title

author
Journal of
Managerial
Psychology
Volume 10 Number 7 1995 30
Perfor
mance
a!!rai
sal
an"
true
!erfor
mance
Mark
Cook
The author
Mark Cook is a
#hartere"
$ccu!ational
Psychologist %ith
the Personnel
&election 'esearch
(rou! at
)ni*ersity
#ollege+ &%ansea+
),-
Abstract
.rgues that the
con*entional
*ali"ation
!ara"igm+ %hich
uses sub/ecti*e
!erformance or
a!!raisal ratings
as criteria+ may
be of "oubtful
*ali"ity- 0iscusses
research into
!erformance
a!!raisal %hich
"ocuments four
sets of
!roblems %hich may
re"uce the
usefulness of
!erformance
ratings as criteria-
1hese !roblems
inclu"e biases+
!olitic23 ing+
im!ression
management an"
un"eser*e"
re!utation-
0escribes the
inaccuracies to
%hich these
!roblems gi*e rise
an" conclu"es that
instea" of selecting
the right
!eo!le for
management+
selection metho"s
*ali"ate" against
a!!raisal %ill
sim!ly !er!etuate
an unsatisfactory
status 4uo-
.
u
t
h
o
r
5
s

a
"
"
r
e
s
1
s

.
u
t
h
o
r
5
s

n
a
m
e

h
e
r
e

.
"
"
r
e
s
s

1
.""ress 6
.""ress 3
.""ress 7
.""ress 5
e mail8
insert here
Journal of
Managerial
Psychology
Volume 10
Number 7
1995 !!- 37
9 M#:
)ni*ersity Press
;&&N 06<=3
397<
Performan
ce
appraisal
and
performan
ce- related
pay are
increasingl
y popular
in the UK
at present,
and both
seem to be
viewed
with great
favour by
the
present
Governme
nt. In most
sectors
appraisal
is
introduced
first at
manageme
nt level,
and then
cascades
down to
less-
e!alted
levels of
staff.
"efore
this hap-
pens, it
might be
timely to
review
research
on
performan
ce
appraisal
which
documents
the many
problems
of
performan
ce ratings
#P$s%,
and casts
serious
doubt on
their
validity as
measures
of true wor&
performance.
:iases
P$s suffer from
many biases,
meaning factors
that should
not influence
raters but do,
or factors that
influence ratings
in ways about
which raters are
unaware.
Age, ethnicity,
and gender
'he pooled results
of () separate
research-
es*+, find a low
negative
correlation
between
age and P$s,
whereas age and
ob-ective
indices of
performance
correlate
positively
.
'his implies that
P$s show
substantial bias
against older
persons.
.imilar
large-scale
analysis*/,
finds a small
but consistent
own race bias
in P$s0 whites
favour whites,
blac&s favour
blac&s, etc.
Gender also
distorts
P$s.
.ometimes
P$s
show simple
gender bias*1,, so
that women
with masculine
characteristics are
rated more
promotable.
.ometimes the
bias is subtler,
involving gender
stereotype of the
occupation,
so that raters who
do not see
management as a
traditionally
female occupation
tend to give
women managers
lower P$s*(,.
Physical
appearance
2ther biases in
P$s are subtler.
3ommon
sense argues that
physical
attractiveness is
not li&ely to
influence P$s,
because beauty is
in the eye of the
beholder and
anyway should
not influence
behaviour in the
wor&place. In
fact there is a very
reliable broad
consensus
about who is and
who is not good-
loo&ing,
and strong
evidence that
appearance does
affect P$s. 'wo
separate
researches*4,5,
find
a strong lin&
between
attractiveness 6
rated on
a five-point scale
6 and later salary
level. "oth
find that each
e!tra scale point of
rated attrac-
tiveness is
worth 7/,)))-
/,5)) more in
2
salary. 'his
implies that
the salary
difference
between the
most and
least
attractive
could be
as great as
7+1,))).
'he effect is
far stronger
3
Performance a!!raisal an" true !erformance
Mar2 #oo2
Journal of Managerial Psychology
Volume 10 Number 7 1995 37
for males, in male -obs, and in older
age groups. P$s are strongly biased by
appear- ance, and the bias is not
confined, as com- mon sense might
argue, to young female ratees. In fact it is
probably strongest in tradi- tional
management circles.
Attitudes and values
8ursing managers9 P$s are shaped more by
the manager9s belief system than by the subor-
dinate9s actual wor& performance*:,. 'he
research uses a laddering techni;ue to
e!plore nursing managers9 concepts of effec-
tiveness. 'his means continuing to as& man-
agers the ;uestion why is that important
until one eventually elicits answers li&e self-
fulfilment or pride in being a good manag-
er, which the researchers argue represent the
managers9 fundamental values. 'he managers
said
that staff wor& behaviour shaped their
P$s, but policy capturing analysis showed
that the managers9 own values were the best
predictor of their P$s. .ubordinates who
helped managers to feel proud of being a good
manager, or helped to ma&e them feel ful-
filled, received better P$s.
In-group/out-group
2ne school of thought argues that most wor&
groups can be divided into an in-group,
whose members en-oy the supervisor9s confi-
dence and concern and get assigned the more
challenging tas&s, and an out-group, whose
members are treated li&e hired hands and
are assigned the mundane tas&s. $esearch on
ban& staff*<, shows that in-group members
get better P$s but do not perform any better
on ob-ective performance indices. 'his
implies that the in-group members achieve
their position not by better wor&, but by some
other path.
Personal like/dislike
In-depth interviews with managers who
regularly ma&e P$s reveal that three-;uarters
freely admit they allow li&ing to inflate P$s
and disli&e to lower P$s*=,. 'he same analy-
sis also reports that <1 per cent of managers
say being in a good or bad mood shifts the
P$s they ma&e 6 probably downwards in most
cases, given that :1 per cent say they hate
having to ma&e P$s.
P$s can be analysed to detect age, race or
gender bias #and usually are, to avoid e;ual
opportunities problems%. P$ data cannot be
analysed to detect bias created by appearance,
Performance a!!raisal an" true !erformance
Mar2 #oo2
Journal of Managerial Psychology
Volume 10 Number 7 1995 37
li&e>disli&e, similarity of outloo&, or
in-group>out-group effects, because
this information is not routinely
recorded, and often it is not readily
accessible.
'his means that many sources of bias
render performance appraisal ratings less
accurate, but in ways which neither
employer nor employee can easily
identify. In technical terms, the
performance ratings are contami-
nated, and the contamination is
invisible.
'his in turn has further implications
for research on personnel selection.
"etween
5) per cent and <) per cent of researches
on selection use supervisor rating as their
criteri- on of true performance. "ias
in ratings, from the sources described
above, means that rating may often be a
fairly poor criterion of true performance,
which in turn implies that much research
may actually underestimate
the true validity of selection methods.
Politic2ing
2rgani?ations usually assume that
managers want to ma&e accurate P$s,
and are prevented only by unconscious
biases which can be
cured by sufficient training or the right
rating format. 'he in-depth interview
research flatly contradicts this and
concludes that political considerations,
or private agendas, nearly always
influence P$s. 'he managers list many
reasons for giving inflated P$s@ to
ma!imi?e merit increases, to protect
persons whose performance is suffering
because of personal problems, to avoid
washing dirty laundry in public, to
avoid creating a badly written record, to
avoid confrontation, and to rid
themselves of people by promoting up
and out.
'he interviews also uncover managers9
reasons for ma&ing deliberately low P$s@
to shoc& someone bac& onto a higher
perfor- mance trac&, to teach a
rebellious subordi- nate who is in charge,
to send someone a message that they
should consider leaving the organi?ation,
and to build a well-documented record
of poor performance to speed up
termination.
Gross leniency in appraisal rating is
very apparent, but subtler politic&ing
may not be. Ai&e biases based on
appearance or atti- tude, subtler
politic&ing contaminates the appraisal
rating, and ma&es it less accurate, both
as a measure of performance, and as a
criterion in research of selection
methods.
;m!ression management
Bany organi?ations set great store by
relative- ly trivial details, that impress
senior manage- ment but have little to do
with effective wor& performance@ the right
mannerisms, the right clothes, or the right
bu??words. Padfield9s history*+), of
battleships describes how promotion in
the $oyal 8avy of the +<=)s went to
officers with the most highly polished
ships, which caused a few to behave as if
they had forgotten what battleships were
for0 they avoided gunnery practice in case
the powder smo&e spoiled their
paintwor&. Cn e!treme case of this trend
may be termed the Dorld Dar I
mentality. 2rgani?ations occasionally
e!ist in which subordinates gain credit for
pushing ahead with management plans
that are absurdly wrong, in pursuit of
aims which are completely pointless,
stifling criticism either of purpose or of
method with cries of commitment and
loyalty.
Ingratiation
English has a rich vocabulary to
describe wor&place ingratiators 6
including words listed in dictionaries as
not in polite use 6 which implies that
the behaviour is widely recogni?ed, but
not widely popular. 2ne
research*++, identified three underlying
types of ingratiating behaviour, or
upwards influ- ence styles@
#+% Job-focused ingratiation@ claiming
credit for things you have done,
claiming credit for things you have not
done, claiming credit for what the
group has done, arriving at wor& early
to loo& good, wor&ing late to loo&
good.
#/% Super visor-focused ingratiation@
ta&ing an interest in the supervisor9s
private life, praising the supervisor,
doing favours for the supervisor,
volunteering to help the supervisor,
complimenting the supervisor on
his>her appearance and dress,
agreeing with the supervisor9s ideas.
#1% Self-focused ingratiation@
presenting self to the supervisor as a
polite and friendly person, wor&ing
hard when results will be seen by the
supervisor, letting the super- visor
&now that you are trying to do a good
-ob.
$esearch suggests however that
ingratiation does not always succeed in
obtaining good P$s. Unsubtle
ingratiation may sometimes be too blatant
to be credible, or palatable.
Ingratiation and other impression man-
agement techni;ues also contaminate
appraisal ratings, and ma&e them less
accu- rate reflectors of true worth to the
organi?a- tion. "esides undermining
performance appraisal, and selection
research, this tends to be bad for morale,
when staff see persons whose true
performance is poor, but who are good at
ingratiating themselves, get merit awards,
or promotion, or other mar&s of favour.
)n"eser*e" re!utation
C good reputation can be earned by
good wor&, but many features of large
organi?a- tions ma&e it possible to
earn one in other ways.
Its who you know, not what you know
C widely voiced observation, which
implies that the employee9s time may
often be as usefully spent creating a
networ& of allies and contacts as in
doing any actual wor&. 'eam
building or being a team player9 are
phrases often used in -ob advertisements
and personnel specifications. 3o-
operating
with colleagues is self-evidently
important in most wor& 6 but could
team-player some- times also refer to
people who are very good at ma&ing
themselves li&ed or good at appearing to
be indispensable, but who are not
necessarily very effective performersF C
closely &nit team is also more li&ely to
share, and stic& closely to, a social
reality #see Dho sets the standards,
below%.
The non-working day
$esearch on U. army infantry*+/,
showed that the average soldier spends
/4 minutes of a typical training day
receiving instruction and 4: minutes
practising the tas&0 the rest of
the day is swallowed up by parades,
being en route, polishing boots, hanging
about, etc. .imilar research*+1, shows
that a substantial part of the manager9s
day is spent doing things that cannot be
unambiguously lin&ed to the
accomplishment of specific tas&s.
Polishing the image
.ome individuals consciously set about
build- ing themselves a reputation, by
conventional public relations
techni;ues of controlling information
about themselves and their doings.
Cover your back
In cautious organi?ations a good reputation is
built largely by
not
doing things@ not ma&ing
controversial decisions, not attracting com-
plaints, not getting bad publicity.
Attributability
3omple! organi?ations and>or long time
scales mean that it is often difficult to assign
true responsibility for outcomes. Dhen it is
difficult to &now who is really responsible for
successes and failures, it is correspondingly
easier for devious individuals to steal the
credit for successes, avoid the blame for fail-
ures, and build an undeserved reputation.
!icking over beehives"
In some sectors people move -obs every year
or so, so that they can ma&e many changes but
escape before any conse;uent problems
emerge. Banagers call this &ic&ing over
beehives. $apid turnover also ma&es it diffi-
cult to &now who is responsible for what.
#ho sets the standards
C company that manufactures gearbo!es has
its success defined e!ternally, and unambigu-
ously, by its sales figures. C university by
contrast constructs its own
social reality
about
success. 'he academic staff decide what
issues are worth researching, and what sub-
-ects should be taught and by implication
whose wor& has merit. Dhere success is
defined by the organi?ation and its staff,
greater scope e!ists for creating undeserved
reputations.
$mpire-building
In many organi?ations, success consists of
increasing the si?e of one9s department or
budget. .ervices are provided for the sa&e of
-ustifying the organi?ation9s e!pansion. 3yril
8orthcote Par&inson*+(, describes how staff
levels in the "ritish 3olonial 2ffice rose
steadily throughout the twentieth century,
while the number of "ritish colonies fell
e;ually steadily.
%eorgani&ations
$eorgani?ations, large or small, create a
perfect form of pseudo-wor&, divorced from
e!ternal standards. Everyone9s efforts are
centred for months on something that has no
end product, and fre;uently serves no useful
purpose, but is an ideal environment for the
person who see&s to build a reputation.
$eorgani?ations also ma&e it easy to
blur responsibility for successes and
failures.
'aking your mark
It has been said of Parliament that no
Binister has really arrived until he>she
has passed an Cct. 3ritics argue that this
e!plains the enact- ment of numerous
badly drafted and unen- forceable laws.
.imilar mar& ma&ing mech- anisms
include opening a new department or new
building #or often, these days, closing
one%, and reorgani?ations #see preceding
paragraph%. 2ften the more outrageous
the mar& ma&ing, the better. Even now,
practical- ly everyone remembers the
name of Gr $ichard "eeching, if not as
the author of Reshaping the Railways,
then as a code name for someone who
was brought in to close things down.
.ome -ob advertisements appear to be
groping towards aspects of this mar&
ma&- ing mechanism with references to
stature #not referring to physical build%
and credibil- ity #not meaning that the
person has not
been caught telling lies recently%.
;m!lications for P's
.omeone whose true performance is
poor, but who is a good self-publicist, or
good at claiming responsibility for others9
wor&, or successful at defining his>her
achievements as useful and important,
will achieve better P$s. 3onversely
someone whose true performance is
good, but who is not a good self-
publicist,
or does not claim responsibility for
others9 successes, or who fails to get
his>her wor& defined as important,
may get poorer P$s. 'he P$ has been
contaminated, and less accurately
reflects true performance.
In blue-collar -obs there is often a
simple ob-ective criterion of
performance, of the units-produced-or-
processed type. Dhen
these ob-ective measures of wor&
performance are compared with
sub-ective appraisals for
the same employee, a very low
correlation results # p H )./:%*+4,,
which means that ob-ective measures
and the sub-ective appraisals are not
measuring the same thing. Perhaps the
same is true of management, which
implies that those who get the good
ratings, and the promotions, and the
perfor- mance-related pay, may not
always, or even often, be the ones who
are really doing a good -ob.
'his has a further implication. Bost
selec- tion research uses performance
rating as its criterion of true
performance. If perfor- mance appraisal
is not a good measure of true
performance, then selection methods
validat- ed against appraisal will not be
selecting the right people for
management, but will simply be
perpetuating an unsatisfactory status
;uo.
'eferences
1 >al"man+ 0-.- an" .*olio+ :-J-+ ?.
meta3analysis of age "ifferences in /ob
!erformance@+ Journal of .!!lie"
Psychology+ Vol- 71+ 19=<+ !!- 3333=-
6 ,raiger+ ,- an" Aor"+ J-,-+ ?. meta3
analysis of ratee race effects in
!erformance ratings?+ Journal of
.!!lie" Psychology+ Vol- 70+ 19=5+ !!-
5<3<5-
3 Bartman+ &-J-+ (riffieth+ '->-+ #rino+
M-0- an"
Barris+ $-J-+ ?(en"er3base" influences8
the !romotion recommen"ation@+ &eC
'oles+ Vol- 65+ 1991+
!!- 6=53300-
7 Dan"y+ A-J- an" Aarr+ J-D-+ ?Performance
rating@+Psycho3 logical :ulletin+ Vol- =7+
19=0+ !!- 763107-
5 ArieEe+ ;-B-+ $lson+ J-F- an" 'ussell+ J-+
?.ttracti*eness an" income for men
an" %omen in management?+
Journal of .!!lie" &ocia; Psychology+ Vol-
61+ 1991+ !!- 1039357-
< 'osEell+ P-+ ,enne"y+ 0- an" (rabb+ F-+
?Physical attracti*eness an" income
attainment among #ana"i3 ans@+ Journal
of Psychology+ Vol- 163+ 19=9+
!!- 577359-
7 Jolly+ J-P-+ 'eynol"s+ 1-J- an" &locum+
J->-+ ?.!!lication of the means3en"
theoretic for un"erstan"ing the cogniti*e
bases of !erformance a!!raisal@+
$rganiEa3 tional :eha*ior an" Buman
0ecision Processes+
Vol- 71+ 19==+ !!- 153379-
= Vecchio+ '-P- an" (ob"el+ :-#-+ ?1he
*ertical "ya"ic lin2age mo"e of
lea"ershi!8 !roblems an" !ros!ects@+
$rganiEational :eha*ior an" Buman
Performance+
Vol- 37+ 19=7+ !!- 5360-
9 Dongenec2er+ #-$-+ &ims+ B-P- an" (ioia+
0-.-+ ?:ehin" the mas28 the !olitics of
em!loyee a!!raisal@+ .ca"e3 my of
Management FCecuti*e+ Vol- 1+ 19=7+ !!-
1=33
93-
10 Pa"fiel"+ P-+ 1he :attleshi! Fra+ Bart30a*is+
Don"on+
1976-
11 >ayne+ &-J- an" Aerris+ (-'-+ ?;nfluence
tactic affect+ an" eCchange 4uality in
su!er*isor3subor"inate interactions8 a
laboratory an" fiel" stu"y@+ Journal of
.!!lie" Psychology+ Vol- 75+ 1990+ !!-
7=7399-
16 :iale2+ B-+ Ga!f+ 0- an" Mc(uire+ >-+
Personnel 1urbulence an" 1ime
)tililiEation in an ;nfantry 0i*ision+
Bum''$ A'3>03#.-773; l+ Buman
'elations 'esearch $rganiEation+
.leCan"ria+ V.+ 1977-
13 #am!bell+ J-P-+ 0unnette+ M-0-+ Da%ler
F-F- an" >eic2 ,-F-+ Managerial
:eha*ior+ Performance+ an" Fffec3
ti*eness+ Mc(ra%3Bill+ Ne% Hor2+ NH+
1970-
17 Par2inson+ #-N-+ Par2inson5s Da%+
John Murray+ Don"on+ 195=-
15 Beneman+ '-D-+ ?1he relationshi!
bet%een su!er*isory ratings an" results3
oriente" measures of !erformance8 a
meta3analysis@+ Personnel Psychology+
Vol- 39+
19=<+ !!- =1136<-