JOURNAL OF PERSONALITY ASSESSMENT, 78(3), 391404

Copyright 2002, Lawrence Erlbaum Associates, Inc.

A New Nonpatient Sample for the

Rorschach Comprehensive System:
A Progress Report

John E. Exner, Jr.

Rorschach Workshops
Asheville, North Carolina

Some concern about recently published nonpatient data, plus the fact that some re-
cords in the published sample of 600 nonpatients (Exner, 2001) were collected more
than 20 years ago, prompted the initiation of a project to develop a new nonpatient
sample. In this article, findings are presented for the first 175 participants tested in the
new project. It is noted that the data from the new project are quite similar to those for
the sample of 600. Results suggest that the similarities between the data sets supports
the notion that the published sample is probably representative and provides a reason-
able basis from which to identify response rates and the proportions of nonpatients ex-
pected to fall in, or outside of, established parameters for various ratios and other
structural indexes.

An understanding of the Rorschach is usually enhanced by an appreciation for the

response rates by which the features of answers occur and the relation of those re-
sponse rates to the numerous structural features of the test. The specific codes as-
signed to responses form the base from which this array of data evolves. The codes
are used to identify the features of responses for which there is some empirical link-
age to interpretation. In turn, the frequencies for the various coded features create
the platform from which the variety of structural data that forms the interpretive
core of the test are derived.
Ordinarily, normative (nonpatient) data that have been established for the test
provide the source from which this information is gleaned. The data are useful in
providing descriptive information about groups and offer reference points against
which individual scores can be compared. Possibly most important is the basis that
392 EXNER

they provide for developing some general interpretive postulates using the devia-
tion principle, which focuses on findings different than expected.
Almost any normative samples will have limitations, and those formed to create
tables for use with the Comprehensive System are no exceptions. The nonpatient
records used to create those samples were collected over a period of more than 10
years (19731986). As the number of available records increased, the tables were
revised three times as various attempts to stratify the sample ensued (Exner, 1978,
1986, 1990; Exner, Weiner, & Schuyler, 1976). A major revision occurred in 1990
because of findings indicating that brief recordsthat is, those containing less
than 14 answersare likely to be invalid (Exner, 1988). Those findings made it
necessary to discard all records with less than 14 answers that had been included in
previously developed samples. By that time, the pool of nonpatient protocols num-
bered more than 1,100 records, and the revised sampleselected using the stratifi-
cation criteria of sex, geographic area, and socioeconomic levelincluded
findings for 700 persons.

FURTHER REVISION OF THE SAMPLE

When the sample of 700 was selected, using the stratification criteria, more than
200 duplicate records were inadvertently included. They were detected in 1999,
and the duplicates were removed from the sample. The majority have been replaced
with records from the nonpatient pool that had not been used previously and that in-
cluded demographic features that would not alter the stratification of the sample
significantly. The current published sample consists of data from the protocols of
600 adult nonpatients (Exner, 2001). The sample represents 300 males and 300 fe-
males, with 120 participants from each of five geographic areasNortheast, South,
Midwest, Southwest, and Westand is partially stratified for socioeconomic level.
Attempts were made to equalize the number of males and females from each region
but that was not always possible. Thus, the number of males and females are nearly
equal for four regions, but the southwest group includes 72 females and 48 males,
whereas the midwest group contains 74 males and 46 females.

GENERAL DESIGN FOR COLLECTING

NONPATIENT DATA

The 600 protocols were collected by 42 examiners working within the constraints
of problems created by participant recruitment and sample sizes. No examiner con-
tributed more than 25 protocols to this sample. All of the participants were, in one
sense or another, volunteers. None had special reasons to be examined, and none
had any significant psychiatric history. About 17% (101) of the participants gave
histories that included having eight or fewer contacts (the maximum permitted for
 NEW NONPATIENT SAMPLE 393

inclusion) in their past with psychologists or educational counselors. Sixty-nine did

so for purposes of academic or vocational counseling. Nineteen had been involved
in brief marital counseling, and 13 had received brief supportive treatment follow-
ing the loss of a family member or friend.
Volunteers were recruited by letters distributed at workplaces or through coop-
erating organizations. The letters discounted any possibility of feedback concern-
ing results and identified the project as one involving the standardization of the
test. The 600 participants in the 2001 sample include 409 who volunteered through
their places of work, usually under conditions of encouragement by supervisors or
union leaders, and typically were provided with time away from work for the test-
ing. An additional 153 persons volunteered through social or interest organizations
to which they belonged, such as the PTA, Audubon groups, bowling leagues, and
so on, and the remaining 38 were recruited through the assistance of social service
agencies. None were financially reimbursed for their participation, although all re-
ceived greeting cards of appreciation.

REPRESENTATIVENESS OF THE SAMPLE

Among the issues that confront those who use normative data as a guide for inter-
pretation are the extent to which the data are truly representative of nonpatients and
whether the sample has been cross-validated. Recently, Shaffer, Erdberg, and
Haroian (1999) used graduate students to test 123 volunteer nonpatients. Many of
their findings are very similar to those in the published sample of 600 nonpatients.
However, there are also some striking differences. The mean for R is 20.83, with a
median of 18 and a mode of 14, as contrasted with a mean R of 22.32, with a median
of 22 and a mode of 23 for the sample of 600. Lambda values greater than 0.99 oc-
curred for 51 (41%) of the participants in the Shaffer et el. sample as compared to 58
(10%) in the sample of 600. Some other substantial differences between the two
samples include means of .48 versus .67 for the Afr, .78 versus .92 for the XA%,
and .82 versus .94 for the WDA%.
The differences between the samples for the XA% and WDA% reflect differ-
ences for the other variables related to form use. The means are .51 versus .77 for
the X + %, .28 versus .15 for the Xu%, and .21 versus .07 for the X %. Only 44
(36%) of persons in the Shaffer et al. sample gave at least one texture answer,
whereas 490 (82%) of those in the larger sample did so. Nearly 30% of the persons
in the Shaffer et al. sample gave at least one reflection response as contrasted with
only 8% in the sample of 600. In addition, the means for WSumC, EA, and es in the
Shaffer et al. sample are about 2 points lower than those for the sample of 600. In-
terestingly, when the Shaffer et al. sample was increased to include 283
nonpatients (Shaffer & Erdberg, 2001), only modest changes in the values for
these variables were noted.
394 EXNER

The differences noted between the Shaffer et al. data and those in the current
Comprehensive System sample of 600, plus the fact that some of those data were
collected more than 20 years ago, prompted the Research Council of Rorschach
Workshops1 to recommend the collection of a new nonpatient sample to ascertain
the ultility of the published sample.

DESIGN USED FOR THE NEW SAMPLE

This project was initiated during the fall of 1999, using essentially the same design
and exclusionary criteria as for the original project during the period from 1973 to
1986. There are three differences in this model. About 75% of the nonpatients
tested between 1973 and 1981 were recruited by persons employed primarily to so-
licit participants for various investigations. In this project, examiners recruited
their own participants through organizations and businesses using essentially the
same procedures previously used by recruiters, including a slightly modified ver-
sion of the solicitation letter (see Appendix). A second difference concerns pre-
scribed medications or illegal drug use. Questions concerning these issues were not
asked of persons tested between 1973 and 1986. However, both seem more impor-
tant at this time. Anyone volunteering was tested, but the records of those having a
prolonged or significant history involving prescribed psychotropics (N = 3) or ad-
mitting to regular use of illegal drugs (N = 0) were excluded from the sample. A
third difference concerns the financial structure. Persons tested in the original pro-
ject were not paid. In this project, volunteers also were not paid, but as a form of en-
ticement, a $25 donation was paid in the name of the participant to any recognized
charity that the person selects.
More than 10 years were required to obtain the pool of more than 1,100 records
from which to form the stratified sample representing 600 nonpatients. This pro-
ject is proceeding at about the same pace. During the first 2 years since the project
was initiated, nearly 200 participants, from 14 states2, have been tested by 13 expe-
rienced examiners. Each examiner codes the records that he or she has collected
and the coding is reviewed for errors in the central office of Rorschach Workshops
at the time it is computer entered. In addition, every fourth protocol is recoded by
the author, and percentages of correct agreement are recorded as a tactic to review
scoring accuracy and provide feedback to examiners when relevant.

1Research Council membership at the time of the recommendation included Thomas Boll, Philip

Erdberg, John Exner, Mark Hilsenroth, Gregory Meyer, William Perry, and Donald Viglione.
2States represented in this sample include Alaska, California, Georgia, Iowa, Kentucky, Maryland,

New Hampshire, New Jersey, North Carolina, Tennessee, Texas, Virginia, Vermont, and West
Virginia.
 NEW NONPATIENT SAMPLE 395

Percentages of correct agreement, calculated for 40 records, yield findings

ranging from 95% to 99% for W, D, S, Dd, DQ+, DQo, M, active and passive
movement, Sum T, Sum V, Pairs, Reflections, P, Zf, and the Special Scores INC,
FAB, and COP. Results ranging from 90% to 94% agreement include DQv/+,
DQv, pure F, FM, m, FC, Sum Y, FD, the Form Quality codes for ordinary, un-
usual, and minus, and the Special Scores AB, AG, DV, and MOR. Agreement lev-
els between 86% and 89% exist for CF, Pure C, Sum C', the Special Score DR, and
the Form Quality coding of plus.

RESULTS

Descriptive statistics for the various codings and some calculated variables for the
first 175 persons included in this project are shown in Table 1. Table 2 includes
some demography information concerning the group, plus frequency data for vari-
ous cutoff points related to some general interpretive principles.
Table 3 includes data for 12 variables from the published sample of 600
nonpatients and those for the first 175 records collected in this project. These are
the variables described earlier for which there were noticably discrepant findings
between the Shaffer et al. sample and the sample of 600 participants. The large
samples sizes yielded very large degrees of freedom when t tests were used to
study differences between the means, and resulting p values can be misleading or
misinterpreted. Thus, it seems more appropriate to evaluate the differences by us-
ing the values for Cohens d and Pearson correlations to provide information about
effect sizes (Cohen, 1992). These are also included in Table 3.
The data for the sample of 600 and the new sample of 175 are reasonably simi-
lar for all 12 variables. The mean for the X + % in the new sample is .09 lower, the
mean for the Xu% is .06 higher, and the mean for the X % is .03 higher than in the
sample of 600. Using Cohens classification for effect sizes, the first two indicate a
large effect, whereas the third represents a medium effect. A medium effect is also
indicated for the XA% and WDA%. These may represent true differences between
the two samples. However, as suggested by Meyer (2001), differences in pre- and
post-1990 data sets for Form Quality scores may simply reflect the revisions and
expansions of the Form Quality table during the 15 year period after the first ver-
sion of the table was published. Regardless, the differences between the two
groups for Form Quality variables are rather modest and do not suggest the need
for any change in the basic rules of interpretation. In fact, the data for all of the
variables shown in Table 1 indicate reasonable similarities with the data published
for the sample of 600.
There are some differences worth noting in some of the frequency data shown
in Table 2. As noted in Table 3, even though the mean values for Lambda are es-
396
TABLE 1
Descriptive Statistics for 175 Nonpatient Adults

Variable M SD Minimum Maximum Frequency Mdn Mode Skewness Kurtosis

Age 35.53 13.23 19.00 86.00 175 32.00 24.00 1.26 1.70
Years of education 14.15 1.73 10.00 20.00 175 14.00 14.00 0.61 0.35
R 22.98 5.51 14.00 51.00 175 22.00 21.00 1.65 5.14
W 9.20 4.23 3.00 37.00 175 8.00 8.00 2.35 10.59
D 12.45 5.11 0.00 32.00 173 13.00 14.00 0.19 1.31
Dd 1.33 [1.47] 0.00 8.00 116 1.00 0.00 1.74 4.31
S 2.23 [1.96] 0.00 17.00 162 2.00 1.00 3.36 19.57
DQ+ 8.33 3.13 1.00 21.00 175 8.00 9.00 0.73 2.05
DQo 13.93 4.59 4.00 36.00 175 14.00 14.00 1.02 3.57
DQv 0.41 [0.70] 0.00 3.00 52 0.00 0.00 1.74 2.47
DQv/+ 0.32 [0.70] 0.00 6.00 44 0.00 0.00 4.04 26.03
FQx+ 0.47 [0.87] 0.00 5.00 52 0.00 0.00 2.27 5.92
FQxo 14.95 3.39 8.00 29.00 175 15.00 16.00 0.33 1.25
FQxu 5.03 2.77 1.00 17.00 175 5.00 3.00 1.64 4.25
FQx 2.33 1.73 0.00 12.00 163 2.00 2.00 2.09 7.51
FQxNone 0.20 [0.49] 0.00 3.00 29 0.00 0.00 2.76 8.54
MQ+ 0.31 [0.62] 0.00 3.00 42 0.00 0.00 1.94 3.15
MQo 3.86 1.88 1.00 9.00 175 4.00 3.00 0.36 0.60
MQu 0.41 0.77 0.00 5.00 49 0.00 0.00 2.46 8.00
MQ 0.17 [0.44] 0.00 2.00 24 0.00 0.00 2.75 7.10
MQNone 0.00 [0.00] 0.00 0.00 0 0.00 0.00
SQual 0.53 [0.82] 0.00 4.00 67 0.00 0.00 1.76 3.12
M 4.75 2.19 1.00 11.00 175 5.00 3.00 0.37 0.11
FM 3.86 1.98 0.00 10.00 169 4.00 4.00 0.42 0.53
m 1.47 1.35 0.00 10.00 138 1.00 1.00 2.08 8.79
FC 2.81 1.79 0.00 8.00 161 2.00 2.00 0.60 0.11
 CF 2.93 1.89 0.00 12.00 165 3.00 2.00 1.49 4.92
C 0.23 [0.54] 0.00 3.00 33 0.00 0.00 2.70 8.24
Cn 0.00 [0.00] 0.00 0.00 0 0.00 0.00
Sum Color 5.97 2.68 0.00 14.00 174 6.00 5.00 0.54 0.42
WSum C 4.69 2.27 0.00 15.00 174 4.50 4.00 1.09 2.87
Sum C' 1.64 [1.31] 0.00 8.00 146 1.00 1.00 1.26 2.90
Sum T 0.90 [0.64] 0.00 2.00 130 1.00 1.00 0.08 0.52
Sum V 0.37 [0.82] 0.00 5.00 42 0.00 0.00 3.23 13.06
Sum Y 0.82 [1.06] 0.00 7.00 95 1.00 0.00 2.44 9.60
Sum Shading 3.72 2.27 0.00 14.00 173 3.00 3.00 1.67 4.21
Fr + rF 0.26 [0.81] 0.00 7.00 26 0.00 0.00 4.81 30.77
FD 1.42 [1.13] 0.00 5.00 136 1.00 1.00 0.67 0.12
F 7.97 3.51 1.00 23.00 175 7.00 7.00 0.96 2.28
(2) 8.51 2.69 2.00 21.00 175 8.00 9.00 0.66 2.50
3r + (2)/R 0.41 0.11 0.13 0.87 175 0.39 0.38 0.93 2.84
Lambda 0.61 0.38 0.06 2.33 175 0.50 0.50 1.43 2.65
FM + m 5.33 2.55 0.00 20.00 174 5.00 6.00 1.34 6.30
EA 9.43 3.39 2.00 24.00 175 9.50 8.00 0.82 2.08
es 9.05 3.99 2.00 34.00 175 9.00 8.00 1.98 9.16
D Score 0.03 0.90 3.00 3.00 60 0.00 0.00 0.45 3.83
AdjD 0.29 0.89 3.00 3.00 73 0.00 0.00 0.05 2.71
a (active) 6.54 2.84 0.00 18.00 174 6.00 6.00 0.44 1.07
p (passive) 3.57 2.09 0.00 13.00 169 3.00 3.00 0.98 2.23
Ma 2.86 1.66 0.00 8.00 164 3.00 3.00 0.50 0.32
Mp 1.90 1.33 0.00 7.00 154 2.00 2.00 0.99 1.84
Intellect 2.26 2.08 0.00 12.00 141 2.00 1.00 1.40 2.83
Zf 13.34 4.55 2.00 41.00 175 12.00 12.00 2.04 8.62
Zd 0.47 3.88 13.50 12.00 164 0.50 0.50 0.16 1.30
Blends 5.55 2.72 0.00 18.00 172 5.00 4.00 0.54 1.58
Blends/R 0.24 0.11 0.00 0.53 172 0.24 0.24 0.00 0.30
(continued)
397
 TABLE 1 (Continued)
398

Variable M SD Minimum Maximum Frequency Mdn Mode Skewness Kurtosis

Col-Shd Blends 0.65 [0.91] 0.00 6.00 82 0.00 0.00 2.43 9.42
Afr 0.61 0.18 0.21 1.26 175 0.63 0.67 0.19 0.35
Populars 6.30 1.58 1.00 11.00 175 6.00 7.00 0.32 1.02
XA% 0.89 0.07 0.69 1.00 175 0.90 0.90 0.75 0.49
WDA% 0.91 0.06 0.69 1.00 175 0.92 0.95 0.98 1.29
X+% 0.68 0.11 0.35 0.95 175 0.69 0.67 0.55 0.25
X% 0.10 0.06 0.00 0.31 163 0.10 0.05 0.77 0.72
Xu% 0.21 0.09 0.05 0.49 175 0.21 0.16 0.50 0.07
Isolate/R 0.21 0.10 0.00 0.60 171 0.20 0.16 0.56 1.11
H 3.12 1.71 0.00 10.00 172 3.00 3.00 0.82 1.00
(H) 1.39 1.09 0.00 6.00 137 1.00 1.00 0.86 1.33
HD 1.06 [1.03] 0.00 5.00 116 1.00 1.00 1.07 1.46
(Hd) 0.59 [0.87] 0.00 4.00 68 0.00 0.00 1.44 1.52
Hx 0.14 [0.50] 0.00 4.00 18 0.00 0.00 4.75 27.45
H + (H) + Hd + (Hd) 6.15 2.53 0.00 18.00 174 6.00 7.00 0.96 3.05
A 8.05 2.65 3.00 25.00 175 8.00 6.00 1.78 8.80
(A) 0.32 [0.58] 0.00 3.00 47 0.00 0.00 1.82 3.20
Ad 2.70 [1.57] 0.00 9.00 169 3.00 2.00 0.79 0.99
(Ad) 0.14 [0.40] 0.00 2.00 22 0.00 0.00 2.85 7.96
An 0.87 [1.12] 0.00 7.00 98 1.00 0.00 2.15 6.67
Art 1.27 1.31 0.00 6.00 112 1.00 0.00 1.09 1.13
Ay 0.59 [0.76] 0.00 4.00 82 0.00 0.00 1.47 2.78
Bl 0.27 [0.55] 0.00 3.00 39 0.00 0.00 2.16 4.79
Bt 2.25 1.54 0.00 6.00 150 2.00 2.00 0.36 0.48
Cg 1.99 1.56 0.00 8.00 143 2.00 2.00 0.92 1.31
Cl 0.22 [0.49] 0.00 2.00 33 0.00 0.00 2.17 4.00
Ex 0.20 [0.51] 0.00 4.00 29 0.00 0.00 3.59 18.27
Fi 0.75 [0.87] 0.00 4.00 95 1.00 0.00 1.36 2.31
Food 0.30 [0.56] 0.00 3.00 45 0.00 0.00 1.90 3.67
 Ge 0.09 [0.33] 0.00 2.00 14 0.00 0.00 3.80 15.15
Hh 1.11 0.97 0.00 5.00 123 1.00 1.00 0.72 0.45
Ls 1.04 1.16 0.00 9.00 112 1.00 1.00 2.51 12.88
Na 0.51 [0.89] 0.00 6.00 62 0.00 0.00 2.70 10.50
Sc 1.55 [1.23] 0.00 6.00 143 1.00 1.00 0.97 1.03
Sx 0.18 [0.49] 0.00 3.00 25 0.00 0.00 3.01 9.74
Xy 0.09 [0.29] 0.00 1.00 16 0.00 0.00 2.86 6.25
Idiographic 0.37 [0.61] 0.00 3.00 53 0.00 0.00 1.60 2.19
DV 0.42 [0.71] 0.00 5.00 58 0.00 0.00 2.43 9.62
INCOM 0.72 [0.90] 0.00 4.00 85 0.00 0.00 1.25 1.28
DR 0.82 [0.93] 0.00 7.00 101 1.00 0.00 2.11 10.20
FABCOM 0.43 [0.68] 0.00 3.00 61 0.00 0.00 1.72 3.11
DV2 0.00 [0.00] 0.00 0.00 0 0.00 0.00
INC2 0.03 [0.18] 0.00 1.00 6 0.00 0.00 5.16 24.94
DR2 0.03 [0.18] 0.00 1.00 6 0.00 0.00 5.16 24.94
FAB2 0.06 [0.23] 0.00 1.00 10 0.00 0.00 3.84 12.96
ALOG 0.06 [0.23] 0.00 1.00 10 0.00 0.00 3.84 12.96
CONTAM 0.00 [0.00] 0.00 0.00 0 0.00 0.00
Sum6 Sp Sc 2.58 1.79 0.00 10.00 155 2.00 2.00 0.78 1.09
Lvl 2 Sp Sc 0.13 [0.37] 0.00 2.00 20 0.00 0.00 2.96 8.70
WSum6 7.08 5.35 0.00 28.00 155 6.00 0.00 1.03 1.66
AB 0.20 [0.51] 0.00 3.00 27 0.00 0.00 2.82 8.26
AG 0.94 1.08 0.00 7.00 103 1.00 0.00 1.73 5.45
COP 2.09 1.35 0.00 6.00 157 2.00 2.00 0.54 0.06
CP 0.00 [0.00] 0.00 0.00 0 0.00 0.00
GOODHR 5.13 2.15 0.00 13.00 173 5.00 4.00 0.42 0.68
POORHR 1.97 1.57 0.00 8.00 147 2.00 1.00 1.10 1.56
MOR 0.92 [0.96] 0.00 4.00 103 1.00 0.00 0.84 0.13
PER 0.87 [1.01] 0.00 7.00 101 1.00 0.00 1.95 7.37
PSV 0.10 [0.33] 0.00 2.00 15 0.00 0.00 3.63 13.73
399

Note. Standard deviations shown in brackets indicate that the value is probably unreliable and/or misleading and should not be used to estimate expected
ranges. Ordinarily these variables should not be included in most parametric analyses.
 TABLE 2
Demography Data and Frequencies for 36 Variables for Nonpatients
DEMOGRAPHY VARIABLES
MARITAL STATUS AGE RACE
Single . . . . . . . . . . . . . . . . . . . 45 26% 1825 . . . . . . 44 25% White . . . . . . . . . . . . . 150 86%
Lives w/S. . . . . . . . . . . . . . . . . . 6 3% 2635 . . . . . . 56 32% Black . . . . . . . . . . . . . . 14 8%
Married . . . . . . . . . . . . . . . . . . 91 52% 3645 . . . . . . 44 25% Hispanic . . . . . . . . . . . . 10 6%
Separated. . . . . . . . . . . . . . . . . . 6 3% 4655 . . . . . . 15 9% Asian . . . . . . . . . . . . . . . 1 1%
Divorced . . . . . . . . . . . . . . . . . 23 13% 5665 . . . . . . 10 6% Other . . . . . . . . . . . . . . . 0 0%
Widowed . . . . . . . . . . . . . . . . . . 4 2% OVER 65 . . . . 6 3% Unlisted . . . . . . . . . . . . . 0 0%
Unlisted . . . . . . . . . . . . . . . . . . . 0 0%
EDUCATION
SEX Under 12 . . . . . . . . . . . . 2 1%
Male . . . . . . . . . . . . . . . . . . . . 85
49% 12 Years. . . . . . . . . . . . 33 19%
Female. . . . . . . . . . . . . . . . . . . . 0
51% 1315 Years . . . . . . . . 101 58%
16+ Years. . . . . . . . . . . 39 22%
RATIOS, PERCENTAGES AND SPECIAL INDEXES
STYLES FORM QUALITY DEVIATIONS
Introversive . . . . . . . . . . . . . . . 59 34% XA% > .89. . . . . . . . . . 93 53%
Pervasive Introversive. . . . . . . . 8 5% XA% < .70. . . . . . . . . . . 1 1%
Ambitent . . . . . . . . . . . . . . . . . 34 19% WDA% < .85 . . . . . . . . 20 11%
Extratensive. . . . . . . . . . . . . . . 54 31% WDA% < .75 . . . . . . . . . 2 1%
Pervasive Extratensive . . . . . . . 7 4% X + % < .55 . . . . . . . . . 19 11%
Avoidant . . . . . . . . . . . . . . . . . 28 16% Xu% > .20 . . . . . . . . . . 88 50%
X % > .20 . . . . . . . . . 15 9%
D-SCORES X % > .30 . . . . . . . . . . 1 1%
D Score > 0 . . . . . . . . . . . . . . . 34 19%
D Score = 0 . . . . . . . . . . . . . . 115 66% FC:CF + C RATIO
D Score < 0 . . . . . . . . . . . . . . . 26 15% FC > (CF + C) + 2 . . . . 25 14%
D Score < 1 . . . . . . . . . . . . . . . 7 4% FC > (CF + C) + 1 . . . . 40 23%
(CF + C) > FC + 1 . . . . 53 30%
Adj D Score > 0 . . . . . . . . . . . 57 33% (CF + C) > FC + 2 . . . . 29 17%
Adj D Score = 0 . . . . . . . . . . 102 58%
Adj D Score < 0 . . . . . . . . . . . 16 9%
Adj D Score < 1 . . . . . . . . . . . 5 3%
S-Constellation Positive . . 0 0%

Zd > +3.0 (Overincorp). . . . . . 39 22% HVI Positive . . . . . . . . 10 6%

Zd < 3.0 (Underincorp). . . . . . 24 14% OBS Positive . . . . . . . . . 2 1%
PTI = 5 . . . . . . . . . . . . . . . . . . . 0 0% DEPI = 7 . . . . 1 1% CDI = 5 . . . . . . . . . . . . . 0 0%
PTI = 4 . . . . . . . . . . . . . . . . . . . 0 0% DEPI = 6 . . . . 5 3% CDI = 4 . . . . . . . . . . . . 11 6%
PTI = 3 . . . . . . . . . . . . . . . . . . . 1 1% DEPI = 5 . . . 22 13%
(continued)

400
 NEW NONPATIENT SAMPLE 401

TABLE 2 (Continued)

MISCELLANEOUS VARIABLES
R < 17 . . . . . . . . . . . . . . . . . . . 12 7% (2AB + Art + Ay) > 5. . 13 7%
R > 27 . . . . . . . . . . . . . . . . . . . 25 14% Populars < 4. . . . . . . . . . 8 5%
DQv > . . . . . . . . . . . . . . . . . . . 23 2% Populars > 7. . . . . . . . . 31 18%
S > 2 . . . . . . . . . . . . . . . . . . . . 58 33% COP = 0 . . . . . . . . . . . . 18 10%
Sum T = 0 . . . . . . . . . . . . . . . . 45 26% COP > 2 . . . . . . . . . . . . 61 35%
Sum T > 1 . . . . . . . . . . . . . . . . 27 15% AG = 0 . . . . . . . . . . . . . 72 41%
3r + (2)/R < .33 . . . . . . . . . . . . 34 19% AG > 2 . . . . . . . . . . . . . 12 7%
3r + (2)/R > .44 . . . . . . . . . . . . 53 30% MOR > 2 . . . . . . . . . . . 11 6%
Fr + rF > 0. . . . . . . . . . . . . . . . 26 15% Level 2 Sp Sc > 0. . . . . 20 11%
PureC > 0 . . . . . . . . . . . . . . . . 33 19% GHR > PHR. . . . . . . . 152 87%
PureC > 1 . . . . . . . . . . . . . . . . . 6 3% Pure H < 2 . . . . . . . . . . 29 17%
Afr < .40 . . . . . . . . . . . . . . . . . 19 11% Pure H = 0 . . . . . . . . . . . 3 2%
Afr < .50 . . . . . . . . . . . . . . . . . 44 25% p > a + 1. . . . . . . . . . . . 12 7%
(FM + m) < Sum Shading. . . . . 30 17% Mp > Ma . . . . . . . . . . . 41 23%

Note. N = 175.

sentially the same, only 58 (10%) of those in the sample of 600 have Lambda val-
ues of 1.0 or greater, whereas the new sample includes 28 (16%) persons who have
high Lambda values (d = .17, r = .08). In addition, the sample of 600 includes 10%
who gave Pure C responses, as contrasted with 19% of the new sample (d = .22, r =
.11). Only 5% of the persons in the sample of 600 have positive DEPI values,
whereas 17% of those in the new sample have such values (d =.17, r = .08). Like-
wise, 66% of those in the sample of 600 gave at least one more FC response than
the composite of CF + C. In the new sample, only 37% did so (d = .48, r =.23). Al-
though these differences are interesting, none cause concern about the basic princi-
ples of interpretation.

DISCUSSION

Although the new sample does not have the broad geographic distribution or the
attempted stratification for socioeconomic status that is represented in the larger
sample, the comparative similarities between the two groups tend to support the
notion that published data for the sample of 600 nonpatients (Exner, 2001) is
probably representative. As such, it provides a reasonable basis to gain some un-
derstanding of the response rates for the coded features of responses. It also pro-
vides a source to identify the proportions of nonpatients that can be expected to
fall within, or outside of, established parameters for various ratios and other in-
dexes that form the core of structural data from which some interpretive princi-
ples have been developed.
402
TABLE 3
A Comparison of Data for 12 Variables for the Published Sample of 600 Nonpatients (Exner, 2001) and the First 175 Nonpatients Tested in
This Project

Variable N M SD Range Frequency Mdn Mode Skewness Kurtosis d r

R 600 22.32 4.40 14 43 600 22 23 0.86 1.90

R 175 22.98 5.51 14 51 175 22 21 1.65 5.14 .14 .06
Lambda 600 0.60 0.31 0.11 2.33 600 0.53 0.50 2.27 8.01
Lambda 175 0.61 0.38 0.06 2.33 175 0.50 0.50 1.43 2.65 .03 .01
Afr 600 0.67 0.16 0.23 1.29 600 0.67 0.67 0.35 0.65
Afr 175 0.61 0.18 0.21 1.26 175 0.63 0.67 0.19 0.35 .36 .15
XA% 600 0.92 0.06 0.57 1.00 600 0.94 0.96 1.34 3.68
XA% 175 0.89 0.07 0.69 1.00 175 0.90 0.90 0.75 0.49 .48 .20
WDA% 600 0.94 0.06 0.54 1.00 600 0.95 1.00 1.42 4.93
WDA% 175 0.91 0.06 0.69 1.00 175 0.92 0.95 0.98 1.29 .50 .20
X+% 600 0.77 0.09 0.35 1.00 600 0.78 0.80 0.86 2.33
X+% 175 0.68 0.11 0.35 0.95 175 0.69 0.67 0.55 0.25 .95 .37
Xu% 600 0.15 0.07 0.00 0.45 600 0.15 0.13 0.54 0.86
Xu% 175 0.21 0.09 0.05 0.49 175 0.21 0.16 0.50 0.07 .80 .32
X% 600 0.07 0.05 0.00 0.43 513 0.05 0.04 1.41 4.56
X% 175 0.10 0.06 0.00 0.31 163 0.10 0.05 0.77 0.72 .57 .23
Sum T 600 0.95 0.61 04 490 1.00 1.00 0.83 3.33
Sum T 175 0.90 0.64 02 130 1.00 1.00 0.08 0.52 .08 .03
Fr + rF 600 0.11 0.43 04 48 0.00 0.00 4.98 30.45
Fr + rF 175 0.26 0.81 07 27 0.00 0.00 4.81 30.77 .28 .12
EA 600 8.66 2.38 2 18 600 9.00 9.50 0.04 0.42
EA 175 9.43 3.39 2 24 175 9.50 8.00 0.82 2.08 .29 .12
es 600 8.34 2.99 3 31 600 8.00 7.00 1.43 6.58
es 175 9.05 3.99 2 34 175 9.00 8.00 1.98 9.16 .22 .09
 NEW NONPATIENT SAMPLE 403

REFERENCES

Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155159.

Exner, J. E., Jr. (1978). The Rorschach: A comprehensive system. Vol. 2: Current research and ad-
vanced interpretation. New York: Wiley.
Exner, J. E., Jr. (1986). A Rorschach workbook for the Comprehensive System (2nd ed.). Bayville, NY:
Rorschach Workshops.
Exner, J. E., Jr. (1988). Problems with brief Rorschach protocols. Journal of Personality Assessment,
52, 640647.
Exner, J. E., Jr. (1990). A Rorschach workbook for the Comprehensive System (3rd ed.). Asheville, NC:
Rorschach Workshops.
Exner, J. E., Jr. (2001). A Rorschach workbook for the Comprehensive System (5th ed.). Asheville, NC:
Rorschach Workshops.
Exner, J. E., Jr., Weiner, I. B., & Schuyler, W. (1976). A Rorschach workbook for the Comprehensive
System. Bayville, NY: Rorschach Workshops.
Meyer, G. J. (2001). Evidence to correct misperceptions about Rorschach norms. Clinical Psychology:
Science and Practice, 8, 389396.
Shaffer, T. W., & Erdberg, P. (2001, March). An International Symposium on Rorschach Nonpatient
Data: Worldwide Findings. Annual meeting, Society of Personality Assessment.
Shaffer, T. W., Erdberg, P., & Haroian, J. (1999). Current nonpatient data for the Rorschach, WAISR,
and MMPI2. Journal of Personality Assessment, 73, 305316.

APPENDIX
Initial Solicitation Letter

Hello,

I am ______________________, a research associate for Rorschach Work-

shops. Currently we are working on a new standardization project concerning the
Rorschach Inkblot Test.

Rorschach Workshops is a not for profit research foundation. We were estab-

lished in 1969 for the main purpose of doing research in the broad area of personal-
ity assessment, with special focus on research and development of the Rorschach
Test. During the past 30 years we have completed more than 400 research projects
in a variety of mental health related areas and members of our staff have written
more than 15 books and over 100 articles in professional journals.

We are currently involved in a project to restandardize the Rorschach Test that,

in effect, will require the administration of the test to several hundred persons who
are not, and never have been, psychiatric patients. The data collected from these
nonpatients serves a baseline that helps us to understand some of the characteris-
tics of people who do have the sorts of psychological difficulties that require
treatment.
404 EXNER

We hope that you might be willing to volunteer to participate in this project. If

you do volunteer, we will arrange to administer the Rorschach Test to you at a time
and place convenient to you. Before the testing, you will be given an informed con-
sent that lists further details of the study. The testing will require approximately
1 hours of your time. In return for your willingness to volunteer, a $25 donation
will be made to a recognized local charity of your choice, in your name.

Sometimes people who are interested in volunteering for a project such as this
hope to obtain feedback about their test performance. Unfortunately, the design of
this project prohibits us from doing this. Thus, the only way in which we are able to
show our appreciation of your assistance is through the donation made in your
name. Identification numbers will be used for all those who volunteer so that no
names will ever be entered on any of the test data collected.

If you are willing to volunteer for this project or if you have any questions about
volunteering, please contact me by phone at ___________________ (If you use e-
mail, you can also include your e-mail address.)
Sincerely,

Rorschach Workshops

John E. Exner, Jr.

Rorschach Workshops
P.O. Box 9010
Asheville, NC 288150010

Received September 10, 2001

Revised November 16, 2001