About the
Alternatives 3 3.99 1.78 .81 .51*
Method About Ways of
Obtaining
Participants. Four hundred and seventy-seven 11th-grade stu- Additional
dents (239 girls, 219 boys, 19 did not report sex) from 17 Information 1 3.33 2.03 — .45*
classes (in 3 high schools) in Israel participated in the study. Inconsistant
Information
In this sample, 89% of the students were native Israelis. Unreliable
Procedure. We adapted the CDDQ to the context of de- Information 4 2.67 1.40 .72 .34*
ciding on a military job preference using the same proce- Internal Conflicts 6 3.09 1.31 .72 .45*
External Conflicts 4 2.35 1.50 .84 .24*
dure as in the previous studies. The content of the items Lack of Readiness 10 3.36 0.83 .56 .05
was revised to match the specific decision; the revised and Lack of Information 11 3.40 1.41 .88 .65*
adapted questionnaire consisted of 35 questions. The data Inconsistent
collection procedure and analyses were similar to those of Information 14 2.70 1.19 .87 .40*
Total CDDQ 35 3.18 0.93 .90 .56*
Studies 1 and 2.
Note. See Table 1 Note.
Results *p < .001.

Table 3 presents the means, standard deviations, and

Cronbach’s alpha reliabilities of the 10 scale scores of the
questionnaire, its three major categories, and its overall score.
As can be seen in Table 3, the lowest internal consistency
was observed for Lack of Motivation (.47). The other scales
had moderate to high reliabilities, ranging from .60 for the
scale of Indecisiveness to .84 for External Conflicts (the
median scale reliability was .72). Among the three major
categories, Lack of Readiness had the lowest reliability (.56);
the reliability of the two other major categories was again
much higher (.88 for Lack of Information, and .87 for In-
consistent Information). The reliability of the questionnaire
categories—Lack of Readiness, Lack of Information, and
Inconsistent Information—and a distinction arises between
the first major category, which includes difficulties arising
prior to the career decision-making process, and the other
two major categories, which include difficulties arising dur-
ing actual engagement in the process.
As in the previous studies, we also computed the correla-
tions between the overall subjective severity reported by the
Lack of Motivation

as a whole was .90. The mean difficulty score of students Dysfunctional Beliefs

who had already considered an alternative (3.04) was sig- Indecisiveness

nificantly lower than that of students who had not yet con- Lack of Information About the Process

sidered one (3.60), t(470) = 5.69, p < .001. Similarly, the Lack of Information About the Self

mean of severity ratings of the former group (3.14) was lower Lack of Information About Alternatives

than that of the latter group (4.96), t(468) = 9.80, p < .001. Lack of Information About Ways of Obtaining Information

The clustering structure derived (by ADDTREE) from Unreliable Information

the intercorrelation matrix is presented in Figure 4. The Internal Conflicts

variance linearly accounted for by the distances in the clus- External Conflicts

tering structure is .94. As can be seen in Figure 4, the em- FIGURE 4

pirical structure is very close to the theoretical model pre-
sented in Figure 1. The scales are clearly grouped into three The Structure of the 10 Difficulty Scales in
clusters corresponding to the hypothesized three major Grade 11 (N = 477)

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 337

Gati and Saka
students and each of the 10 difficulty scales. These correla-
tions are presented in the right-hand column of Table 3, and
were similar in pattern to those found in the previous studies.
A COMPARISON AMONG THE GRADES AND AN ACROSS-GRADE
ANALYSIS
Gender Differences
Boys reported greater difficulties than girls in all grades re-
garding external conflicts, t(558) = 2.53, p < .01, t(750) =
4.42, p < .0001, and t(456) = 2.13, p < .05; and in two grades
regarding dysfunctional beliefs, t(561) = 2.88, p < .01, t(763)
= 1.91, p = .06, and t(456) = 3.82, p < .001, respectively for
Studies 1, 2 and 3. In addition, greater difficulties were re-
ported by boys in Grades 9 and 10 in the scales Lack of
Information About Ways of Obtaining Additional Informa-
tion, t(560) = 4.53, p < .001, t(756) = 3.92, p < .001, respec-
tively; and Internal Conflicts, t(560) = 2.48, p < .01, t(757) =
2.76, p < .01, respectively. Differences were also noted in
the major category Inconsistent Information, t(560) = 3.06,
p < .01, t(757) = 3.59, p < .001, respectively; and in the
overall difficulty score, t(561) = 3.60, p < .001, t(763) = 3.17,
p < .001, respectively. Girls reported more difficulties than
boys in the 11th grade in three of the scales related to lack of
information: Lack of Information About the Process, t(454)
= 3.06, p < .01; Lack of Information About Alternatives, t(454)
= 5.46, p < .0001; and Lack of Information About Ways of
Obtaining Additional Information, t(454) = 2.68, p < .01.
These differences were also reflected in the major category
Lack of Information, t(454) = 4.04, p < .001.
The Relative Salience of Difficulties Involving External
Conflicts
We also tested the hypothesis that the relative salience of
difficulties involving external conflicts (whose sources are
“significant others,” e.g., parents) decreases with age. To test
this hypothesis, we first computed the standard score of
the scale External Conflicts for each student. The mean of
the standard score of the External Conflicts scale was nega-
tive and statistically significant in all three grades (–0.33),
t(572) = 8.86; (–0.56), t(769) = 16.34; and (–0.64), t(466) =
15.90; for Grades 9, 10, and 11, respectively. This indicates
that these difficulties are relatively mild. As anticipated, we
found that the degree of difficulty involving external
conflicts decreased with age: A planned contrast revealed
that external conflicts were smaller in the 11th grade than in
the 9th and 10th grades, t(1808) = 4.22, p < .001, and smaller
in the 10th grade than in 9th grade, t(1341)= 4.75, p < .001.
The Structure of the Scales
Figure 5 presents the clustering structure derived from an
across-samples intercorrelation matrix (the linearly ac-
counted for variance is .96). This empirical structure pro-
vides strong support for the theoretical taxonomy because
it is highly similar to the theoretical structure presented in
338 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Lack of Motivation
Dysfunctional Beliefs
Indecisiveness
Lack of Information About the Process
Lack of Information About the Self
Lack of Information About Alternatives
Lack of Information About Ways of Obtaining Information
Unreliable Information
Internal Conflicts
External Conflicts
FIGURE 5
The Structure of the 10 Difficulty Scales Across All
Students (N = 1,843)
Figure 1. As can be seen, a distinction clearly emerges
between difficulties arising prior to the career decision-making
process and those arising during the actual process. The
latter difficulties are divided into those involving lack of
information and those involving inconsistent information.
As anticipated, the scales Lack of Information About Alter-
natives and Lack of Information About Ways of Obtaining
Additional Information are grouped together, as are the
scales Internal Conflicts and External Conflicts. This struc-
ture suggests that the observed deviations in the separate
analyses of the grades may reflect grade-dependent specific
factors, whereas on the whole the structure holds.
DISCUSSION
The present study examined career-related decision diffi-
culties among adolescents. To assess these difficulties, we
used revised versions of the CDDQ, adapted to suit three
decision situations encountered by most Israeli adolescents.
We collected and analyzed the responses of 9th-grade stu-
dents choosing a high school, 10th-grade students choos-
ing high school elective courses, and 11th-grade students
deciding on their military job preference.
The structures of the adolescents’ decision-making difficul-
ties in the three studies were compatible with the hypoth-
esized structure (presented in Figure 1), and the across-grade
structure replicated the findings of the Israeli sample in Gati
et al. (1996). The small deviations found among grades may
be accounted for by particular, grade-specific characteristics;
however, only further research can tell whether these devia-
tions are indeed consistent, and if they are, what these specific
characteristics may be. Finally, in the general analysis, the three
major categories emerged as expected, and a distinction was
observed between the major category Lack of Readiness and
the other two, Lack of Information and Inconsistent Informa-
tion. The great similarity between the across-grade structure
and the theoretical structure proposed by Gati et al. (1996)
provides clear support for the robustness of the latter.
The pattern of scale reliabilities obtained in this research
replicated previous findings (Gati et al., 1996; Gati et al.,
2000; Osipow & Gati, 1998). In particular, the consistently
low reliability of the Dysfunctional Beliefs scale is a matter
 High School Students’ Career-Related Decision-Making Difficulties
of concern, and further research is needed to better under-
stand this category of difficulties. The low reliability of the
scale Lack of Motivation in the present studies also deserves
further attention. In general, as could be expected with a
small number of items per scale, the reliabilities of the scales
are only moderate; further research may contribute to im-
proving these reliabilities. Nevertheless, the questionnaire
was clearly able to distinguish between students who had
already considered an alternative and those who had not,
in terms of their overall difficulties (with the former hav-
ing significantly fewer difficulties), a finding which pro-
vides support for its concurrent validity.
We also examined to what extent differences in the pat-
tern of relationships between adolescents and their parents
in the three grades are expressed in the relative difficulties
related to significant others’ influence. As could be expected,
the severity of external conflicts decreased from 9th to 11th
grade. This finding indicates that difficulties stemming from
the influence of significant others is positively correlated with
the significant other’s influence: the lower such influence is,
the lower are the difficulties attributed to that source.
Possible gender differences were also explored in the
present research. In external conflicts, boys reported greater
difficulties than girls consistently in all three grades. A pos-
sible explanation may be that parents and society have higher
educational and career aspirations for boys than for girls and
that they generally expect boys to be more career-oriented
(Hoffman, 1977). Furthermore, boys are subjected to more
peer pressure regarding occupational choices (McMahon &
Patton, 1997). This may lead to more intense conflicts with
significant others regarding boys’ choices. In other catego-
ries, 9th- and 10th-grade students revealed a similar pattern
of gender differences, wherein boys reported greater diffi-
culties involving internal conflicts, unreliable information,
and lack of information about ways of obtaining additional
information. In the 11th grade, however, girls reported more
difficulties involving lack of information than boys did. The
finding that 9th- and 10th-grade boys have more difficulties
related to lack of information than girls have may be attrib-
uted to the fact that boys find it more difficult to seek help
(Gross & McMullen, 1983), whereas the reversal of the
pattern in Grade 11 may be attributed to the specific char-
acteristics of the decision regarding army service.
Limitations
This research was carried out in Israeli junior high schools
and high schools. In countries with different characteristics
(e.g., a different high school education system) or with dif-
ferent decision tasks, the relative salience of the various dif-
ficulty categories will probably be different. In addition,
the reliability of the scale Dysfunctional Beliefs was lower
than acceptable, even given the small number of items.
Another limitation is that we could not compare the de-
gree of the difficulty reported for the various scales be-
tween the three grades, due to the confounding between
the grade and the decision task. Finally, the assessment of
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
the career decision-making difficulties was based on a par-
ticular theoretical approach and on the use of the revised
CDDQ as an instrument; further research could use other
theoretical models and other instruments.
Counseling Implications
Despite these limitations, the present research provides addi-
tional support for the theoretical taxonomy of career decision-
making difficulties proposed by Gati et al. (1996). Furthermore,
the findings suggest that the CDDQ may be used to adequately
assess systematic categories of difficulties that share common
features such as cause, timing, impact, or required intervention.
Specifically, the revised CDDQ may be used for needs assess-
ments of particular groups. For example, school counselors can
use the CDDQ to facilitate the identification of groups of
students who have difficulties related to one of the three major
categories and who may benefit from the same intervention.
The CDDQ also makes it possible to assess outcomes of inter-
ventions aimed at reducing general or specific career decision-
making difficulties. However, further research aimed at
improving the scales’ reliability is needed before the revised
CDDQ can be introduced as a potential diagnostic instrument.
This research has direct implications for counseling and
general intervention programs: Students should be taught
basic decision-making skills. Such interventions may ease
students’ decision-making process and help them make
better career decisions. Further research should aim at cross-
cultural validation of the taxonomy and its relative relevance
to adolescents in other countries. It should also investigate
whether the need for instruction in decision-making skills
is a common phenomenon in other countries and whether
teaching such skills to students indeed helps them make
better decisions. Further research should also examine
whether the major difficulties contributing to the subjec-
tive feeling of severity of difficulty are similar to those that
emerged in this study.
Finally, the present research suggests that the general struc-
ture of difficulties consisting of Lack of Readiness, Lack of
Information, and Inconsistent Information underlies career-
related decisions made in high school. Recent research (Shefer,
2000) found that the taxonomy of difficulties used in this
research might also underlie difficulties in other types of deci-
sions faced by young adults, such as choosing a partner. Future
research might further test the hypothesis that the same tax-
onomy can be applied to difficulties involved in other, nonca-
reer-related decisions faced by adolescents and adults.
REFERENCES
Betz, N. E. (1992). Career assessment: A review of critical issues. In S.
D. Brown & R. W. Lent (Eds.), Handbook of counseling psychology
(pp. 453–484). New York: Wiley.
Bibby, R. W., & Postersky, D. C. (1985). The emerging generation. Toronto,
Canada: Irwin.
Brown, D. (1990). Models of career decision making. In D. Brown &
L. Brooks (Eds.), Career choice and development (2nd ed., pp. 295–421).
San Francisco: Jossey-Bass.
339
Gati and Saka
Campbell, R. E., & Cellini, J. V. (1981). A diagnostic taxonomy of adult
career problems. Journal of Vocational Behavior, 19, 175–190.
Collins, J., & Harper, J. (1974). Problems of adolescents in Sydney. Jour-
nal of Genetic Psychology, 125, 189–194.
Crites, J. O. (1978). Career Maturity Inventory theory and research hand-
book. Monterey, CA: CTB/McGraw-Hill.
Eme, R., Maisiak, R., & Goodale, W. (1979). Seriousness of adolescent
problems. Adolescence, 14, 93–194.
Fouad, N. A. (1994). Annual Reviews 1991–1993: Vocational choice,
decision-making, assessment, and intervention. Journal of Vocational
Behavior, 45, 125–176.
Friedman, A. I. (1991). Areas of concern and sources of advice for Israeli
adolescents. Adolescents, 26, 967–976.
Friedman, A. I., & Mann, L. (1993). Coping patterns in adolescent deci-
sion making: An Israeli-Australian comparison. Journal of Adolescence,
16, 187–199.
Gati, I. (1986). Making career decisions: A sequential elimination ap-
proach. Journal of Counseling Psychology, 33, 408–417.
Gati, I., & Asher, I. (2001). The PIC model for career decision making:
Prescreening, in-depth exploration, and choice. In F. T. L. Leong & A.
Barak (Eds.), Contemporary models in vocational psychology (pp. 7–54).
Mahwah, NJ: Erlbaum.
Gati, I., Krausz, M., & Osipow, S. H. (1996). A taxonomy of difficulties in
career decision making. Journal of Counseling Psychology, 43, 510–526.
Gati, I., Osipow, S. H., Krausz, M., & Saka, N. (2000). Validity of the Career
Decision-Making Difficulty Questionnaire: Counselees’ versus career
counselors’ perceptions. Journal of Vocational Behavior, 56, 99–113.
Gross, A. E., & McMullen, P. A. (1983). The help-seeking process. In V.
J. Derleja & J. J. Grzelak (Eds.), Cooperation and helping behavior: Theo-
ries and research (pp. 305–326). New York: Academic Press.
Hoffman, L. W. (1977). Changes in family roles, socialization and sex
differences. American Psychologist, 32, 644–657.
Jepsen, D. A., & Dilley, J. S. (1974). Vocational decision-making models:
A review and comparative analysis. Review of Educational Research,
44, 331–349.
Katz, M. R. (1966). A model for guidance for career decision making.
Vocational Guidance Quarterly, 15, 2–10.
Keating, D., & Bobbitt, B. (1978). Individual and developmental differ-
ences in cognitive processing component of ability. Child Develop-
ment, 52, 538–544.
Lancaster, P. L., Rudolph, C., Perkins, S., & Patten, T. (1999). Difficulties
in career decision making: A study of the reliability and validity of the
Career Decision Difficulties Questionnaire. Journal of Career Assess-
ment, 4, 393–413.
Larson, M. L., & Majors, M. S. (1998). Application of the Coping with
Career Indecision Instrument with adolescents. Journal of Career As-
sessment, 6, 163–179.
Lewis, C. (1981). How adolescents approach decisions: Changes over
grades seven to twelve, and policy implications. Child Development,
52, 538–544.
Mann, L., Harmoni, R., & Power, C. (1989). Adolescent decision making:
The development of competence. Journal of Adolescence, 12, 265–278.
Mann, L., Harmoni, R., Power, C., Beswick, G., & Ormond, C. (1988).
Effectiveness of the GOFER course in decision making for high school
students. Journal of Behavioral Decision Making, 1, 159–168.
McMahon, M., & Patton, W. (1997). Gender differences in children and
adolescents’ perceptions of influence on their career development.
The School Counselor, 44, 368–376.
Mitchell, L. K., & Krumboltz, J. D. (1984). Research on human decision
making: Implications for career decision making and counseling. In S.
D. Brown & R. W. Lent (Eds.), Handbook of counseling psychology (pp.
238–280). New York: Wiley.
Nevo, O. (1987). Irrational expectations in career counseling and their con-
fronting arguments. The Career Development Quarterly, 35, 239–250.
Nicholson, S. I., & Antill, J. K. (1981). Personal problems of adolescents
and their relationship to peer acceptance and sex role identity. Jour-
nal of Youth and Adolescence, 10, 309–325.
340 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Ormond, C., Luszcz, M. A., Mann, L., & Beswick, G. (1991). A
metacognitive analysis of decision making in adolescence. Journal of
Adolescence, 14, 275–291.
Osipow, S. H. (1987). Manual for the Career Decision Scale. Odessa, FL:
Psychological Assessment Resources.
Osipow, S. H. (1999). Assessing career indecision. Journal of Vocational
Behavior, 55, 147–154.
Osipow, S. H., Carney, C. G., & Barak, A. (1976). A scale of educational-
vocational undecidedness: A typological approach. Journal of Voca-
tional Behavior, 9, 233–324.
Osipow, S. H., & Fitzgerald, L. F. (1996). Theories of career development
(4th ed.). Boston: Allyn & Bacon.
Osipow, S. H., & Gati, I. (1998). Construct and concurrent validity of
the Career Decision-Making Difficulties Questionnaire. Journal of
Career Assessment, 6, 347–364.
Osipow, S. H., & Winer, J. (1996). Career assessment and the Career
Decision Scale. Journal of Career Assessment, 4, 117–130.
Owens, T. R. (Ed.). (1983). Helping youth become more responsible. Wash-
ington, DC: National Institute of Education (EDN00001).
Phillips, S. D. (1994). Choice and change: Convergence from the decision-
making perspective. In M. L. Savikas & R. W. Lent (Eds.), Convergence
in career development theories (pp. 155–163). Palo Alto, CA: Consult-
ing Psychologists Press.
Piaget, J., & Inhelder, B. (1969). The psychology of the child. New York:
Basic Books.
Pitz, G., & Harren, V. A. (1980). An analysis of career decision making
from the point of view of information processing and decision theory.
Journal of Vocational Behavior, 16, 320–346.
Rounds, J. B., & Tinsley, H. E. A. (1984). Diagnosis and treatment of
vocational problems. In S. D. Brown & R. W. Lent (Eds.), Handbook of
counseling psychology (pp. 137–177). New York: Wiley.
Rutter, M. (1980). Changing youth in a changing society. Cambridge, MA:
Harvard University Press.
Sattath, S., & Tversky, A. (1977). Additive similarity trees. Psychometrika,
42, 319–345.
Scott, E. S., Reppucci, N., & Woolard, J. (1995). Evaluating adolescent
decision making in legal contexts. Law and Human Behavior, 19,
221–244.
Shefer, M. (2000). Difficulties of young adults in choosing a partner. Un-
published master’s thesis. Department of psychology, Tel-Aviv Uni-
versity, Israel.
Silverman, B., & Wells, S. M. (1987). Facing the facts regarding teenage
pregnancy. Social Work in Education, 9, 191–196.
Slaney, R. B. (1988). The assessment of career decision making. In W. B.
Walsh & S. H. Osipow (Eds.), Career decision making (pp. 33–76).
Hillsdale, NJ: Erlbaum.
Spokane, A. R., & Jacob, E. J. (1996). Career and vocational assessment
1993–1994: A biennial review. Journal of Career Assessment, 4, 1–32.
Steinberg, L. (1985). Adolescence. New York: Knopf.
Sternberg, R., & Nigro, G. (1980). Developmental patterns in the solu-
tions of verbal analogies. Child Development, 51, 27–38.
Sternberg, R., & Rifkin, B. (1979). The development of analogical
reasoning process. Journal of Experimental Child Psychology, 27,
195–232.
Super, D. E. (1953). A theory of vocational development. American Psy-
chologist, 8, 185–190.
Taveira, M. D. C., Silva, M. C., Rodriguez, M. L., & Maia, J. (1998).
Individual characteristics and career exploration in adolescents. Brit-
ish Journal of Counselling and Guidance, 26, 89–104.
Taylor, K. M., & Betz, N. E. (1983). Applications of self-efficacy theory
to the understanding and treatment of career indecision. Journal of
Vocational Behavior, 22, 63–81.
Violato, C., & Holden, W. B. (1988). A confirmatory factor analysis of a
four-factor model of adolescent concerns. Journal of Youth and Adoles-
cence, 17, 101–112.
Walsh, W. B., & Osipow, S. H. (Eds.). (1988). Career decision making.
Hillsdale, NJ: Erlbaum.
Assessing the Impact of a Prejudice Prevention Project

Michael Salzman and Michael D’Andrea

This article reports on the effectiveness of a prejudice prevention intervention that was used among a culturally diverse group of
students in Hawaii. Implications of the results of this primary prevention effort for counseling practice and interculturally adaptive
social development are discussed.

T
he increasing racial and cultural diversity in our
schools challenges counselors and educators. The
potentials inherent in current demographic
trends are multiple. For instance, diversity may
fuel conflict or mutual enrichment, fear or de-
lightful and fruitful interaction, appreciation or the dero-
gation of different “others.” School counselors can and
should be involved in maximizing the positive potentials
inherent in our diverse school communities while minimiz-
ing the potentials for stereotyping, prejudice, and intergroup
conflict. In the absence of intentional and effective educa-
tional programming designed to prepare young people to
live and work productively in a multicultural, diverse soci-
ety and global economy, the negative aforementioned po-
tentials are more likely to become manifest. The promo-
tion of respect and appreciation for human diversity is an
educational imperative and an ethical responsibility of
school counselors (Salzman, 1995). Ethnocentrism and
prejudice can degrade the human experience and may mo-
tivate the destructive behavioral expressions of the dynam-
ics inherent in cross-cultural and inter-group interactions
(Gazda, Ginter, & Horne, 2001).
Ethnocentrism, prejudice, and racism are maladaptive
responses to the anxieties provoked by human differences
on central existential problems such as the core ontological
concerns of how one is to live and what kind of person one
is to be in order to achieve a sense of anxiety-buffering self-
esteem. Diverse cultural responses to these concerns may
produce anxieties and provoke intolerant defensive responses
(Greenberg, Solomon, & Pyszczynski, 1997). Education,
counseling, and guidance can and should play a significant
role in reducing the probability of racist and intolerant re-
sponses to human anxieties. There is empirical evidence that
worldviews imbued with a sense of tolerance and respect for
diversity are resilient to defensive manifestations of intoler-
ance evoked by anxiety-provoking stimuli. People with such
worldviews may actually become more tolerant under such
conditions (Greenberg et al., 1997).
There is substantial evidence suggesting that the type of
intergroup contact people from diverse cultural or racial back-
grounds have with one another influences the cognitive, af-
fective, and behavioral outcomes of that contact (Gazda et al.,
2001). In this regard, it has been noted that equal status and
intimate and cooperative contact in the service of a shared
superordinate goal support the development of positive inter-
group interactions (Allport, 1982). Counselors can work with
teachers and administrators to promote these conditions within
the school setting in order to facilitate the positive potentials
of intergroup interactions and thereby reduce the negative
potentials inherent in such encounters. By cultivating these
intergroup characteristics and behavioral contexts, counselors
are able to positively contribute to the construction of a more
peaceful, respectful, and collaborative society. In doing so, coun-
selors can help to build a more just world.
FOSTERING PREJUDICE REDUCTION
Primary, secondary, and tertiary efforts toward prejudice
reduction and respectful appreciation of human diversity
help to contribute to the development of a healthy, peace-
ful, and productive human community. Primary prevention
refers to intentional programs and services that target groups
of currently unaffected people for purposes of helping them
continue functioning in healthy ways, free from disturbance
(Conyne, 1987). Secondary prevention refers to an early iden-
tification of problems followed by interventions designed
to minimize their further continuation, development, or
escalation. Tertiary prevention refers to efforts to remediate
or diminish the intensity of serious problems that have al-
ready emerged (Ponterotto & Pedersen, 1993).
D’Andrea and Daniels (1995) tested the effectiveness of
a multicultural developmental framework that was designed
to foster prejudice reduction. This prejudice reduction
framework comprised 10 guidance activities that were origi-
nally developed by Omizo and D’Andrea (1995). The ini-
tial research that was done by these investigators involved

Michael Salzman is an assistant professor, and Michael D’Andrea is a professor, both in the Department of Counselor Education at the University of Hawaii,
Manoa. The authors thank Salynn Ancheta, Liane Saito, and Michelle Tesoro for their invaluable contributions to this project. Correspondence regarding this
article should be sent to Michael Salzman, Department of Counselor Education, University of Hawaii, 1776 University Avenue, Honolulu, HI 96822 (e-mail:
msalzman@hawaii.edu).

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 341

Salzman and D’Andrea
a group of culturally diverse third-grade students who ex-
hibited signs of ethnic conflict and stress. More specifically,
the students who participated in this school-based prejudice
reduction intervention demonstrated negative aggressive
behaviors (i.e., arguing and fighting in class and on the play-
ground during recess) and were referred to the school coun-
selor by their teachers. These incidents were almost always
reported to be initiated by some student making deriding
and derogatory comments about another child’s cultural,
ethnic, or racial background (D’Andrea & Daniels, 1995).
The focus of the initial prejudice reduction intervention was,
therefore, on the level of secondary prevention.
Significant differences were reported in pre- and posttest
measures on a number of essential social skills including
“cooperation,” “assertiveness,” and “self-control” among the
students who participated in the Multicultural Guidance
Project (D’Andrea & Daniels, 1995; Omizo & D’Andrea,
1995). Teachers also rated participating students significantly
lower on a measure of “total problem behaviors” (Gresham
& Elliott, 1990). However, because no control group was
used for comparison in this study, causal inference was com-
promised by the researchers’ inability to control for his-
torical and maturation threats to internal validity.
EXTENDING THE EXISTING RESEARCH ON SCHOOL-BASED
PREJUDICE REDUCTION INTERVENTIONS
The present study extends D’Andrea and Daniels’s (1995)
initial research in the area of school-based prejudice reduc-
tion interventions in four fundamental ways. First, a com-
parison group of students was used to enhance the validity
of causal inference. Second, the same set of 10 classroom
guidance activities was applied to a group of students who
had not been experiencing interpersonal problems that re-
flected intercultural and interracial conflicts with their peers.
This study, then, was an effort at primary prevention. Third,
the present study added a self-esteem measure (Coopersmith,
1981) to assess the effects of the intervention on student
self-perception as well as social skill development. Fourth,
the current study tested the effectiveness of the interven-
tion on fourth-grade students at a different school site in
Hawaii, thereby extending the external validity of the treat-
ment. Thus, the overall goal of this study was to replicate
and extend D’Andrea and Daniels’s (1995) initial research
by testing the effectiveness of a primary prevention inter-
vention designed to promote social skills and attitudinal char-
acteristics that are necessary to reduce conflict and increase
the potential for positive interpersonal interactions in
multicultural and diverse school environments.
METHOD
Participants
Fifty (fourth-grade) students attending a public school in
Hawaii participated in this study. Twenty-eight of these stu-
dents were assigned to the treatment group and 22 students
342 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
were assigned to the control group. The group assignments
were made based on class membership in which the classes
were essentially equivalent in ability range, age, students’
socioeconomic backgrounds, and ethnic diversity. Both the
treatment and control groups were ethnically diverse, in-
cluding children identified as Hawaiian or part Hawaiian,
Filipino, Asian, Caucasian, African American, and Hispanic.
Instruments
The Social Skills Rating System (SSRS) used in the
D’Andrea and Daniels (1995) study was used to assess stu-
dents’ social skills in all conditions. The SSRS provides a
broad measure of students’ social competencies. Teacher and
student forms were used in this study. The SSRS provides a
measure of youngsters’ positive behaviors or proactive
social skills. It also includes an assessment of potential prob-
lem behaviors students may exhibit in their interactions with
others. National norms have been compiled on a diverse
sample (multiracial, handicapped, and nonhandicapped male
and female students consisting of more than 4,000 chil-
dren and adolescents; Gresham & Elliott, 1990). The
Teacher Form of the SSRS provides a total Social Skills
score and various subscale scores. Three of the subscale
scores measure students’ Cooperation, Assertiveness, and
Self-Control. The Teacher Form also includes the follow-
ing subscales:
1. An Externalizing subscale, which measure a student’s
tendency to exhibit verbal or physically aggressive
behaviors toward others
2. An Internalizing subscale, which measures a student’s
level of anxiety, sadness, and reluctance to interact
with others
3. A Hyperactivity subscale, which measures the degree
to which a student is easily distracted when interact-
ing with others
The Student Form of the SSRS also provides a total So-
cial Skills score but includes subscale ratings. Four subscale
scores measure students’ Cooperation, Assertiveness, Em-
pathy, and Self-Control. These proactive social skills are
thought to be vital for young people’s development in or-
der that they might experience a sense of well-being in their
interpersonal interactions (Selman & Byrne, 1974). These
skills are also considered to be especially important in cul-
turally and racially diverse school settings (D’Andrea &
Daniels, 1995).
Test–retest reliability coefficients for the teacher’s version
of the SSRS were reported to be .85 for the Social Skills
scales and .84 for the scales measuring Problem Behaviors.
The student forms had a test–retest reliability coefficient
of .68 for their social skills (Gresham & Elliott, 1990).
Stephens (1978) compared the SSRS with the Social Behav-
ior Assessment Scale and reported correlation coefficients
ranging from –.15 to –.73 with the Social Skill scales and
.01 to .57 with the Problem Behaviors scales. The samples
used to assess the reliability and validity of the SSRS

consisted of a diverse group of students and teachers. These
results suggest that the SSRS is a reasonably reliable and
valid way to measure positive and negative aspects of stu-
dents’ social development and interpersonal competency.
Intercorrelations among the subscales of the Teacher Form
of the SSRS ranged from .46 to .64 on the Social Skills
measures and .20 to .70 on the Problem Behaviors subscales.
The subscales of the Students Form of the SSRS have
intercorrelations ranging from .49 to .65.
The Coopersmith Self-Esteem Inventory (SEI; Coopersmith,
1981) is a self-report measure consisting of 58 items. The items
assess self-attitudes in four domains: Social, Home, School,
and General. A total self-esteem score is derived by summing
the high self-esteem responses on the four subscales and dou-
bling the score. The SEI is considered suitable for use within
the 8- to 15-year-old age range. The reliability of the SEI has
been tested and produced coefficients greater than .80 (Spatz
& Johnson, 1973). Concurrent validity of the Coopersmith
SEI was assessed by correlating the SEI and the SRA Achieve-
ment Series, and the SEI’s predictive validity was supported
by the finding that it predicted reading achievement (Post &
Robinson, 1998). Intercorrelations among the subscales of the
SEI range from .29 to .52.
Procedure
The SSRS and the SEI were administered to two fourth-
grade classes in an Oahu, Hawaii, public school by an advanced
graduate student doing her school counseling internship at
the school. One of these classes served as the control group
and the other as the treatment group. The student then
conducted the guidance activities with the treatment group
under the supervision of the two authors. Each guidance class
lasted 40 minutes. The classes were held once a week for 10
consecutive weeks. The instruments were readministered at
the conclusion of the treatment among students in both the
control and the experimental groups. What follows is a brief
description of the guidance activities that were used to pre-
vent the development of various cultural, racial, and ethnic
prejudices by fostering positive social skills and self-esteem
among students in the experimental group.
The Multicultural Guidance Activities
Class 1. During the initial session, the group facilitators
administered the SSRS and the SEI to the students in their
homeroom classes. The facilitators also discussed the rules
of the group meetings, the goals of the project, and their
expectations of the students.
Class 2. Name Tag Activity: After having the students write
their names, favorite foods, songs and music, places they like
to visit, and most enjoyable holidays on a blank piece of pa-
per, the facilitators led a class discussion that focused on the
youngsters’ responses to this exercise. The leaders empha-
sized any cultural differences that were associated with the
responses the students gave to these issues.
Class 3. The “I am” Activity: The students were asked to
draw a picture of themselves and to finish the statement
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Assessing the Impact of a Prejudice Prevention Project
“I am . . .” on the bottom of a blank piece of paper. The
purpose of this activity was to provide positive feedback
regarding any cultural issues/statements that emerged when
the students completed the “I am” statement.
Class 4. “Labeling”: The boys and girls in the class were
divided into two groups. The girls were arbitrarily identi-
fied as the “good” group in the class and the boys were re-
ferred to as the “bad” group. The facilitators announced vari-
ous privileges the “good” group would have from that point
on, whereas several punitive restrictions were assigned to
members of the “bad” group. Following these announce-
ments, the students were asked to discuss their reactions to
being in either group. The purpose of this session was to
introduce the concepts of “prejudice” and “stereotyping” to
the youngsters and to explore the various ways people feel
when they are the victims of cultural, ethnic, or racial preju-
dice and stereotyping.
Class 5. Videotaping Activity: While one of the facilitators
videotaped this session, the other one summarized the last
class meeting and continued to process students’ reactions.
The facilitator then helped the youngsters explore various
ways that elementary school students demonstrate negative
prejudices and stereotyping when interacting with people
from diverse backgrounds. After exploring these issues for
about 20 minutes, the facilitators asked the students to view
the video and to comment on their own interactions with
the facilitators and other students during that session.
Class 6. Multicultural People Bingo: The students were
provided a “Multicultural Bingo Sheet” that included boxes
containing various cultural-specific statements (i.e., “Enjoys
eating sushi,” “Has tried adobo,” “Is Samoan or part Samoan,”
etc.). They were instructed to move around the room and
find individuals who could “sign off” in one of the boxes on
the sheet. To “sign off,” students had to possess the quality
or characteristic that was listed in one of the boxes on the
bingo sheet. After the students had all of their boxes “signed
off” by other students, the facilitators led the class in a dis-
cussion about the positive aspects of cultural diversity.
Class 7. Abstract Concept Activity: The students were pre-
sented with several abstract concepts like “love” and “fairness”
and encouraged to discuss the different meanings these words
might have for people from culturally and racially diverse back-
grounds. They were then asked to draw pictures of what these
words looked like in their class and at their school. The stu-
dents were asked to show their pictures to the rest of the class
and explain what they represented.
Class 8. Rainbow Poster Activity: The facilitators led the
students in a discussion about “rainbows.” Next, they were
asked to paint a large rainbow on poster board paper using
the following colors: brown, black, white, red, and yellow.
The children were then asked to paint a picture of them-
selves next to the color that represented their own racial
background. As the students were involved in this activity,
the facilitators walked around the room asking individual
students to talk briefly about their racial background. The
facilitators provided positive feedback on various comments
the students made about their different backgrounds.
343
Salzman and D’Andrea

Class 9. The Hands Activity: After being arranged in small TABLE 1

groups, the students were asked to place their hands in the
middle of their desks and to look at their neighbor’s hands. Adjusteda Means for Subscales of SSRS
While asking them to keep their hands in the middle of their (Teacher Form)
desks, the facilitators led a group discussion that addressed the Group
following questions: “What do you notice about your own
Measure n Treatment n Control
and your neighbor’s hands?” “Are they all the same color/shape/
size?” and “Do you think any hand is better than another?” Assertiveness 25 11.45 21 11.07
Class 10. Summary and Posttesting: The facilitators asked Cooperation 25 14.81 21 12.38
the students if they could recall all the activities that they Externalizing 25 0.92 21 0.37
had done together during the past several weeks. They were Hyperactivity 25 1.85 27 2.34
Internalizing 25 1.73 21 2.11
also asked to discuss one thing they learned that they could Self-Control 25 14.73 21 12.07
use when working or playing with other students in the Social Skills 25 40.94 21 35.53
class. Finally, the SSRS and the SEI were readministered to
Note. SSRS = Social Skills Rating System.
students in both the control and the treatment groups at a
Adjusted for pretest scores.
the end of this session.

Data Analysis
nalizing, F(1, 36) = .86, p = 360; Self-Control, F(1, 36) =
A multivariate analysis of covariance (MANCOVA) was 21.46, p = .000; and Social Skills, F(1, 36) = 12.88, p = .001.
used to assess if significant differences existed between The results of the MANCOVA on the Students Form of
the posttest scores on the dependent measures that were the SSRS did not yield significant results (Wilks’s lambda =
generated by the students in the control and experimen- .79. Exact F(5, 32) = 1.71, p = .16. The power at the .05-
tal groups after completing the SSRS and the SEI. This level was calculated at .51. Adjusted means for the com-
procedure was deemed necessary due to the multiple depen- parison groups on the subscales are reported in Table 2.
dent measures used and their modest-to-moderate Although the Wilks’s lambda (p = .16) did not reach the
intercorrelations (e.g., SSRS subscale intercorrelations level of significance, it is important to note the relatively
ranging from .20 to .70). The pretest scores were treated substantial differences between the adjusted means on the
as covariates thereby equalizing the pretest means and control- Cooperation subscale scores. In view of these differences and
ling for the pretest scores. Means were calculated after the relatively modest power of the test, univariate F tests
adjusting for pretest scores and then inspected. One-way were done. Although it did not reach the .05 level for signifi-
analyses of covariance (ANCOVAs; univariate F tests) were cance, the Cooperation subscale F-test score (p = .059) did
then performed, and the results were interpreted in consid- approach this level of statistical significance. This is note-
eration of intercorrelations and overlapping variances among worthy given the relatively substantial differences between
subscales. An alpha level of .05 was established to deter- the adjusted posttest means on the Cooperation subscale
mine whether significant differences existed between the scores that are noted earlier. The results of the univariate
treatment and control group scores. The F-test scores reported tests for each of the variables on the Student Form of the
in the following tables therefore represent a comparison SSRS are as follows: Assertiveness, F(1, 36) = 0.63, p = .434;
between the treatment and the control groups’ posttest Cooperation, F(1, 36) = 3.80, p = .059; Empathy, F(1, 36) =
scores on the SSRS and SEI. .06, p = .807; Self-Control, F(1, 36) = .41, p = .528; and
Social Skills, F(1, 36) = .24, p = .628.
RESULTS
The results of the MANCOVA on the Teacher Form of the
SSRS indicated a significant omnibus effect, Wilks’s lambda TABLE 2
= 4.64, Exact F(7, 30) = 4.65, p = .001. The observed power a
Adjusted Means for Subscales of SSRS
at the .05 level was .99. Adjusted posttest means for the (Student Form)
comparison groups on the subscales are reported in Table
1. Subsequent univariate F tests on the Teacher Form of Group
the SRSS indicated significant differences between the Measure n Treatment n Control
teachers’ scores on the Cooperation, Self-Control, Social
Skills, and Externalizing SSRS subscales for the students Assertiveness 26 12.00 19 13.11
Cooperation 26 14.58 19 12.90
who were assigned to the treatment and control groups. Empathy 26 14.56 19 14.79
The results of the univariate tests for each of the variables Self-Control 26 11.86 19 11.24
are as follows: Assertiveness, F(1, 36) = 0.68, p = .417; Co- Social Skills 25 53.08 19 51.93
operation, F(1, 36) = 8.50, p = .006; Externalizing, F(1, 36) Note. See Table 1 Note.
= 6.63, p = .014; Hyperactivity, F(1, 36) 1.72, p = .198; Inter- a
Adjusted for pretest scores.

344 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79


The MANCOVA for the Coopersmith SEI also did not yield
a significant omnibus effect, Wilks’s lambda = .70, Exact F(6,
27) = 1.93, p = .11. Adjusted means were calculated and are
reported in Table 3. Follow-up univariate analyses for each of
the variables indicated one significant difference at p < .02
between the self-esteem scores that were generated by the stu-
dents in the treatment group in comparison with the students
in the control group. Specifically, this involved a significantly
lower Social Self-Esteem subscale score among the students in
the treatment group. The results are as follows: General,
F(1, 33) = .77, p = .388; Home, F(1, 33) .77, p = .399; School,
F(1, 33) = .04, p = .843; Lie (F = .94, p = .339); and Social
(F = 7.27, p = .011).
DISCUSSION AND IMPLICATIONS FOR SCHOOL COUNSELORS
The posttest scores on the SSRS indicated that teachers
observed significant improvement in the students’ coop-
erative social skills as a result of participating in the
multicultural guidance activities that were designed by
Omizo and D’Andrea (1995). The students’ ratings on the
Cooperation subscale, although not quite achieving signifi-
cance at the .05 level did yield a p = .059. In fact the F
value on this subscale seems dramatically higher than the
other F values reported on the Students Form of the SSRS.
The fact that both the teachers and the students ratings seemed
to converge on this essential dimension is worth noting. Gains
in this social skill enhances the validity of the inference that
this essential intercultural skill was fostered by the multicultural
guidance intervention. These results also provided the research-
ers with converging evidence that led to the conclusion that
the students in the treatment group benefited from partici-
pating in the prejudice reduction intervention as demonstrated
by their improved cooperation scores.
The importance of the converging findings on the Coop-
eration subscale scores is particularly impressive because
of the centrality that this interpersonal factor has been given
in prejudice reduction theory. In his seminal work on preju-
dice, Allport (1982) explained that the development of
cooperative skills and attitudes not only helps prevent the
promotion of inaccurate and prejudicial views toward oth-
ers, but it also underlies the sense of solidarity that people
TABLE 3
Adjusteda Means for Coopersmith Self-Esteem
Inventory
Group
Measure n Treatment n Control
General 25 17.04 19 16.38
Home 25 5.32 19 4.77
School 25 5.00 19 5.21
Social 25 5.00 19 6.28
Lie 25 2.59 19 3.11
Total Self-Esteem 25 64.86 19 64.81
a
Adjusted for pretest scores.
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Assessing the Impact of a Prejudice Prevention Project
from diverse cultural-racial backgrounds are able to achieve
when they interact with one another. Given the rapid cul-
tural transformation of the United States and the degree to
which cultural prejudices and racial tensions continue to
exist in our society, it is important from a preventive per-
spective that school counselors think of ways to intention-
ally and effectively promote the interpersonal solidarity that
Allport described in his work among students from diverse
groups and backgrounds. In part, this can be accomplished
by implementing school-based interventions that effectively
foster the development of cooperative social skills, much
like those of the prejudice prevention guidance classes de-
scribed in this article.
It should be further noted that significant negative find-
ings were observed on the teachers’ ratings of the students
in the control and treatment groups on the Externalizing
subscale. Initially, these findings seemed to represent a con-
tradiction in the other positive findings that were reflected
in the treatment group’s scores. However, a close scrutiny of
Table 1 shows that the adjusted mean posttest scores for
both the treatment and the control groups on the External-
izing subscale are less than 1 (i.e., control group, M = .37;
treatment group, M = .92). This means that the teachers
reported very few times in which the students in both the
treatment and the control groups manifested negative
“externalizing” behaviors at the time posttesting was done.
However, this finding was difficult to interpret and war-
rants further study.
One of the common problems that school counselors en-
counter when working with students of all ages involves
difficulties that occur when youngsters fail to demonstrate
self-control and instead act in an impulsive manner in in-
terpersonal situations in which conflicts or differences are
apparent. Given the frequency with which interpersonal
conflicts occur within schools that comprised students from
diverse cultural-racial backgrounds (D’Andrea & Daniels,
1996), the significant difference reported between the teach-
ers’ posttest Self-Control subscale scores may be particularly
interesting and relevant for the challenges school counselors
are likely to experience when working within the context of a
culturally diverse school setting. Recognizing that (a) this
subscale provides a measure of students’ ability to “control
one’s temper in conflict situations with peers” (Gresham &
Elliott, 1990, p. 3) and (b) noting that the students in the
treatment group received higher ratings in this area by their
teachers than the control group youngsters, counselors are en-
couraged to consider using Omizo and D’Andrea’s (1995)
guidance lessons as a way both to prevent prejudice and to
promote greater self-control among students.
Another interesting finding involved the significant differ-
ence that was observed on the students’ Social Self-Esteem
subscale scores. As reported, the students in the treatment
group scored significantly lower (result of F test) on social
self-esteem than their control group peers did. Rather than
interpreting this finding in negative terms (i.e., that young-
sters have a reduced sense of social self-esteem as a result
of participating in this intervention), this finding may be
345
Salzman and D’Andrea
interpreted as potentially positive. That is, the lower Social
Self-Esteem subscale scores that were reported by the stu-
dents in the treatment group may correlate with an emerg-
ing sense of cultural relativism and dissonance that the stu-
dents might have experienced as a result of participating in
the prejudice prevention intervention.
Like many adults, many students possess ethnocentric ways
of looking at themselves, others, and the world in which they
live. This sort of ethnocentric thinking is likely to help fuel
positive feelings individuals have about themselves and other
culturally and racially similar people as well as to contribute
to the negative attitudes individuals hold toward people who
come from different cultural-ethnic groups. Such an ethno-
centric perspective, if left unchallenged, represents fertile
terrain for the development of various types of cultural and
racial prejudices during one’s childhood and adolescence.
When thinking about the complex psychological factors
that contribute to prejudice prevention, it may be impor-
tant for school counselors to implement strategies that are
intentionally designed to disrupt students’ ethnocentric con-
struction of self-esteem. This can be accomplished, in part,
by providing youngsters with alternative cultural construc-
tions of reality and human “being-ness.” In doing so, a self-
esteem structure that may have been built on ethnocentric
notions of superiority and inferiority may well give way to a
more enduring and constructive sense of personal value. We
also suggest that this would help promote a sense of per-
sonal value that is derived from a profound sense of equality
and belonging to the diverse human community. This con-
struction of self-esteem offers a more enduring and less de-
structive foundation for the achievement of this necessary
psychological resource. Such a sense of belonging and equal-
ity was concretely manifested by the students who took part
in the prejudice prevention project that is described in this
article. This was particularly evident when the youngsters
worked together on the Rainbow Poster Activity (Class 8)
and when discussing the Hands Activity (Class 9).
LIMITATIONS
The limitations of this research include the relatively small
sample size that inhibits the external validity of the study.
In addition, although covariance analysis may, in a limited
sense, adjust for a failure of random assignment to groups
within the design, it does not adjust for a failure to sample
from portions of the population to which one wishes to
generalize. The negative finding on the Externalizing
subscale of the Teacher Form of the SSRS is somewhat
puzzling and warrants further study.
Another limitation in this study relates to the possible
confounding effect that attention may have played in this
study. In this regard, the researchers acknowledge that the
attention that was given to the students in the treatment
group as a result of participating in the multicultural guid-
ance activities in comparison with the absence of similar
attention directed to the control group students may have
affected their test scores.
346 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
CONCLUSION
In conclusion, counselors can make a difference in the con-
struction of a more respectful, harmonious, and enriching
multicultural environment if they intentionally choose to
do so. By not doing so they run the risk of having students
maintain various types of ethnocentric biases that contrib-
ute to the perpetuation of divisive cultural and racial preju-
dices. All counselors should work to infuse a sense of re-
spect and appreciation for human diversity into the
worldviews of young people, thereby making tolerance more
likely in times of stress and heightened anxiety.
The potentials inherent in the increasing diversity of com-
munities and schools across the United States range from
horrible conflict to a genuine appreciation of the richness
and vitality of a multicultural society. Counselors can choose
to nourish the latter awareness and make its behavioral
manifestations more probable if we act intentionally to
develop the positive potentials that exist in our increas-
ingly diverse society and interdependent world. The
Multicultural Guidance Project is one such intentional act.
REFERENCES
Allport, G. (1982). The nature of prejudice (4th ed.). Reading, MA: Adios-
Wesley.
Conyne, R. K. (1987). Primary preventative counseling: Empowering people
and systems. Muncie, IN: Accelerated Development.
Coopersmith, S. (1981). Self-esteem inventories. Palo Alto, CA: Consult-
ing Psychology Press.
D’Andrea, M., & Daniels, J. (1995). Helping students learn to get along:
Assessing the effectiveness of a multicultural guidance project. El-
ementary School Guidance & Counseling, 30,143–154.
D’Andrea, M., & Daniels, J. (1996). Promoting peace in our schools:
Developmental, preventive, and multicultural considerations. The
School Counselor, 44, 55–64.
Gazda, G., Ginter, E., & Horne, A. (2001). Group counseling and group
psychotherapy. Boston: Allyn & Bacon.
Greenberg, J., Solomon, S., & Pyszczynski, T. (1997). Terror manage-
ment theory of self-esteem and cultural world views: Empirical as-
sessments and conceptual refinements. In M. P. Zanna (Ed.), Ad-
vances in experimental social psychology (pp. 61–139). San Diego, CA:
Academic Press.
Gresham, F. M., & Elliott, S. N. (1990). The Social Skills Rating System
(SSRS). Circle Pines, MN: American Guidance Service.
Omizo, M. M., & D’Andrea, M. (1995). Multicultural classroom guidance.
In C. C. Lee (Ed.), Counseling for diversity: A guide for school counselors
and related professionals (pp. 143–158). Boston: Allyn & Bacon.
Ponterotto, J. G., & Pedersen, P. (1993). Preventing prejudice: A guide for
counselors and educators. Newbury Park, CA: Sage.
Post, P., & Robinson, B. E. (1998). School-age children of alcoholics and
non-alcoholics: Their anxiety, self-esteem, and locus of control. Profes-
sional School Counseling, 5, 36–45.
Salzman, M. (1995). Attributional discrepancies and biases in cross-
cultural interactions. The Journal of Multicultural Counseling and
Development, 23, 181–193.
Selman, R. L., & Byrne, D. A. (1974). A structural-developmental
analysis of role-taking in middle childhood. Child Development, 45,
803–806.
Spatz, K., & Johnson, J. (1973). Internal consistency of the Coopersmith
Self-Esteem Inventory. Educational and Psychological Measurement, 33,
875–876.
Stephens, T. (1978). The Social Behavior Assessment Scale (SBA). Circle
Pines, MN: American Guidance Service.
Training AIDS and Anger Prevention Social Skills in
At-Risk Adolescents

Melbourne F. Hovell, Elaine J. Blumberg, Sandy Liles, Linda Powell, Theodore C. Morrison,
Gabriela Duran, Carol L. Sipan, Susan Burkham, and Norma Kelley

Youth in alternative schools often engage in high rates of risk behaviors for AIDS and violence. This study included 42 youth, ages
13 to 18, from alternative schools in San Diego, CA, and tested the effectiveness of behavioral skills training based on the
Behavioral-Ecological Model. Two interventions were evaluated: one teaching condom use skills and the other teaching anger
management skills. Changes in most skills were significant at postintervention but were not maintained at 6 months. Few risk-
related attitudes or behaviors improved at 6 months for either group.

A
mong adolescents, violence (Centers for Dis-
ease Control [CDC], 1992a, 1993, 1997; Na-
tional Institutes of Health [NIH], 1997; Se-
lect Committee on Children, Youth, and
Families, 1992) and sexually transmitted
diseases (STDs), including AIDS (Bureau of the Census,
1993; CDC, 1996; Children’s Defense Fund, 1994;
Fingerhut & Kleinman, 1990; National Center for Health
Statistics, 1993; Whitaker & Bastian, 1991), are major causes
of morbidity and mortality having behavioral determinants.
A significant proportion of adolescents engage in risk
behaviors that can lead to STD infection (CDC, 1992b,
1992c) or violent injury (Valois, McKeown, Garrison, &
Murray, 1995). These behaviors increase the opportunities
for transmission of disease or for physical injury and in-
clude multiple sexual partners, non-use of condoms, alco-
hol/substance use or abuse, fights, weapon carrying, and
gang membership. According to the 1995 Youth Risk Be-
havior Survey, more than half (53.1%) of students in Grades
9–12 have had sexual intercourse, and 17.8% reported
sexual intercourse with four or more partners during their
lifetime (CDC, 1996). Of sexually active students, only
54.4% reported condom use at last intercourse (CDC,
1996). In the 30 days prior to the survey, 20.0% of students
had carried a weapon (such as a gun, knife, or club); 7.6%
had carried a gun specifically. In the past year, 38.7% had
been in a physical fight (CDC, 1996).
Certain higher risk groups of adolescents are more likely
to demonstrate risk behaviors for HIV infection and vio-
lence than the general population of youth. The propor-
tions of youth engaging in AIDS risk or violent behaviors
are higher among youth in detention (Shafer et al., 1993),
those involved with the criminal justice system (Melchert
& Burnett, 1990), out-of-school youth (CDC, 1994a), and
youth in alternative school programs (Bjerregaard & Smith,
1993; CDC, 1994b; Grunbaum & Basen-Engquist, 1993).
In 1994, 5.3% of high school students enrolled the previ-
ous year dropped out (National Center for Educational
Statistics, 1997); over 2 million youth under the age of 18
were arrested in 1995 (Federal Bureau of Investigation,
1996); and over 100,000 youth resided in juvenile institu-
tions in 1990 (Bureau of the Census, 1994). Youth in alter-
native school settings represent a sizable portion of the
adolescent population. Nonetheless, few studies of risk re-
duction have been conducted among adolescents attending
alternative schools (Grunbaum & Basen-Engquist, 1993)
or those outside the mainstream school system (O’Hara,
Messick, Fichtner, & Parris, 1996). Limited data exist on
adolescents in alternative school settings in terms of their
receptivity to or lack of training regarding AIDS and anger
or violence prevention. Given statistics on risk behaviors in
these respective areas for these high-risk youth, it is par-
ticularly important to target such groups for risk-behavior
reduction training.

Melbourne F. Hovell is a professor and director, Elaine J. Blumberg is a senior research associate, Sandy Liles is a research associate, Gabriela Duran is
a research assistant, Carol L. Sipan is a senior research associate, and Norma Kelley is a data coordinator, all in the Center for Behavioral Epidemiology and
Community Health, Graduate School of Public Health, San Diego State University, San Diego, California. Linda Powell is a graduate student in the
Department of Human Development and Family Life, University of Kansas, Lawrence. Theodore C. Morrison is a graduate student at the University of
Alabama at Birmingham, School of Public Health, Department of Health Behavior. Susan Burkham is a special projects director/epidemiologist at the Bureau
of Information Resources, Health Care Financing, Texas Department of Health in Austin. This research was supported by grants from the California State
Office of AIDS (93-18444) and the University -Wide Aids Research Program, University of California (R98-SDSU-092 and IS99-SDSUF-206). The authors
thank the students, teachers, and other participating personnel from the San Diego County Juvenile Court and Community Schools for their contributions to
this work. Correspondence regarding this article should be sent to Melbourne F. Hovell, Center for Behavioral Epidemiology and Community Health, Graduate
School of Public Health, San Diego State University, 9245 Sky Park Court, Suite 230, San Diego, CA 92123 (e-mail: hovell@mail.sdsu.edu).

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 347

Hovell, Blumberg, Liles, Powell, Morrison, Duran, Sipan, Burkham, and Kelley
The Behavioral-Ecological Model (Hovell et al., 1994) and
related community applications of social learning theory
(Biglan & Hayes, 1996; Biglan et al., 1990) support the use
of behavioral skills training techniques for interventions tar-
geting violence-risk and AIDS-risk behaviors. These models
integrate learning theories with an ecological approach to
explain risk behaviors. The models assume a developmental
process of learning in which social competencies and ongo-
ing social influences, including reinforcement contingencies
and other motivational variables, are responsible for routinely
practiced behaviors. Although skills training studies have dem-
onstrated significant increases in condom use (Main et al., 1994;
O’Hara et al., 1996; Rotherum-Borus, Koopman, Haignere, &
Davies, 1991; Stanton et al., 1996; St. Lawrence et al., 1995)
and significant decreases in aggressive behaviors among
children and adolescents (Grossman et al., 1997; Larson, 1992),
the degree to which social skills are sufficient to alter risk prac-
tices remains unclear. Theoretically, one would assume that
social contingencies, including peer models (norms), influence
risk practices directly as well as through skills acquisition
(Catania, 1998; Hovell et al., 1994).
Comparative studies of approaches to aggressive behavior
have found that interventions targeting internal processes, such
as cognitions, have been less effective than those targeting
observable behaviors (e.g., Gudjonsson & Drinkwater, 1986;
Schneider, 1991). A recent meta-analysis also demonstrated
the relative success of social learning theory-based behavioral
interventions among diverse populations for HIV risk reduc-
tion (Kalichman, Carey, & Johnson, 1996).
The present study tests the effectiveness of behavioral
skills training methods in two separate, parallel interven-
tions—one teaching skills in condom use and negotiation
related to AIDS risk reduction, the other teaching anger
management skills related to violence risk reduction—in
youth placed in an alternative school by either the courts
or the mainstream school system. The purpose of this in-
vestigation was to determine whether skills training, with-
out formal contingencies directed to risk practices, increased
Anger/Aggression and HIV prevention skills, whether ac-
quired skills were sustained at the 6-month follow-up, and
whether changes in AIDS or violence-related risk behav-
iors occurred. This feasibility study sets the stage for larger
scale studies of skills training as well as informing the need
for skills plus contingency interventions.
METHOD
Participants
Adolescents ages 13 to 18 were recruited from four Juve-
nile Court and Community School (JCCS) sites in San
Diego County. Students who attend these sites live with
their parents, in foster care, or in group homes and attend
school during the day. Educational programs are designed
to meet the needs of students who are detained due to court
action, either as a delinquent or neglected ward of the court,
or placed by local school districts through the probation
department (Juvenile Court & Community Schools, 1995).
348 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Adolescents were recruited in the classroom after an oral
presentation detailing study requirements. All participants
as well as a parent of each participant under the age of 18
reviewed and signed informed consent forms. More than 40
adolescents (N = 42) from two sequential cohorts volun-
teered and received baseline interviews for the present study.
All descriptive statistics and analyses were based on the
consistent cohort of participants retained across the three
measurement periods (N = 29). The mean age of the sample
at pretest was 16.3 years; 28% were female. The majority
of study participants identified themselves as ethnic mi-
norities: 31% were African American, 38% were Latino, and
21% were White. There was one (3%) American Indian,
and 2 (7%) participants who did not classify themselves.
See Table 1 for additional demographic information and
baseline risk profiles.
Procedure
Design. This investigation implemented two interventions that
served as alternate controls for one another. Kellam and Rebox
(1992) have noted the advantage of testing two interventions
in a parallel design: “Such a dual intervention strategy allows
direction of effects among the variables to be tested along with
the specificity of effects of the interventions” (p. 172). How-
ever, because it was not possible to randomly assign students to
condition, this analysis remains quasi-experimental. Following
advice reported by Slymen and Hovell (1997) for cluster de-
signs and small samples, the unit of analysis was the individual.
Intervention. Trainings were simultaneously conducted for
the two conditions in sequential cohorts. Each condition
TABLE 1
Characteristics and Risks of Sample (N = 29)
Measure %
Participant’s family receiving government assistance 37
Participant’s family renting home 72
Sexually active (vaginal or anal sex), past 6 months 83
Lifetime history of 4 or more vaginal sexual partners 68
Never used a condom during vaginal sex, past 6 monthsa 25
Lifetime history of having had an STD 14
Knows someone with AIDS or who died from AIDS 21
Alcohol use, past 6 months 79
Marijuana use, past 6 months 66
Cocaine use, past 6 months 14
Intravenous drug use, past 6 months 0
Sex under the influence of alcohol, past 6 monthsa 42
Currently a gang member 31
Hurt someone, past 6 months
Using a gun 7
Using a knife 17
Using a stick or bottle 38
Using one’s body 52
Lifetime history of one or more arrests 76
Lifetime history of one or more convictions 59
Currently on probation 50
Note. Descriptive statistics are given for the consistent cohort of youth
who were retained for all measures.
a
Percentage based on the sexually active subsample (participants who
reported having vaginal or anal sex in the past 6 months, n = 24).

received the intervention program at widely separated JCCS
school sites, minimizing the chance of cross-contamination.
The goal of the HIV prevention intervention was to
increase competencies in four skills: condom negotiation,
discussing a friend’s risk of AIDS, condom application
technique, and teaching AIDS risk reduction skills to a friend.
The goal of the Anger Management intervention was to
increase competencies in four skills: negotiating an agree-
ment, resisting taunts, dealing with authority figures, and
teaching anger management skills to a friend. Skills were
chosen based on interventions previously used with nonde-
linquent Latino and Anglo adolescents (Blumberg et al., 1997;
Hovell et al., 1998).
Both groups were trained using the same 10-step behav-
ioral rehearsal training approach:
1. Presentation of skill and rationale
2. Explicit modeling by teacher and peer trainers
3. Directed role play practice in small groups
4. Instructor feedback
5. New modeling with corrections
6. Repeated role play with corrections
7. Pairs demonstration of skills in front of class
8. Instructor, peer trainer, and class feedback
9. Repeated role play with corrections
10. Group discussion
Graduate students and peer trainers who had received 8 hours
of training conducted twelve 1½- to 2-hour training sessions
over a 4-week period for each condition. Training for the
two conditions differed in content only (see Table 2).
During Sessions 2 through 6, a given skill was introduced,
discussed, and divided into specific objectives. The trainer
or peer trainer, or both, modeled incorrect (passive or ag-
gressive) and correct (assertive) forms of a given skill, and
guided students to practice skills during structured role play
exercises. Each participant role-played each skill repeatedly
with another student or the peer trainer serving as the per-
son who pressured the participant. During role plays, each
incorrect response by the participant was given immediate
verbal feedback and the correct response was modeled by
the trainer. Role-playing continued with each participant
until the trainer judged the performance satisfactory. In the
remaining sessions, time was set aside for continued role
play practice. Participants were encouraged to create sce-
narios with increasing degrees of difficulty. In Session 10,
they were encouraged to design their own risk reduction
skill and demonstrate teaching the skill to peers. In addi-
tion to verbal feedback and praise, raffle tickets for a weekly
prize of a $25 gift certificate were awarded for each ob-
served correct response or behavior. At the end of the in-
tervention, tickets from each weekly raffle were combined
for a grand prize drawing of a $100 gift certificate.
Variables
Skills measures. Videotaped role plays were used to assess
skills. Due to cost constraints, participants were measured
for only one HIV prevention skill (Condom Negotiation)
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Tr a i n i n g A I D S a n d A n g e r P r e v e n t i o n S o c i a l S k i l l s
and one anger management skill (Resisting Taunts). Role
play testing occurred at pretest, posttraining, and at 6-month
follow-up. The Condom Negotiation (CN) vignette assessed
the ability to request condom use and the ability to refuse
sexual intercourse if a partner refused to use condoms.
The Condom Negotiation Vignette: You and__________are at a
party. You both like each other and have been hanging out for a few
weeks. While at the party, you and________start making out. You
both want to have sex. You go off into one of the bedrooms. You
want to use a condom, but_________does not. Show what you
would do.
For anger management, the Resisting Taunts (RT) vignette
assessed the ability to resist taunts from an “authority fig-
ure” (confederate) while being incited to anger by a “friend”
(confederate), as well as the ability to exit the situation
without escalation.
The Resisting Taunts Vignette: You and your friend_________have
just walked into a store and are looking around to buy something.
The storekeeper, ____________, the brother or sister of a rival gang
member, keeps on looking at you and then comes up to you and
asks what you are doing in the store. Show what you would do.
Measurement role plays were videotaped at the JCCS
sites in rooms separate from the classroom. Each measure
was approximately 15 minutes in length. Role plays were
performed with an opposite-sex confederate for CN. RT
role plays were performed with two confederates, a same-
sex confederate as the store manager and an opposite-sex
confederate as a “friend.”
The setting for CN was two chairs side by side. For RT a
simulated store counter was placed between the store man-
ager confederate and the participant with his or her “friend.”
A research assistant gave instructions before each role play.
Each measurement session began with a practice role play
intended to put the participant at ease. CN was measured
first, followed by RT, and concluding with another nonsen-
sitive role play to diffuse aggressive feelings potentially
raised by the RT role play.
Confederates were undergraduate and graduate students
selected for their youthful appearance so that the measure-
ment scenario more closely resembled a real-life encounter
with participants’ peers. Confederates were provided 20
hours of training. In CN, they were trained to resist con-
dom use for five interchanges and to “give in” on the sixth
interchange. In RT, the storekeeper confederate was trained
to taunt the teen to anger within seven interchanges. The
confederate acting as the friend was trained to further pro-
voke the teen when the storekeeper’s taunts proved unsuc-
cessful. For example, the confederate might say, “C’mon
Mary, are you gonna let him talk to you like that?”
Coding. The videotape of each role play was consecu-
tively evaluated for Verbal Content, Nonverbal Behavior,
and overall levels of Aggression (RT skill only), Assertiveness,
and Anxiety. Coders reviewed videotapes repeatedly to score
the presence of each of these components of HIV risk re-
duction and anger resistance skills, respectively. Coders were
349
Hovell, Blumberg, Liles, Powell, Morrison, Duran, Sipan, Burkham, and Kelley

TABLE 2
Summary of Training Sessions

Session AIDS Risk Reduction Anger Reduction

1 Introduction; ground rules; focus group; role play process
2 Opening discussion; passive, aggressive, and assertive
behaviors; condom negotiation skill
3 Opening discussion; continued condom negotiation skill; correct
condom use technique
4 Opening discussion; introduce discussing a friend’s risk of AIDS skill
5 Opening discussion (focus on peer norms and skills
dissemination); introduce teaching condom application
technique
6 Opening discussion; introduce teaching AIDS risk reduction skills
to a friend
7 Opening discussion; how to get friends to think in terms of long-term
consequences; continued role play practice; film: Teen AIDS in
Focus; discussion
8 Opening discussion; peer training skills; discussion of long-term
consequences of sexual behavior (teen pregnancy); guest speakers:
teen mom & dad
9 Opening discussion; continued role play practice; Long-term
consequences of sexual behavior: AIDS; guest speaker: person
living with AIDS
10 Opening discussion; continued role play practice; view and critique
government ad campaign promoting safer sex; discussion: “How to
Get the Message Across”; participants design their own risk
reduction skill and demonstrate teaching
11 Opening discussion; continued role plays; STD film: A Million Teens
12 Values clarification exercise; presentation of teen-developed skills; Continued role play practice; closing focus group and
evaluation of training
Introduction; ground rules; focus group; role play process;
Video: Kids Killing Kids
Opening discussion; passive, aggressive, and assertive
behaviors; negotiating an agreement skill
Opening discussion; relaxation techniques; resisting taunts
skill
Opening discussion; continued resisting taunts skill
Opening discussion; Introduce dealing with authority
figures
Opening discussion; introduce teaching anger management
skills to a friend; discussion of importance of skill
dissemination
Opening discussion; understand the effects of violence on
others; continued role play practice
Opening discussion; continued role play practice; discussion
of long-term consequences of behavior on others; interview
with a panel of “victims” (Bureau of Victims)
Opening discussion; continued role play practice; discussion
of long-term consequences of violent behavior (jails,
institutions, and death or dismemberment) and “How to
Change”; interview with former gang member
Opening discussion; continued role play practice; discussion
of long-term consequences of violent behavior; interview
with ex-offenders; participants design their own risk
reduction skill and demonstrate teaching
Opening discussion; continued role play practice; discuss
behavioral expectations of employers; guest speaker: Job
Corp.
evaluation

Note. Bold text indicates skills that were taught in various sessions.

blind to condition. Components of verbal behavior devel-
oped for the CN skill were distinct from those developed
for the RT skill. In CN, coding for verbal behavior was
broken down into measures for Content and Style. The
Content score assessed levels of negotiating skill using a
10-point scale ranging from 0 (giving in to sex without a
condom) to 9 (both requesting condom use and strongly say-
ing no to sex without a condom). Style score assessed level
of social communication skill using a 10-point scale rang-
ing from 0 (not admitting to responsibility) to 9 (both taking
responsibility and acknowledging a partner’s feelings). In RT,
codings for verbal behavior were broken down into measures
for appropriate and inappropriate responses. The sum of
positive attributes in the vignette tallied by the coder comprised
the score for positive responses, such as attempting to diffuse
the situation, remaining assertive in the situation, or asser-
tively stating one’s rights as a customer. The score for nega-
tive responses was the sum of negative attributes tallied by
the coder, such as interrupting, insulting, cussing, or intimi-
dating the storekeeper confederate. For Nonverbal Behavior,
coders alternated 15-second intervals of observing the
participant’s behavior with 15 seconds of coding occurrence
or nonoccurrence of 15-component Nonverbal Behaviors,
such as eye contact and body posture. For overall levels of
Aggression, Assertiveness, and Anxiety, coders judged the
degree to which each characteristic represented low to high
Aggression, Assertiveness, and Anxiety, using a 1-to-5 scale
in increments of 0.5.
Reliability. The videotapes for each participant were coded
independently by two coders. A criterion-referenced stan-
dard of 80% agreement was used. Reliability between cod-
ers for Verbal and Nonverbal Behaviors was determined for
each measure by calculating a point-to-point percent agree-
ment. Reliability of the three overall measures for the RT
skill—Aggression, Assertiveness, and Anxiety—were deter-
mined by calculating percentage agreement for the three
scales combined; reliability for the two overall measures
for the CN skill—Assertiveness and Anxiety—were deter-
mined by calculating percentage agreement for the two
scales combined. For overall measures, agreements were
defined as ratings within + .5 between coders. All coding
that failed to reach 80% agreement was recoded by one of
the original coders and a third coder independently but
simultaneously to avoid any variation in timing of video-
tape viewing. Despite the 80% criterion, final reliability
coefficients always exceeded this standard. In all cases,
mean reliability percentages were greater than 90, and in
several instances reached 100%.

350 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79


Interview measures. Participants were administered a ques-
tionnaire at pretest and at 6-month follow-up, conducted
by trained interviewers in private settings at JCCS sites.
For those students no longer attending JCCS at the time of
the 6-month follow-up, the interview was conducted at the
research offices. Each interview took approximately 45
minutes and assessed demographics, drug use history, sexual
history (including history of condom use and history of sex
under the influence of alcohol/drugs), attitudes and knowl-
edge pertaining to AIDS, and other variables related to be-
havioral risk for HIV infection. The instrument also assessed
types of situations and people that elicit anger responses;
type and frequency of participants’ responses to those situ-
ations; type of responses to various authority figures (e.g.,
police officers); peer influences on prosocial and antisocial
behaviors; frequency of verbally or physically aggressive
behaviors; frequency of weapon use; and history of arrest,
conviction, and incarceration. In the anger domain, several
scaled variables were created. Questions about arrest, convic-
tion, and detention were combined into a Crime scale (α =
.68). Harm was a composite of seven questions of how often a
participant had used various sorts of weapons, ranging from
hands to guns, to hurt another person (α = .64). An Argue scale
summed four items assessing instances of various sorts of pro-
vocative speech, such as cussing and yelling (α = .76). Actout
comprised eight measures of verbally or physically aggressive
behaviors, such as shoves, slaps, or threats of injury (α = .66).
RESULTS
All data were analyzed on a Pentium-90 PC using SPSS for
Windows 95, Version 6.1.3 (SPSS, 1995). One-way
ANOVAs or t tests (for interval level variables) and chi-
square tests or Fisher’s exact tests (for dichotomous vari-
ables) were used to test for differences on pretest variables
by retention status, experimental condition, gender, and
ethnicity. All tests were two-tailed.
For the skills measures, separate 2 × 2 (group-by-time)
repeated measures analyses of variance (ANOVAs) were
conducted for the pretest to posttest and the posttest to
follow-up periods. This was done to more clearly distin-
guish the effects due to training in the 4-week intervention
period (pretest to posttest) from the maintenance of ef-
fects during the 6-month follow-up period (posttest to fol-
low-up). As a check on possible influences of pretest group
differences on the results in the pre-to-post period, analy-
ses of covariance (ANCOVAs) were run on the posttest
measures using the pretest scores as covariates. Finally, one-
way ANOVAs were conducted on the follow-up skills mea-
sures to ascertain posttraining effects.
Repeated measures analyses of variance, using 2 × 2 (group-
by-time) ANOVAs, were conducted on interview data (pre-
test to follow-up), supplemented by ANCOVAs run on the
follow-up measures, using pretest scores as covariates.
Cohort retention. Pretest measures (both interview and
role play) were completed by 42 youth. Attrition occurred
for several reasons: incarceration in Juvenile Hall, reloca-
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Tr a i n i n g A I D S a n d A n g e r P r e v e n t i o n S o c i a l S k i l l s
tion out of San Diego County, running away from home,
and evasion of authorities (i.e., hiding to elude warrant for
arrest). In addition, some youth elected not to complete
selected skill measures. As a result, skill and interview data
were obtained for a consistent cohort of 29 participants
(69% of original sample) across the three measurement
periods. All descriptive and analytical statistics were based
on this sample (N = 29).
Preliminary Analyses
A series of t tests on demographic and baseline risk factors
were used to compare the 13 excluded teens who failed to
complete all parts of the study and the 29 teens retained
for all measures. Significant differences were found for four
behaviors during the past 6 months: times used marijuana,
times friends encouraged you to fight, times damaged prop-
erty, and “Harm” (a composite measure of the frequency of
physically injuring another person). The excluded teens
were more likely to exhibit each of these behaviors. Thus
the 29 participants on whom the descriptive and analytical
statistics were based may represent a less extreme risk pro-
file than is typical of the JCCS population.
Risk conditions. Nonetheless, large proportions of teens
in the sample reported at-risk behavior on a number of mea-
sures. Sixty-eight percent had had at least four vaginal sex
partners. Seventy-nine percent had used alcohol in the past
6 months. Seventy-six percent had been arrested. Other
indicators of risk are given in Table 1.
Group comparability. The AIDS (n = 14) and anger (n =
15) groups were compared at pretest to assess potential
differences on demographics, skill levels, AIDS-related risk
behaviors and attitudes, and violence-related risk behaviors
and attitudes. There was only one significant difference at
pretest: 21% of the AIDS group and 57% of the anger group
reported a history of HIV antibody testing (Fisher’s exact
test, two-tailed, p = .02).
Main Analyses
Role play skill change—repeated measures MANOVAs. For
both the AIDS prevention and the anger management in-
terventions, separate Group × Time repeated measures
ANOVAs were used to analyze changes in skill over time
within each of two time periods—pretest to posttest, and
posttest to 6-month follow-up. Analyses were conducted
on five measures (Style, Content, Assertiveness, Anxiety,
and Nonverbal) of the CN skill and on six measures (Nega-
tive Responses, Positive Responses, Aggression,
Assertiveness, Anxiety, and Nonverbal) of the RT skill. To
rule out possible confounding due to differences at baseline,
ANCOVAs, using baseline scores as covariates, were used
as secondary analyses. In general, ANCOVAs confirmed all
ANOVA findings for both the AIDS-related training and
the anger-related training outcomes; the single exception is
noted in the Condom Negotiation results description in the
following section. With a sample size of 29 youth, results
should be interpreted based on both the consistency across
351
Hovell, Blumberg, Liles, Powell, Morrison, Duran, Sipan, Burkham, and Kelley

measures and the experimental conditions as illustrated in Mean Content scores increased from pretest to posttest
Table 3, as well as by formal significance testing. for the AIDS group and decreased for the anger group. This
Condom Negotiation: AIDS prevention training versus an- differential change did not reach significance. No signifi-
ger management controls. A significant group-by-time (pre/ cant effects were found for Content scores for the posttest
posttest) interaction was found for the Condom Negotia- to 6-month follow-up period.
tion Style score, F(1, 27) = 4.50, p = .04. Mean Style scores Assertiveness for the AIDS group increased from pretest
increased dramatically for the AIDS group and only slightly to posttest, while remaining essentially unchanged for the
for the anger group. However, ANCOVA results were non- anger group, and this differential reached significance, F(1,
significant, suggesting that differential change was not sus- 27) = 5.82, p = .02. There was also a significant group-by-
tained after adjustment for baseline differences in Style. time interaction from posttest to 6-month follow-up, F(1,
During the posttest to 6-month follow-up period, Style 27) = 4.78, p = .04. Assertiveness decreased for the AIDS
scores for the AIDS group declined, while those for the group and increased for the anger group.
anger group remained essentially unchanged. The posttest A significant, F(1, 27) = 8.19, p = .01, group-by-time (pre/
to 6-month follow-up changes did not reach significance. posttest) interaction was found for Anxiety. Anxiety for the
AIDS group decreased from pretest to posttest, while mean
Anxiety scores for the anger group increased. The posttest
TABLE 3 to 6-month follow-up changes did not reach significance.
There were no significant effects for the Condom Nego-
Change in Condom Negotiation Skills and tiation Nonverbal score for pretest to posttest or for posttest
Resisting Taunts Skills by Groupa and Time to follow-up measures.
Resisting Taunts: Anger management training versus AIDS
Time of Measure prevention controls. A significant, F(1, 27) = 11.91, p = .01,
Pretest Posttest 6-mo. Follow-up group-by-time (pre/posttest) interaction was found for the
Skill Component
and Group M SD M SD M SD Negative Responses score. Mean Negative Responses scores
decreased from pretest to posttest for the anger group and
Condom Negotiation
(CN) Skills
increased for the AIDS group. Negative Responses scores
Style for the posttest to 6-month follow-up period increased for
AIDS 0.43 1.34 2.62 2.02 1.86 2.14 the anger group, while they decreased for the AIDS group.
Anger (control) 1.33 1.98 1.67 1.99 1.60 2.69 This differential change did not reach significance, F(1, 27)
Content
AIDS 4.93 3.22 6.15 2.94 6.07 2.27 = 3.96, p = .06.
Anger (control) 4.60 3.22 3.87 3.14 5.33 2.64 A significant, F(1, 27) = 6.03, p = .02, group-by-time (pre/
Assertiveness posttest) interaction was found for the Positive Responses
AIDS 2.89 0.72 3.65 0.80 3.46 0.60 score. Positive Responses increased for the anger group from
Anger (control) 2.77 0.94 2.73 0.75 3.27 0.56
Anxiety pretest to posttest, while Positive Responses for the AIDS
AIDS 2.21 0.72 1.85 0.75 1.75 0.58 group decreased. A significant, F(1, 27) = 4.72, p = .04, group-
Anger (control) 1.83 0.41 2.00 0.33 1.77 0.32 by-time interaction for this measure also was found for the
Nonverbal
AIDS n/a b n/ab 2.85 0.69 2.79 0.43
posttest to 6-month follow-up period. Positive Responses
Anger (control) 2.73 0.59 2.80 0.41 2.67 0.72 decreased for the anger group, while Positive Responses for
Resisting Taunts the AIDS group increased.
(RT) Skills A significant, F(1, 27) = 10.54, p = .01, group-by-time (pre/
Negative Responses
AIDS (control) 1.43 1.99 3.77 3.09 3.00 2.80 posttest) interaction was found for the Aggression score. The
Anger 2.73 2.25 1.93 2.15 3.87 3.23 mean score decreased slightly from pretest to posttest for
Positive Responses the anger group, while the mean score increased for the AIDS
AIDS (control) 4.14 3.46 3.15 3.65 3.64 2.41
Anger 3.13 1.77 5.07 3.43 3.07 2.49
group. No significant effects were found for Aggression scores
Aggression for the posttest to 6-month follow-up period.
AIDS (control) 1.89 1.15 2.85 1.36 2.93 1.60 The group-by-time interaction for Assertiveness for the
Anger 2.60 1.30 2.33 1.29 2.90 1.71 pretest to posttest period was significant, F(1, 27) = 4.75,
Assertiveness
AIDS (control) 2.64 0.82 2.46 0.92 3.00 1.02 p = .04. The mean Assertiveness score for the anger group
Anger 2.73 0.80 3.17 1.03 2.80 1.18 increased from pretest to posttest, while the mean score
Anxiety for the AIDS group decreased. The group-by-time interac-
AIDS (control) 2.00 0.55 1.85 0.72 1.68 0.37 tion for Assertiveness for the posttest to 6-month follow-up
Anger 1.80 0.37 1.70 0.32 1.83 0.36
Nonverbal period was also significant, F(1, 27) = 4.64, p = .04. The mean
AIDS (control) 3.21 0.80 2.85 0.80 3.36 1.74 Assertiveness score for the anger group declined while that
Anger 3.13 0.64 3.27 0.70 3.13 0.35 for the AIDS group increased.
a
AIDS group, n = 14; anger group, n = 15. bUnable to compute M The group-by-time effect for the Resisting Taunts Anxiety
and SD because the CN Nonverbal measure had no variance at score in the pretest to posttest period was not significant.
pretest for the AIDS condition. The group-by-time interaction for this measure in the

352 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79


posttest to 6-month follow-up period was also not signifi-
cant, F(1, 27) = 3.15, p = .09. Mean Anxiety scores for the
anger group increased, returning to pretest level, while
scores for the AIDS group continued to decline.
There were no significant effects for pretest to posttest
or for the posttest to 6-month follow-up time period for
Resisting Taunts Nonverbal scores.
Behavior/Attitude Change. Self-report items from the
behavior/attitudes interview were used to assess changes
from pretest to 6-month follow-up. Using repeated mea-
sures ANOVAs and confirmed by ANCOVA, significant
group-by-time effects were found for four variables. Agree-
ing with the statement, “Using a condom is more trouble
than it’s worth,” F(1, 26) = 6.08, p = .02, decreased for the
AIDS group and increased for the anger group (5-point
ordinal scale ranging from strongly disagree = 0 to strongly
agree = 4). Having used alcohol in the past 6 months, F(1,
27) = 6.67, p = .02, and having used marijuana in the past 6
months, F(1, 27) = 4.44 p = .04, decreased for the AIDS
group and increased for the anger group (7-point ordinal
scale ranging from not at all = 0 to daily = 6). Complying
with requests by authority figures without arguing or fight-
ing back in the past week, F(1, 23) = 4.29, p = .05, decreased
for the anger group and increased for the AIDS group, con-
trary to expectation (5-point ordinal scale ranging from never
= 1 to always = 5). Changes by group were also examined
for all other behavioral or attitudinal risks related to AIDS
and violence, including frequency of vaginal or anal sex;
number of sex partners; pregnancy; condom use; sex under
the influence of alcohol and other drugs; sex with high-risk
partners; discussion with peers about their AIDS-related
risk behavior; arrest, incarceration, or probation history; gang
membership; weapon use; and verbally and physically aggres-
sive acts. There were no significant group-by-time effects
for these measures.
DISCUSSION
Despite obstacles to recruitment, training, measurement, and
follow-up typical for high-risk adolescents, the encouraging
participation rate in this study demonstrates the feasibility
of conducting this type of intervention in an alternative school
setting. Given the pronounced difficulty of carrying out in-
terventions within this population, the outcome is notably
encouraging and suggests that similar skills-training interven-
tions could be successfully implemented among other groups
of high-risk youth, such as runaways (Rotherum-Borus et
al., 1991) or the economically disadvantaged (DiClemente
& Wingood, 1995).
However, given the volunteer nature and subsequent at-
trition of the sample, and the small sample size, the results
should be cautiously interpreted as indicative of the gen-
eral population of alternative school students, particularly
those who are most severely at risk for AIDS or violence.
The attrition rate also suggests that more attention to in-
centives aimed at retaining high-risk students in training
would be important for future studies or services.
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Tr a i n i n g A I D S a n d A n g e r P r e v e n t i o n S o c i a l S k i l l s
For both skills targeted and the dimensions assessed, re-
sults tended to be consistent across conditions. Measures of
Condom Negotiation Style and Assertiveness increased and
Anxiety decreased after HIV-prevention social skills train-
ing, relative to controls. Similarly, an increase in Positive
Reactions, a decrease in Negative Reactions to Taunts, a de-
crease in Aggression and an increase in overall Assertiveness
followed anger management social skills training, relative to
controls. Most of these tests proved significant, even in the
context of a relatively small sample size and limited statis-
tical power. The consistency of effects across dimensions
within condition and the replication of effects across groups
suggest that role playing procedures can be effective for
increasing both HIV and anger prevention social skills for
delinquent adolescents.
The results showed essentially no generalization across
conditions for the reciprocal skills and dimensions. With
the possible exception of anxiety, which tended to decrease
in both groups (although not significantly with the anger
group), skill taught in one group did not improve in the
alternate group at posttest. Similarly, change was minimal
or nonexistent for specific dimensions, such as Anxiety and
Nonverbal Responses, respectively, which were the subset
dimensions of the target skills given the least amount of
explicit training. These findings add discriminant validity
for the attribution of response-specific change due to train-
ing. These results in total suggest that training increased
target skills; they also suggest that increasing assertiveness
for one or a few specific skills will not reliably produce a
general increase in assertiveness for all social skills. Whether
such a generalized assertiveness repertoire can be established
without training component skills individually remains to
be determined.
These results also show that almost all skills trained were
not sustained during the 6-month follow-up period. This
suggests that training needs to be continued, skills need to
be increased to still higher levels, or some other system of
support must be in place to sustain training effects. This is
consistent with the Behavioral-Ecological Model; without
ongoing practice and feedback and changes in culture such as
peer norms, obtained skill level can be expected to decrease.
Educational interventions may need to be sustained to main-
tain achievements and contribute to HIV or anger prevention.
Interview measures taken at 6-month follow-up were
examined with respect to changes in risk practices or atti-
tudes. Although three of these variables changed in the
expected direction, one changed in a contrary direction,
and the majority of variables in both interventions did not
change significantly. A reasonable and conservative inter-
pretation of these results is that neither type of training
was associated with reliable change in behavior. This is con-
sistent with our inability to sustain skills for the 6-month
follow-up. It seems unlikely that temporary skill improve-
ments would lead to changes in risk practices or attitudes.
Training a few skills in a short time may be insufficient
to increase adolescents’ frequent use of these skills and
ultimately reduce their risk practices. Future studies will
353
Hovell, Blumberg, Liles, Powell, Morrison, Duran, Sipan, Burkham, and Kelley
need to assess conditions required to sustain skills and
their application in order to test their effects on risk be-
havior. On the basis of the Behavioral-Ecological Model,
we presume that skills may be prerequisite but not suffi-
cient to change risk practices. In other words, condom use
may be dependent not only on having the requisite skills
for correctly using a condom and convincing a partner to
do so but on other factors including social support to apply
those skills. The same can be assumed for practicing anger
management skills. Targeting peer norms, as suggested by
Winslow, Franzini, and Hwang (1992), may be necessary to
affect motivation sufficiently to change behavior. The use
of a “booster” intervention, such as that used by Levy et al.
(1995), may be required to sustain those changes.
The results from this study need be interpreted with caution
for a number of reasons: inability to use random assignment,
small sample size, limited training time, and the potential for
unreliable self-reported information concerning risk practices.
Technically, this study is quasi-experimental, in absence of
random assignment. However, examination of group similari-
ties and the use of alternate interventions for different skills,
with replication of training effects, provide strong evidence that
effects were real. The risk of confounding due to uncontrolled
alternate independent variables seems limited.
Perhaps the most damaging limitations for this study are
limited statistical power due to small sample size and lim-
ited effects due to limited training time. These limitations
tend to bias the study in the direction of null outcomes. For
the few skills or dimensions that did not reach significance,
these null outcomes could have been true null effects of
training or artifacts of limited power or effect size. In fact,
having attained significant results posttraining for most vari-
ables evaluated in the context of limited power strongly
suggests that more complete training for larger samples
could produce even more substantial changes in skills. Re-
sults support the replication of these procedures under con-
ditions of greater support and more complete control.
For assessing the risk practices of adolescents, especially
youth with a history of less than complete honesty, self-
reports of criminal activity, or sexually sanctioned practices,
or both should be considered very cautiously. Future studies
of high-risk youth should include objective outcomes in
order to validate possible prevention effects. These might
include assessment of arrests, physical examinations of injury
due to violence, pregnancy, sexually transmitted disease
infections, and so on, as outcome indicators of skills train-
ing or other behavioral interventions.
Given the continued rise in AIDS cases among youth and
young adults and the continued epidemic of violence and
criminal activity among low income youth, additional study
and the identification of effective preventive interventions
is urgently needed. Although only representative of a small
part of the needed research and development process, the
present study suggests that role play technology may con-
tribute to skill acquisition for preventing HIV infection and
violence among delinquent and high-risk youth. The present
study also has raised serious concerns about the factors
354 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
needed to sustain skills and whether skills training alone is
sufficient to change risk practices.
REFERENCES
Biglan, A., & Hayes, S. C. (1996). Should the behavioral sciences become
more pragmatic? The case for functional contextualism in research on
human behavior. Applied and Preventive Psychology, 5(1), 47–57.
Biglan, A., Metzler, C. W., Wirt, R., Ary, D., Noell, J., Ochs, L., French,
C., & Hood, D. (1990). Social and behavioral factors associated with
high-risk sexual behavior among adolescents. Journal of Behavioral
Medicine, 13, 245–262.
Bjerregaard, B., & Smith, C. (1993). Gender differences in gang partici-
pation, delinquency, and substance use. Journal of Quantitative Crimi-
nology, 9, 329–355.
Blumberg, E. J., Hovell, M. F., Werner, C. A., Kelley, N. J., Sipan, C. L.,
Burkham-Kreitner, S. M., & Hofstetter, C. R. (1997). Evaluating AIDS-
related social skills in Anglo and Latino adolescents: Focus on assess-
ment. Behavior Modification, 21, 281–307.
Bureau of the Census. (1993). Statistical abstract of the United States
(113th ed.). Washington, DC: U.S. Department of Commerce, Eco-
nomics and Statistics Administration.
Bureau of the Census. (1994). Statistical abstract of the United States
(114th ed.). Washington, DC: U.S. Department of Commerce, Eco-
nomics and Statistics Administration.
Catania, A. C. (1998). Learning (4th ed.). Upper Saddle River, NJ: Prentice Hall.
Centers for Disease Control. (1992a). Preventing risk behaviors among
students, HIV/AIDS Prevention, Morbidity and Mortality Weekly Re-
port, 3, 1–2.
Centers for Disease Control. (1992b). Selected behaviors that increase
risk for HIV infection among high school students—United States, 1990.
Morbidity and Mortality Weekly Report, 41, 231–240.
Centers for Disease Control. (1992c). Sexual behavior among high school
students—United States, 1990. Morbidity and Mortality Weekly Report,
40, 885–888.
Centers for Disease Control. (1993). 1993 Sexually transmitted disease
guidelines. Morbidity and Mortality Weekly Report, 42(RR-14), 9.
Centers for Disease Control. (1994a). Health risk behaviors among ado-
lescents who do and do not attend school—United States, 1992. Mor-
bidity and Mortality Weekly Report, 43, 129–132.
Centers for Disease Control. (1994b). Sexual behaviors and drug use among
youth in dropout-prevention programs—Miami, 1994. Morbidity and
Mortality Weekly Report, 43, 873–876.
Centers for Disease Control. (1996). Youth risk behavior surveillance—United
States, 1995. Morbidity and Mortality Weekly Report, 45, No. SS-4.
Centers for Disease Control. (1997). HIV/AIDS surveillance report, Morbid-
ity and Mortality Weekly Report, 9(1), 13.
Children’s Defense Fund. (1994). The state of America’s children: Yearbook
1994. Washington, DC: Author.
DiClemente, R. J., & Wingood, G. M. (1995). A randomized controlled trial
of an HIV sexual risk-reduction intervention for young African-American
women. Journal of the American Medical Association, 274, 1271–1276.
Federal Bureau of Investigation, U.S. Department of Justice. (1996). Uni-
form Crime Reports for the United States, 1995. Washington, DC: Author.
Fingerhut, L. A., & Kleinman, J. C. (1990). International and interstate
comparison of homicide among young males. JAMA, 263, 3292–3295.
Grossman, D. C., Neckerman, H. J., Koepsell, T. D., Liu, P. Y., Asher,
K. N., Beland, K., Frey, K., & Rivara, F. P. (1997). Effectiveness of a
violence prevention curriculum among children in elementary school.
A randomized controlled trial. JAMA, 277, 1605–1611.
Grunbaum, J. A., & Basen-Engquist, K. (1993). Comparison of health
risk behaviors between students of a regular school and students in an
alternative high school. Journal of School Health, 63, 421–425.
Gudjonsson, G. H., & Drinkwater, J. (1986). Intervention techniques for
violent behavior. Issues in Criminological and Legal Psychology, 9, 37–47.
Hovell, M., Blumberg, E., Sipan, C., Hofstetter, C. R., Kreitner, S., Atkins, C.,
& Felice, M. (1998). Skills training for pregnancy and AIDS prevention in
Anglo and Latino youth. Journal of Adolescent Health, 23, 139–149.

Hovell, M. F., Hillman, E. R., Blumberg, E., Sipan, C. L., Atkins, C. J.,
Hofstetter, C. R., & Myers, A. (1994). A behavioral-ecological model
of adolescent sexual development: A template for AIDS prevention.
Journal of Sex Research, 31(4), 1–15.
Juvenile Court & Community Schools. (1995). School accountability re-
port card. San Diego, CA: San Diego County Office of Education.
Kalichman, S. C., Carey, M. P., & Johnson, B. T. (1996). Prevention of sexu-
ally transmitted HIV infection: a meta-analytic review of the behavioral
outcome literature. Annals of Behavioral Medicine, 18(1), 6–15.
Kellam, S. G., & Rebox, G. W. (1992). Building developmental and etio-
logical theory through epidemiologically based preventive interven-
tion trials. In J. McCord & R. E. Tremblay (Eds.), Preventing antisocial
behavior: Interventions from birth through adolescence (pp. 162–195).
New York: Guilford Press.
Larson, J. D. (1992). Anger and aggression management techniques
through the Think First curriculum. Journal of Offender Rehabilita-
tion, 18(1–2), 101–117.
Levy, S. R., Perhats, C., Weeks, K., Handler, A. S., Zhu, C., & Flay, B. R.
(1995). Impact of a school-based AIDS prevention program on risk
and protective behavior for newly sexually active students. Journal of
School Health, 65, 145–151.
Main, D. S., Iverson, D. C., McGloin, J., Banspach, S. W., Collins, J. L.,
Rugg, D. L., & Kolbe, L. J. (1994). Preventing HIV infection among
adolescents: Evaluation of a school-based education program. Preven-
tive Medicine, 23, 409–417.
Melchert, T., & Burnett, K. F. (1990). Attitudes, knowledge, and
sexual behavior of high-risk adolescents: Implications for coun-
seling and sexuality education. Journal of Counseling & Develop-
ment, 68, 293–298.
National Center for Educational Statistics. (1997). The condition of edu-
cation 1996 (NCES 96-304). Washington, DC: U.S. Department of
Education, Office of Educational Research and Improvement/U.S.
Government Printing Office.
National Center for Health Statistics. (1993). Advance report of final
mortality statistics, 1991. (Monthly Vital Statistics Report, 42[2,
Suppl.]). Hyattsville, MD: Public Health Service.
National Institutes of Health. (1997). NIH consensus statement: February
11–13. Bethesda, MD: Department of Health and Human Services.
O’Hara, P., Messick, B. J., Fichtner, R. R., & Parris, D. (1996). A peer-led
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Tr a i n i n g A I D S a n d A n g e r P r e v e n t i o n S o c i a l S k i l l s
AIDS prevention program for students in an alternative school. Jour-
nal of School Health, 66, 176–182.
Rotherum-Borus, M. J., Koopman, C., Haignere, C., & Davies, M. (1991).
Reducing HIV sexual risk behaviors among runaway adolescents. Jour-
nal of the American Medical Association, 226, 1237–1241.
Schneider, B. H. (1991). A comparison of skill-building and desensitiza-
tion strategies for intervention with aggressive children. Aggressive
Behavior, 17, 301–311.
Select Committee on Children, Youth, and Families. (1992). A decade of
denial: Teens and AIDS in America. Washington, DC: U.S. Govern-
ment Printing Office. (House of Representatives, 102nd Congress,
ISBN 0-16-039006-0)
Shafer, M. A., Hilton, J. F., Ekstrand, M., Keogh, J., Gee, L., DiGiorgio-
Hag, L., Shalwitz, J., & Schachter, J. (1993). Relationship between
drug use and sexual behaviors and the occurrence of sexually trans-
mitted diseases among high-risk male youth. Sexually Transmitted
Diseases, 20, 307–313.
Slymen, D. J., & Hovell, M. F. (1997). Cluster versus individual random-
ization in adolescent tobacco and alcohol studies: A basis for design
decisions. International Journal of Epidemiology, 26, 766–771.
SPSS for Windows (Version 6.1.3) [Computer software]. (1995). Chi-
cago: SPSS Inc.
Stanton, B. F., Li, X., Ricardo, I., Galbraith, J., Feigelman, S., & Kaljee, L.
(1996). A randomized, controlled effectiveness trial of an AIDS pre-
vention program for low-income African-American youths. Archives of
Pediatrics and Adolescent Medicine, 150, 363–372.
St. Lawrence, J. S., Brasfield, T. L., Jefferson, K. W., Alleyne, E., O’Bannon,
R. E., III, & Shirley, A. (1995). Cognitive-behavioral intervention to
reduce African American adolescents’ risk for HIV infection. Journal
of Consulting and Clinical Psychology, 63, 221–237.
Valois, R. F., McKeown, R. E., Garrison, C. Z., & Murray, M. L. (1995).
Correlates of aggressive and violent behaviors among public high school
adolescents. Journal of Adolescent Health, 16, 26–34.
Whitaker, C., & Bastian, L. (1991). Teenage victims: A national crime sur-
vey report. Washington, DC: U.S. Department of Justice, Bureau of
Justice Statistics (Publication NCJ-128129).
Winslow, R. W., Franzini, L. R., & Hwang, J. (1992). Perceived peer norms,
casual sex, and AIDS risk prevention. Journal of Applied Social Psy-
chology, 22, 1809–1827.
355
Adolescent Strategies for Coping With Cultural Diversity

Hardin L. K. Coleman, Sherry B. Casali, and Bruce E. Wampold

This investigation tested H. L. K. Coleman’s (1995b) hypotheses that the strategies adolescents use to cope with cultural diversity
will be organized in a sequential manner and that adolescents will use different strategies depending on the situation. To test
these hypotheses, the authors had 398 adolescents rate the likelihood of using 6 strategies for coping with cultural diversity:
Separation, Assimilation, Acculturation, Alternation, Integration, and Fusion. The authors make suggestions for future research ,
and implications for counseling are suggested.

A
group of adolescent African American girls,
while sharing their feelings about “wannabees”
(G. Lindsay, personal communication, March
31, 1995), were clear in their absolute disdain
for those who tried to be something they were
not. They did not like African Americans who acted “White”
with European Americans and “Black” with African Ameri-
cans. They did not like European Americans who tried to
talk or act “Black” when they were with African Americans.
After describing several situations in which adolescents
could be considered “wannabees,” these girls concluded that
they respected people who acted like themselves no mat-
ter where they were. The European Americans they liked
were ones who used their own vernacular when spending
time with others from different cultures. What these girls
were describing are some strategies individuals use when
they have contact with members of a different culture.
What an individual does when he or she comes into con-
tact with a second culture is assumed to have an effect on his
or her sense of emotional well-being (LaFromboise, Coleman,
& Gerton, 1993; Padilla, 1980). It has been hypothesized
that how one copes with that contact will affect such
socioemotional factors as self-esteem (e.g., Phinney, 1991),
social adjustment and academic performance (e.g., Gomez
& Fassinger, 1994; Oliver, Rodriguez, & Mickelson, 1985),
counseling process and outcome (e.g., Atkinson, Casas, &
Abreu, 1992; Coleman, Wampold, & Casali, 1995), and mental
health (e.g., Rogler, Cortes, & Malgady, 1991). Furthermore,
Coleman (1995a, 1995b) has hypothesized that the strategies
an individual uses to cope with second culture contact will
have an effect on that individual’s academic or job per-
formance, sense of social competence, and psychological
well-being. This article briefly reviews the ways in which
the process of second culture contact has historically been
described and the conceptualizations used to explain this
process and then presents a model that integrates previous
theories concerning second culture contact. We then present
an attempt to empirically validate that model.
Berry, Poortinga, Segall, and Dasen (1992), Coleman
(1995b), and LaFromboise et al. (1993) have observed that
the process of coping with second culture contact has gen-
erally been called “acculturation.” LaFromboise et al. pointed
out that the term acculturation has been used to describe
both a particular strategy for coping with second culture
contact and the process for coping with second culture con-
tact. LaFromboise et al. suggested that the phrase second
culture acquisition be used to describe the process of coping
with second culture contact and reserve the term accul-
turation for a way to explain or describe the process. This
convention is used throughout the article.
In a review of the psychological impact of biculturalism,
LaFromboise et al. (1993) found six ways that have been
used to describe the process of second culture acquisition.
These six descriptions represent different conceptualizations
of what happens when individuals come into contact with
people from a different culture. Three of the descriptions—
assimilation, acculturation, and separation—reflect a tradi-
tional assumption (Stonequist, 1935, 1937) that the process
of second culture acquisition is essentially linear: individuals
either let go of their culture of origin and join the second
culture or they remove themselves from contact with the
second culture. A person who assimilates or acculturates is
one who attempts to join the second culture. A person who
separates is one who withdraws from, or avoids contact with,
persons who are not members of his or her culture of origin.
As Rogler et al. (1991) have indicated, most instruments

Hardin L. K. Coleman is an associate professor of counseling psychology, and Bruce E. Wampold is a professor of counseling psychology, both at the
University of Wisconsin–Madison. Sherry B. Casali is a doctoral candidate in counseling psychology at the University of Wisconsin–Madison and a psychi-
atric nurse at Meritter Hospital in Madison, Wisconsin. This investigation was made possible, in part, by a grant from the Institute for Race and Ethnicity,
University of Wisconsin-System. The authors thank the staff, parents, and students of the Cherokee Heights Middle School for participating in this project, and
Miguel Ybarra, Alfiee Breland, and Marcia Moody for their help in collecting data. Correspondence regarding this article should be sent to Hardin L. K.
Coleman, Department of Counseling Psychology, 321 Education Building, 1000 Bascom Mall, University of Wisconsin–Madison, Madison, WI 53706 (e-
mail: hcoleman@Facstaff.wisc.edu).

356 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79


designed to measure second culture acquisition have used
this linear model. Most of these measurements assess the
degree to which a person has become a member of a new
culture (has assimilated or acculturated) or has stayed in-
volved with his or her culture of origin (usually being la-
beled as “low-acculturated” or “traditional”), even if he or
she has migrated to a new culture. Rogler et al. observed
that most conceptualizations of the process of second cul-
ture acquisition included the “assumption that increments
of involvement in the . . . host society necessarily entail
corresponding decrements of disengagement from the
immigrant’s traditional culture” (p. 587). A variation on this
theme is represented by efforts to include a bicultural di-
mension between the poles of high and low acculturation
(e.g., Landrine & Klonoff, 1994; Szapocznik, Kurtines, &
Fernandez, 1980).
The other three descriptions of second culture acquisition—
alternation, integration, and fusion—represent a different
conceptualization of the process. According to these
conceptualizations, it is possible to both maintain involve-
ment with one’s culture of origin and develop competence
in a second culture. Each conceptualization includes
orthogonal dimensions (LaFromboise et al., 1993; Oetting
& Beauvais, 1991; Rogler et al., 1991) that represent a dis-
tinct behavioral pattern operating independently of the
other behavior patterns. Each of these three descriptions
suggests there are different ways in which an individual
could manage this process. Cuellar, Arnold, and Maldonado
(1985), in a revision of the Acculturation Rating Scale for
Mexican Americans (ARSMA-II), used this approach to
capture the degree to which Mexican Americans integrate
Anglo and Mexican cultures and assimilate, remain sepa-
rate from, or become marginalized in either the Anglo or
Mexican cultures. The alternation conceptualization (Ogbu
& Matute-Bianchi, 1986) assumes that it is possible to al-
ternate between two cultures in the same manner as one
alternates the use of language in different contexts. The inte-
gration conceptualization (Berry, Kim, Power, Young, & Bujaki,
1989) assumes that it is possible to have individuals from
different cultures coexist without compromising their cul-
tural identities. In other words, each person maintains their
culture of origin while interacting with others from differ-
ent cultures. The fusion conceptualization (LaFromboise et
al., 1993; Coleman, 1995b) assumes that individuals from
different cultures who are in consistent contact with each
other will eventually fuse to create a new culture that sub-
sumes individuals’ cultures of origin.
Research on second culture acquisition has attempted
either to determine which description best describes the
behavior of a particular individual or group of individuals
when coming into contact with a second culture or to deter-
mine which reflects the most effective way to act when
experiencing second culture contact. The focus on individual
behavior is common to the literature that has investigated
the relationship between psychological variables (e.g., mental
health) and second culture acquisition. The focus on group
behavior is common to the literature that has addressed
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Adolescent Strategies for Coping With Cultural Diversity
issues related to second culture acquisition and policy or
program development (e.g., curriculum development for cul-
turally diverse schools). These traditional conceptualizations
of second culture acquisition have also focused on the out-
comes of the process. These approaches to conceptualiz-
ing second culture acquisition suggest that the individual
becomes “something” as the result of coming into contact
with a second culture. For example, a person who lets go of
his or her culture of origin to gain competence in a second
culture becomes “assimilated.” A person who creates a
living and working situation in which he or she maintains
his or her culture of origin while simultaneously interacting
with individuals from other cultures becomes “integrated.”
The assessment tools used to measure this process of second
culture acquisition reflect this perspective. These measures
are often used to determine whether an individual is accul-
turated, assimilated, biculturated, separated, fused, integrated,
or separated. An individual can either score high, medium,
or low in these categories (e.g., acculturated, bicultural, low-
acculturated).
Coleman (1995b) proposed an alternate paradigm for
understanding how individuals manage the process of sec-
ond culture acquisition. Coleman argued that traditional
conceptualizations focused on the outcomes of the second
culture acquisition. He hypothesized that the process or
strategies individuals use to achieve particular ends within
particular contexts is an equally important focus for in-
vestigation. Coleman suggested that the context in which
an individual is coping with cultural difference and the
goals that individual wants to achieve in that context will
influence the individual’s choice of strategy. This
conceptualization suggests that individuals manage their
behavior in cross-cultural situations by developing what
Ford (1992) called “behavioral episode schemata.” This
means an individual determines his or her goals in a particu-
lar situation and will use a particular behavioral strategy
that he or she believes will achieve those goals. The strat-
egy an individual uses, therefore, will be related to his or
her goals, the context in which the behavior is to be displayed,
and the skills the individual possesses to use a particular
strategy. Coleman hypothesized that the negative and
positive outcomes associated with the process of second
culture acquisition (LaFromboise et al., 1993) reflect how
effective the individual’s strategy for coping with cultural
diversity is at helping that individual achieve his or her
goals within a particular cultural context. Coleman sug-
gested that the strategies a person uses to cope with sec-
ond culture contact will affect his or her performance in
particular contexts and how the individual feels about him
or herself.
Coleman (1995b) also hypothesized that these strategies
may be organized in a sequential rather than a linear or
orthogonal manner. As Rogler et al. (1991) have indicated,
theories of second culture acquisition have described the
process in either linear or orthogonal terms. These theories
suggest that an individual is either at one end of a mutually
exclusive bipolar relationship (e.g., American or Hispanic)
357
Coleman, Casali, and Wampold
or the individual has, or does not have, orthogonal compe-
tence in both cultures (e.g., American and Hispanic). Fig-
ure 1, building on a suggestion by I. Pagan (personal com-
munication, January 12, 1995), represents a hypothesized
sequential relationship among these strategies.
The model, depicted in Figure 1, suggests that an indi-
vidual who comes into contact with a second culture will
have to make a series of choices, consciously or uncon-
sciously, about how he or she wants to associate with the
new and old cultures, and those goals will be reflected in
his or her behavioral response. As Ford (1992) suggested,
the goals an individual has will affect his or her motivation
to perform certain behaviors. This sequential model sug-
gests that the first choice an individual makes is whether or
not he or she will associate with more than one cultural
group. If the individual chooses to associate with only one
group, then he or she needs to decide whether to associate
only with his or her culture of origin. If the individual makes
that choice, the model proposes that he or she will use a
separation strategy when coming into contact with indi-
viduals from another culture. In contrast, if the individual
does not associate with his or her culture of origin, then a
third choice may be made. If the individual wants to be-
come a full member of the second culture, then he or she
will use the assimilation strategy. If the individual does not
intend to become a full member of the second culture to
achieve a particular goal (e.g., economic stability) but only
associates with that culture, then he or she is using the ac-
culturation strategy. (See Coleman, 1995b, for a more ex-
tensive discussion of these strategies.)
If, at the beginning of this process, the individual chooses
to associate with more than one culture at a time, then
there is another range of strategies for coping with cultural
diversity available to him or her. If the individual intends to
associate with more than one culture, but not at the same
Do you seek to
Do you attempt to associate combine two or
with your culture of origin more cultures?
and a second culture at
the same time?
Yes
Do you associate No
with more than one
Alternation
cultural group?
Yes
No
Do you associate
with only your
Separation
cultural group?
Do you try to
Yes become a full
No member of the
second culture? Yes
FIGURE 1
Decision Tree of Strategies for Coping With Cultural
Diversity
358 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
time, then he or she will use an alternation strategy. If the
individual does intend to combine his or her culture of origin
with the second culture, there are two more strategies he or
she can use. If the person intends to blend the two cultures, he
or she will be using the fusion strategy. If the person intends to
have his or her culture of origin coexist with the second cul-
ture, then he or she is using the integration strategy.
Rather than conceptualizing the process of second cul-
ture acquisition as linear or orthogonal, this model suggests
that the strategies individuals use to cope with cultural di-
versity will be determined by the goals individuals have
within particular contexts. This suggests that an individual
will possess a behavioral episode schemata for each strat-
egy but may be more likely to use one than another as a
result of his or her socialization, the context in which he or
she is coming into contact with a second culture, and the
goal he or she is attempting to achieve within that context.
The purpose of this investigation is to test Coleman’s
(1995b) hypothesis that the strategies that individuals use
to cope with cultural diversity reflect a sequential process
and that those choices are context dependent. To achieve
this purpose, we present a method for identifying these strat-
egies within an adolescent population. Adolescents were
selected because they are a group that is most likely to have
contact with individuals from different cultures within their
schools. This investigation determines the reliability of the
method developed to identify the strategies, examines the
sequential relationship among these strategies, and tests
whether the choice of strategy is dependent on context. It
also reports on the degree to which use of these strategies
may be affected by the ethnicity, gender, or socioeconomic
status of the participant.
METHOD
Participants
Fusion Three hundred and ninety-eight students in a midwestern
middle school (Grades 6–8) completed at least part of the
Yes questionnaire. Of the students, 237 were European Ameri-
No can, 41 were African American, 33 were Asian American, 8
Integration were Hispanic American, 3 were Native American, 15 were
biracial, and 2 did not report their ethnicity. All the partici-
pants were volunteers and were treated in accordance with
the American Counseling Association’s (1995) Code of Ethics
and Standards of Practice.
Materials
Assimilation Stimulus. A questionnaire was developed to help us identify
the strategies adolescents use to cope with cultural diversity.
This questionnaire, referred to as the Coping With Cul-
No
tural Diversity Scale (CCDS), consists of nine descriptions
Acculuration of situations that involve coping with cultural diversity. After
each situation, there are six responses an individual can
make. Each response represents one of the strategies
described earlier. On the CCDS, the participant is ask to
state, on a scale of 0 to 6, how likely he or she is to choose

each of the responses to the situation. A 0 indicates that the
participant is not at all likely to use that response in that situa-
tion; a 6 means he or she would definitely use that response in
that situation. This questionnaire yields 7 scores. There is a
score for each strategy (Assimilation, Acculturation, Alterna-
tion, Integration, Separation, and Fusion). The higher the score,
the more likely it is that the participant may use that strategy.
There is also a total score. This score represents the range of
strategies an individual has at his or her disposal. The higher
the total score, the more strategies the individual believes he
or she is likely to use. The Appendix presents an example of
one of the situations and the possible response. (Details on
the development of the questionnaire are available on request
from the first author.)
Demographic information. At the end of the CCDS was a
section that solicited demographic information concerning the
participant. Participants were asked to identify their ethnicity
from a list and to state their mother’s and father’s employ-
ment. These latter two pieces of information were used to
determine the participant’s socioeconomic status (SES) using
Duncan’s Revised Socioeconomic Index (Featherman &
Stevens, 1982; Stevens & Cho, 1985).
Procedure
A notice was sent to parents in the school newsletter in-
forming them about the purpose and timing of this inves-
tigation. Over a 2-day period, students completed the scale
in their social studies class. They watched a video explain-
ing the purpose of the investigation, and then they com-
pleted the questionnaire.
RESULTS
To test Coleman’s (1995b) hypothesis that these strategies
are organized in a sequential manner and that an individual’s
responses are context dependent required a series of analy-
ses. The internal consistency of the CCDS was established
using Cronbach’s alpha. The data were then submitted to a
cluster analysis to test the sequential organization of the
strategies. To test the context-dependent nature of these
strategies, structural equation modeling was used. Finally,
group differences in the use of the strategies were initially
determined by submitting the data to a series of 1 × 6 fac-
torial analysis of variance to determine the effects of group
membership (i.e., ethnicity, gender, socioeconomic status
as independent variables) on the use of the six strategies as
the dependent variables. Univariate analysis of variance was
completed on the total scale score and those strategies that
indicated significant differences in the multivariate analy-
sis of variance. A general linear model was used to correct
for uneven cell numbers.
Reliability Estimates
The internal reliability of the whole questionnaire and strat-
egy-specific questions using Cronbach’s alpha is as follows:
total questionnaire = .90, Assimilation = .62, Acculturation
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Adolescent Strategies for Coping With Cultural Diversity
= .57, Separation = .69, Alternation = .66, Integration =
.75, and Fusion = .81. These coefficients suggest that the
total scale has strong internal reliability with adequate-to-
strong internal reliability for most of the subscales. The
reliability of the strategy-specific questions ranged from
adequate to strong.
Relationship among and between strategies. The zero-order
correlation matrix among these strategies is presented in
Table 1. These correlations suggest a very strong relationship
among the Fusion, Integration, and Alternation strategies,
and a strong relationship among the Acculturation, Assimi-
lation, and Alternation strategies. The Separation strategy
seems to have a low correlation with the other five strate-
gies. This matrix supports the hypothesis that strategies that
involve affiliating with a single culture are distinct from
strategies that involve affiliating with one or more cultures.
The fact that the Alternation strategy is highly correlated
with both groups is further support for this hypothesis
because using that strategy demands affiliating with one
culture at a time.
To further understand the relationship between and among
these strategies, the data were submitted to an average link-
age agglomerative method of cluster analysis. As represented
in Figure 2, the six strategies seemed to reflect the hypoth-
esized sequential relationship. Fusion and Integration seem
to share the most similar properties. They then joined with
Alternation as a group. At the same time, Acculturation and
Assimilation seem to cluster into a second set. These two
sets eventually cluster together and then, at the end of the
solution, join with Separation. This solution closely matches
the proposed model in that it indicates a distinct relation-
ship between those strategies that involve affiliation with
more that one culture (i.e., Fusion and Integration) and those
that involve affiliating with only one culture (i.e., Assimila-
tion and Acculturation). The solution also suggests that the
behaviors associated with the Alternation strategy share simi-
larities with both groups, whereas the Separation strategy is
distinctly different from the other five.
Tests of context dependence and adequacy of construct mea-
surement. One of the important aspects of Coleman’s
(1995b) conceptualization of cultural coping strategies is
that the choice of strategy depends on the context. That is,
TABLE 1
Correlations Among Strategies for Coping With
Cultural Diversity
Strategy 1 2 3 4 5 6
1. Acculturation —
2. Alternation .62 —
3. Assimilation .56 .55 —
4. Fusion .46 .78 .48 —
5. Integration .45 .75 .39 .84 —
6. Separation .35 .21 .24 .14 .11 —
Note. All correlations were significant at p < .05.
359
Coleman, Casali, and Wampold

Rescaled Distance Cluster Combine

CASE 0 5 10 15 20 25
Label Number + -------------- + --------------- + --------------- + --------------- + ------------ +

FUSION 4

INTEGRATION 5

ALTERNATION 2

ACCULTURATION 1

ASSIMILATION 3

SEPARATION 6

FIGURE 2
Dendrogram Using Average Linkage (Between Groups)

an individual’s choice of strategy will depend, in part, on the
situation; a person’s exhibited strategy will be a function of
the environment as well as the person. The present data pro-
vided the opportunity to test whether or not the context ac-
counts for variance in reported use of strategies.
To test the context dependence hypothesis, we used struc-
tural equation modeling. To obtain multiple measures of each
strategy, the nine situations for each strategy were distributed
into two groups of situations. Group 1 included Situations 3,
2, 5, 1, and 9; and Group 2 included Situations 8, 4, 7, and 6.
This strategy provided two subscales for each strategy. The
variance/covariance matrix for the 12 subscales (two for each
strategy) was computed and served as input for the modeling.
The test of context dependence was accomplished by
comparing two structural equation models, one of which
modeled context dependence and one of which did not.
Although the typical method in structural equation mod-
eling is to specify a model and modify it by adding or delet-
ing paths, this method capitalizes on chance (MacCallum,
Roznowski, & Necowitz, 1992). As explained by MacCallum
et al., the preferable way to use structural equation model-
ing is to compare and test two competing models, where
the superiority of one model has theoretical implications.
In the present study, if the context-dependent model fits
the data significantly better than the context-independent
model, then dependence on context is confirmed. The con-
text independence model is the one in which each of the
strategies loads on the two corresponding subscales (i.e.,
the subscales are the observed variables and the strategies
are the latent variables), the strategies are allowed to corre-
late, and the error variances of the subscales are indepen-
dent. The context-dependent model is identical except that
the error covariances for the subscales with the same situ-
ations are allowed to correlate. The reasoning is that if choice
of strategy is dependent on the context, then some of the
variance of each of the subscales is unique to the situations

360 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79


that compose those subscales, and consequently the error
variances would be correlated (see Cole, 1987; Jöreskog &
Sörbom, 1993).
The fit statistics for the two models are presented in Table
2. The difference between the models was tested by calcu-
lating the difference in chi-square values for the fit of each
model and comparing the difference to a chi-square distri-
bution with degrees of freedom equal to the difference in
degrees of freedom. Because the chi-square for the differ-
ence, χ2(26) = 109.60, p < .001, is sufficiently large to reject
the null hypothesis that the two models fit the data equally
well, we conclude that the context-dependent model is
more explanatory of the phenomenon under study. That is,
choice of strategy was dependent on the situation.
It should be noted that the context-dependent model pro-
vides evidence that the CCDS adequately measures the
strategies it was intended to measure. Although the chi-
square for this model, χ2(13) = 55.25, is large enough to
reject the model, this is due to the large number of partici-
pants (330); rejection of the model is not uncommon for
studies with large samples. However, both the root mean
square residual (1.35) and the goodness-of-fit index (.97)
are within the range of cut-off statistics suggested for ad-
equate models (Jöreskog & Sörbom, 1993).
Correlations among the latent variables (i.e., the strate-
gies) of the context-dependent model were also examined.
These correlations are the disattenuated correlations in that
they are adjusted for the reliability of the latent variables
(and consequently differ from those in Table 1, which are
attenuated by the unreliability of the measurement). These
correlations, as well as the standard errors, are presented in
Table 3. A correlation between two latent variables that is
more than two standard errors less than 1.00 implies that
the latent variables represent distinct, although related,
constructs (Jöreskog & Sörbom, 1993); on the other hand,
correlations within two standard errors of 1.00 indicated
that the two constructs, as measured, are identical. Accord-
ingly, this analysis suggests that Fusion, Alternation, and
Integration are essentially identical, as measured.
TABLE 2
Comparison of Fit Statistics for Context-Dependent
and Context-Independent Models
Fit Statistics
Model χ 2
df RMSR GFI
Context Independent 164.85 39 1.47 .93
Context Dependent 55.25 13 1.35 .97
Difference in Modelsa 109.60 26
Note. n = 330. RMSR = root mean square residual; GFI = goodness-
of-fit index.
a
Chi-square (.26) = 109.60, p < .001, indicating that the
context-dependent model accounts for significantly more variance in
the data than does the context-independent model.
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Adolescent Strategies for Coping With Cultural Diversity
Group Differences
To determine the effect of ethnicity, gender, and SES on
the degree to which participants were likely to use a par-
ticular strategy, we performed a series of multivariate and
univariate analyses. Post hoc analysis of mean differences
were computed using a Bonferroni test with a significance
level of .05
Effect of ethnicity of total scale score and strategy scores. A
multivariate analysis of variance (MANOVA) found a sig-
nificant main effect for ethnicity, F(5, 299) = 2.52, p < .001.
Univariate analysis of the effect of ethnicity on the strategies
for coping with cultural diversity found the following
effects: on the Separation strategy, F(5, 299) = 2.91, p = .01,
with African Americans using this strategy significantly more
than European Americans; on Acculturation F(5, 299) = 2.64,
p = .02, with Asian Americans more likely to use this strategy
than European Americans; on Alternation, F(5, 299) = 3.58,
p = .004, with no significant group contrasts; and on Inte-
gration, F(5, 299) = 4.99, p < .001, with Asian Americans
more likely to use this strategy than European Americans.
There were no significant effects for ethnicity on the
Assimilation or Fusion strategies. A test of the effect of
ethnicity on the Total scale score revealed a significant
effect, F(5, 299) = 3.95, p = .002, with Asian Americans
having a significantly higher total score than European
Americans. The means and standard deviations for the
effects of group membership on the total score and strate-
gies are available from the first author on request.
Effect of gender on total scale and strategy scores. A
MANOVA showed a significant main effect for gender,
F(1, 321) = 5.47, p < .001. Univariate analysis of the effect
of gender on the strategies for coping with cultural diversity
were significant for every strategy—Separation, F(1, 321)
= 3.58, p = .05; Acculturation, F(1, 321) = 10.25, p = .002;
Alternation F(1, 321) = 10.11, p = .002; Integration, F(1, 321)
= 9.71, p = .002; and Fusion, F(1, 321) = 10.64, p = .001—
except Assimilation, in which there were no significant
differences. Gender had a significant effect on the Total
scale score, F(1, 321) = 6.65, p = .01, with women, as a
group, having significantly higher scores than men.
Effect of SES on total scale and strategy scores. A MANOVA
found a significant main effect for SES, F(2, 285) = 3.17, p
< .001. Univariate analysis of the effect of SES on the strat-
egies for coping with cultural diversity being significant on
every strategy: Assimilation, F(2, 285) = 4.41, p = .01, with
the middle group more likely to use this strategy than the
upper SES group; Acculturation, F(2, 285) = 3.59, p = .03,
with the lower SES group more likely than the either the
middle or the upper groups to use this strategy; Alterna-
tion, F(2, 285) = 5.42, p = .005, with the lower group more
likely than the upper group to use this strategy; Integra-
tion, F(2, 285) = 5.17, p = .006, with the lower SES group
more likely than either the middle or the upper group to
use this strategy; and Fusion, F(2, 285) = 3.64, p = .03, with
no significant post hoc group contrasts) except separa-
tion. SES had a significant effect on the Total scale score,
361
Coleman, Casali, and Wampold
TABLE 3
Correlations Among Latent Variables (i.e., Strategies) in the Context-Dependent Model
Latent Variable 1 2
1. Acculturation — —
2. Alternation 0.89 (.06) — —
3. Assimilation 0.80 (.06) 0.81 (.05) — —
4. Fusion 0.63 (.06) 1.03 (.03) 0.64
5. Integration 0.59 (.06) 0.98 (.03) 0.52
6. Separation 0.68 (.08) 0.49 (.08) 0.41
Note. Standard errors in parentheses.
F(2, 285) = 5.19, p = .006, with the lower SES group having
significantly higher total scores than the middle and higher
SES groups.
These results indicate a main effect for ethnicity, gender,
and class in the likelihood that an individual will choose a
particular strategy when coping with cultural diversity. What
this pattern of usage means is discussed as follows.
DISCUSSION AND IMPLICATIONS FOR COUNSELING
The purpose of this investigation was to test Coleman’s
(1995b) hypotheses that the strategies individuals use to cope
with cultural diversity are organized in a sequential manner
and that the use of these strategies is context dependent. To
test these hypotheses, we successfully developed a question-
naire that reliably measures the strategies adolescents use to
cope with cultural diversity. The CCDS allowed us to test
the sequential relationship among the strategies and the ef-
fect of context on the use of these strategies.
Both the correlation matrix of the scores for the strate-
gies and the subsequent cluster analysis supported the hy-
pothesis that the strategies individuals use to cope with
cultural diversity have a sequential rather than a linear or
orthogonal relationship to each other. Coleman’s (1995b)
model hypothesized that the strategies would cluster to-
gether in a particular manner. He suggested that those strat-
egies that involve association with only one culture would
be strongly related to each other as would those that in-
volved associating with more that one culture. The cluster
analysis supports this hypothesis. It shows that the monoc-
ultural strategies, Assimilation and Acculturation, are closely
aligned, as are the multicultural strategies, Fusion and Inte-
gration. These groups are joined through the Alternation
strategy, which shares characteristics of both groups. The
Separation strategy seems to form a distinct group. This
makes conceptual sense because it is the one strategy that
allows individuals to remove themselves from contact with
persons from other cultures. The cluster analysis suggests
that it remains within the monocultural grouping but is
still distinct.
If these strategies represented typologies of ways to en-
gage in the second culture acquisition process, that is, if
they were distinct and nonoverlapping behavior patterns,
there would have been no interpretable solution to the clus-
ter analysis. This solution reveals that the coping strategies
362 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
3 4 5 6
(.05) — —
(.06) 1.01 (.02) — —
(.08) 0.26 (.07) 0.28 (.07) — —
are distinct behavior patterns, but they are not mutually
exclusive. In other words, one individual may be able to use
all the strategies at different times. This is the core assump-
tion in Coleman’s (1995b) hypothesis: that the use of the
strategies is context dependent. The structural equation
modeling analysis of this data supports this hypothesis.
That analysis confirmed, as well, that the situations in the
CCDS had a significant effect on the participants’ choice of
strategy. In other words, the strategy an individual will use is
dependent on the social context in which it will be used. For
example, an adolescent client may use an assimilation or ac-
culturation strategy in a classroom situation and then use a
separation or integration strategy in a social situation. This
further supports the hypothesis that the strategies an indi-
vidual client uses to cope with cultural diversity are not
mutually exclusive, as is suggested by linear and orthogonal
models of second culture acquisition, but are simultaneously
available to an individual, as is hypothesized in a sequential
model. Which strategy an individual client will use is depen-
dent on the setting, his or her goals in the setting, and his or
her ability to achieve those goals in that setting.
Coleman (1997) has suggested that understanding a
client’s strategy for coping with cultural diversity will
have an important influence on the counseling process.
How an adolescent copes will influence how well they
work within a school environment. A student of color
who attempts to assimilate into a predominately White
school environment and is rejected will have one set of
concerns and a student of color who uses a separation strat-
egy in that context will present with a very different set of
issues. Furthermore, the strategy a counselor uses to cope
with cultural diversity can either converge or conflict with
a client’s, which will affect the development of a working
alliance. The results of this investigation reinforce the idea
that counselors need to look beyond their own or their
clients’ group membership and focus on how within-group
variations lead to problems for clients and solutions within
counseling.
The sequential and context-dependent nature of these
coping strategies are further supported by analysis of group
differences in the use of these strategies. It would be a
mistake to overinterpret these data, but there are some
important trends that need to be noted. One significant
trend is the fact that those individuals who most closely
represented the mainstream values of the school in which

this investigation was completed, middle and upper SES
European American boys, had the lowest Total scale score
and, often, lower subscale scores, than their lower SES,
ethnic minority, and female peers. This suggests that in a con-
text in which one is the dominant group, one needs less of a
repertoire for coping with cultural diversity, and vice versa. What
the results of this investigation suggest is that factors such as
ethnicity, gender, social class, and social skills will affect how a
client behaves in different situations.
The fact that these strategies seem to be context depen-
dent highlights the distinction between Coleman’s model
(1995b) and models of ethnic identity (Cross, 1991; Helms,
1993; Phinney, 1991). These results suggest that the strate-
gies a client uses to cope with cultural diversity will vary
depending on the social context. These strategies, therefore,
seem to be exogenous variables in contrast to ethnic iden-
tity, which is an endogenous variable. Future research is
needed to better establish the relationship between ethnic
identity, these strategies for coping with cultural diversity,
and counseling outcomes.
There are certain limitations to this investigation that need to
be addressed in future research in this area. Given the limited
number of culturally diverse participants, we were unable to
examine the multivariate effects of group membership on the
use of these strategies. Given the support of the context-de-
pendent nature of these strategies, it will also be necessary to
examine this process in settings where ethnic minorities are the
majority to see how context affects these patterns.
Although the psychometric properties of the Coping With
Cultural Diversity Scale were adequate for the needs of
this investigation, the results revealed several issues that
need to be examined further. The first is that the reliability
for the Acculturation subscale is marginally adequate. Ef-
forts should be made to improve those items if subsequent
investigations using the questionnaire do not produce
better results. The structural equation modeling analysis
also suggests that there may be fewer than six distinct strat-
egies. Specifically, that analysis suggests that Alternation,
Integration, and Fusion may actually be a single strategy. If
that is the case, then it would suggest that Acculturation
and Assimilation make a second group, and Separation
remains as a distinct strategy or third behavior pattern.
Further examination of the psychometric properties of
the CCDS, particularly its construct validity, is needed to
clarify these relationships.
In this investigation, we have observed that the strategies
adolescents use to cope with cultural diversity possess a se-
quential relationship to each other and that their use is con-
text dependent. This suggests that the wider the range of
strategies an individual has to cope with cultural diversity,
the better he or she will be able to effectively manage a cul-
turally diverse or monocultural environment. If that is the
case, helping individuals to effectively use a wide range of
strategies may have a beneficial effect on their school or job
performance and their sense of psychological well-being.
Future research in this area needs to examine that possi-
bility, especially in the context of working with clients. We
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Adolescent Strategies for Coping With Cultural Diversity
need to know more about the role of context in the
development and use of these strategies and how they are
related to similar constructs, such as ethnic identity. We
also need to identify the degree to which the likelihood
to use particular strategies within particular settings is pre-
dictive of such significant outcomes as academic achieve-
ment, social competence, and mental status. The results
of this investigation lay the groundwork for these future
investigations.
REFERENCES
American Counseling Association. (1995). Code of ethics and standards
of practice. Alexandria, VA: Author.
Atkinson, D. R., Casas, A., & Abreu, J. (1992). Mexican-American accul-
turation, counselor ethnicity and cultural sensitivity, and perceived
counselor competence. Journal of Counseling Psychology, 39, 515–520.
Berry, J. W., Kim, U., Power, S., Young, M., & Bujaki, M. (1989). Accul-
turation attitudes in plural societies. Applied Psychology: An Interna-
tional Review, 38, 185–206.
Berry, J. W., Poortinga, Y. P., Segall, M. H., & Dasen, P. R. (1992). Cross-
cultural psychology: Research and applications. New York: Cambridge
University Press.
Cole, D. A. (1987). Utility of confirmatory factor analysis in test validation
research. Journal of Consulting and Clinical Psychology, 55, 584–594.
Coleman, H. L. K. (1995a). Cultural factors and the counseling process:
Implications for school counselors. The School Counselor, 42, 180–185.
Coleman, H. L. K. (1995b). Strategies for coping with cultural diversity.
The Counseling Psychologist, 23, 722–740.
Coleman, H. L. K. (1997). Conflict in multicultural counseling relation-
ships: Source and resolution. Journal of Multicultural Counseling and
Development, 25, 195–200.
Coleman, H. L. K., Wampold, B. E., & Casali, S. B. (1995). Ethnic mi-
norities’ ratings of ethnically similar and European American counse-
lors: A meta-analysis. Journal of Counseling Psychology, 42, 55–64
Cross, W. E. (1991). Shades of black. Philadelphia: Temple University
Press.
Cuellar, I., Arnold, B., & Maldonado, R. (1985). Acculturation Rating
Scale for Mexican Americans-II: A revision of the original ARSMA
Scale. Hispanic Journal of Behavioral Sciences, 17, 275–304.
Featherman, D. L., & Stevens, G. (1982). A revised socioeconomic
index of occupational status: Application in analysis of sex differ-
ences in attainment. In M. G. Powers (Ed.), Measures of socioeco-
nomic status: Current issues (pp. 83–127). Boulder, CO: Westview
Press.
Ford, M. E. (1992). Motivating humans. Newbury Park, CA: Sage.
Helms, J. E. (1993). Black and White racial identity: Theory, research, and
practice. Westport, CT: Praeger.
Gomez, M. J., & Fassinger, R. E. (1994). An initial model of Latina achieve-
ment: Acculturation, biculturalism, and achieving styles. Journal of
Counseling Psychology, 41, 205–215.
Jöreskog, K. G., & Sörbom, D. (1993). LISREL 8: Structural equation
modeling with SIMPLIS command language. SSI Scientific Software In-
ternational. Mahwah, NJ: Erlbaum.
LaFromboise, T. D., Coleman, H. L. K., & Gerton, J. (1993). Psychological
impact of biculturalism: Evidence and theory. Psychological Bulletin, 114,
395–412.
Landrine, H., & Klonoff, E. A. (1994). The African-American Accultura-
tion Scale: Development, reliability, validity. Journal of Black Psychol-
ogy, 20, 104–127.
MacCallum, R. C., Roznowski, M., & Necowitz, L. B. (1992). Model
modifications in covariance structure analysis: The problem of capi-
talization on chance. Psychological Bulletin, 111, 490–504.
Oetting, E. R., & Beauvais, F. (1991). Orthogonal cultural identification
theory: The cultural identification of minority adolescents. The Inter-
national Journal of the Addictions, 25, 655–685.
363
Coleman, Casali, and Wampold

Ogbu, J. U., & Matute-Bianchi, M. A., (1986). Understanding sociocul- APPENDIX

tural factors: Knowledge, identity, and social adjustment. In Language
Minority Children/Bilingual Education Office, Beyond language: Social Example of Items From the Coping With Cultural
and cultural factors in schooling (pp. 73–142). Los Angeles: California Diversity Scale: Adolescent Version
State Department of Education Evaluation, Dissemination and Assess-
ment Center, California State University. Rating Scale:
Oliver, M. L., Rodriguez, C. J., & Mickelson, R. A. (1985). Brown and Not at all likely to do Likely to do Definitely would do
Black in White: The social adjustment and academic performance of 0 1 2 3 4 5 6
Chicano and Black Students in a predominately White university. The
Urban Review, 17, 3–23. Situation 4
Padilla, A. M. (1980). Acculturation: Theory, models, and some new find- If you were with someone from a different race or culture than you,
ings. Boulder, CO: Westview Press. and someone from your race or culture started to bother both of
Phinney, J. S. (1991). Ethnic identity and self-esteem: A review and inte- you for being together, how likely are you to:
gration. Hispanic Journal of Behavioral Sciences, 13, 193–208.
1. Stop hanging out with the person from the other race or culture?
Rogler, L., Cortes, D. E., & Malgady, R. G. (1991). Acculturation and
0 1 2 3 4 5 6
mental health among Hispanics. American Psychologist, 46, 585–597.
2. Just laugh if off?
Stevens, G., & Cho, J. H., (1985). Socioeconomic indexes and the new
0 1 2 3 4 5 6
1980 census occupational classification scheme. Social Science Research,
14, 142–168. 3. Pretend not to hear them?
Stonequist, E. V., (1935). The problem of marginal man. American Jour- 0 1 2 3 4 5 6
nal of Sociology, 7, 1–12. 4. Ask the person who is bothering you what their problem is?
Stonequist, E. V., (1937). The marginal man: A study in personality and 0 1 2 3 4 5 6
culture conflict. New York: Scribner. 5. Ask him or her to leave the two of you alone?
Szapocznik, J., Kurtines, W. M., & Fernandez, T. (1980). Bicultural in- 0 1 2 3 4 5 6
volvement in Hispanic American youths. International Journal of In- 6. Tell the person bothering you that she or he is wrong?
tercultural Relations, 4, 353–365. 0 1 2 3 4 5 6

364 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79

Rational Suicide: An Empirical Investigation of
Counselor Attitudes

James R. Rogers, Christine M. Gueulette, Jodi Abbey-Hines, Jolynn V. Carney, and James L. Werth, Jr.

This study investigated attitudes of mental health counselors toward rational suicide. Over 80% of respondents were moderately
supportive of the idea that people can make well-reasoned decisions that death is their best option, and further, they identified a
consistent set of criteria to evaluate such decisions. Additionally, analyses revealed that as the scenario moved from client, to
friend, to self, rational suicide seemed to gain in acceptability for these professionals. Implications for training and practice in light
of the apparent responsibilities with suicidal clients defined in the 1995 American Counseling Association’s Code of Ethics and
Standards of Practice are discussed.

I
n recent years, the highly publicized issue of rational
suicide has generated considerable debate in both the
popular (e.g., Quill, 1993) and professional literature
(e.g., Werth, 1996, 1999a). Although the concept of
rational suicide has historical roots reaching back cen-
turies (e.g., Previn, 1996) and is clearly embedded in the
existential literature (e.g., Camus, 1955), contemporary at-
tention to this highly charged issue has been fueled by
multiple sources. For example, concern over the issue of
rational suicide has resulted from such realities as the on-
going advancement in life-saving medical technology and
the related aging of society in the United States (Rogers,
1996), the AIDS pandemic (Rogers & Britton, 1994; Werth,
1992), and general changes in the way human beings die
(Battin, 1994). According to Battin, the way we die has
changed drastically over the past century from acute death
at an early age resulting from opportunistic infections to
death from chronic and degenerative diseases later in life.
Thus, in our contemporary society, rational suicide is an
attempt to control the timing and process of death and,
hereby, secure what has been termed by Battin as “the least
worst death” (p. 36).
The issue of rational suicide is especially relevant in the
United States because it underlies the passionate debate
regarding physician-assisted suicide (Werth, 1999a). For
example, in the last decade this debate was stimulated on
the West Coast by voter approval of the Oregon Death
With Dignity Act (1995), its subsequent legislative chal-
lenge, and the 1996 ruling by the Ninth Circuit Court of
Appeals that state laws prohibiting physician-assisted sui-
cide were unconstitutional (Compassion in Dying v. Wash-
ington). Similarly, in that same year on the East coast, the
Second Circuit Court of Appeals drew the same conclu-
sion, albeit by a different constitutional route, in its ruling
in the case of Quill v. Vacco.
In response to legislative challenges to the rulings of the
Second and Ninth Circuit Courts of Appeals, the United
States Supreme Court agreed to hear the two cases and in
1997 ruled that terminally ill people do not have a consti-
tutional right to physician-assisted suicide (Brody, 1998).
In this ruling, the Supreme Court declared that the deci-
sion whether to allow physician-assisted suicide was under
the purview of the individual states.
Of particular relevance to counselors in this legislative
process was the fact that the American Counseling Asso-
ciation (ACA) signed an amicus curiae (“friend of the
court”) brief focused on mental health issues associated
with hastened death decisions and in support of the de-
cisions of the circuit courts (“Brief of the Washington
State Psychological Association,” [“Brief–WSPA”] 1996).
ACA’s decision to sign onto this brief was made unilater-
ally based on the results of discussions among the
organization’s leadership and management (Morrissey,
1997b) and with limited input from the organization’s
membership. After the announcement of ACA’s involve-
ment in the amicus curiae brief in that organization’s

James R. Rogers is an assistant professor in the Collaborative Program in Counseling Psychology and the Counselor Education and Supervision programs,
and Christine M. Gueulette is a Marriage and Family doctoral student, both in the Department of Counseling and Special Education at The University of
Akron, Akron, Ohio. Jodi Abbey-Hines is a licensed professional counselor working in community mental health in Golden, Colorado. Jolynn V. Carney is
an assistant professor in the Department of Counseling at Youngstown State University, Youngstown, Ohio. James L. Werth, Jr., is an assistant professor in the
Collaborative Program in Counseling Psychology, Department of Psychology, The University of Akron. This research was supported by a grant from the
University Research Council of the School of Graduate Studies of Youngstown State University. The authors thank Angel Olsen and Jeanette Higgins for their
assistance in the data collection and coding activities related to this research. Correspondence regarding this article should be sent to James R. Rogers,
Department of Counseling and Special Education, Carroll Hall 127, The University of Akron, Akron, OH 44325-5007 (e-mail: jrrogers@uakron.edu).

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 365

R o g e r s , G u e u l e t t e , A b b e y - H i n e s , C a r n e y, a n d W e r t h , J r.
professional newsletter (Morrissey, 1997a), members of
ACA registered protests through letters to the editor (e.g.,
Battersby, 1997; Duggan, 1997). Although some litera-
ture does exist related to the attitudes of psychotherapists
toward the issue of rational suicide (e.g., Werth, 1996; Werth
& Cobia, 1995: Werth & Liddle, 1994), there has been no
research investigating these attitudes specifically for mem-
bers of a counseling association, and thus no empirically
based support for the position (i.e., “Brief–WSPA,” 1996)
supported by ACA.
Given that the Supreme Court has left the question of
the legalization of physician-assisted suicide up to the indi-
vidual state legislatures and the prediction of legislative
experts that the question of legalization will continue to be
argued at the state level in the foreseeable future, it seems
important to have an empirical sense of where counselors
stand on the underlying issue of rational suicide. The pur-
pose of this study was to begin to develop a knowledge
base regarding counselor attitudes toward rational suicide
by modifying the methodology used by Werth and Liddle
(1994) and Werth and Cobia (1995) in their investigations
of the attitudes of psychotherapists toward rational suicide
(see also Werth, 1996).
RATIONAL SUICIDE AND ATTITUDES OF MENTAL HEALTH
PROFESSIONALS
Although there has been some work investigating attitudes
toward rational suicide in the sociological literature (Stack,
1999), with the exception of Werth and Liddle (1994),
Werth and Cobia (1995), and Werth (1996), this specific
issue as it relates to attitudes of mental health practitio-
ners has received little empirical attention in the published
counseling literature.
Werth and Liddle (1994; see also Werth, 1996) reported
the quantitative results of a survey of 400 members of
Division 29 (Psychotherapy) of the American Psychological
Association. On the basis of their 50% response rate, these
authors concluded that psychotherapists were “differen-
tially accepting of suicidal ideation” (Werth & Liddle, 1994,
p. 440) and reported differing amounts of preventative
action as a function of contextual circumstances. Specifically,
their results suggested that psychotherapists were most
accepting of suicidal ideation and least prone to inter-
vene in cases of terminal illness and were progressively
less accepting and more prone to intervene in situations
reflecting physical pain, psychological pain (e.g., depression),
and bankruptcy, respectively. These authors also reported
that 81% of their sample indicated a belief in the concept
of rational suicide.
In 1995, Werth and Cobia (see also Werth, 1996) pre-
sented the results of their analysis of qualitative data col-
lected in the same survey of psychotherapists focused on
identifying “empirically based criteria for rational suicide”
(p. 231). According to Werth and Cobia, 88% of the re-
spondents in the qualitative data indicated a belief in the
366 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
concept of rational suicide that, collectively, included the
following components:
(1) The person considering suicide has an unremitting hopeless
condition. Hopeless conditions include, but are not necessarily
limited to, terminal illnesses, severe physical pain, and/or psy-
chological pain, physically or mentally debilitating and/or de-
teriorating conditions, or quality of life no longer acceptable to
the individual.
(2) The person makes the decision as a free choice (i.e., is not
pressured by others to choose suicide).
(3) The person has engaged in a sound decision-making process.
This process should include the following:
(a) Consultation with a mental health professional who can
make an assessment of psychological competence.
(b) Nonimpulsive consideration of all of the alternatives.
(c) Consideration of the congruence of the act with one’s
personal values.
(d) Consideration of the impact of the act on significant others.
(e) Consultation with objective others (e.g., medical and
religious professionals) and with significant others. (p. 238)
In tandem, these reports have provided some sense of the
attitudes of APA member psychotherapists toward the con-
cept of rational suicide and have resulted in an initial set of
empirically based criteria for differentiating rational from
nonrational suicide. It remains an empirical question whether
similar attitudes exist among members of a counseling orga-
nization. To address this question, we developed a modified
version of the survey instrument used in Werth and Liddle
(1994), Werth and Cobia (1995), and Werth (1996) to in-
vestigate counselor attitudes toward rational suicide.
METHOD
Participants
To investigate counselor attitudes toward the issue of ratio-
nal suicide, a random sample of 1,000 professional members
of the American Mental Health Counselors Association
(AMHCA) were mailed the survey instruments and invited
to participate in the study. AMHCA members were chosen
with the assumption that members of this organization (i.e.,
specialty area) are more likely than some types of counselors
to work in settings where they would encounter suicidal in-
dividuals and would have professional experiences to inform
their reported attitudes (see Rogers, 1996, for a discussion of
the importance of participant experience in assessing right-
to-die attitudes). Usable responses were obtained from 241
(24%) of the surveyed professionals representing 40 states.
Of the 241 respondents, 73% (176) were women, 93.4%
(225) reported their race as White, 17% had a doctoral-
level education, and 44% were trained at the master’s level.
In terms of discipline, 40.7% were professional counselors,
34.4% counseling psychologists, 16.2% social workers, and
8.71% clinical psychologists. Reported work settings included
private practice (53.1%), community agencies (22.8%),
hospitals (7.1%), academic settings (6.2%), and others
(10.8%). Sixty-six percent of the respondents were inde-
pendently licensed and the mean years of experience was
12.24 (SD = 8.74), ranging from 0 to 44 years. Religious
 Rational Suicide: An Empirical Investigation

affiliations were reported as Protestant (22.0%), Catholic of the respondents. There were no follow-up mailings or
(19.5%), Jewish (10.8%), Methodist (6.2%), Agnostic/Atheist other attempts to increase the response rate for the survey.
(5.8%), Lutheran (5.0%), and Other (30.7%) (see Table 1
for a precise breakdown of demographics; due to rounding, Materials
percentages may not total 100%).
Demographics. A one-page demographic questionnaire was
Procedure constructed to include a number of respondent factors that
could potentially influence mental health professionals’ atti-
The random sample of 1,000 professional members (i.e., tudes toward suicide. In addition to sex and race, the demo-
excluding student members) of AMHCA was obtained from graphic form collected information on participant religious
the organization’s national headquarters. Each member iden- affiliation, discipline, work setting, experience, and licensure.
tified in this process received a cover letter explaining the Finally, participants were asked to respond to the following
purpose of the study, the questionnaire materials, and a self- three questions: (1) Have you ever had a client attempt sui-
addressed, stamped envelope for return of the completed cide? (2) Have you ever had a client commit suicide? and
survey. Surveys were constructed to ensure the anonymity (3) Have you ever seriously considered suicide?
Survey. The survey instrument was adapted from the
TABLE 1 questionnaire used in Werth and Liddle (1994), Werth and
Cobia (1995), and Werth (1996). First, because our inter-
Sample Demographic and Informational Results est was not in counselors’ judgments of the personal char-
acteristics of the individual depicted in the vignette, we
Variable na % dropped a 10-item semantic differential scale from the
Sex original survey that focused on those characteristics. Sec-
Women 176 73.0 ond, on the basis of literature suggesting that attitudes
Men 65 27.0 vary as a function of the personal relevance of the issues
Race
White 225 93.4 being investigated (e.g., Ableson, 1988; Rogers, 1996), we
Other 16 6.6 added two items that progressively increased the personal
Religious Affiliation relevance of the issue of rational suicide to the respondents.
Catholic 47 19.5
Methodist 15 6.2
The first of these items changed the individual in the vignette
Protestant 53 22.0 from a client to a friend, and the second item asked the
Jewish 26 10.8 respondents to place themselves in the context of the scenario.
Agnostic/Atheist 14 5.8 For each of these new items as well as for the original client-
Lutheran 12 5.0
Other 74 30.7 focused item, respondents were asked to provide a narrative
Discipline rationale for their responses. Also, to investigate possible gen-
Clinical psychology 21 8.7 der effects, we varied the sex of the individual depicted in
Professional counseling 98 40.7
Counseling psychology 83 34.4
the case vignette. Finally, on the basis of Maltsberger (1994),
Social work 39 16.2 we added an item to investigate the effects of changing the
Primary Work Setting word rational to reasonable on the respondents’ attitudes
Community agency 55 22.8 toward the suicidal intent expressed in the vignette. These
Academic setting 15 6.2
Hospital 17 7.1 adaptations resulted in a survey instrument that was more
Private practice 128 53.1 clearly focused on the attitudes and perceptions that we
Other 26 10.8 were interested in investigating.
Licensure Our survey presented the following vignette describing
Independent 159 66.0
Licensed 46 19.1 the context and condition of an individual who has made a
Not licensed 36 14.9 decision to commit suicide. The sex of the individual de-
picted in the vignette was randomly varied.
Experienceb
Have you ever had a client attempt suicide?
Yes 170 72.0 John/Joan, a 45-year-old factory worker who is currently a client of
No 66 28.0 yours, has been suffering from severe physical pain related to ma-
Have you ever had a client commit suicide? lignant bone cancer. Despite having received a wide variety of thera-
Yes 68 28.7 pies, the cancer is spreading. John/Joan feels that both you and his/
No 169 71.3 her physicians have already done all you can and he/she has no
Have you ever seriously considered suicide? hope that his/her symptoms will be reduced. John/Joan is currently
Yes 48 20.0 experiencing a great deal of pain and is very upset over the fact
No 192 80.0 that his/her condition is draining the emotional and financial re-
sources of his/her family and friends. John/Joan feels that the qual-
Note. N = 241. ity of his/her life now is very poor and will only get worse. He/she
a
Number of participants responding to the open-ended questions also fears that he/she will be an increasingly large burden to his/
differed from one question to another. b Mean = 12.24 years, N = her loved ones. With the support of his/her family and friends, John/
237, SD = 8.74 years, Range = 0–44 years. Joan has decided to kill him-/herself.

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 367

R o g e r s , G u e u l e t t e , A b b e y - H i n e s , C a r n e y, a n d W e r t h , J r.
Participants were asked to respond to the following three
quantitative questions based on their reading of the case
vignette using a 7-point rating scale:
1. As a professional, how much action would you
take to prevent John/Joan from killing him-/her-
self? Anchors—“Nothing” (1) and “Everything/Any-
thing including involuntary hospitalization” (7).
2. As a friend (in this case John/Joan is not a client),
how much action would you take to prevent John/
Joan from killing him-/herself? Anchors—“Nothing”
(1) and “Everything/Anything including steps that
may lead to involuntary hospitalization” (7).
3. If you were in John/Joan’s circumstance, how viable
would you view suicide as an option? Anchors—“Not
viable at all” (rated as 1) and “Extremely viable” (7).
After the first two questions, respondents were asked to
present a rationale for their ratings.
Finally, participants were asked to respond to three open-
ended items:
1. Do you believe in the idea of “rational suicide?” Why
or why not?
2. Under what circumstances would you consider a
decision to commit suicide to be rational?
3. If the word “rational” in the above questions was re-
placed with the term “reasonable,” would that affect
your responses? If so, in what ways?
RESULTS
Seventy-one percent (n = 170) of the survey respondents
reported having at least one client who had made a suicide
attempt. In addition, 28% (n = 68) reported having at least
one client who committed suicide, and 20% (n = 48) indi-
cated that they themselves had seriously considered suicide.
Responses to the three quantitative questions were analyzed
first for the total sample and then as a function of respon-
dent sex, race, religious affiliation, discipline, primary work
setting, vignette sex, having clients who have attempted sui-
cide, having clients who have committed suicide, and having
personally considered suicide. Qualitative responses were
coded by two researchers who worked to achieve consensus
on the coding categories. In cases in which consensus could
not be reached in categorizing the responses as either against
rational suicide, in support of rational suicide, or ambivalent,
responses were dropped from the analyses.
Quantitative Data
For Item 1, “As a professional, how much action would you
take to prevent John/Joan from killing him-/herself?” the
total sample mean was 3.89 (SD = 2.1). There were 245
individual responses to the question “What is the basis for
your decision?” Eight of those responses were ambiguous
and were dropped from the analysis for a final total of 237
responses. Of the remaining 237 responses, 24% (n = 57)
368 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
were clearly not in support of a decision to commit suicide,
citing moral and ethical positions as the basis for their
decision. An additional 40% (n = 94) of the responses were
supportive of a suicide decision citing rationality and client
autonomy as the primary considerations. The remaining 36%
(n = 86) of the responses were either ambivalent in terms of
the basis for their quantitative responses (n = 22) or wanted
more specification regarding issues of pain control and pal-
liative care (n = 40) or more information as to whether or
not all other options had been exhausted (n = 24).
For Item 2, “As a friend, how much action would you take
to prevent John/Joan from killing him-/herself?” the total
mean score was 3.38 (SD = 2.1). There were 239 responses
to the question “What is the basis for your decision?” Six-
teen of those responses were ambiguous and dropped from
the analysis for a final total of 223 responses. Of the remain-
ing 223 responses, 16% (n = 36) were not supportive of a
decision to commit suicide with issues of morality, hope for
condition improvement, and the impact of suicide on sig-
nificant others as the most cited bases for their decision. Fifty-
nine percent (n = 132) were supportive of a decision to com-
mit suicide citing autonomy, hopeless conditions, pain, and
rationality as the primary considerations. The remaining 25%
(n = 55) of the responses were ambivalent or in some way
wanting to have more information than provided in the vi-
gnette to support the response to the quantitative item.
Finally, for Item 3, “If you were in John/Joan’s circum-
stance, how viable would you view suicide as an option?”
the total sample mean score was 4.79 (SD = 2.3). Narrative
responses were not solicited for this item.
To explore possible differences in responses to the three items
as a function of respondent characteristics, a multivariate
analysis of variance (MANOVA) was performed with each
characteristic as the independent variable and the three items
as the dependent variables. MANOVA results were not statisti-
cally significant as a function of the respondent characteristics,
with the exception of religious affiliation (Pillai’s Trace = .218,
F = 2.99, df = 18, 687, p < .0001, η2 = .074) and vignette sex
(Pillai’s Trace = .047, F = 3.85, df = 3, 237, p < .01, η2 = .047).
Religious affiliation. Follow-up analyses of variance (ANOVAs)
with religious affiliation as the independent variable revealed
statistically significant differences on all three quantitative items.
Table 2 lists the means and standard deviations for the three
quantitative items as a function of religious affiliation. For Item
1, “As a professional, how much action would you take to pre-
vent John/Joan from killing him-/herself?” differences as a func-
tion of religious affiliation were found at the p = .0002 level
(F = 4.62, df = 6, 234). Tukey’s post hoc standardized range
tests were used in the follow-up analyses to control the experi-
ment-wise error rate. Tukey’s procedure is a conservative post
hoc test analogous to a Bonferroni correction with alpha di-
vided by the number of all possible paired comparisons
(Marascuilo & Serlin, 1988). Tukey’s post hoc standardized
range tests indicated statistically significant (p < .05) differ-
ences between Catholic and Jewish respondents, between
Catholic and Agnostic/Atheist respondents, and between Meth-
odist and Jewish respondents with the Catholic and Methodist

TABLE 2
Means and Standard Deviations for the Three Quantitative Items as a Function of Religious Affiliation
Religious Item 1
Affiliation M SD M SD
Catholic 4.91 a 2.1 4.77a
Methodist 4.93 1 2.1 4.12
Protestant 3.76 2.1 2.91b
Jewish 2.85 b2 1.9 2.50b
Agnostic/Atheist 2.79 b 1.6 2.57b
Lutheran 4.42 1.8 3.17
Other 3.88 2.1 3.17b
Note. Within columns, means with nonsimilar subscripts and superscripts are significantly different at the p ≤ .05 level.
respondents scoring higher (i.e., adopting a more preventa-
tive stance) than the Jewish and Agnostic/Atheist respondents.
For Item 2 (“As a friend, how much action would you
take to prevent John/Joan from killing him-/herself?”), dif-
ferences were statistically significant at the p < .0001 level
(F = 6.12, df = 6, 234). Again, Tukey’s post hoc standard-
ized range tests indicated statistically significant (p < .05)
differences between the means of the Catholic respondents
and those of the Protestant, Jewish, Agnostic/Atheist, and
Other respondents with the Catholic respondents scoring
higher (i.e., adopting a more preventative stance).
For Item 3 (“If you were in John/Joan’s circumstance, how
viable would you view suicide as an option?”), differences were
statistically significant at p < .0001 (F = 4.92, df = 6, 234).
Similarly, Tukey’s post hoc standardized range tests indi-
cated statistically significant (p < .05) mean differences between
the Catholic respondents and the Protestant, Jewish, Agnostic/
Atheist, and Other respondents, with Catholic respondents scor-
ing lower (i.e., seeing suicide as a less viable personal option).
Vignette sex. Follow-up ANOVAs with vignette sex as the
independent variable revealed statistically significant dif-
ferences for Item 3 only (F = 9.61, df = 1, 239, p < .002). In
this analysis, the mean score for respondents exposed to
the vignette depicting a woman was higher (M = 5.22, SD
= 2.2, n = 126) than the mean for the respondents pre-
sented with the vignette depicting a man (M = 4.31, SD =
2.4, n = 115). In an attempt to explore this finding further,
we conducted an ANOVA with respondent sex and vignette
sex as main effects and tested the interaction term to inves-
tigate a possible gender effect. The interaction term in the
model was not statistically significant (F = .61, df = 1, 239,
p = .435). Thus, the meaning of this result is unclear given
that Item 3 asked the respondents to rate the viability of a
rational suicide option for themselves. Any guess as to why
those responses would vary systematically as a function of
the sex of the person depicted in the vignette would be
pure speculation given the data.
Open-Ended Questions
Two-hundred twenty-seven individuals provided 234 re-
sponses to the first open-ended question, “Do you believe
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Rational Suicide: An Empirical Investigation
Item 2 Item 3
M SD n
2.1 3.45 a 2.4 47
2.4 4.27 2.7 15
2.0 4.83 b 2.2 53
1.6 5.96 b 1.2 26
1.6 5.57 b 2.2 14
1.6 4.33 2.2 12
2.0 5.16 b 2.3 69
in the idea of “rational suicide? Why or why not?” Of the
participants, 81% (n = 184) indicated some level of belief
in the concept of rational suicide. Of the remaining 42 re-
spondents (one participant’s response could not be inter-
preted) who indicated that they did not believe in the con-
cept of rational suicide, 22 responses indicated that the belief
was grounded in their moral values. The remaining 20 indi-
viduals either gave no rationale for their position (n = 14)
or in some form restated their belief that suicide ipso facto
could not be rational.
The 184 participants indicating some level of belief in
rational suicide provided 195 responses. Of these, 33 re-
sponses indicated a degree of qualified ambivalence; that
is, these respondents reported a belief in rational suicide
but qualified their reported belief in ways that could not
be unambiguously coded as either supportive or
unsupportive. An additional 54 responses repeated their
affirmative position but did not provide any further ex-
planation for their response. The remaining responses were
categorized based on the criteria identified by Werth and
Cobia (1995; see also Werth, 1996). Within this framework,
66 responses clearly mirrored Criterion 1: “The person con-
sidering suicide has an unremitting hopeless condition.” Of
these responses, 26 cited an unacceptable quality of life, 20
cited severe physical pain, 19 cited terminal illness, and 1 cited
physical disability as providing an acceptable rationale for
suicide. An additional 27 responses fit within Criterion 2:
“The person makes the decision as a free choice.” Fifteen
responses fit within Criterion 3: “The person has engaged in
a sound decision-making process.” Of the 15 responses
matching Criterion 3, 13 cited a consideration of all pos-
sible alternatives, 1 cited congruence with personal values,
and 1 cited a consideration of the impact on others as
defining standards.
For the second open-ended question (“Under what circum-
stances would you consider a decision to suicide to be ratio-
nal?”), there were 349 responses relevant to the question. Of
those responses, 38 indicated that there were no circum-
stances that would lead to a rational decision to suicide. The
remaining responses are listed in Table 3 by category.
Finally, 41 individuals responded to one or both of the
questions “If the word ‘rational’ in the above question was
369
R o g e r s , G u e u l e t t e , A b b e y - H i n e s , C a r n e y, a n d W e r t h , J r.
TABLE 3
Responses to “Under What Circumstances Would
You Consider a Decision to Commit Suicide to
Be Rational?”
Category
Terminal illness
Severe physical pain
Nonimpulsive consideration of alternatives
Unacceptable quality of life
Consultation with significant others
Consultation with other professionals
Physical debilitation
Clear state of mind
Free choice
Consultation with mental health professional
Severe psychological pain
Congruence with values
Mental disability
Total Responses
replaced with the term ‘reasonable,’ would that affect your
responses? If so, in what ways?” Only 51% (n = 21) indi-
cated that the change in terminology would have affected
their responses to the previous question. Of those respon-
dents, 17 indicated that they would be less accepting of the
decision to commit suicide and 1 indicated that he or she
would be more favorable if the qualifier “rational” were re-
placed by “reasonable.” The other 3 respondents indicated
that the change in terminology would affect their responses,
but the direction of the effect could not be determined
from the written narratives. The remaining 20 respondents
did not indicate whether the change in terminology would
affect their responses but provided narratives discussing
their perceptions of the differences between the two terms
as they would relate to a decision to commit suicide. In
most of those narratives, “rational” was seen as the more
objective and definable term, whereas “reasonable” was seen
as more subjective and emotionally laden.
DISCUSSION
The purpose of the present study was to provide empiri-
cally based data related to attitudes of counselors toward
rational suicide. Our decision to survey members of
AMHCA with the assumption that members of this coun-
seling organization would have professional experiences to
inform their reported attitudes was supported in the data,
which indicated that 71% of the respondents reported
working with clients who had attempted suicide and 28%
reported having at least one client who committed suicide.
These high percentages of experience with suicide ideation
and completion indicate that training in the detection of
suicidality and interventions to prevent suicidal deaths
should be a part of counselor training programs (see
Westefeld et al., 2000). The 28% figure adds to the body of
370
research that has consistently shown that about one fourth
to one third of nonpsychiatric mental health professionals
can expect to have at least one client die by suicide during
their counseling career (Chemtob, Bauer, Hamada, Pelowski,
& Muraoka, 1989; Werth, 1996).
One surprising finding was that one fifth of our sample
Number of
Responses of counselors reported that they had personally considered
suicide as an option at some point in their lives. Although
85 there may be additional implications of this last finding, for
53 our purposes, these data provide further validation for our
37
36
decision to survey members of AMHCA and strengthen
23 our confidence in the responses to our survey as reflecting
20 the respondents’ attitudes. Additional support for the va-
12 lidity of the responses to the survey can be interpreted from
11
10 our results related to religious affiliation. As indicated in
9 Table 2, attitudes toward rational suicide reported on the
6 three quantitative items evidenced a somewhat predictable
6
3
pattern regarding religious doctrine (Anderson & Caddell,
1993; Rogers, 1996; Ross & Kaplan, 1993–94). In general,
311 Catholic respondents were least accepting of the concept
of rational suicide across the three items, and religious af-
filiation responses varied as a function of a belief that one’s
life belongs to God.
In terms of the quantitative results, the average response
to Item 1 (“As a professional, how much action would you
take to prevent John/Joan from killing him-/herself?”) sug-
gests that these professionals would not apply a great deal
of action. The mean in this study of 3.89 is between 3.41
(Werth & Liddle, 1994) and 4.65 (Werth, 1996) obtained
in two studies of psychologists. Given that the present mean
is near the middle of the scale, this finding clearly indicates
a moderate acceptance of rational suicide. In fact, the re-
lated qualitative data that were generated indicated that
only 24% of the respondents were unambiguously opposed
to the concept of rational suicide and 40% were clearly sup-
portive. With 76% of the respondents indicating some level
of acceptance of the concept of rational suicide in working
with an individual in a professional context, it seems that
there may be support for the position taken by ACA re-
garding the 1997 Supreme Court decisions on the issue of
physician-assisted suicide. It would be interesting to explore
how these moderately accepting attitudes for rational sui-
cide in a professional relationship fit for these mental health
professionals in relationship to the relatively unambiguous
guidelines embedded in the ACA Code of Ethics (ACA, 1995;
see Werth, 1999b, for an analysis of how the ACA ethics
code can be reconciled with ACA’s decision to sign onto
the amicus brief). Of course, one possibility is that the ex-
pressed attitudes reflected in these data do not translate
into actual professional behavior (see Rogers, 1996 for a
discussion of the attitude/behavior issue as it relates to right
to die concerns).
Similarly, for Item 2 (“As a friend, how much action would
you take to prevent John/Joan from killing him-/herself?”),
the average response of this sample of counselors indicated
a moderate acceptance of rational suicide. Although the av-
erage score on Item 2 was indicative of slightly less action
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79

in preventing suicide when the individual was a friend than
when the individual was a client, the qualitative data sug-
gested that only 16% of the respondents were unambigu-
ously against the idea of rational suicide in this situation
compared with 24% when the individual was a client. Thus,
84% of the respondents reported at least some level of sup-
port for rational suicide, with 59% unambiguously support-
ive when the individual involved was a friend.
Finally for the quantitative data, the average response
to Item 3 (“If you were in John/Joan’s circumstances, how
viable would you view suicide as an option?”) indicated
that, in circumstances similar to those depicted in the sce-
nario, rational suicide would be viewed as a moderately viable
personal option. It is intriguing that as the scenario moved
from depicting client, to friend, to self, rational suicide seemed
to gain in acceptability for these mental health professionals.
It is interesting that the responses to the quantitative
items in the survey instrument did not vary significantly
as a function of respondent sex, race, discipline, work set-
ting, experience, licensure, on the basis of whether the
respondents had clients who attempted or committed suicide,
or as a function of the clinicians’ history of personal
suicidality. The two areas for which the responses did vary
significantly were religious affiliation and vignette sex.
As with the total sample results for the three attitude
items, the religious affiliation data generally suggest that as
the issue of rational suicide moves across the contexts of
client, friend, and self, attitudes become more accepting.
These data seem to make intuitive sense especially because
they may reflect a progressive move from the role of the
professional counselor (where the profession emphasizes
protecting clients from inflicting harm on others or self) to
a personal value position. From the professional role as re-
flected in the item “As a professional, how much action
would you take to prevent John/Joan from killing him-/
herself?” however, these results may indicate the influence
of religious doctrine on therapists’ judgment processes re-
lated to rational suicide communications. This suggests that
counselors choosing to work with individuals about has-
tened death decisions such as rational suicide may need to
engage in personal value clarification exercises in this area
to ensure that they provide appropriate unbiased services
to their clientele (Fenn & Ganzini, 1999; Werth, 1996).
The 81% affirmative response to the question “Do you
believe in the idea of “rational suicide?” provides some
empirical support for the ACA’s decision to support the
WSPA amicus curie brief to the Supreme Court. It is also
consistent with the levels of support for rational suicide
found in research with psychologists (77%, Aulbach, 1997;
81%, Werth & Liddle, 1994; 86%, Werth, 1996) as well as
support for physician-assisted suicide found among psy-
chologists and psychiatrists in Oregon (78% of psycholo-
gists, Fenn & Ganzini, 1999; 66% of psychiatrists, Ganzini,
Fenn, Lee, Heintz, & Bloom, 1996) and social workers in
British Columbia (78%, Ogden & Young, 1998).
The qualitative data generated by participants to support
their affirmative responses were easily classified within the
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Rational Suicide: An Empirical Investigation
criteria reported by Werth and Cobia (1995) and Werth
(1996) and, thus, provide some measure of support for these
classification schemes. In addition, the most typical circum-
stances under which these counselors would consider a
decision to commit suicide to be rational were reported as
(a) terminal illness, (b) severe physical pain, (c) a nonimpulsive
consideration of alternatives, and (d) an unacceptable quality
of life.
Although the findings are consistent with previous
research, they must still be considered preliminary because
of the limitations associated with this study. The relatively
low response rate is a concern; however, given the combi-
nation of funding restraints and the desire to avoid raising
fears about the loss of anonymity by using tracking num-
bers to allow the targeting of nonrespondents, the rate may
be as good as could be expected. Another limitation is the
fact that only AMHCA members were surveyed. Although
the demographic results indicate that the participants were
experienced counselors, nonmembers of this counseling
association may differ systematically from those who choose
to join AMHCA. Thus, it is possible that the current find-
ings, even though empirical in nature, are limited in terms
of generalizability.
However, notwithstanding these limitations, the overall
findings of this study are important for the profession in
that the responses of the sample population indicated a
moderate (based on ratings) acceptance of rational suicide
for clients and for friends, and an even greater acceptance
of rational suicide for the therapists themselves. The re-
sults indicating support for rational suicide are particularly
interesting in light of the ACA (1995) ethical guidelines
that promote clinical involvement, at all costs, in instances
in which suicide is a possible outcome (see also Werth,
1999b). Again, these findings provide valuable insight into
the beliefs and attitudes of therapists regarding rational sui-
cide and may be useful for ACA to consider as individual
states continue to grapple with the issue in the future.
REFERENCES
Ableson, R. P. (1988). Conviction. American Psychologist, 43, 267–275.
American Counseling Association. (1995). Code of ethics and standards
of practice. Alexandria, VA: Author.
Anderson, J. G., & Caddell, D. P. (1993). Attitudes of medical profes-
sionals toward euthanasia. Social Science Medicine, 37, 105–114.
Aulbach, P. (1997). Attitudes of psychologists in regard to prevailing argu-
ments for and against rational suicide as an option for the terminally ill
(Unpublished doctoral dissertation). Spalding University, Louisville, KY.
Battersby, A. (1997, April). Physician-assisted suicide [Letter to the edi-
tor]. Counseling Today, p. 4.
Battin, M. P. (1994). The least worst death: Essays in bioethics on end-of-
life. New York: Oxford University Press.
Brief of the Washington State Psychological Association (WSPA); The Ameri-
can Counseling Association; The Association for Gay, Lesbian, and Bisexual
Issues in Counseling; and the Coalition of Mental Health Professionals Sup-
porting Individual Self-Determination in Decisions to Hasten Death. (1996,
December). Filed with the Supreme Court of the United States, Octo-
ber, 1996 term for Washington v. Glucksberg and Vacco v. Quill.
Brody, H. (1998). Physician-assisted suicide in the courts: Moral equivalence,
double effect, and clinical practice. Minnesota Law Review, 82, 939–963.
371
R o g e r s , G u e u l e t t e , A b b e y - H i n e s , C a r n e y, a n d W e r t h , J r.
Camus, A. (1955). The myth of Sisyphus and other essays. New York:
Knopf.
Chemtob, C. M., Bauer, G. B., Hamada, R. S., Pelowski, S. R., & Muraoka, M.
Y. (1989). Patient suicide: Occupational hazard for psychologists and
psychiatrists. Professional Psychology: Research and Practice, 20, 294–300.
Compassion in Dying v. Washington, 79 F.3d 790 (9th Circuit 1996) (en banc).
Duggan, D. D. (1997, March). Opposed to physician-assisted suicide
[Letter to the editor]. Counseling Today, p. 4.
Fenn, D. S., & Ganzini, L. (1999). Attitudes of Oregon psychologists
toward physician-assisted suicide. Professional Psychology: Research and
Practice, 30, 235–244.
Ganzini, L., Fenn, D. S., Lee, M. A., Heintz, R. T., & Bloom, J. D. (1996).
Attitudes of Oregon psychiatrists toward physician-assisted suicide.
American Journal of Psychiatry, 153, 1469–1475.
Maltsberger, J. T. (1994). Calculated risk-taking in the treatment of sui-
cidal patients: Ethical and legal problems. In A. A. Leenaars, J. T.
Maltsberger, & R. A. Neimeyer (Eds.), Treatment of suicidal people (pp.
195–205). Washington, DC: Taylor & Francis.
Marascuilo, L. A., & Serlin, R. C. (1988). Statistical methods for the social
and behavioral sciences. New York: Freeman.
Morrissey, M. (1997a). ACA, AGLBIC file brief supporting physician-
assisted suicide. Counseling Today (On-line), 1–4. Available: http://
www.counseling.org/ctonline/archives/suicide297.htm
Morrissey, M. (1997b, March). Editor’s note. Counseling Today, p. 4.
Ogden, R. D., & Young, M. G. (1998). Euthanasia and assisted suicide: A
survey of registered social workers in British Columbia. British Jour-
nal of Social Work, 28, 161–175.
Oregon Death With Dignity Act. (1995). Or. Rev. Stat. 127.800–127.995.
Previn, M. P. (1996). Assisted suicide and religion: Conceptions of the
sanctity of human life. The Georgetown Law Journal, 84, 589–616.
Quill, T. E. (1993). Death and dignity: Making choices and taking charge.
New York: Norton.
372 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Quill v. Vacco, 870 F. Supp. 78 (S.D.N.Y. 1994), rev’d, 80 F.3d 716 (2nd
Circuit, 1996).
Rogers, J. R. (1996). Assessing right to die attitudes: A conceptually guided
measurement model. Journal of Social Issues, 52, 63–84.
Rogers, J. R., & Britton, P. J. (1994). AIDS and rational suicide: A coun-
seling psychology perspective or a ride on the slippery slope. The Coun-
seling Psychologist, 22, 171–178.
Ross, L. T., Kaplan, K. J. (1993-94). Life ownership orientation and atti-
tudes toward abortion, suicide, doctor-assisted suicide and capital
punishment. Omega, 28, 17–30.
Stack, S. (1999). The influence of rational suicide on nonrational sui-
cide: A sociological analysis of attitudes. In J. L. Werth, Jr. (Ed.), Con-
temporary perspectives on rational suicide (pp. 41–47). Philadelphia:
Taylor & Francis.
Werth, J. L., Jr. (1992). Rational suicide and AIDS: Consider-
ations for the psychotherapist. The Counseling Psychologist, 20,
645–659.
Werth, J. L., Jr. (1996). Rational suicide? Implications for mental health
professionals. Washington, DC: Taylor and Francis.
Werth, J. L., Jr. (Ed.). (1999a). Contemporary perspectives on rational
suicide. Philadelphia: Taylor & Francis.
Werth, J. L., Jr. (1999b). Mental health professionals and assisted death:
Perceived ethical obligations and proposed guidelines for practice.
Ethics and Behaviors, 9, 159–183.
Werth, J. L., Jr., & Cobia, D. C. (1995). Empirically based criteria for
rational suicide: A survey of psychotherapists. Suicide and Life-
Threatening Behavior, 25, 231–240.
Werth, J. L., Jr., & Liddle, B. J. (1994). Psychotherapists’ attitudes to-
ward suicide. Psychotherapy, 31, 440–448.
Westefeld, J. S., Range, L. M., Rogers, J. R., Maples, M. R., Bromley, J. L,
& Alcorn, J. (2000). Suicide: An overview. The Counseling Psychologist,
28, 445–510.
The Impact of Group Work on Offender Adolescents

Linda L. Viney, Rachael M. Henry, and Joanne Campbell

The impact of personal construct and psychodynamic group work on the psychosocial functioning of offender adolescents was
evaluated. The repeated measures design involved 3 data collections. Data from 102 participants included measures of 5 pairs of
maturational processes and 5 psychological states. Group work was effective immediately after terminating treatment in increas-
ing helpful maturational processes and in reducing the less helpful maturational processes. The group work was also effective to
some extent in reducing less helpful psychological states. The implications of this research for future interventions and its evalu-
ation are discussed.

T
he aim of this research was to evaluate the im-
pact of psychological interventions on offender
adolescents. The significance of this research lies
in the need for effective interventions for ado-
lescents who have repeatedly committed crimes
against property and people (Darden, Gazda, & Ginter, 1996;
Fine, Forth, Gilbert, & Haley, 1991; Quay, 1987). These
interventions should enhance the maturational functioning
and improve the psychological states of these adolescents.
Offender adolescents, being susceptible to peer pressure, are
particularly responsive to interventions that involve peers
(Zimpfer, 1992). Group work may also be the preferred
intervention of offender adolescents because they frequently
operate in gangs (i.e., to make plans, finalize decisions;
Macedo, 1955). The important social processes of group work
for these adolescents include building trust, testing the sta-
bility of their environments, settling struggles for leadership,
exploring interpersonal issues, extending self-understanding,
and accepting responsibility for self (Leader, 1991; MacLennan
& Dies, 1992). Group work was therefore judged appropriate
for the interventions of this research.
Previous research has shown that the psychosocial func-
tioning of offender adolescents is poor, placing them at risk of
major problems. These adolescents have been shown to have
relatively high levels of uncertainty, anxiety, depression, and
anger (Burton, Foy, Bwanausi, & Johnson, 1994; Coolidge,
Reilman, Lee, & Cass, 1992). High levels of traumatic stress
have also been reported in this population (Burton et al., 1994).
The positive affect of the participants is rarely investigated in
outcome research but is included here to tap into their
strengths and resources, as well as their distress.
There is also evidence that offender adolescents’ ability
to cope with negative emotional states is influenced by the
levels of maturity of those adolescents (Erikson, 1959).
Erikson asserted that human development is a life span
process passing through eight stages in a fixed order, each
stage presenting conflicts that become crises. The resolu-
tion of each conflict is also seen as the basis for “tackling”
the next developmental stage. Later theorists dealing with
adolescence have focused on identity foreclosure and dif-
fusion (Marcia, 1980), symbiosis and differentiation (Blos,
1979), and whether the balance that is maintained at this
stage is imperial or interpersonal (Kegan, 1983). Only five
of the Eriksonian stages are seen as relevant to the psy-
chosocially immature adolescents of this study. The levels
of maturity of the participants are to be assessed in terms
of whether they are interacting according to Erikson’s
meanings of trust, autonomy, initiative, industry, and af-
filiation, and whether they have been able to give up, to
some extent, the meanings of mistrust, constraint, hesi-
tancy, inferiority, and isolation.
Evaluation studies of group work (cognitive behavioral,
behavioral, and multicomponent) with such adolescents
yielded encouraging results. Offender adolescents who had
group work have showed improvement in self-esteem
(Fashinger & Harris, 1987), problem solving (Hains,
Herrman, & Balker, 1986), school achievement (Fashinger
& Harris, 1987), empathy with and support of others (Car-
penter, 1984; Darden et al., 1996; Fashinger & Harris, 1987;
Madonna & Caswell, 1991) and reoffending on release
(Fashinger & Harris, 1987; Larson, 1990; Leeman, Gibbs, &
Fuller, 1993) in contrast with offender adolescents who did
not participate in group work. Also, a meta-analysis of nine
evaluation studies that have compared the effects of indi-
vidual and group interventions has indicated consistently
better effects for group than for individual interventions
with adolescents, including offender adolescents, but no such
differential effects occur with children (Tillitski, 1990).

Linda L. Viney is a personal construct therapist and researcher and was director of Clinical Psychology Training Programs at the University of Wollongong,
Wollongong, New South Wales, Australia, when this article was written. Rachael M. Henry is a psychodynamic therapist and researcher and was director of
the Graduate Diploma of Psychodynamic Psychotherapy at the University of Wollongong. Joanne Campbell is a psychologist in New South Wales, Australia.
Correspondence regarding this article should be sent to Linda L. Viney, Department of Psychology, University of Wollongong, Northfields Avenue, Wollongong,
NSW 2500, Australia (e-mail: linda_viney@uow.edu.au).

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 373

V i n e y, H e n r y, a n d C a m p b e l l
Hoag and Burlingame (1997), using a set of meta-analyses,
concluded that group work was effective for adolescents.
The forms of group work to be tested, personal construct
and psychodynamic group work, were those that were se-
lected as alternatives to the prevailing cognitive behavioral
and behavioral group work (e.g., Carpenter, 1984; Fine et
al., 1991; Larson, 1990) for many reasons. They were seen
to be sufficiently complex to do justice to the psychosocial
functioning of offender adolescents (Klein, 1957; Neimeyer,
1993). These approaches have a wide range of strategies
with which to assess clients, the assessment presented here
focusing on both the psychological and the social processes
(Klein, 1948; Neimeyer, 1993). They provide the benefits
of acknowledging the role of unconscious processes in the
functioning of adolescents and of the history of adolescents
as important to their present functioning (Klein, 1948;
Neimeyer, 1993). Finally, both approaches also take the
history of the therapists into account so that the concepts
of transference and countertransference are used by both
of them (Klein, 1957; Viney, 1996). These concepts help to
greatly strengthen the understanding of group work pro-
cesses with offender adolescents.
PERSONAL CONSTRUCT GROUP WORK WITH OFFENDER
ADOLESCENTS
Adolescents, as do others, continually try to predict and
control their worlds through interpretation and anticipa-
tion (Kelly, 1955). For this purpose, they create systems of
constructs, or personal meanings. These meanings develop
through individual experiences and yet are subject to change.
Because of the limited range of the meaning systems of
adolescents, their meanings about self can be invalidated by
others. Anxiety, anger, and threat may then result (Viney,
1996). Such emotions may be especially strong for offender
adolescents whose illegal behavior often seems designed to
elicit invalidation of meanings about self from others. More
mature adolescents who are functioning normally will be
able to discriminate in terms of others and be responsive in
interacting with them. They will also be able to be commit-
ted to and forgiving with others and to respect their own
intimate relationships (Leitner & Pfenninger, 1994). Poorly
functioning and less mature adolescents will develop fewer
of these capacities and develop them to a lesser extent. Their
problematic meanings can be changed through structured
group work, which provides a laboratory in which these prob-
lematic meanings about people can be tested, with different
types of structure providing different experiments for them
(Kelly, 1955). The validating or confirming climate of the
group minimizes, most importantly, the threat to the most
influential meanings about self of the members during this
experimentation (Viney, 1996). It should also reduce their
anxiety and anger because more of their meanings are
effective in predicting and controlling their personal and
interpersonal worlds. Personal construct group work has
proved effective with a range of goals and with clients of
different ages (Winter, 1992). For example, it increased
374 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
the ability of acting out adolescents to understand one an-
other (Jackson, 1990).
PSYCHODYNAMIC GROUP WORK WITH OFFENDER ADOLESCENTS
According to theory, with the onset of puberty, there is an
intensification of sexual impulses, producing a regression
to more immature stages of development and a return of
early infantile impulses (Klein, 1932). Earlier anxieties and
other negative emotions are aroused as earlier unresolved
conflicts threaten to reemerge. The defensive processes of
projection and splitting are readily apparent. Both of these
developments contribute to adolescents’ struggles to achieve
their sense of identity (the central task of adolescence), as
does the young person’s attempt to reestablish the disturbed
emotional equilibrium (Klein, 1957). It seems reasonable to
assume that these struggles will be particularly severe for
offender adolescents because it is more likely that they
will have more unresolved losses and conflicts and less
well developed ego mechanisms of coping and defense. Psychody-
namic group work provides opportunities for understanding
the complex emotional episodes making up group life (Bion,
1961). How the group is used, whether as “waste bin” or
“creative container,” will determine whether a developmental
transformation to a higher level of organization is achieved
(Bion, 1970). Through interpretation, it is possible for the
leaders to demonstrate the following in the here and now:
the emotional attitudes the group has to individuals; the
members’ perceptions of the attitude of the group to each
member; the attitude of the individual to the group; and the
attitude of members to the leader or authority figure (Shapiro,
Zimmer, Berkowitz, & Shapiro, 1963). These processes of
demonstration and interpretation facilitate integration and
reduce splitting and ego fragmentation.
AIM AND HYPOTHESES
The intent of the present research was to build on the prom-
ising research with offender adolescents by evaluating two
forms of group work using the rarely assessed but key matu-
rational processes and psychological states of the participants.
The purpose of this research was to evaluate the effects
of group work on offender adolescents. We investigated the
following hypotheses.
1. Offender adolescents who experience group work com-
pared with those who have no group work will show greater
immediate improvement, in terms of the following:
a. Five pairs of helpful and less helpful maturational
processes——trust and mistrust, autonomy and constraint,
initiative and hesitancy, industry and inferiority, and affilia-
tion and isolation
b. Five psychological states——uncertainty, anxiety, depres-
sion, anger and good feelings (positive affects).
2. Offender adolescents who experience group work will
show more gains 6 months after group work than those who
have no group work.

Tests of some secondary hypotheses about the impact of
participant sex, race, and language use and group leader sex
and age on the effects of the group work were also to be
carried out.
METHOD
Participants and Design
One hundred and two 14- to 18-year-olds volunteered to
participate in the research in three juvenile justice centers.
The offences with which adolescents in the juvenile justice
center had been successfully charged were all repeated of-
fences. The offences were breaking and entering (39%),
unlawful possession and driving of a vehicle (34%), break-
ing and entering with a weapon (19%), and attacks against
persons (8%). Of the sample, 63% were male and 37% were
female. Ages were 14 years (21%), 15 years (27%), 16 years
(38%), 17 years (10%), and 18 years (4%). Sixteen percent
were aboriginal. Nineteen percent reported a language other
than English being spoken at home. The recruitment rate
was 91%. The participants were assigned randomly in this
design to one of two conditions: group work (personal con-
struct/psychodynamic) and control with no group work.
Groups
Both types of group work occurred 1½ hours weekly for
10 weeks. Twelve 10-session groups were run, 6 of personal
construct group work for a third of the sample and 6 of
psychodynamic group work for another third of the sample.
The group work was conducted to follow the manuals that
have been published for the personal construct group work
(Viney, Truneckova, Weekes, & Oades, 1999) and the psy-
chodynamic group work (Henry, 1996). Two group leaders
conducted each group, which had from 5 to 7 members.
All prospective members were screened by appropriate ju-
venile justice center staff to exclude any who were likely to
be physically violent in the group setting. The 16 leaders
volunteered for training in the group work of their choice.
They were all trained mental health professionals, including
registered psychologists and drug and alcohol counselors,
with appropriate degrees for their positions. All of them
were experienced in leading therapeutic groups, with years
of experience ranging from to 2 to 16 years. Ten were women
and 6 were men. Two of the leaders were aboriginal. Their
ages ranged from 23 to 45 years. The leaders attended a
1-day training workshop and read and discussed the train-
ing manuals, together with selected background material.
Under the guidance of the manuals’ authors, they also had
2 hours of peer and expert supervision weekly for 14 weeks
in which conceptualization of the group processes in each
session took place, followed by planning for the next ses-
sion. (Linda L. Viney supervised the personal construct
group work and Rachael M. Henry the psychodynamic
group work.)
The personal construct group work. The personal construct
group work used a structured group format (Viney et al.,
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
The Impact of Group Work on Offender Adolescents
1999) because higher structure of such groups has been
linked with more therapeutic movement (G. J. Neimeyer
& Merluzzi, 1981). The tasks used aimed to extend the range
of interpretations of the offenders. Tasks were selected ini-
tially by the leaders, but, as the group developed, they were
often selected jointly by leaders and group members. Such
tasks included the interpersonal transaction group exercise
(Landfield & Rivers, 1975; R. A. Neimeyer, 1988), involv-
ing responses to some carefully chosen questions: “Who am
I?” “How are my expectations of the opposite sex like their
expectations of ours, and how are they different?” “How
can I give help to others and seek help from them?” “When
do I feel safe, and when do I not?” and “What would I most
like to say to each group member?” Yet another structuring
task involved role play of common but problematic interac-
tions with other adolescents, such as negotiating a sexual
interaction, and then one that is safe from sexually transmit-
ted diseases. Such an enactment, followed by role reversal,
aided in helping group members to better understand the
meanings used by their potential sexual partners in this area,
as well as their own. In the second to last group session,
members created, alone or jointly, a certificate to represent
what they had learned from the group.
The psychodynamic group work. The psychodynamic group
work used an unstructured discussion or activity group (Henry,
1996; Henry, Wesley, Jones, Cohen, & Fairhall, 1997). This
group work involved (a) rapid establishment of a working
collaboration between members and leaders by providing struc-
ture and fostering group cohesiveness; (b) intense, focused
work on a carefully delineated problem area; (c) careful,
interpretive attention to group process about the problem
theme and to processes in the self; and (d) a productive
and manageable termination. Although technically unstruc-
tured, subtle but essential forms of structure were supplied
within the group. For example, both physical and psycho-
logical boundary control required careful management of
what belonged psychologically inside or outside the group.
Drawing boundaries also included protecting time and com-
municating its importance, as well as making the setting
physically secure. Psychological continuity was provided
through elements of ritual (e.g., beginning and ending a
session and other procedural matters) to reduce the anxi-
ety of members to a tolerable level where intense feelings,
impulses, and thoughts could be expressed in ways that
were not overwhelming or inhibiting. Firm boundaries
fostered a sense of group cohesiveness in which mem-
bers could gain or regain a sense of kinship, which then
allowed optimism and facilitated group and individual
development. Working with the dynamics, process, and
emotional realm of groups, rather than with content
alone, required the leaders to be clear about the instru-
mental purposes of group work, to develop and refine an
understanding of each person in the group, to keep in touch
with the dynamics of the group as a whole, to keep in
touch with their own feelings, to note their own behavior
and its consequences, and to see connections between
group and individual dynamics.
375
V i n e y, H e n r y, a n d C a m p b e l l
Treatment validation. Given the more active overtly struc-
turing role of the personal construct leader compared with
that of the psychodynamic leader that is apparent in our
accounts of these groups, we hypothesized that if these group
processes were representative of the two types of group work,
then members of the personal construct group would report
a higher level of activity for leaders than would members
of the psychodynamic group but that there would be no
differences for other contributions of the leaders in both
groups. Samples of 5 participants in personal construct
group work and another 5 in psychodynamic group work
were asked to report their impressions of the contributions
of leaders and other members to the group process. They
used a 3-point ranking scale to report on the activity level
(choice of topic and talk) of leaders and members and of
their other contributions (their levels of thinking and feel-
ing). Agreement of rankings between each of the two sets
of group members was considerable, averaging 88% for both
types of group work. These hypotheses about the differen-
tial rankings of the two types of participants were supported.
Outcome Measures
Content analysis scales. The maturational processes and psy-
chological states of these research participants were assessed
using content analysis scales applied to the following open-
ended request (the responses were recorded with permis-
sion): “Now I’d like you to talk for a few minutes about
your life at the moment, the good things and the bad, what
it’s like for you.” Content analysis scales, based on thematic
analysis, result in continuous measurement with normal-
ized distributions of scores (Gottschalk, Lolas, & Viney,
1986; Viney, 1983a). Such scales were preferred to ques-
tionnaire and inventory measures of the variable assessed
for these reasons. The content analysis of free responses in
these scales overcomes many of their problems, such as the
ambivalence of the emotions of offender adolescents and
the effects of social desirability on their responses. It also
avoids the use of selected words to identify for the partici-
pant the area of interest to the researcher. Scales based on
that content analysis also make possible an ethical, honest
approach to offender adolescents while giving them the
opportunity to deal with what is important to them. The
results from content analysis scales are less influenced by
interviewer characteristics than are those from other types
of measures (Gottschalk, 1982). Content analysis scales are
also conceptually and methodologically compatible with both
the personal construct and the psychodynamic approaches
and have proved useful in other studies of adolescents (Viney,
1983b, 1987; Wang & Viney, 1996, 1997).
Maturational processes. Maturational processes were
measured by applying the Content Analysis Scales of Psy-
chosocial Maturity (CASPM; Viney & Tyche, 1985) to answers
to the open-ended question provided. Five of the eight
pairs of positive and negative scales making up CASPM
were used because of their appropriateness for this type of
participant (i.e., offender adolescents). These scales assess
the helpful and less helpful meanings people use to deal
376 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
with Erikson’s (1959) epigenetic tasks: Trust and Mistrust,
Autonomy and Constraint, Initiative and Hesitancy, In-
dustry and Inferiority, and Affiliation and Isolation. Follow-
ing are examples of statements from research participants
that would be scored on the first pair of these scales: for
the Trust scale—“I get on good with my mum now”; for the
Mistrust scale—“I try to stay out of the house when my
stepfather is home.” A manual to facilitate the scoring of
these scales is available (Viney, Rudd, Grenyer, & Tyche,
1995). These CASPM scales have appropriate levels of
interjudge reliability, with a range of Pearson’s product–
moment correlation coefficients from .80 to .95 and no sig-
nificant differences in the mean score sizes of independent
judges. They have good test–retest reliability ranging from
.65 to .71 (Viney & Tyche, 1985). There has been an exten-
sive network of evidence for their construct validity. Most
important, they have shown the expected discriminations
for 813 participants of the age group 7 to 96, both in Aus-
tralia and in the United States, and with both White and
African American youth (Viney, 1987). The scales have been
shown to discriminate as predicted between children and
adolescents of different ages (Wang & Viney, 1997) and cul-
tures in the People’s Republic of China and in Australia
(Wang & Viney, 1996). The scales have discriminated between
people with different health, religious, and employment
status, as well as relating differentially as predicted to the
measures of psychological states subsequently described
(Viney & Tyche, 1985). As measures of psychotherapy pro-
cess, the scales have successfully predicted outcome
(Grenyer, Viney, & Luborsky, 1996).
Psychological states. The psychological states of the of-
fender adolescents were also assessed using content analy-
sis scales applied to answers to the same open-ended inter-
view question. Five content analysis scales were selected to
measure the psychological states of uncertainty, anxiety,
depression, and anger, as well as positive affect. The Cogni-
tive Anxiety Scale (Viney & Westbrook, 1976) is a measure
of the uncertainty that occurs when a person has difficulty
in making sense of their experiences, as for example in a
new, unfamiliar situation (e.g., “I was not sure what was
happening”). The Total Anxiety Scale (Gottschalk, 1982)
consists of six subscales, measuring different types of anxi-
ety, examples of which follow for the first two of these
subscales: Death (“She is scared that I am going to get
killed”), Bodily Mutilation (“I broke my knuckles trying to
belt the crap out of somebody”). The Hostility In Scale
(Gottschalk, 1982) measures depression, focusing on self-
critical responses in which anger is directed to the self (“I
am not good at anything”); and the Hostility Out Scale as-
sesses anger directly expressed to others or to the external
world (“Most of the guys in charge are nerds”). The Posi-
tive Affect Scale (Westbrook, 1976) taps good, positive feel-
ings associated with happy, enjoyable experiences, which
are contrasted with disruptive and disconcerting experiences
(“I enjoy hanging out with my friends”). The Cognitive
Anxiety, Total Anxiety, Hostility In and Hostility Out, and
Positive Affect Scales have appropriate levels of interjudge

reliability, with a range of Pearson’s product–moment corre-
lation coefficients from .71 to .99, and with no significant
differences in mean score size for the two independent
judges. Good test–retest reliability would not be appro-
priate for these measures of transient psychological states.
Evidence of their criterion and construct validity has been
provided by many studies, showing the correlation of the
scale scores with self reports and psychiatric ratings of
these states and with other measures of the same psy-
chological states, including psychophysiological indices,
discrimination of appropriate samples of research partici-
pants and situations, and responsivity to psychoactive drugs
(Gottschalk, 1982; Gottschalk et al., 1986; Viney, 1983a;
Viney, Henry, Walker, & Crooks, 1989; Viney et al., 1994).
Procedure
The data were collected during standardized interviews in
the juvenile justice centers before and after the group work
and at 9-months follow-up. The interviews took ½ hour to 1
hour. One female interviewer, experienced in working with
adolescents, conducted all of the interviews. This interviewer
was not one of the group leaders and was blind to the treat-
ments to which the research participants had been assigned.
The responses of the participants to the previously pro-
vided open-ended question in the interview were tran-
scribed and divided into clauses, each one containing an
active verb. The responses were content analyzed follow-
ing the standard scoring instructions for each of the con-
tent analysis scales (Viney, 1983a; Viney et al., 1995). Each
clause was compared with the sets of content analysis cat-
egories, which provided verbal cues for each statement that
could be scored. Clauses matching these verbal cues were
then summed, and the total score thus calculated was mul-
tiplied by a weight representing the verbal productivity of
each participant. The final total score consisted of the square
root of this multiplied score, the square root procedure
designed to normalize each distribution of scores. The con-
tent analysis was conducted by a trained and experienced
coder, a female psychologist, who remained blind to the
treatment group to which each research participant had
been assigned. For a subsample of 30 transcripts, interjudge
reliability (Pearson’s product–moment correlation coeffi-
cients) with an independent blind scorer (also a female psy-
chologist) varied from .82 to .94 for the 15 content analy-
sis scales. There were no significant differences between
the mean scores of the two judges.
RESULTS
The design and analyses of the study are described, then
the preliminary data analyses dealing with retention rates
and analyses of bias, together with analyses of the possible
secondary effects of both participant and leader variables,
are provided. Comparisons of data at Times 1 and 2 are
then presented for the maturational processes and psycho-
logical states. The results of comparisons of the same vari-
ables at Times 1 and 3 are presented later.
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
The Impact of Group Work on Offender Adolescents
The sample of offender adolescents had been assigned
randomly to the two conditions of the treatment factor of
the design: group work and no-group-work. The similarity
of the distributions for age, for aboriginal status, and for
use of language other than English in the home across the
group work section of the design was apparent, although
the no-group-work condition proved to have a few more
female adolescents than did the group work condition. Sex
of the leader was balanced over the intervention cells of
the design. For the measures of maturational processes
(trust, autonomy, initiative, industry and affiliation on the
one hand, and mistrust, constraint, hesitancy, inferiority, and
isolation on the other) and psychological states (uncertainty,
anxiety, depression, anger, and good feelings), two-factor
repeated measures multivariate analyses of variance
(MANOVAs) were used to test for differences according
to treatment versus no treatment (group). Because of the
large number of follow-up comparisons, alpha levels were
set at p = .01.
Preliminary Analyses
The retention rate in the group from Time 1 to Time 2
three months later was high for such mobile populations at
78%, leaving a subsample at Time 2 of 78. From Time 1 to
Time 3 nine months later the retention rate was consider-
ably lower at 55%, leaving subsample of 55. This retention
rate was low, but not unexpected, for a sample of offend-
ers. Only 17% of those who had dropped out of the group
at Time 3 refused to participate. The other 83% had relo-
cated to distant towns in New South Wales, Queensland,
or Victoria, distances that were not covered by the research
funding. When MANOVAs were conducted to test differ-
ences between Time 1 scores for the two groups using each
of the four sets of their variables, no significant differences
were found. MANOVAs also tested for bias in these data
by contrasting those who had dropped out of the group
with those who had remained at Time 2 for their sets of
demographic variables, together with their two sets of five
CASPM scales and their set of five other content analysis
scales measuring psychological states at Time 1. They all
showed no significant differences. The same comparisons
were also made for those who had dropped out of the group
with those who had remained at Time 3, on their Time 1
data, with the same consistent lack of any significant results.
The retained samples at both Times 2 and 3 were demo-
graphically comparable then, as well as psychologically so.
Also, tests of several effects showed no significance for any
of the sets of content analysis scale scores: participant sex,
race, language usage, and leader sex and leader age.
Time 1 Versus Time 2 Analyses
Maturational processes. These results are described in terms
of the positive and then the negative CASPM scores. The
means and standard deviations for five positive CASPM
scale scores dealt with first are available in Table 1.
MANOVA showed significant effects for group (multivari-
377
V i n e y, H e n r y, a n d C a m p b e l l

TABLE 1

Means and Standard Deviations for Five Positive Content Analysis Scales of Psychosocial Maturity
(CASPM) for Offender Adolescents With Group Work and No-Group-Work at Times 1, 2, and 3
Positive CASPM
Trust Autonomy Initiative Industry Affiliation
Variable M SD M SD M SD M SD M SD
Group Work
Time 1 1.82 0.73 1.05 0.59 1.86 1.22 1.86 1.22 1.41 0.95
Time 2 1.82 0.60 0.85 0.47 1.41 0.72 2.66 1.12 1.64 0.96
Time 3 1.30 0.72 0.95 0.61 1.39 0.81 3.01 1.10 1.32 0.89
No-Group-Work
Time 1 1.79 0.62 1.18 0.65 1.93 0.97 1.93 0.97 1.99 1.17
Time 2 1.87 0.64 1.01 0.61 1.34 0.78 1.87 1.04 2.00 1.10
Time 3 1.55 0.57 0.81 0.36 1.28 0.61 2.61 1.28 1.66 1.15

ate F = 10.76, df = 5, 96, p < .001) in comparisons of group
work and no-group-work for Time 1 and Time 2 scores.
The subsequent univariate analyses for these scores showed
Industry (univariate F = 7.10, df = 1, 98 p < .01) and Affili-
ation (univariate F = 7.86, df = 1, 98, p < .01) to make
significant contributions to the group effect. Inspection of
the means of Table 1 indicated that those offender adoles-
cents who experienced interventions in the treatment con-
dition showed more gains in Industry and Affiliation at Time
2 than did those who had the no-group-work condition.
The group (multivariate F = 10.13, df = 5, 96, p < .01)
effect proved significant for comparisons of group work
with no-group-work for the five negative CASPM scale
scores for Times 1 and 2. The means and standard devia-
tions for five negative CASPM scales are available in Table
2. The subsequent univariate analyses for the group effect
on these scores showed Hesitancy (univariate F = 8.04, df =
1, 98, p < .01) to make the main significant contribution to
this effect. The participants in the treatment condition
showed more reduced Hesitancy at Time 2 than those in
the control condition.
Psychological states. The group (multivariate F = 3.98, df
= 5, 96) showed significance (assuming p = .10) for com-
parisons of group work with no-group-work for the five
psychological states. Although it did not achieve significance
at .01, we believe this finding has important clinical impli-
cations. The means and standard deviations for five scales
measuring the psychological states are available in Table 3.
The subsequent univariate analyses for the group effect in
these scale scores revealed Uncertainty (univariate F = 8.33,
df = 1, 98, p < .01) to make the main significant contribu-
tion to this effect.
Time 1 Versus Time 3 Analyses
MANOVA analyses identified no significant differences in
the sets of maturational processes and psychological states
data from Time 3 (9 months after the end of the group
work) and Time 1.
DISCUSSION
The confirmation and lack of confirmation for the hypoth-
eses about the impact of the group work for the four sets
of data immediately after its conclusion are first consid-
ered, together with the results for these data sets 9 months

TABLE 2

Means and Standard Deviations for Five Negative Content Analysis Scales of Psychosocial Maturity
(CASPM) for Offender Adolescents With Group Work and No-Group-Work at Times 1, 2, and 3
Negative CASPM
Mistrust Constraint Hesitancy Inferiority Isolation
Variable M SD M SD M SD M SD M SD
Group Work
Time 1 1.39 0.55 2.12 0.81 1.42 0.92 0.87 0.46 1.12 0.75
Time 2 1.00 0.37 1.86 0.86 1.13 0.88 0.78 0.46 0.89 0.52
Time 3 1.12 0.62 1.33 0.42 1.30 0.82 0.79 0.40 0.92 0.62
No-Group-Work
Time 1 1.31 0.61 2.07 1.14 1.42 1.03 0.89 0.61 1.05 0.64
Time 2 1.01 0.39 1.67 0.63 1.57 1.14 1.02 0.63 1.05 0.63
Time 3 1.07 0.65 1.23 0.65 1.13 0.81 0.69 0.34 1.16 0.77

378 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79

 The Impact of Group Work on Offender Adolescents

TABLE 3

Means and Standard Deviations for Five Scales Measuring Psychological States for Offender Adolescents
With Group Work and No-Group-Work at Times 1, 2, and 3
Psychological State Scales

Uncertainty Anxiety Depression Anger Positive Affect

Variable M SD M SD M SD M SD M SD
Group Work
Time 1 1.46 0.86 2.11 0.93 2.16 0.97 1.91 0.72 1.21 0.60
Time 2 1.26 0.77 1.52 0.81 1.26 0.67 1.33 0.54 1.38 0.66
Time 3 1.25 0.88 1.34 0.76 1.65 0.60 1.36 0.75 1.28 0.59
No-Group-Work
Time 1 1.39 0.81 2.26 0.93 2.33 1.17 1.67 0.87 1.24 0.53
Time 2 1.59 0.93 1.55 0.77 1.53 0.65 1.25 0.66 1.34 0.68
Time 3 1.18 0.81 1.39 0.91 1.48 0.84 1.26 0.63 1.41 0.51

later. Finally, some criticisms of and recommendations from
the research are considered.
The first set of hypotheses was concerned with the im-
mediate impact of the group work on offender adolescents.
The hypothesis that those adolescents who had experienced
group work compared with those who had no group work
would experience gains immediately after was confirmed
overall for the five positive, helpful maturational processes.
In particular, the offender adolescents showed greater gains
in industry and affiliation with others after group work com-
pared with those who had no group work. The linked hy-
pothesis about reductions in the five negative, less helpful
processes was also confirmed overall and, in particular, offender
adolescents who had group work showed less hesitation than
those who had no group work. This first hypothesis about the
immediate gains of group work was also confirmed to some
extent for the five psychological states; in particular, offender
adolescents undergoing group work reported less uncertainty
than did those who had no group work. Effect sizes (Cohen,
1988) were calculated for the samples with interventions
from Time 1 to Time 2 (.55 for Industry, .61 for Affinity, and
.25 for Hesitancy), showing only small to moderate effects.
These effects of the group work did not vary according to
the sex, aboriginal background, or language usage of the group
members or to the sex or age of the group leaders. These
overall gains for offender adolescents support the use of the
group work with them.
These offender adolescents showed higher levels of mis-
trust, constraint, and hesitancy, together with anxiety and
depression than did those in other samples of adolescents
(Viney, 1983b, 1987). The types of offenses for which the
offender adolescents were sentenced are meaningfully re-
lated to these findings. It was often violence, mainly against
property and sometimes against people. The fact that of-
fender adolescents derived benefit from group work is con-
sistent with findings in the United States that offender ado-
lescents are responsive to interventions (Fashinger & Har-
ris, 1987; Leeman et al., 1993; Madonna & Caswell, 1991).
No research into the potential for interventions for this vul-
nerable group of adolescents has been conducted previously
in Australia. The results of these short-term group work
evaluations are encouraging. Counselors planning to use
group work with offender adolescents first need to assess
their maturational processes and their psychological states.
Given that at least short-term goals were accomplished
by this group work, it is our intent to briefly focus now on
the findings from another examination of these data that
looks at differences in the effects of personal construct and
psychodynamic group work described in detail elsewhere
(Henry, Viney, & Campbell, 1996; Viney, Henry, &
Campbell, in press). The personal construct group work
proved to be particularly effective in reducing less helpful
maturational processes such as mistrust, as well as distressing
psychological states, especially anger, for offender adolescents.
The psychodynamic group work proved to be effective in
increasing more helpful maturational processes such as ini-
tiative, as well as distressing psychological states such as
uncertainty and depression. One of the main implications of
these findings for practice with offender adolescents is that
they highlight the need to select the goals for group work
to match them with forms of group work that are likely to
achieve these goals.
Although the immediate gains for offender adolescents
from the group work were encouraging, the gains in matu-
rational processes and psychological states at the conclu-
sion of the group work of the second set of hypotheses
were not maintained at 9 months after its conclusion. Two
possible causes for this finding are considered. First, one
possible reason for the failure to maintain gains over the
longer term may be that the brief interventions that we
were able to offer simply did not match the level of depri-
vation our participants encountered. The poor psychoso-
cial functioning apparent in the scale scores of offender
adolescents in these samples, when they were compared
with those of adolescents in other samples, supports this
explanation. More sustained and longer term group work,
or perhaps even more costly individual or multicomponent
interventions, may be needed. That interpretation of the
pattern of findings here received informal support from
reflections of group leaders and professionals responsible

JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79 379

V i n e y, H e n r y, a n d C a m p b e l l
for the welfare of the offender adolescents. Second, a com-
mon criticism of outcome measures that demonstrate short-
term gains is that impact measures, selected to tap directly
into the content of the interventions, falsely exaggerate the
impact of the interventions. This possibility can be ruled
out in the present study, which used instruments with in-
dependent content, with demonstrated reliability and va-
lidity, and that were sensitive to underlying processes rather
than to surface content. The instruments used, although
experienced-based, were more subtle and so more demand-
ing than self-reports of gains by participants.
The aim of this research, to evaluate the impact of two
forms of group work with offender adolescents, was
achieved. The findings from the interviews collected im-
mediately after the intervention were encouraging. The
group work was effective in increasing helpful maturational
processes, especially ability to work in a sustained way and
to achieve a sense of relationship with others, and in reduc-
ing the less helpful maturational processes, especially lack
of confidence. The interventions were also effective in re-
ducing less helpful psychological states, especially uncer-
tainty. However, the gains were not found months later. It
is possible that group work with more than 10 sessions, say
15 or 20, would show gains over the follow-up period. Be-
cause that solution could, in fact, double the cost of the
intervention, the use of a few booster group sessions, or
even a telephone session with a group leader, would be
worth evaluating. The encouraging findings can probably
be generalized to offender adolescent populations because
those who had dropped out of the group were not distin-
guishable demographically or psychologically from those
who stayed. Yet, some criticism of the study must be made.
Its sampling has been based on volunteers, limiting the
generalizability of the findings only to those who would
choose to be involved in such work. The retention rate im-
mediately after the group work was just 78%, and at follow
up, only 55%. Also, the design of the research could have
been improved by the inclusion of more extensive treat-
ment validation data and a better nontherapeutic group
work control.
Further research examining the impact of group work on
offender adolescents with larger samples, better retention
rates, better validated interventions, a better nontherapeutic
group work contrast condition, a wider range of appropri-
ate outcome measures, closer examination of group pro-
cess with these and other models of adolescent functioning
is recommended. We have already begun the examination
of group process in this area (Henry, 1996; Henry et al.,
1997; Viney et al., 1999). Promoting psychological matu-
ration and more helpful psychological states in offender
adolescents still remains a goal with high priority, and the
findings of this research make some contribution to achiev-
ing that goal using two promising forms of group work.
Author note. This research was funded by a grant from
the Australian Commonwealth Committee on AIDS Re-
search Grants (CARG) to the two senior authors. Permis-
sion was kindly given to conduct the research by the New
380 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
South Wales Department of Juvenile Justice, Australia. The
support of administrative and professional staff in this or-
ganization has been very much appreciated. We would also
like to thank the senior staff of the Keelong, Mount Penang
and Reiby Juvenile Justice Centres for making this work
possible. Furthermore, the gifts made to us of time and at-
tention from the busy lives of our adolescent research par-
ticipants was invaluable.
Among the leaders of the preventive group work, the
seminal contributions of two should be specially acknowl-
edged: Patricia Weekes and Lindsay Oades. The important
contributions of the other group leaders who trained in and
conducted the group work is also very gratefully acknowl-
edged: Craig Perkins, Tony Weaver, Ann Harwood, Ann
Rodum, Hany Gayed, Jenny Barton, Annalisa Dezarnaulds,
Nasim Wesley, Harry Smith, Monique Cohen, Greg Konza,
Chris Fairhall, Renata Kautz, Paul Whetham, Cassandra Mc
Naught and Alexandra Jones. Peter Caputi, Ross Colquhoun,
Marilyn Rudd and Lindsay Oades also aided us in our analy-
ses of the data.
REFERENCES
Bion, W. R. (1961). Experiences in groups. London: Tavistock Publications.
Bion, W. R. (1970). Attention and interpretation. London: Tavistock
Publications.
Blos, P. (1979). The adolescent passage: Developmental issues. New York:
International Universities Press.
Burton, A. R., Foy, D. W., Bwanausi, C., & Johnson, P. (1994). The rela-
tionship between traumatic exposure, family dysfunction and post-
traumatic stress symptoms in male juvenile offenders. Journal of Trau-
matic Stress, 7, 83–93.
Carpenter, P. (1984). “Green Therapy” revisited: The evolution of 12
years of behaviour modification and psychoeducationist techniques
with young delinquent boys. Psychological Report, 54, 99–111.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences.
Hillsdale, NJ: Erlbaum.
Coolidge, F. L., Reilman, B. J., Lee, L. A., & Cass, V. J. (1992). Emotional
problems and neuropsychological symptoms in juvenile non-offenders.
Journal of Personality and Clinical Studies, 8, 7–13.
Darden, C. A., Gazda, G. M., & Ginter, E. J. (1996). Life- skills and mental
health counseling. Journal of Mental Health Counseling, 18, 134–141.
Erikson, E. (1959). Growth and crises in the healthy personality. Psycho-
logical Issues, 1, 50–100.
Fashinger, E., & Harris, L. T. (1987). Social work at 30MPH: Mini-bike
rehabilitator groups for juvenile delinquents. Social Work with Groups,
10, 33–45.
Fine, S., Forth, A., Gilbert, M., & Haley, G. (1991). Group therapy for
adolescent depressive disorder: A comparison of social skills and thera-
peutic support. Journal of the Academy of Child and Adolescent Psy-
chiatry, 30, 79–85.
Gottschalk, L. A. (1982). Manual of uses and applications of the
Gottschalk-Gleser verbal behaviour scales. Research Communications
in Psychiatry, 7, 273–327.
Gottschalk, L. A., Lolas, F., & Viney, L. L. (Eds.). (1986). Content analysis
of verbal behaviour in clinical medicine. Heidelberg, Germany: Springer.
Grenyer, B. F. S., Viney, L. L., & Luborsky, L. (1996, August). Changes in
Eriksonian psychosocial maturity levels over the course of psycho-
therapy and their relationship to interpersonal mastery and psycho-
logical health-sickness. Twenty-ninth Annual Meeting of the Society
for Psychotherapy Research, Amelia Island, FL.
Hains, A. A., Herrman, L. P., & Balker, K. L. (1986). The development of
a psycho-education group program for adolescent sex offenders. Jour-
nal of Offender Counselling Services and Rehabilitation, 11, 63–76.

Henry, R. M. (1996). Psychodynamic group psychotherapy with adoles-
cents: Exploration of HIV-related risk taking. International Journal of
Group Psychotherapy, 46, 229–253.
Henry, R. M., Viney, L. L., & Campbell, J. (1996). Mechanisms for change
in adolescent group interventions for HIV risk-taking: A contrast of
psychodynamic and personal construct models. Unpublished manu-
script, University of Wollongong, Wollongong, New South Wales,
Australia.
Henry, R. M., Wesley, N., Jones, A., Cohen, M., & Fairhall, C. (1997) AIDS
and the unconscious. Journal of Child Psychotherapy, 23, 373–398.
Hoag, M. J., & Burlingame, G. M. (1997). Child and adolescent group
psychotherapy: A narrative review of effectiveness and the case for meta-
analysis. Journal of Child and Adolescent Group Therapy, 7, 51–68.
Jackson, S. (1990). A personal construct therapy group for adolescents.
In P. Maitland (Ed.), Personal construct theory, deviancy and social work
(pp. 1–72). London: Centre for Personal Construct Psychology.
Kegan, R. (1983). A neo-Piagetian approach to object relations. In B. Lee
& G. G. Noam (Eds.), Developmental approaches to the self (pp. 34–
48). New York: Plenum.
Kelly, G. A. (1955). The psychology of personal constructs. New York:
Norton.
Klein, M. (1932). The psychoanalysis of children. London: Hogarth.
Klein, M. (1948). Contributions to psychoanalysis 1921–1945. London:
Hogarth.
Klein, M. (1957). Envy and gratitude. London: Tavistock.
Landfield, A. W., & Rivers, P. C. (1975). An introduction to interper-
sonal transaction and rotating dyads. Psychotherapy, 12, 366–374.
Larson, J. D. (1990). Cognitive-behavioural group therapy with delin-
quent adolescents: A cooperative approach with the juvenile courts.
Journal of Offender Rehabilitation, 16, 47–64.
Leader, F. (1991). Why adolescent group therapy? Journal of Child and
Adolescent Group Therapy, 1, 81–93.
Leeman, L. W., Gibbs, J. C., & Fuller, D. (1993). Evaluation of a multi-
component group treatment program for juvenile delinquents. Ag-
gressive Behaviour, 19, 281–292.
Leitner, L. M., & Pfenninger, D. T. (1994). Sociality and optimal func-
tioning. Journal of Constructivist Psychology, 7, 119–136.
Macedo, G. (1955) Group psychotherapy in juvenile criminology. Inter-
national Journal of Group Psychotherapy, 5, 54–60.
Madonna, J. M., & Caswell, P. (1991). The utilization of flexible tech-
niques in group therapy with delinquent adolescent boys. Journal of
Child and Adolescent Group Therapy, 1, 147–157.
Marcia, J. E. (1980). Identity in adolescence. In J. Adelson (Ed.), Hand-
book of adolescent psychology (pp. 97–108). New York: Wiley.
MacLennane, B. W., & Dies, K. R. (1992). Group counseling and psycho-
therapy with adolescents. New York: Columbia University.
Neimeyer, G. J. (Ed.). (1993). Constructivist assessment: A casebook. New
York: Sage.
Neimeyer, G. J., & Merluzzi, T. V. (1981). Group structure and group
process: Explorations in therapeutic sociality. Small Group Behavior,
13, 150–164.
Neimeyer, R. A. (1988). Clinical guide-lines for conducting Interper-
sonal Transaction Groups. International Journal of Personal Construct
Psychology, 1, 181–190.
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
The Impact of Group Work on Offender Adolescents
Neimeyer, R. A. (1993). An appraisal of constructivist psychotherapies.
Journal of Consulting and Clinical Psychology, 61, 221–234.
Quay, H. C. (1987). Handbook of juvenile delinquency. New York: Wiley.
Shapiro, R. L., Zimmer, J., Berkowitz, D. A., & Shapiro, E. R. (1963). The
impact of group experiences on adolescent development. In M. Sugar
(Ed.), The adolescent in group and family therapy (pp. 145–161). New
York: Brunner Mazel.
Tillitski, C. J. (1990). A meta-analysis of estimate sizes for group versus
individual versus control treatments. International Journal of Group
Therapy, 40, 215–224.
Viney, L. L. (1983a). The assessment of psychological states through content
analysis of verbal communications. Psychological Bulletin, 94, 542–563.
Viney, L. L. (1983b). Psychological reactions of young people to unem-
ployment. Youth and Society, 14, 457–474.
Viney, L. L. (1987). A sociophenomenological approach to life span de-
velopment complementing Erikson’s psychodynamic approach. Hu-
man Development, 30, 125–136.
Viney, L. L. (1996). Personal construct therapy: A handbook. Norton, NJ:
Ablex.
Viney, L. L., Henry, R. M., & Campbell, J. (in press). Evaluating personal
construct and psychodynamic group work with offenders and non-
offenders. In G. J. Neimeyer & R. A. Neimeyer (Eds.), Advances in
personal construct psychology, Vol. 4. New York: JAI Press.
Viney, L. L., Henry, R. M., Walker, B. M., & Crooks, L. (1989). The emotional
reactions of HIV positive men. British Journal of Medical Psychology, 62,
153–161.
Viney, L. L., Rudd, M., Grenyer, B., & Tyche, A. M. (1995). Content
Analysis Scales to Measure Psychosocial Maturity (CASPM): A manual.
Wollongong, New South Wales, Australia: University of Wollongong.
Viney, L. L., Truneckova, D., Weekes, P., & Oades, L. (1999). Personal
construct group work for adolescent offenders: Dealing with their
problematic and more helpful meanings. Journal of Child and Adoles-
cent Group Work, 9, 169–185.
Viney, L. L., & Tyche, A. M. (1985). Content analysis scales measuring
psychological maturity in the elderly. Journal of Personality Assess-
ment, 49, 311–317.
Viney, L. L., Walker, B. M., Robinson, T., Pincombe, J., Lilley, B., & Ewan, C.
(1994). Dying in palliative care units and hospital: Quality of life of terminal
cancer patients. Journal of Consulting and Clinical Psychology, 40, 157–165.
Viney, L. L., & Westbrook, M. T. (1976). Cognitive anxiety: A method of
content analysis for verbal samples. Journal of Personality Assessment, 40,
140–150.
Wang, W., & Viney, L. L. (1996). A cross-cultural comparison of
Eriksonian psychosocial development: Chinese and Australian chil-
dren. School Psychology International, 17, 33–48.
Wang, W., & Viney, L. L. (1997). The psychosocial development of chil-
dren and adolescents in the People’s Republic of China. International
Journal of Psychology, 32, 139–153.
Westbrook, M. T. (1976). The measurement of positive affect using content
analysis scales. Journal of Consulting and Clinical Psychology, 12, 85–86.
Winter, D. (1992). Personal construct psychology in clinical practice. Lon-
don: Routledge.
Zimpfer, D. F. (1992). Group work with delinquents. Journal for Special-
ists in Group Work, 17, 116–126.
381
 TRENDS
Finding Happiness for Ourselves and Our Clients
Geri Miller

D. G. Myers (2000) examined the contributing factors of happiness: money, relationships, and religion. The implications of these factors for
counseling are discussed with specific recommendations made for counselors regarding their own self-care and their work with their clients.

F
inding Happiness for Ourselves and Our Clients”
is a review of Myers (2000) article “The Funds,
Friends, and Faith of Happy People.” Myers pro-
vided an analysis of the current research on hap-
piness and made a plausible argument for the
necessary follow-up research regarding happiness and its
causes. The impact of culture, personality, and activities on
happiness were examined. In addition, Meyers emphasized
the relationship of three factors (money, relationships,
religion) with regard to happiness. The main points of the
article are reviewed here, followed by implications for the
professional counselor.
ARTICLE REVIEW
Myers (2000) began the article with an overall view of happi-
ness: “our working definition is simply whatever people mean
when describing their lives as happy” (p. 57). He noted that
according to research, age and gender do not seem to be con-
tributing factors to happiness. He also discussed factors that
do seem to contribute to happiness. One of these is culture,
especially those cultures that offer individuals opportunities
for political freedom and affluence. Others are personality traits
and temperament such as extroversion. Finally, he cited the
work of Csikszentmihalyi (1999), who has studied what he
calls “flow.” This experience of flow, described as an activity
that involves “skills, concentration, and perseverance”
(Csikszentmihalyi, 1999, p. 825), is enjoyed for its own sake
even though there may not be a specific result from the
activity; a person “loses oneself” in a mindful challenge. Happi-
ness, then, is relevant to an ability to experience flow during
activities, with individuals experiencing happiness because of
“how” they do an activity, not “what” they do.
In the article, Myers examined in depth three factors that
potentially contribute to happiness: money, relationships, and
religion. Regarding money, there seems to be a connection to
well-being: some wealthy countries have more satisfied people.
The very poor tend not to be happy, but as people become
more comfortable, money does not seem to make a substan-
tial difference. Once individuals have the necessities of life
and are not extremely poor, then income does not seem to
make a difference in happiness (people can adapt to
changes so that an earlier extravagance becomes a neces-
sity). Myers (2000, p. 61) cited the example of happiness
not being related to the economy in what he calls “the
American paradox”: Americans are prosperous materially,
but have a recession socially.
When discussing relationships, Myers highlighted the impor-
tance of belonging and how the need to belong can result in
healing or pain. The research indicates that close relationships
seem related to good health, whereas a lack of meaningful rela-
tionships can be harmful. For example, social support of close
friends is related to happiness; attached people (married) are
happier overall than unattached people. Yet there are some
exceptions to these findings. Although people in happy mar-
riages tend to be some of the happiest people, people in
unhappy marriages are even less satisfied than unmarried
or divorced people. The happiness in marriage may be a result
of happy people tending to marry (they are more enjoyable to
be around than unhappy people) or possibly due to the
benefits of intimacy.
The third area, religious faith, also seems to have an impact
on happiness. Individuals with faith seem better able to handle
crises in their lives. The well-being related to faith may be a
result of social support because religious practices usually oc-
cur within the context of a community. The well-being may
also be related to meaning and purpose because faith provides
individuals with a framework from which they can view their
lives as making a difference. Finally, faith can provide people
with hope as they face the existential questions connected to
death and suffering in their lives. Therefore, faith seems to
provide people a cushion when coping with life’s struggles.

Geri Miller is an associate professor in the Department of Human Development and Psychological Counseling at Appalachian State University in Boone,
North Carolina, and a licensed psychologist and certified substance abuse counselor in North Carolina. Correspondence regarding this article should be sent
to Geri Miller, Department of Human Development and Psychological Counseling, Edwin Duncan Hall, Appalachian State University, Boone, NC 28608 (e-
mail: millerga@appstate.edu).

382 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79


In summary, the author noted that to understand happiness
better, researchers need to examine in more depth personality
traits, work and leisure activities, relationships, and religious
faith. A deeper understanding of these general factors can fa-
cilitate how our culture looks at its priorities and encourages
the well-being of humans. So how can counselors make use of
such conclusions? What are the implications of Myers’s analy-
sis of happiness research?
IMPLICATIONS FOR COUNSELORS
Initially counselors may simply say, “Of course” to reviews of
research on happiness: “I already knew that.” “I already en-
courage that with my clients.” However, a challenge to the
counseling profession is to suspend such typical responses and
wonder how much of this information is truly taken in and
known by counselors. How much do counselors practice these
factors related to happiness in their own lives, and how much
do counselors encourage the presence of these factors in the
lives of their clients?
A similar challenging process may occur in counselor edu-
cation training in which students learning about the impor-
tance of listening (i.e., of hearing someone’s story in counsel-
ing) may say, “I know, I know” to their professor “Now, teach
me some counseling techniques, something I can do. Listen-
ing is easy. I already know how to do that.” And the professor
responds, “Do you already know how to listen? Do you really
listen to others in your life? To your clients? Do you stop
everything, all thoughts of the past and future and focus to-
tally on the present and the person with you? Listening is the
hardest, most important activity done with a client. Listening
is hard work. It begins with slowing oneself down. By slowing
down, the counselor can hear and care for self better, thereby
allowing the counselor to hear and care for others better.”
The happiness research findings involve similar dynamics:
what sounds simple and easy may be simple but not easy
to achieve. Just as clients can sense if counselors are really
listening to them, they may also sense the counselor’s level of
happiness. It is difficult for counselors to encourage happiness
in clients if they are not happy themselves. So the questioning
begins with counselors’ happiness: Are counselors happy? Do
they experience the flow as described by Csikszentmihalyi
(1999)? Just as counselors can pass on happiness to their clients,
so can they pass on unhappiness. “Happiness choice points” are
critical for counselors. For example, do counselors emphasize
aspects of counseling such as the restrictions of managed care
and the possibility of liability or do counselors emphasize their
participation in the healing of others where they witness and
honor the sacredness of hearing another’s story? Do counselors
nurture their own personal happiness and that of their clients?
Implications of these happiness findings for counseling, then, begin
with counselors’ lives and extend to the lives of their clients.
Money
In the last few years, managed care issues have placed tre-
mendous constraints on the counseling profession. These
constraints unfortunately result in the counselor’s need to make
JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
Finding Happiness for Ourselves and Our Clients
professional concessions regarding techniques, for example,
using more behavioral focused techniques and having the
general feeling of “needing to do more with less.” These added
stresses cause some counselors to struggle financially, and these
stresses place heavier administrative burdens on them (reim-
bursement of services, accountability, additional paperwork,
ethical dilemmas related to diagnosis and delivery of services).
These additional stresses can result in negativity and burn-
out for counselors. In addition, given the important work of
counselors, they may legitimately view themselves as sub-
stantially underpaid, resenting their financial difficulties.
Yet the research regarding happiness does not indicate that
a change in income is necessarily related to happiness. Perhaps
the struggle here is related more to the loss of the flow, in that
the activity of counseling has changed to the point that one
cannot become “lost” in the skillful process. One must always
be mindful that “others” are controlling when and how
services are delivered or that one’s work is not valued in terms
of financial reimbursement within the culture.
Our clients may struggle with the same types of thoughts
about money. If only they had more money they would
have less stress, fewer problems. This misconception may
be quite strong for our clients, causing them to stay in rela-
tionships or jobs that are very difficult and draining. The
exception may be for those counselors who work directly
with the extremely poor and witness the trauma poverty
causes. Our work with low-income individuals entails help-
ing them locate resources that ease the burden of their pov-
erty. However, those clients whose basic needs are met may
be caught in the illusions of money equalling happiness. In
this instance, the counselor can approach the illusion as
any other addressed in the session: with a sensitive, focused
inquiry that encourages the client to examine his or her
own life regarding that aspect.
Relationships
Although all human beings vary in terms of the number
and type of intimate relationships they require, the under-
lying need to belong is a core, common one in relationships
that counselors must examine in their lives. What are our
own needs for belonging? I remember a respected clinician
telling me one time, “I need fewer intimacies outside of
being a counselor because I have many of my intimacy needs
met with my clients.” Although I am unaware of any time
that this counselor acted inappropriately regarding client
involvement, I was concerned for the well-being of both
counselor and client if the need for closeness was being
almost entirely met through the counseling relationship.
Yet there is a genuine danger for all counselors, personally
and professionally, of relying on closeness with our clients
in lieu of a life outside the counseling session. It is critical
that we have a life away from clients’ problems—that we
have relationships in which we can simply be, without trans-
ference issues influencing them.
The social support of friends is one avenue for our self-care.
Individuals who know us and are willing to be honest with us
provide the needed support to survive our demanding work.
383
Miller
This does not mean that these relationships should be en-
tirely focused on a serious processing of thoughts and feelings
but that they should also allow us to play and to live fully—
to be real, genuine, and spontaneous in our interactions. Such
a network provides a life away from our work that sustains us.
Social support is also critical for our clients. Who in his or
her life is a natural healer? With whom, after sitting in his or
her presence, does the client simply feel better about life? Is it
a religious leader, a hair stylist, a neighbor? Assisting clients in
identifying their natural healers allows them to structure ac-
tivities to involve those individuals. Counselors need to en-
courage clients to seek people who are naturally good matches
for them, people that make them feel as though the world is a
good, safe, nurturing place in which to live. Once this is ac-
complished (and in the case of some of our clients, they may
truly be starting from “scratch”), they need to work at main-
taining those contacts. This may be especially important in a
culture where the client receives conflicting messages as to
the importance of being independent and “pulling oneself up
by the bootstraps.” Because the reality of needing and relying
on another individual may be difficult for the client, the coun-
selor should continue to encourage the client to make and main-
tain these connections.
Regarding marriage or other forms of intimate commitment,
the counselor and client may again face the same struggles. If the
counselor has a difficult partnership, how much happiness does
this drain from the counselor? Where does the counselor draw
support? Counselors who are involved in a happy intimate
relationship can use this relationship as a resource of sustenance
both personally and professionally, but the counselor in an
unhappy relationship may need a better honed support network
to compensate for the unmet needs of the committed relation-
ship. As in other difficult situations, counselors themselves
might need counseling to determine the best long-range plan.
Finally, counseling may be beneficial for those counselors who
anguish over the lack of an intimate relationship.
Clients go through a process similar to that of counse-
lors. Those who are in unhappy relationships may use
counseling as a safe place to process their options; a place
not affected by family, friends, or others invested in one
option over another. The nonjudgmental counselor offers
a safe, healing place for the client with the time and space
to sort. Those clients who are single may need to review
their conceptions of marriage or a committed relation-
ship and process their suffering over the lack of such a
support. They may also need assistance in examining where
and how they are looking for a partner. Finally, those cli-
ents who have a sustaining intimate relationship may use
that resource as one that helps them cope with the other
areas in which they struggle.
Religion
Of the factors related to happiness as reviewed by Myers,
this one may be the most uncomfortable for some counse-
lors. If so, reframe it by discussing religion in the sense of
384 JOURNAL OF COUNSELING & DEVELOPMENT • SUMMER 2001 • VOLUME 79
what force(s) outside of the counselor helps him or her feel
that living has meaning and purpose, experience a sense of
hopefulness, and simply enjoy being alive. For some counse-
lors, this evolves from a religious perspective, whereas for
others, it may be nature, for example. It is related once again
to Csikszentmihalyi’s (1999) sense of flow: an activity that
revitalizes us, that provides us with a sense of being alive.
These same resources need to be explored with our cli-
ents. Depending on the client and their spiritual views, this
avenue toward happiness may take different shapes. Our
responsibility is to carefully and respectfully access with our
clients those views that provide him or her a sense of mean-
ing, purpose, hope, and life enjoyment. Then, for our clients
as well as for ourselves, we should encourage participation
in activities related to these resources. Some of our clients
have such narrow life views that they are not aware of avail-
able resources. For example, clients caught in an addiction
cycle may not know of other avenues that could be fulfilling
for them because they are so lost in their addiction. Here the
counselor may need to educate the client about options and
encourage a process of self-exploration. The importance of
this area of faith, no matter how uncomfortable, cannot be
neglected by the counselor due to its potentially significant
impact on assisting the client in coping with life.
SUMMARY
Myers’s examination of three factors related to happiness
(money, relationships, and religion) is important for counselors
and their clients. To be qualitatively helpful to their clients,
counselors need to examine their own lives in relation to
these factors and remedy those areas where possible. With-
out examining our own wounds and strengths we reduce
our clients’ possibilities for happiness through our intentions
or our ignorance. Through self-examination and a commit-
ment to self-care, we can diminish our suffering and increase
our happiness. This process can serve as both a role model
for our clients and a wellspring of happiness and joy that can
have a contagious effect on our clients. We can counsel cli-
ents from a place of balance within ourselves that is main-
tained by both internal and external resources.
By improving our chances of happiness, we increase the likeli-
hood of our clients “catching” happiness from us. Such a focus on
happiness in life can also influence how we punctuate our clients’
stories and assist them in enhancing the happiness of their lives, the
functional aspects, and reducing the unhappiness, the dysfunctional
aspects. A commitment to the happiness, the well-being of our-
selves and our clients results in a tremendous amount of healing
merely as a natural part of the counseling process. This healing is
deep, lasting, and meaningful in the context of our clients’ lives.
REFERENCES
Csikszentmihalyi, M. (1999). If we are so rich, why aren’t we happy?
American Psychologist, 54, 821–827.
Myers, D. G. (2000). The funds, friends, and faith of happy people. Ameri-
can Psychologist, 55, 56–67.