Keywords: validation, personality, competency models, meta-analysis

Correspondence concerning this article should be addressed to Dave

Bartram, SHL Group plc, The Pavilion, !t"ell Place, Thames Ditton,
Surrey #T$ %&', 'ngland( '-mail) dave(bartram*shlgroup(com
!bility and Personality as Predictors o+ ,ob Per+ormance
!bility measures have been ac-no"ledged as good predictors o+ .ob
per+ormance and even better predictors o+ training per+ormance( The early
meta-analyses /e(g(, Hunter 0 Hunter, 1234 sho"ed the generali5ability
o+ this +inding( 6ore recently, the
This research was funded by the SHL Group. Dave Bartram is
research director with the SHL Group.
I thank Rainer ur! and He"en Baron# who contributed to the work
reported here whi"e they were emp"oyees of SHL Group. Rainer ur!
made a ma$or contribution to the deve"opment of the Great %i&ht
mode"# and He"en Baron and my research team co""ected and co""ated
a"" the va"idity studies# which were subse'uent"y ana"y!ed for this
study. I a"so thank (eter )arr for his va"uab"e comments and
su&&estions on a draft of this artic"e and *erard ehoe for his
thorou&hness and he"pfu" comments and su&&estions.
The Great 'ight Competencies) ! Criterion-Centric !pproach to 7alidation
Dave Bartram SHL
Group
The author presents results o+ a meta-analysis o+ 81 validation studies /N 3,294 that uses the Great
'ight competency +actors /#ur5 0 Bartram, 8%%84 as the criterion measurement +rame"or-( Predictors
o+ the Great 'ight competencies based only on personality scales sho" moderate to good correlations
"ith line-manager ratings +or all 2 o+ the competencies( :n their o"n, ability tests correlate "ith 3 o+ the
2 competencies, and together ability and personality data yield operational validities ranging +rom %(8%
to %(33 +or the 2 competencies( :perational validities +or aggregated predictors "ith aggregated criteria
"ere estimated to be %(;<( The value o+ di++erentiating the criterion space and o+ relating predictor
variables to criterion variables in a one-to-one +ashion is discussed(
This study presents a model o+ per+ormance in the "or-place that de+ines
eight broad competency +actors, "hich "e re+er to as the Great 'ight
/Bartram, =obertson, 0 Callinan, 8%%8> #ur5 0 Bartram, 8%%84( The
Great 'ight have emerged +rom +actor analyses and multidimensional
scaling analyses o+ sel+-and manager ratings o+ "or-place per+ormance,
not +rom the predictor domain /i(e(, ability tests, motivation or personality
?uestionnaires4( Thus, they provide a criterion-centric model +rom "hich
to e@plore the validity o+ various potential predictors o+ "or-place
per+ormance( This model and its associated predictorAoutcome
relationships are e@plored through a meta-analysis o+ 81 validity studies(
The intention behind this research is not .ust to add to the body o+ data
relating to the validity o+ personality and ability tests as predictors o+
"or-place behavior but also to demonstrate the value o+ an approach that
uses a model o+ the criterion domain as the organi5ing +rame"or- +or
meta-analysis rather than the more usual predictor domain models /e(g(,
the Big Bive personality +actors model4(
academic literature has begun to support the vie" that personality
measures also predict per+ormance at "or-( Studies concentrating on the
Big Bive personality +actors have sho"n that Conscientiousness and
'motional Stability have broadly generali5able relationships "ith overall
.ob per+ormance /:,P> Barric- 0 6ount, 11> Hough, 118> Salgado,
11$, 1124( Barric- and 6ount /114 report corrected mean validities o+
r (88 +or Conscientiousness and r (< +or '@traversion( Tett, ,ac-son,
and =othstein /114 ?uote corrected mean validities ranging +rom r (
9 +or '@traversion to r (<< +or !greeableness( Salgado /11$, 1124
replicated Barric- and 6ountCs results "ith 'uropean data sets and also
+ound evidence +or the validity o+ 'motional Stability /corrected r (14(
6ore recently, Hurt5 and Donovan /8%%%4 have reported corrected
correlations o+ r (88 +or Conscientiousness, r (3 +or 'motional
Stability, and r (%1 +or '@traversion( Thus, the current evidence is
generally supportive o+ some o+ the Big Bive in providing moderate
predictions o+ relatively gross .ob per+ormance measures(
Predictor-7ersus Criterion-Centric !pproaches
The traditional approach to validation has been predictor centric(
=esearchers have as-ed ?uestions li-e DEhat does instrument X
predictF,G DHo" "ell does personality predict .ob per+ormanceF,G and
DEhat do ability tests predictFG !s a conse?uence, "e have seen separate
literatures develop regarding the validity o+ personality scales and ability
scales even though both types o+ instruments are used to predict aspects o+
behavior in the "or-place( Through meta-analyses, there has been some
pulling together o+ the +indings "ithin each literature( Ho"ever, this has
tended to be at the e@pense o+ a loss o+ detail, because there has been a
relative lac- o+ +ocus on the nature or appropriateness o+ the criterion
measures /"hich are typically supervisor ratings o+ :,P4 "hen studies
have been dra"n together +or meta-analysis(
The argument presented here is that "e should re+ocus our ?uestions( Ee
should be as-ing DHo" can "e best predict YFG "here Y is some
meaning+ul and important aspect o+ "or-place behavior( Competency
+rame"or-s, "hen de+ined in terms o+ observable "or-place behaviors,
provide the basis +or a di++eren
2;
tiated criterion measurement( They support investigation o+ di++er
ent aspects o+ per+ormance, promoting a more sophisticated un
derstanding o+ the +actors underlying :,P( Hnconsistencies in predictors o+
overall per+ormance bet"een .obs might be e@plained by di++erences
bet"een .obs in the relative importance o+ di++erent aspects o+
per+ormance( Di++erentiation o+ the criterion space "ould allo" better
prediction o+ .ob per+ormance +or a particular role once the competency
re?uirements +or the role "ere understood(
6odels o+ the Criterion 6easurement Domain
H+ "e are to sho" a more complete and consistent pattern o+ relationships
bet"een predictors and "or-place per+ormance, "e need to di++erentiate
the criterion measures in a meaning+ul "ay( Ehen this more di++erentiated
approach has been adopted, a richer picture o+ the relationship bet"een
predictors and per+ormance in the "or-place emerges( =obertson and
#inder /11<4 e@amined the average validity o+ a range o+ personality
scales .udged to be relevant +or each o+ 8 di++erent per+ormance areas in a
metaanalysis and +ound generali5able validity +or % o+ them( &y+ield,
Gibbons, Baron, and =obertson /11;4 reported consistent patterns o+
correlations bet"een each o+ the Big Bive personality dimensions and
di++erent aspects o+ .ob per+ormance as measured using managerCs ratings(
=obertson, Baron, Gibbons, 6acHver, and &y+ield /8%%%4 demonstrated
that ratings o+ :,P that correlated highly "ith planning and organi5ing
competency did sho" a positive correlation "ith Conscientiousness, as
e@pected based on the metaanalysis literature(
The =obertson et al( study illustrates the importance o+ considering
moderator variables in "hich arti+act corrections leave substantial amounts
o+ variance in the validity coe++icients( !s a +urther e@ample, ,udge, Bono,
Hlies, and Gerhardt /8%%84 have sho"n ho" the correlations bet"een
conscientiousness and ratings o+ leadership ?ualities are moderated by
sample e++ects) Bor student samples Conscientiousness is correlated "ith
ratings o+ leadership ?ualities /r (<94, but this drops to r ($ +or
people in government and military organi5ations and to r (%; +or
people in business and commercial settings(
Studies li-e these illustrate the importance o+ doing more than .ust loo-ing
at correlations "ith overall measures o+ .ob per+ormance( Ee need to have
a "ell-articulated model o+ the domain o+ "or-place behaviors and then
evaluate the utility o+ our predictor instruments in terms o+ ho" "ell they
enable us to account +or variance in this domain( Campbell /11%4, +or
e@ample, de+ined per+ormance as +ollo"s)
Per+ormance is behavior( Ht is something that people do and is re-
+lected in the actions that people ta-e( ( (Per+ormance is not the con-
se?uence/s4 or result/s4 o+ action> it is the action itsel+( ( ( Bor any .ob,
there are a number o+ ma.or per+ormance components,
distinguishable in terms o+ their determinants and covariation patterns
"ith other variables( The correlations among their true scores are less
than one(
/p( $%34
These vie"s are e@tended in a later publication)
Per+ormance is( ( (something that people actually do and can be ob-
served( By de+inition, it includes only those actions or behaviors that
are relevant to the organi5ationCs goals and that can be scaled /mea-
sured4 in terms o+ each personCs pro+iciency /e(g(, level o+ contribu
tion4( Per+ormance is "hat the organi5ation hires one to do, and do
"ell( Per+ormance is not the conse?uence or result o+ action, it is the
action itsel+( ( (Per+ormance consists o+ goal-relevant actions that are
under the control o+ the individual, regardless o+ "hether they are
cognitive, motor, psychomotor, or interpersonal( /Campbell, 6cCloy,
:ppler, 0 Sager, 11<, pp( 3% A34
Similarly, =otundo and Sac-ett /8%%8, p( 994 suggest that D.ob
per+ormance is conceptuali5ed as those actions and behaviors that are
under the control o+ the individual and contribute to the goals o+ the
organi5ation(G
There have been a number o+ attempts to develop models o+ the criterion
domain( Campbell, 6cHenry, and Eise /11%4 described the per+ormance
o+ entry-level I(S( !rmy soldiers in terms o+ +ive dimensions) core
pro+iciency, general soldier pro+iciency, e++ort and leadership, personal
discipline, and physical +itness and military bearing( Campbell et al(
/11<4 also describe a general model o+ "or- per+ormance consisting o+
eight +actors) .ob-speci+ic tas- pro+iciency, non-.ob-speci+ic tas-
pro+iciency, "ritten and oral communication, demonstrating e++ort,
maintaining personal discipline, +acilitating team and peer per+ormance,
supervision and leadership, and management and administration(
Higher level categori5ations o+ models have been proposed by Borman
and 6oto"idlo /11<4, "ho distinguish bet"een conte@tual and tas-
per+ormance, and by ,( Hogan and Holland /8%%<4, "ho tal- o+ Dgetting
alongG and Dgetting aheadG competencies(
,( Hogan and HollandCs /8%%<4 "or- is particularly relevant to the current
research in that they emphasi5ed the need to align predictors and criteria
and e@plored the hypothesis that validities "ould increase as one moved
+rom broad multiple-construct criteria /li-e :,P ratings4 to more narro"
single-construct criteria( Hn their research, this "as done using predicted
relations bet"een seven personality constructs measured using the Hogan
Personality Hnventory /HPH4 and t"o criteria) getting along and getting
ahead( The HPH scales !d.ustment, Prudence, and !mbition all provided
good prediction o+ both criteria( HPH Li-ability predicted getting along
/conte@tual per+ormance4, "hereas HPH Hntellectance "as more related to
getting ahead /tas- per+ormance4(
Scullen, 6ount, and ,udge /8%%<4 +ound that general-level +actor models
/li-e the tas- vs( conte@tual per+ormance model4 +it <9% ratings data less
"ell than more di++erentiated models( Hn particular, they sho"ed a better
+it +or a +our-+actor model /technical s-ills, administrative s-ills, human
s-ills, and citi5enship behaviors4( This supports the vie" that it may be
necessary to use more di++erentiated models to +ully capture the variance
in the criterion( Clearly, a balance is needed bet"een highly di++erentiated
models that may not be generali5able and overly broad constructs that +ail
to capture relevant general dimensions o+ per+ormance( Con+irmatory
+actor analyses have supported the distinction bet"een tas- and conte@tual
or citi5enship behaviors /Con"ay, 119> ,ohnson, 8%%4, and other "or-
has sho"n that, although tas- per+ormance is better predicted by ability
than personality, the reverse is the case "ith citi5enship behavior /Borman,
Penner, !llen, 0 6oto"idlo, 8%%> Hurt5 0 Donovan, 8%%%4( Eor- by
Borman, Buc-, Hanson, 6oto"idlo, Star-, and Drasgo" /8%%4 classi+ied
a "ide range o+ citi5enship behaviors into three main subcategories) per-
sonal support /helping, motivating, cooperating "ith and sho"ing
consideration +or others4, organi5ational support /sho"ing loyalty,
+ollo"ing rules and regulations, representing the organi5ation in a positive
light4, and conscientious initiative /engaging in sel+
development, ta-ing initiatives, persistence, and ma-ing e@tra e+
+ort to complete tas-s4( The latter is clearly related to conscien
tiousness and is li-ely to +acilitate both tas- and conte@tual per+ormance
/6oto"idlo, Borman, 0 Schmit, 11$4(
Bac
tor
Competency
domain title
Competency domain de+inition Hypothesi5ed Big Bive, motivation, and
ability relationshipsa



8



<

3


;




9


$


2
Leading and
Deciding
Supporting and
Cooperating


Hnteracting and
Presenting
!naly5ing and
Hnterpreting

Creating and
Conceptuali5ing



:rgani5ing and


'@ecuting !dapting
and Coping

'nterprising and
Per+orming
Ta-es control and e@ercises leadership( Hnitiates action, gives direction,
and ta-es responsibility(

Supports others and sho"s respect and positive regard +or them in social
situations( Puts people +irst, "or-ing e++ectively "ith individuals and
teams, clients, and sta++( Behaves consistently "ith clear personal values
that complement those o+ the organi5ation(
Communicates and net"or-s e++ectively( Success+ully persuades and
in+luences others( =elates to others in a con+ident, rela@ed manner(
Sho"s evidence o+ clear analytical thin-ing( Gets to the heart o+ comple@
problems and issues( !pplies o"n e@pertise e++ectively(
Juic-ly ta-es on ne" technology( Communicates "ell in "riting
Eor-s "ell in situations re?uiring openness to ne" ideas and e@periences(
See-s out learning opportunities( Handles situations and problems "ith
innovation and creativity(
Thin-s broadly and strategically( Supports and drives organi5ational
change(
Plans ahead and "or-s in a systematic and organi5ed "ay( Bollo"s
directions and procedures( Bocuses on customer satis+action and delivers a
?uality service or product to the agreed standards(
!dapts and responds "ell to change( 6anages pressure e++ectively and
copes "ell "ith setbac-s(

Bocuses on results and achieving personal "or- ob.ectives( Eor-s best
"hen "or- is related closely to results and the impact o+ personal e++orts
is obvious( Sho"s an understanding o+ business, commerce, and +inance(
See-s opportunities +or sel+-development and career advancement(
&eed +or po"er and control,
e@traversion

!greeableness



'@traversion, general mental ability

General mental ability, openness to ne"
e@perience

:penness to ne" e@perience, general
mental ability



Conscientiousness, general mental
ability

'motional stability


&eed +or achievement, negative
agreeableness
The -ey point +rom all this "or- is the demonstration that di++erentiating
the criterion, i+ only into t"o broad areas, provides a considerable gain in
the clarity o+ ho" personality-based predictors relate to per+ormance( By
doing this, one may capture criterion speci+icity e++ects that are lost "hen
:,P is used as the sole criterion and treated as i+ it represented the same
construct +rom study to study(
The current study proposes disaggregating the criterion +urther than this(
Clearly, the degree to "hich it is possible to articulate the criterion space
into distinct +actors "ill be dependent on the range o+ measures available(
H+ the only datum available is an :,P rating, then all "e can do is as-
about ho" reliable or relevant that is( Competency assessment by line
managers and others typically involves a "ide range o+ measures and,
there+ore, provides the potential +or ma@imi5ing di++erentiation o+ the
criterion space(
! Generic Competency Brame"or-
The process adopted in developing the generic competency +rame"or-
used in this study /Bartram et al(, 8%%8> #ur5 0 Bartram, 8%%84 "as
similar to that described in Tett, Guterman, Bleier, and 6urphy /8%%%4,
"ho identi+ied ;< dimensions o+ .ob per+ormance in managerial .obs +rom
8 published and practitioner
Table
Titles and High-Level Definitions of the Great Eight
Cometen!ies
models( The "or- done by mysel+ and my colleagues "as based on an
analysis o+ a "ide range o+ published and practitioner models( Eith a
de+inition o+ competencies as Dsets o+ behaviors that are instrumental in
the delivery o+ desired results or outcomesG /Bartram et al(, 8%%8, p( $4,
the resulting +rame"or- distinguishes 8 component competencies at the
+inest level o+ detail( These components are clusters o+ similar "or-place
behavior, "hich, in practice, are not +ound to be +urther di++erentiated in
competency models /see !ppendi@4( These components can be thought o+
as building bloc-s that can be aggregated together to produce com-
petencies( Sets o+ competencies, in turn, +orm competency models(
Eithin the current +rame"or-, one general purpose model is de+ined, "ith
8% competencies( Ho"ever, the important level +or the current research is
the more general model, "hich aggregates the 8 components under eight
general +actors /#ur5, Bartram, 0 Baron, 8%%34( Ht is this +rame"or- on
"hich the current research is based( These +actors have been labeled the
Great 'ight because they appear to occupy a position "ithin the "or-
per+ormance domain /Table 4 similar to the Big Bive in the personality
predictor domain(
The Great 'ight structure provides an articulation o+ the "or-
per+ormance domain that is consistent "ith a "ide range o+ models used
by practitioners in competency practice and supported empirically by the
"ay in "hich competency ratings cluster "hen sub.ected to +actor analysis
/e(g(, Gotoh, 111> #ur5, 111> #ur5 et al(, 8%%34( Bor e@ample, #ur5 et
al( /8%%34 report an analysis o+
Note" 6ore detailed de+initions o+ each o+ the Great 'ight are provided by the competency component level o+ the SHL Iniversal Competency
Brame"or-K /see !ppendi@4(
a
Ehere more than one predictor is sho"n, the second is e@pected to be o+ lesser importance than the +irst( The competency
titles and de+initions are ta-en +rom the SHL Iniversal Competency Brame"or-K Pro+iler and Designer Cards /copyright L 8%%3 by SHL Group plc,
reproduced "ith permission o+ the copyright holder4( These titles may be +reely used +or research purposes sub.ect to due ac-no"ledgment o+ the copyright
holder(
data +rom <9; managers dra"n +rom +our
I(#(
organi5ations in di++erent industry sectors and +rom a "ide range o+
+unctional areas( T"o aptitude tests +rom the SHL 6anagement and
Graduate Htem Ban- "ere used) 7erbal Critical =easoning /76G84 and
&umerical Critical =easoning /&6G84( Hn addition, t"o ?uestion
naires "ere administered) the :ccupational Personality Juestion
naire /Concept 6odel, :PJ C63(8> SHL Group, 11<b4, "hich measures
<% scales in an ipsative /+orced-choice4 +ormat, and the Hnventory o+
6anagement Competencies /H6C> SHL Group, 11<a4, "hich measures
9 generic competency dimensions using a combined Li-ert-type and
+orced-choice +ormat( ! composite 2%
M
per+ormance score +or each H6C competency "as calculated by averaging
sel+-and boss ratings( These 9 per+ormance scores "ere entered together
"ith selected mar-er scales +rom the :PJ and the ability scales into a
principal-components +actor analysis to ascertain the underlying +actor
structure o+ the competency per
+ormance and competency potential variables( 'ight +actors reached an
eigenvalue greater than , accounting +or 92N o+ the variance( The +actor
structure "as largely in line "ith that pre
dicted by the Great 'ight model( Bactor scores generated +rom these data
"ere relatively independent> the strongest correlations "ere .ust over %(<(
:ther analyses on di++erent data sets have produced similar degrees o+ +it
to the model /e(g(, Gotoh, 111> #ur5, 1114( 6ultidimensional Scaling
analyses o+ criterion H6C competency ratings produce a circumple@
pattern, "ith pairs o+ scales clustering together according to their
pro@imity as de+ined by their e@pected Great 'ight content loadings(
Bactor analyses o+ Great 'ight predictor or criterion data sets tend to
produce higher level solutions( Hn the predictor domain, these re+lect a
broad motivational +actor, general ability, and t"o broad personality
+actors /Digman, 11$4( Bor the criterion domain, t"o +actors tend to
emerge corresponding generally to HoganCs getting along and getting
ahead distinction or BormanCs conte@tual and tas- per+ormance
distinction(
The choice o+ the Great 'ight as the level o+ analysis /rather than either
higher level constructs or more detailed competency models4 "as driven
by the need to provide a degree o+ di++erentiation o+ the criterion space that
re+lected the range o+ attributes that managers and practitioners distinguish
in practice, "hile retaining su++icient generality to enable the same model
to be applied across a "ide range o+ studies involving diverse competency
models and predictor instruments(
:ur approach to validation vie"s predictor instruments as valuable inso+ar
as they can provide valid measures o+ competency potential( The current
research concerns the relationship bet"een measures o+ competency
potential /based on personality and ability tests4 and competencies /as
assessed through supervisor ratings o+ per+ormance on various aspects o+ a
competency model4, "ith both sets o+ measures being mapped onto the
generic Great 'ight level o+ description(
Table gives the DheadlineG de+initions o+ the Great 'ight competency
+actors( Burther details can be obtained by studying the competency
components that underpin each high-level +actor /see !ppendi@4( Table
also summari5es a range o+ hypotheses concerning the e@pected
relationships bet"een high level +actors in the predictor domain and the
Great 'ight competencies( These hypotheses are based on an analysis o+
the content o+ the competencies and .udged relevance o+ the various
underlying traits(
General mental ability is e@pected to relate most strongly to !naly5ing 0
Hnterpreting competencies, because the content o+ these are heavily loaded
on general mental ability /DgG4( :n the basis o+ the content o+ the
competencies, "e "ould also e@pect general mental ability measures to
correlate "ith the ad.acent +actors o+ Presenting 0 Hnteracting as "ell as
Creating 0 Conceptuali5ing( Ht is these three areas o+ competency,
especially !naly5ing 0 Hnterpreting, that contain competency components
that are underpinned by .ob -no"ledge and s-ills and hence are most
li-ely to be "ell predicted by measures o+ general mental ability /see
!ppendi@, especially the !pplying '@pertise 0 Technology and Eriting 0
=eporting components4( :rgani5ing 0 '@ecuting is also li-ely to be
correlated "ith general ability, but more "ea-ly than the other three
competency +actors(
Hn relation to personality, the relationship bet"een the Great 'ight and the
Big Bive is not e@act( ! clearer pattern o+ results can be obtained i+ some
aspects o+ the Big Bive are di++erentiated( &otably, Big Bive
Conscientiousness combines both dependability and achievement> and
'@traversion combines aspects o+ interacting "ith others on the one hand
and dominance or potency on the other /Hough, 118> Hough, :nes, 0
7is"esvaran, 1124( Hn the Great 'ight competency model, "e "ould
e@pect the trait o+ dependability to relate to :rgani5ing 0 '@ecuting "hile
achievement relates to 'nterprising 0 Per+orming> the trait o+ sociability
should relate to Hnteracting 0 Presenting, "hile dominance and potency
aspects o+ e@traversion should relate to Leading 0 Deciding and
Hnteracting 0 Presenting competencies(
!lthough there is clearly some overlap bet"een the current model and that
described by Campbell et al( /11<4, the current model has the advantage
o+ being elaborated in terms o+ the 8 component competencies, "hich,
in turn, are lin-ed both to competency assessment measures and to
personality, motivation, and ability scales in the predictor domain( !s
such, the model provides a single +rame"or- +or ma-ing predictions +rom
measures o+ competency potential /ability, personality, and motivation4 to
ratings o+ actual "or- per+ormance(
Hypotheses
:n the basis o+ the rationale underlying the mapping o+ competency
ratings and competency potential measures to the Great 'ight model and
+rom the results o+ previous meta-analysis research, summari5ed
previously, the data "ere e@amined to e@plore a number o+ hypotheses)
Hyothesis #) Correlations bet"een matched pairs o+ Great 'ight
competency ratings and competency potential scores "ill be higher
than +or unmatched pairs(
Bor e@ample, the correlation bet"een the prediction o+ Leading 0
Deciding, based on personality measures, and managerial ratings o+
Leading 0 Deciding competencies should be higher than the correlations
bet"een this predictor and managerial ratings on any o+ the other seven
Great 'ight competencies(
Hyothesis $) Personality-based predictors "ill sho" non5ero
relationships "ith all eight o+ the areas o+ line managerCs competency
ratings, "hereas ability-based predictors "ill only relate to those areas
o+ competency that are underpinned by .ob -no"ledge and s-ill
ac?uisition(
Hn terms o+ the Great 'ight model, "e e@pect the strongest association
"ith ability measures to be +ound +or !naly5ing 0 Hnterpreting and
Creating 0 Conceptuali5ing( Less strong relation
ships "ould be e@pected +or Hnteracting 0 Presenting and :rga
ni5ing 0 '@ecuting(
1% B!=T=!6
Table <
%&mmary of 'ndivid&al %t&dy Details
Criteria-manager
Predictors ratings
Personality 7erbal &umerical Competency Type o+
=e+( no( Country
N
inventory ability ability model :,P study
I# 82 :PJ C63(8 76G8 &6G8 H6C O Concurrent
8 I# <; :PJ C63(8 76G8 &6G8 H6C O Concurrent
< I# 18 :PJ C63(8 76G8 &6G8 H6C O Concurrent
3 I# <1 :PJ C63(8 76G8 &6G8 H6C O Concurrent
; I# 92 :PJ C63(8 76G8 &6G8 H6C O Concurrent
9 Belgium 2< :PJ C63(8 H6C Predictive
$ I# 22 :PJ C63(8 7!< &!3 Client Concurrent
2 I# 331 :PJ C63(8 H6C O Predictive
1 S !+rica $8 :PJ C63(8 7C( &C8( H6C Predictive
% IS! $; :PJ C63(8 76G H6C O Concurrent
Tur-ey ;%< :PJ C63(8 76G &6G H6C O Concurrent
8 Brance 31 :PJ<8i H6C Concurrent
< #orea <99 :PJ C63(8 76G &6G H6C O Concurrent
3 Germany <8 CCSJ $(8 CCCH O Concurrent
; S !+rica 9; CCSJ ;(8 7CC< CCCH Concurrent
9 I#PIS! 8<9 :PJ<8n P6C Concurrent
$ I# ;3 CCSJ ;(8 CCCH Concurrent
2 I# ;< CCSJ ;(8 CCCH Concurrent
1 I# 9 ESJn ESCH &o details
8% I# 93 ESJn ESCH Concurrent
8 I# % :PJ C6;(8 Client Concurrent
88 I# 38 :PJ C63(8 7C( &6G Client Predictive
8< Belgium 99 :PJ C63(8 Client Predictive
83 I#P'gypt 9% :PJ<8i 76G &6G Client O Concurrent
8; I# << :PJ C6;(8 76G< &6G< Client O Concurrent
89 &etherlands ;3< :PJ C63(8 7! &HT8 Client O Predictive
8$ I# 33 CCSJ ;(8 7CC< &CC3 Client Predictive
82 I# 1< :PJ C63(8 76T &6G8 Client Concurrent
81 I# 8; :PJ C63(8 76G &6G Client Concurrent
6ethod
! total o+ 81 studies "ere collected +rom various client organi5ations(
They came +rom the Inited #ingdom and a number o+ other 'uropean
countries, Tur-ey and the 6iddle 'ast, South !+rica, the Bar 'ast, and the
Inited States and cover a "ide range o+ di++erent industry sectors and .obs
/although supervisory and managerial positions predominate4( The 81
studies, only ; o+ "hich have been included in any previous published
validation reports /&y+ield et al(, 11;4, have a total sample si5e o+ 3,29
people /(dn 8;Pstudy4( Criterion data consist o+ line-manager ratings
on either standardi5ed "or- per+ormance competency instruments or
client-speci+ic measures(
&ineteen o+ the 81 studies used one o+ several standardi5ed competency
rating instruments( The Hnventory o+ 6anagement Competencies /H6C>
SHL Group, 11<a4 includes 9 competency scales and 9% items( The
items are presented in sets o+ +our( 'ach item has to be rated on a scale
ranging +rom to ;, and the rater also has to select "hich o+ each set o+
+our items is most true and least true o+ the target sub.ect( H6C scoring
involves a "eighted sum o+ the scores obtained +rom the rating and +orced-
choice measures( The Customer Contact Competency Hnventory /CCCH>
SHL Group, 11$4 also has 9 competency scales and +ollo"s the same
mi@editem style +ormat as H6C( Ht has a total o+ 82 items arranged in <8
?uads( The Eor- Styles Competency Hnventory /ESCH> SHL Group,
111b4 has 9 competency scales based on 19 items, "hich are responded
to on a rating scale ranging +rom to ;( Perspectives on 6anagement
Competencies /P6C> SHL Group, 1134 has <9 competency scales based
on 33 items, "hich are responded to on a rating scale ranging +rom to
;( These "ere all developed +or use as sel+-and other-rating instruments o+
per+ormance at "or-( Bor the current study, only line-manager ratings
"ere used as data( Ten o+ the 81 studies used client-speci+ic rating
instruments(
The predictor measures include a number o+ di++erent instruments +rom the
:ccupational Personality Juestionnaire /:PJ4 +amily( :PJ Concept
6odel /<% scales plus Consistency or Social Desirability scale> SHL
Group, 11<b4 is the precursor to :PJ<8 /<8 scales plus Consistency or
Social Desirability scale> SHL Group, 111a4( :PJ C63(8 and :PJ<8i
are the +orced-choice +ormat versions o+ :PJ Concept 6odel and :PJ<8,
"hereas :PJ C6;(8( and :PJ<8n are the Li-ert-type versions( The
Customer Contact Styles Juestionnaire /CCSJ> SHL Group, 11$4 is a
version o+ :PJ designed +or use in customer service and sales settings and
is available in both a Li-ert-type rating scale /CCSJ;(8) 9 scales plus
Social Desirability scale4 and a combined rating and +orced-choice
/CCSJ$(8) 9 scales plus Consistency scale4 +orm( Binally, the Eor-
Styles Juestionnaire /SHL Group, 111b4 is a Li-ert-type ?uestionnaire
+or use in production and lo"er level service positions /ESJn) $ scales
plus Social Desirability scale4( Hn addition to these personality
instruments, a "ide range o+ ability tests /both verbal and numerical
reasoning4 "ere used( Bor the purposes o+ this study, these are treated as
e?uivalent measures o+ either verbal or numerical reasoning(
!ll the studies have personality data +rom one or another o+ the instru-
ments .ust discussed( 'ighteen o+ the 81 studies have data relating to one
or more ability tests( Seven o+ the studies "ere predictive validity designs,
and 8 "ere concurrent> no details "ere available +or ( Total samples
si5es and data about the studies, .ob types, industry sectors, and other
demographics are presented in Tables 8A9(
(aing %!ales to the Cometen!ies
! standard set o+ mappings o+ personality and ability scales onto the Great
'ight competencies had been de+ined +or each o+ the predictor
Table 8
N&m)er of %t&dies and Total %amle %i*es Having
Criterion Data and +redi!tor Data for Ea!h of the
Great Eight Cometen!ies
Criteria Predictors Competency &o( studies n !bility Personality
LeadingPDeciding 82 <,291 8,91$ 3,3;; SupportingPCooperating 89
<,$$2 8,91$ 3,3;1 HnteractingPPresenting 82 <,2$3 8,91$ 3,3;2
!naly5ingPHnterpreting 8$ <,$$ 8,91$ 3,3;$ CreatingPHnnovating 8
<,82% 8,91$ 3,3;; :rgani5ingP'@ecuting 81 <,1$ 8,91$ 3,3;2
!daptingPCoping 8; <,993 8,91$ 3,3;< 'nterprisingPPer+orming 8$
<,$38 8,91$ 3,3;2 &o( studies 2 81
instruments /personality and ability tests4 and +or the standardi5ed compe-
tency rating instruments /H6C, CCCH, P6C, and ESCH4( These mappings
are in the +orm o+ linear "eighted composites and "ere de+ined a priori on
the basis o+ ratings o+ construct overlap bet"een traits /individual scales4
and each o+ the 8 competency components in the +rame"or-( !s de-
scribed earlier, in this +rame"or- each o+ the Great 'ight competency
+actors is composed o+ a subset o+ the 8 competency components( The
relevance o+ each competency component +or each personality scale had
been initially estimated using independent .udgments o+ three sub.ect
matter e@perts( 'ach possible scaleAcomponent relationship "as rated on a
scale ranging +rom % /not relevant4to3/highly relevant4( /The scale
"as anchored to e@pectations o+ correlations bet"een scales and items o+
bet"een % and %(3% +or the % to 3 ratings(4 ,udgments "ere revie"ed and a
+inal set o+ component relevancies agreed on( The outcome o+ this process
is a large matri@ o+ components by scales, populated by relevance ratings
in the range o+ 3to 3( Scale "eights +or producing Great 'ight scores
+rom, +or e@ample, :PJ<8 scales "ere then de+ined by aggregation o+ the
relevant sets o+ component "eights into a single linear e?uation(
The validation o+ this approach is described in Bartram /8%%4( Burther
research has also demonstrated the robustness o+ this approach o+ aggre-
gating measures at the component level( Bor the current study, aggregation
o+ scaleAcomponent relations up to the Great 'ight +actor level "as used
to provide the basis +or identi+ying a small number o+ mar-er scales
/generally no more than three scales4 +or each o+ the eight +actors( Typi-
cally, this aggregation process results in a small number o+ scales having
large "eights /i(e(, aggregated relevancies4 and a larger number having
small "eights( Discarding the lo"er "eighted scales reduces the correla-
tions bet"een composites "ithout adversely a++ecting their validities( The
outcome o+ this process "as that +or the :PJ<8, +or e@ample, each o+ the
Great 'ight competencies is measured by three scales /see Table $ +or list
o+ mar-er scales used in predictor measures o+ each o+ the Great 'ight
competencies4, "ith di++erent scales being used +or each o+ the Great 'ight
/i(e(, 83 o+ the <8 available scales are used in all4( Bor ESJ and CCSJ, it
"as not possible to avoid some overlap o+ scale content, given the need to
ensure good coverage o+ the content o+ each competency domain( Ho"-
ever, this had relatively little impact on correlations among the +inal
overall set o+ eight composites /see Table 24(
Ht is important to note that the personality instruments used in these studies
have a broader coverage than Big Bive instruments( Hn particular, the :PJ
model covers aspects o+ motivation and cognitive style as "ell as the more
traditional areas o+ personality( Bor e@ample, Table $ lists scales such as
Controlling, !chieving, and Competitive( :thers that load on
competencies li-e Leading 0 Deciding or 'nterprising 0 Per+orming
include scales such as Persuasive and 'nergetic( Hn the area o+ cognitive
style, the :PJ model includes scales such as 'valuative and Data
=ational(
,ob type Bre?uency Percentage
!ccount e@ecutives 33 8(19
Bro-er consultants 9; <(<1
Collections sta++ ;3 (
'@ecutives 99 <(3
6anagers <,83< 99($8
Sales reps <8 8($8
Sales sta++ ;< (%1
Shop +loor sta++ 93 (<8
System developers 8<9 3(29
Trainees ;3< ($
In-no"n 9 (8;
Total 3,29 %%(%%
Bor client competency models, mapping to the Great 'ight "as carried out
post hoc by t"o sub.ect matter e@perts on the basis o+ the content and the
de+initions given to the competencies in the client model( Brom these
de+initions, it "as possible to map these models to the 8% dimension levels
o+ the +rame"or- and then aggregate the results up to the Great 'ight
level( There "as generally insu++icient detail available on the nature o+ the
competencies +rom client models to permit an analysis in terms o+ the 8
Table 3
%amle ,rea-down )y Co&ntry
Country Bre?uency Percentage
Belgium 831 ;(8 'gypt 9% (8< Brance 31 %(% Germany <8 8($8
Holland ;3< ($ #orea <99 $(;< Saudi !rabia 8<$ 3(2$ Tur-ey ;%<
%(<; Inited #ingdom ,291 <2(3; Inited States 3 2(39 Total 3,29
%%(%%
components( 6ost o+ the client models covered a subset o+ the Great 'ight
competencies( !s a conse?uence, the total number o+ studies having data
on each competency +actor varies +or each o+ the Great 'ight
competencies /see Table 84(
!ll scale scores /both predictor and criterion4 "ere trans+ormed into *
scores be+ore being "eighted and combined into Great 'ight scores( The
Table ;
%amle ,rea-down )y .o) Tye
Note" See te@t +or +urther in+ormation on the competency models and predictor measures( Complete data "ere not available +or all samples( :,P
overall .ob per+ormance ratings> N ma@imum number in each sample> O yes> H6C Hnventory o+ 6anagement Competencies> CCCH Customer
Contact Competency Hnventory> P6C Perspectives on 6anagement Competencies> ESCH Eor- Styles Competency Hnventory(
Table 9
%amle ,rea-down )y 'nd&stry %e!tor
Hndustry sector Bre?uency Percentage
Ban-ing 29 <(2< Call center %$ 8(8% Distribution <1 8(29
'ngineering <99 $(;< Binance 18 (21 Bood manu+acturing 31 %(%
Hospitality <82 9($; Hnsurance <%1 9(<9 Hn+ormation technology ;3<
($ 6anu+acturing 2$% $(1% Pharmaceuticals 22 (2 Public sector
9; <(<1 =etail 89; ;(3; Telecommunications 8<9 3(29 In-no"n 9
(8; 7arious 9; 8(9; Total 3,29 %%(%%
Great 'ight scores "ere then trans+ormed into * scores( Hn each case,
standard score trans+ormations "ere carried out using all those people in
the combined set o+ studies having data on the respective measures(
/.+ 0atings
Some o+ the studies also had data on ratings o+ :,P /< studies4( Hn most
cases, these ratings used a si@-item rating instrument used in the Hnterna-
tional 7alidation Study reported by &y+ield et al( /11;4( The reliability o+
this scale "as %($%(
(eta-1nalysis +ro!ed&res
The procedures used in the meta-analysis "ere as described by Hunter and
Schmidt /11%4> individual corrections "ere carried out on each study(
6eta-analysis ad.ustments included corrections +or criterion reliability and
predictor range restriction( The criterion reliability o+ the Great 'ight
competencies based on the standardi5ed competency instruments "as set
at %($;, "hereas that +or client competency models "as set at %(;8( The
latter value is ta-en +rom Hermelin and =obertsonCs /8%%4 arti+act
distribution estimates, "hereas the +ormer ta-es account o+ published data
on the reliability o+ ratings using the standardi5ed instruments H6C /SHL
Group, 11<a4, P6C /SHL Group, 1134, and CCCH /SHL Group, 11$4(
=ange restriction "as calculated +or each study on the basis o+ di++erences
bet"een the composite score predictor standard deviations +or the selected
samples and the standard deviations +or the relevant applicant groups
"hen these data "ere available( Ehen applicant data "ere not available
/about ;%N o+ the cases4, a relevant norm group "as used /i(e(, one based
on similar applicant samples rather than on the general population4 to
estimate degree o+ range restriction +or the underlying scales o+ the
composites( Because range restriction tended to be uni+orm across scales
+or personality, range restrictions +or composites in this case "ere
estimated on the basis o+ the average range restrictions across the relevant
scales( The average ratio o+ restricted standard deviations to unrestricted
standard deviations "as %(2% /%(%3; %D across studies4 +or the personality
measures and %(2% /%(%98 %D across studies4 +or ability tests(
=esults
0elationshis ,etween 2aria)les 3ithin the
+ersonality-,ased +redi!tor %et and the
Criterion %et
The average correlation bet"een the eight predictor composites
"as (%$, "hereas +or the criteria it "as %(3; /see Table 2 +or the
Table $
+ersonality
'nventory %!ales
4sed to
Generate
Cometen!y
+otential %!ores
G2
+actor
Compe
tency
domain
title
:PJ<8
:PJ
C6
CCSJ
ESJ
Leading 0 Deciding Controlling,
Persuasive, Decisive Controlling,
Persuasive, Decisive Persuasive, =esults
:riented !ssertive, !chieving, Decisive8
Supporting 0 Cooperating Caring,
Democratic, !++iliative Caring,
Democratic, !++iliative 'mpathic,
Participative, Sel+-Considerate, Team-
:riented,Control Dependable<
Hnteracting 0 Presenting Socially
Con+ident, :utgoing,Socially Con+ident,
:utgoing,Sociable, 6odest /-ve4,
Persuasive Socially Con+ident,
!ssertive,6odest /-ve4 6odest /-ve4
!daptable 3 !naly5ing 0 Hnterpreting
'valuative, Data =ational,Critical, Data
=ational, Conceptual !nalytical,
Structured, Hnnovative Practical,
:rgani5ed, HnnovativeConceptual;
Creating 0 Conceptuali5ing Hnnovative,
Hndependent,Hnnovative,
Hndependent,Hnnovative, Ble@ible,
!nalytical Hnnovative, !daptable,
!chievingConventional /-ve4 Traditional
/-ve4 9 :rgani5ing 0 '@ecuting
Conscientious, Detail
Conscious,Conscientious, Detail
Conscious,Conscientious, Detail
Conscious,Dependable, Detail
Conscious,Bor"ard Planning Bor"ard
Planning Structured :rgani5ed $
!dapting 0 Coping Tough-minded,
=ela@ed, :ptimistic Tough-minded,
=ela@ed,=esilience, Sel+-Control,
Ble@ible =esilient, 'motionally:ptimistic
Controlled, :ptimistic2 'nterprising 0
Per+orming !chieving, Competitive,
7igorous !chieving, Competitive, !ctive
=esults :riented, Competitive,!chieving,
Competitive, !ctive 'nergetic
Note" Hn each case, the
+irst scale received a
"eight o+ t"o and the
other scale/s4 "ere unit
"eighted( Hn some
instances +or the ESJ
and CCSJ, the same
scales load on t"o Great
'ight
competencies(!lthough
this introduces a
necessary degree o+
correlation bet"een
these Great 'ight
predictors /see Table 24,
the overlap "as
included to ensure
su++icient breadth o+
coverage o+ the relevant
constructs(:PJ
:ccupational
Personality
Juestionnaire> C6
Concept 6odel> CCSJ
Customer Contact
Styles Juestionnaire>
ESJ Eor- Styles
Juestionnaire(
Table 2
Predictors) rotated +actor loadings Criteria)
rotated
+actor
loadings
7ariable 8 < 8
LeadingPDeciding
SupportingPCooperatin
g
HnteractingPPresenting
!naly5ingPHnterpreting
CreatingPConceptuali5i
ng
:rgani5ingP'@ecuting
!daptingPCoping
'nterprisingPPer+ormin
g
($1
(9 (
91 (8% (
9%
($ (
(9%
(%$
(%
(8$
($3 (
<9 ($<
(%2 (
8
(%8 (
$; (<3
(2
(%<
(; (
;1
(<8
(92 (
; (
;$ (
28 (
2% (
$< (
8< (
$<
(
33
(
29
(
3;
(
%2
(
(
89
(
2%
(
<<
1verage 'nter!orrelations 54n!orre!ted6 for Great Eight Cometen!y (anager
0atings 51)ove Diagonal6 and Great Eight +redi!tors ,ased on +ersonality Data
/nly 5,elow Diagonal6
7ariable LPD SPC HPP !PH CPC :P' !PC 'PP
LPD Q %(3 %(;$ %(3$ %(;8 %(;9 %(3$ %(9 N <,$12 <,21; <,2$< <,<1 <,118 <,923 <,$91 SPC %(%2 Q %(3<
%(82 %(89 %(<$ %(; %(<9 N 3,239 3,%9 <,23 <,;88 3,%8 <,232 <,$2< HPP %(33 %(< Q %(;% %(3$ %(<$ %(3%
%(32 N 3,23; 3,231 <,18 <,9%9 3,1< <,12$ <,228 !PH %(%1 %(9 %(%3 Q %(;2 %(;9 %(< %(;% N 3,233
3,232 3,23$ <,3%< 3,%%2 <,$%% <,21$ CPC %(8$ %( %(8% %(<< Q %(39 %(<; %(;; N 3,238 3,239 3,23;
3,233 <,9%% <,;9 <,<$8 :P' %(%$ %(% %(9 %(8% %(%; Q %(<; %(;9 N 3,23; 3,231 3,232 3,23$ 3,23;
<,12< <,12% !PC %(%3 %(%3 %(%2 %(%1 %(%; %(%% Q %(3< N 3,23% 3,233 3,23< 3,238 3,23% 3,23< <,$;< 'PP
%(<9 %(8< %(8% %(%1 %(8< %(%1 %(%3 Q N 3,23; 3,231 3,232 3,23$ 3,23; 3,232 3,23<
Note" LPD LeadingPDeciding> SPC SupportingPCooperating> HPP HnteractingPPresenting> !PH
!naly5ingPHnterpreting> CPC CreatingPConceptuali5ing> :P' :rgani5ingP'@ecuting> !PC
!daptingPCoping> 'PP 'nterprisingPPer+orming(
complete matri@4( The overall positive correlation +or the criteria
is not surprising because these are line-manager rating data( Ho"-
ever, this analysis does sho" that there is a good degree o+
independence bet"een the eight predictors( Bor the predictors, the
main correlations are bet"een !naly5ing 0 Presenting and Cre-
ating 0 Hnnovating on the one hand /r (<<4 and bet"een
Leading 0 Deciding, Hnteracting 0 Presenting, and 'nterprising
0 Per+orming on the other /average r (<<4(
Hn line "ith previous "or-, principal-components analyses o+ the
predictor and criterion matrices produced three-and t"o-+actor
solutions, respectively( Bor the predictors the three +actors ac-
counted +or ;;(32N o+ the variance, and +or the criteria the t"o
+actors accounted +or 9;(%<N( 7arima@-rotated loadings are
sho"n in Table 1( Bor the criteria, these +actors broadly represent
tas- /Bactor 4 versus conte@tual behaviors /Bactor 84, "hereas +or
the predictors the meaning o+ the +actors is less clear(
Table 1
2arima7 0otated Loadings 8rom +rin!ial-
Comonents 1nalyses of the Eight +redi!tors
and Eight Criteria
The 2alidity of +ersonality Data for the
+redi!tion of %&ervisor-0ated Cometen!ies
Sample-"eighted average predictorAcriterion correlations +or each
o+ the cells o+ the 2 2 matri@ o+ predictors and competencies
across all the studies are sho"n in Table %( The diagonal cells are
the hypothesi5ed relationships /Hypothesis 4( Ht can be seen that
these are all non5ero and are the largest values in each ro" or
column in all cases but one( Bor the personality-based predictor o+
Leading 0 Deciding, the correlation "ith Leading 0 Deciding
competency ratings is e?ual to that "ith Hnteracting 0 Presenting
and only slightly higher than that "ith 'nterprising 0 Per+orming(
The results presented in Table % support Hypothesis , because
the average correlation o+ the hypothesi5ed relationship is (9,
"hereas the average correlation o+ the nonhypothesi5ed relation-
ships /the o++-diagonal cells in Table %4 is (%8(
These results can be compared to use o+ the Big Bive as a
+rame"or- +or structuring the predictor scales( Standard Big Bive
e?uations +or :PJ<8 and :PJ C6 "ere applied to these instru-
ments to produce Big Bive predictor scores( These e?uations are
based on correlations bet"een :PJ and the &': Personality
HnventoryA=evised /&':-PH-=> SHL Group, 111a4( Table
sho"s the correlations bet"een Big Bive and Great 'ight predic-
tors o+ competency potential and provides an empirical underpin-
ning to the assignments noted in Table ( There is close concor-
dance among '@traversion, !greeableness, Conscientiousness,
and &euroticism /negative4 on the one hand and Hnteracting 0
Presenting, Supporting 0 Cooperating, :rgani5ing 0 '@ecuting,
and !dapting 0 Coping on the other( '@traversion is also corre-
lated "ith Leading 0 Deciding, "hereas !greeableness correlates
negatively "ith 'nterprising 0 Per+orming( :penness to &e"
'@perience correlates "ith both !naly5ing 0 Hnterpreting and
Creating 0 Conceptuali5ing( Hn general, these results sho" that
the Big Bive relate to the Great 'ight in the manner e@pected /as
described in Table 4(
Table %
Big Bive predictors
Personality-based Great
'ight predictor '@ample :PJ<8 mar-er scales '@traversion !greeableness :penness Conscientiousness &euroticis
m
LeadingPDeciding Controlling, Decisive +.,- %(; %(%3 %(8 %(<
SupportingPCooperatin
g
Caring, !++iliative %($ +..+ %(%; %(%9 %(%9
HnteractingPPresenting Socially Con+ident, :utgoing +./. %(% %(%1 %(88 %($
!naly5ingPHnterpreting 'valuative, Data =ational %(% %(2 +.,. %(9 %(%9
CreatingPConceptuali5
ing
Hnnovative, Hndependent %($ %(2 +.01 %(8 %(%
:rgani5ingP'@ecuting Conscientious, Bor"ard Planning %(8< %(%% %(% +..0 %(%9
!daptingPCoping Tough-minded, =ela@ed %(%8 %(%8 %( %(%9 +./0
'nterprisingPPer+ormi
ng
!chieving, Competitive %(<
+.21
%(% %(%9 %(%
Competency ratings
'@ample :PJ<8
Predict
or
mar-er scales LPD SPC HPP !PH CPC :P' !PC 'PP Hypothesi5ed &onhypothesi5ed
LPD Controlling, Decisive %(2 %(%8 %($ %(%3 %( %(%8 %(%% %($ %(2 %(%$
SPC Caring, !++iliative %(%8 %( %(%< %(% %(% %(%3 %(%< %(%$ %(% %(%3
HPP Socially Con+ident,
:utgoing %( %(%9 %(1 %(%% %(%2 %(%< %(%3 %(%2 %(1 %(%;
!PH 'valuative, Data
=ational %(%8 %(%$ %(%8 %(9 %(8 %(%$ %(%< %(%8 %(9 %(%8
CPC Hnnovative, Hndependent %(%$ %(%
<
%(%2 %(%1 %(2 %(%% %(%8 %(%9 %(2 %(%3
:P' Conscientious, Bor"ard
Planning %(%% %(%
<
%(%
$
%(%; %(%
3
%(; %(%8 %(%< %(3 %(%
!PC Tough-minded, =ela@ed %(%% %(%8 %(%
8
%(%

%(%

%(%

%(8 %(%

%(8 %(%
'PP !chieving, Competitive %( %(%
<
%(% %(%$ %( %(%3 %(%8 %(1 %(1 %(%9
(
%(9 %(%8
Note" &umber o+ studies 81> sample si5e range <,82%A<,1$( :PJ<8 :ccupational Personality Juestionnaire-<8( LPD LeadingPDeciding> SPC
SupportingPCooperating> HPP HnteractingPPresenting> !PH !naly5ingPHnterpreting> CPC CreatingPConceptuali5ing> :P' :rgani5ingP'@ecuting>
!PC !daptingPCoping> 'PP 'nterprisingPPer+orming(
Table 8 indicates that Big Bive predictors have the e@pected pattern o+
correlations "ith the Great 'ight competencies( !s one "ould e@pect,
ho"ever, they do not +it as "ell as the predictors designed speci+ically to
measure the competencies on a one-to-one basis( The average o+ the
absolute values o+ the correlations predicted +or the +ive Great 'ight
criteria hypothesi5ed to relate most strongly to the Big Bive is (, "hich
is some"hat lo"er than the average correlation +or the Great 'ight /(94
+or the pair"ise predictions sho"n in Table %( The Big Bive sho" much
"ea-er predictions o+ Leading 0 Deciding, 'nterprising 0 Per+orming,
and !naly5ing 0 Hnterpreting( !s noted earlier, this is li-ely because :PJ
instruments provide coverage o+ aspects o+ motivation /speci+ically, need
+or achievement and need +or po"er and control4 and cognitive style that
lie outside the Big Bive domain(
(eta-1nalysis of the Great Eight +redi!tor9
Criterion +airs for +ersonality-,ased
+redi!tors Corrections +or criterion reliability and range restriction "ere carried out
only on the hypothesi5ed correlations /i(e(, those sho"n
Table
in the diagonal o+ Table %4( &one o+ the o++-diagonal /see Table %4
correlations "ere included in the meta-analyses( The resulting corrected
correlations /Table <4 +or the personality-based predictions o+ the Great
'ight are moderate to high +or all eight +actors, ranging +rom %(9 to %(82(
The largest values obtained are, in +act, +or the motivation-and
e@traversion-driven +actors /Leading 0 Deciding, Hnteracting 0
Presenting, and 'nterprising 0 Per+orming competencies4(
The 2alidity of 1)ility Test Data as
+redi!tors of %&ervisor-0ated
Cometen!ies !bility tests "ere predicted to correlate most strongly "ith !naly5ing 0
Hnterpreting competencies and to sho" relationships "ith Creating 0
Conceptuali5ing and "ith Hnteracting 0 Presenting( The results o+ the
meta-analysis carried out on ability tests only are presented in Table <
and clearly sho" the e@pected pattern( The strongest relationship / (
3%4 is, as e@pected, "ith !naly5ing 0 Hnterpreting competencies, "hich
are strongly underpinned in the competency +rame"or- by components
relating to
Correlations ,etween ,ig 8ive and Great Eight +ersonality-,ased Comosite
+redi!tors
%amle-3eighted 1verage Correlations 54n!orre!ted6 ,etween Great Eight Cometen!y ,oss 0atings and
+redi!tors ,ased on +ersonality Data /nly: 3ith Hyothesi*ed 5Diagonal6 and Nonhyothesi*ed 5(ean /ff-
Diagonal6 2al&es
Note" N varies by ro"s +rom 3,888 to 3,889( :PJ<8 :ccupational Personality Juestionnaire-<8( Bold+ace numbers indicate hypothesi5ed Great
'ightABig Bive relationships(
13
B!=T=!6
Table 8
Predictor &o( studies
N r

%D
Nvar %N
CrH
Personality only
LeadingPDeciding 82 <,;1; %(2 %(89$ %(%;3 23(12 %(1$
SupportingPCooperating 89 <,3$% %(%1 %(9 %(;1 <2(1$ %(%3
<
HnteractingPPresenting 82 <,;92 %(2$ %(8$$ %(<3 3$(8 %(%;
!naly5ingPHnterpreting 8$ <,;;3 %(93 %(833 %(8; ;%(93 %(%23
CreatingPConceptuali5ing 8 8,113 %(28 %(89% %(%$1 9$(1< %(;1
:rgani5ingP'@ecuting 81 <,9$% %(;% %(82 %($$ <;(%9 %(%%
1
!daptingPCoping 8; <,<99 %(88 %($; %(%2 ;$(9$ %(%<9
'nterprisingPPer+orming 8$ <,;9 %(21 %(8$; %(%9 2%(;3 %(1$
!bility only
LeadingPDeciding 2 8,99 %(%33 %(%$; %(1% <;(98 %(9
2
SupportingPCooperating 9 8,82 %(%%2 %(%83 %(%8 93( %(%
$
HnteractingPPresenting 2 8,$% %(3$ %(89 %(;1 38($; %(%8
!naly5ingPHnterpreting $ 8,< %(8$9 %(3%3 %(12 89($% %(;%
CreatingPConceptuali5ing 8 ,$<% %($8 %(83 %(%19 ;1(22 %(2
:rgani5ingP'@ecuting 2 8,$% %(%; %(;2 %(9% 3<($9 %(%3
$
!daptingPCoping ; ,129 %(%;% %(%$; %(%33 1%(< %(%2
'nterprisingPPer+orming 9 8,%;$ %(%81 %(%;< %(9< 3%(9< %(;
9
LeadingP SupportingP HnteractingP !naly5ingP CreatingP :rgani5ingP !daptingP 'nterprisingP
7ariable Deciding Cooperating Presenting Hnterpreting Conceptuali5ing '@ecuting Coping Per+orming
N
<,8<9 <,38 <,<%1 <,88
8,$;$ 8,23; <,%< <,8%%
Big Bive predictors
'@traversion +.+. %(%9 +.1/ %(%% %(%$ %(%; %(%% %(%1
!greeableness %(% +.+. %(%; %(%9 %(%2 %(% %(% +.+-
:penness %(% %(%< %(%3 +.+. +.1, %(%8 %(%8 %(%
Conscientiousness %(%% %(%9 %(%$ %(%9 %(%9 +.13 %(%< %(%
&euroticism %(% %(% %(% %(%% %(%8 %(%3 +.+. %(%8
N
o
t
e
"
Bold+ace numbers indicate hypothesi5ed Great eightABig +ive relationships(
1verage Correlations ,etween ,ig 8ive +ersonality-,ased +redi!tors and Great Eight Cometen!y
Criteria
Great 'ight competency criterion ratings
.ob -no"ledge and .ob s-ills( :verall, there are relationships bet"een
ability and the middle +our competencies /+rom Hnteracting 0 Presenting to
:rgani5ing 0 '@ecuting4( !s e@pected /Hypothesis 84, the +irst and last
pairs o+ Great 'ight competencies /see Table 4 sho" no relationships "ith
ability(
The 2alidity of Com)ined +ersonality and
1)ility Test Data as +redi!tors of %&ervisor-
0ated Cometen!ies Personality and ability test predictions "ere combined using regression
"eights /see Table 34( =egression analyses, "ith composite personality
and composite ability measures as the t"o predictors and competency
ratings as the criterion, "ere per+ormed +or each o+ the Great 'ight +or
those cases in "hich there "ere data
Table <
on both personality and ability tests( Hn cases in "hich there "ere verbal
and numerical reasoning tests /rather than .ust one or the other4, an e?ual-
"eighted composite o+ these "as used as the estimate o+ general ability(
Bor ability tests there is, o+ course, only one predictor +or each o+ the Great
'ight competencies, "hereas +or the personality data each o+ the Great
'ight has a distinct personality-based predictor( Ht "as decided not to use
the standardi5ed regression "eights given in Schmidt and Hunter /1124,
because these relate to the e++ect o+ adding measures o+ conscientiousness
to general mental ability and relate to validity studies in "hich the criteria
"ere overall measures o+ .ob per+ormance( The ratio o+ ability to
personality regression "eights +rom Schmidt and Hunter is (9;( Bor the
current study, it is (2; +or !naly5ing 0 Hnterpreting /the most g-loaded
competency4 but less than (% +or the rest o+ the Great 'ight(
(eta-1nalysis 0es&lts for Great Eight Cometen!ies ,ased on +ersonality-/nly +redi!tors
and 1)ility-/nly +redi!tors for 1ll Those Cases Having Data on Either /ne or the /ther
or ,oth %ets of (eas&res
Note" r sample-"eighted average o+ uncorrected correlations> sample-"eighted average o+
correlations corrected +or arti+acts> %D standard deviation o+ corrected correlations> %NCrH lo"er
%N boundary o+ 2%N credibility interval> Nvar percentage variance accounted +or by arti+act corrections +or
corrected correlation distribution(
Table 3
7ariable &o( studies
N r

%D
Nvar %N
CrH
Personality only
LeadingPDeciding 2 8,;8 %(93 %(83; %(%$8 $2($$N %(;<
SupportingPCooperating 9 8,3 %(<% %(1$ %(;1 3%(%%N %(%%
$
HnteractingPPresenting 2 8,;$ %(88 %(<81 %(%9 9%(<9N %(18
!naly5ingPHnterpreting $ 8,8 %($1 %(893 %(33 3;(92N %(%$1
CreatingPConceptuali5in
g
8 ,$8$ %(8< %(<%; %(%98 $$(3N %(889
:rgani5ingP'@ecuting 2 8,;9 %(9< %(8<2 %(<1 31(9;N %(%9%
!daptingPCoping ; ,1$$ %(; %(93 %(3 ;$(%9N %(%1
'nterprisingPPer+orming 9 8,%; %(98 %(8<$ %(%3$ 22(9N %($$
!bility only
LeadingPDeciding 2 8,;8 %(%3< %(%$3 %(22 <9(<%N %(9
$
SupportingPCooperating 9 8,3 %(%%$ %(%88 %(8% ;;(1$N %(<
8
HnteractingPPresenting 2 8,;$ %(;% %(81 %(;; 33(%1N %(%8
!naly5ingPHnterpreting $ 8,8 %(8$9 %(3%3 %(12 89(9$N %(;%
CreatingPConceptuali5in
g
8 ,$8$ %($< %(838 %(%13 9%(12N %(88
:rgani5ingP'@ecuting 2 8,;9 %(%3 %(;9 %(;$ 33($%N %(%3
;
!daptingPCoping ; ,1$$ %(%; %(%$9 %(%;3 29(3N %(%%$
'nterprisingPPer+orming 9 8,%; %(%82 %(%; %(9% 3($2N %(;
3
Personality and ability
LeadingPDeciding 2 8,;8 %($ %(8;$ %(%<3 13(8;N %(8<
SupportingPCooperating 9 8,3 %(<< %(8% %(93 <2(1N %(%%
1
HnteractingPPresenting 2 8,;$ %(8$% %(<1$ %(<3 39(81N %(88;
!naly5ingPHnterpreting $ 8,8 %(811 %(3<2 %($$ <%(<N %(8%
CreatingPConceptuali5in
g
8 ,$8$ %(8;< %(<;$ %(%;3 2%(32N %(82$
:rgani5ingP'@ecuting 2 8,;9 %(8%; %(<%8 %(%9; 2(<%N %(81
!daptingPCoping ; ,1$9 %(82 %(2% %(%11 9<(;1N %(%;<
'nterprisingPPer+orming 9 8,%; %(9< %(83% %(%82 1;(32N %(8%3
%tandardi*ed 0egression 3eights 4sed to Com)ine +ersonality-and 1)ility-,ased
+redi!tors of the Great Eight
Correlation) Personality !bility Great 'ight criterion personality =atio competency "ith ability
o+ s 0
LeadingPDeciding %(%$ %(9% (%% %(%;< (%; %(<< %(91 SupportingPCooperating %(<1 %(8$ (%% %(%82
ns %(88% %(8$ HnteractingPPresenting %(%%2 %(88 (%% %(;2 (%% %($; %(8$ !naly5ingPHnterpreting %(8;1
%(88 (%% %(889 (%% (2;8 %(82< CreatingPConceptuali5ing %(29 %(29 (%% %(;< (%% %(28< %(898
:rgani5ingP'@ecuting %(%1 %(2< (%% %($ (%% %(9<1 %(8%9 !daptingPCoping %(%;9 %(8$ (%% %(%31 (
%; %(<29 %(<< 'nterprisingPPer+orming %(<2% %(99 (%% %(% ns %(%99 %(99
Note" N ,$8$A8,;$( &umber o+ studies 8A2(
The results o+ meta-analysis o+ predictorA criterion relation-Creating 0 Conceptuali5ing, and, to a lesser degree, :rgani5ing ships conducted using only
those studies that included both 0 '@ecuting is increased by the addition o+ ability test in+orpersonality and ability predictors are presented in Table ;(
mation( Bor the others, addition o+ ability test data results in This sho"s the outcome o+ the analysis o+ combining person-little or no change o+ validity(
The +inal sample-"eighted avality and ability data and also the results +or each alone +or the erage corrected validities +or the combined measures range
same subset o+ cases +rom the data set( The validity o+ predic-+rom %(8% +or Supporting 0 Cooperating to %(33 +or !naly5ing tions o+ Hnteracting 0
Presenting, !naly5ing 0 Hnterpreting, 0 Hnterpreting(
Table ;
(eta-1nalysis 0es&lts for Great Eight Cometen!ies ,ased on +ersonality-/nly
+redi!tors: 1)ility-/nly +redi!tors: and Com)ined +ersonality and 1)ility
+redi!tors for 1ll Those Cases Having Data on ,oth +ersonality and 1)ility
Note" r sample-"eighted average o+ uncorrected correlations> sample-"eighted average o+
correlations corrected +or arti+acts> %D standard deviation o+ corrected correlations> %NCrH lo"er
%N boundary o+ 2%N credibility interval> Nvar percentage variance accounted +or by arti+act corrections +or
corrected correlation distribution(
2alidity Generali*a)ility and %it&ational %e!ifi!ity
'@amination o+ Table ; sho"s considerable variance in the corrected
correlations +or some competencies( Bor the combined personality and
ability predictors, the percentage o+ variance accounted +or passes the $;N
rule in ;%N o+ the cases, "hereas +or personality alone this is only true +or
three o+ the eight competencies /Leading 0 Deciding, Creating 0
Hnnovating, and 'nterprising 0 Per+orming4 and +or ability alone .ust one
/!dapting 0 Coping4 o+ the eight( Ho"ever, in the case o+ ability and
!dapting 0 Coping competencies, the average validity is close to 5ero,
implying that ability is never a valid predictor +or this aspect o+
competency( Hn their study o+ meta-analyses, Hermelin and =obertson
/8%%4 noted that the average percentage o+ variance accounted +or across
studies tends to be nearer to ;%N than $;N, indicating the presence o+
genuine moderating e++ects( The average variance accounted +or across the
eight competencies /+or the results presented in Table ;4 is 99(83N,
"hich is higher than the Hermelin and =obertson estimate but "ell belo"
the $;N cuto++(
Combining personality and ability as predictors has some une@pected
e++ects on generali5ability( !lthough the three competencies that are above
the $;N cuto++ +or Personality alone remain so +or the combined
predictors, :rgani5ing 0 '@ecuting passes the $;N cuto++ +or the
combined predictors but not +or either on its o"n( !dapting 0 Coping,
"hich "as generali5able according to the ability data, is not generali5able
according to the combined data( Ht is di++icult to e@plain these anomalies
+rom e@amination o+ the data( :ne possibility is that they arise through
instability o+ the estimates o+ variance accounted +or( This is li-ely to be
especially true "hen the average correlation is small and the bet"een-
study variance is also small(
0elationshis ,etween 1ggregated
Criteria and 1ggregated +redi!tors
!ll the analyses reported previously +ocus on the individual pair"ise
relationships bet"een each o+ the Great 'ight predictorA criterion pairs( Hn
practice, selection decisions are made on the basis o+ aggregation o+
in+ormation +rom multiple sources /such as predictors o+ each o+ the Great
'ight competencies4 and are best evaluated in terms o+ their overall
relationship "ith aggregated criteria( !n estimate o+ the upper limit on the
overall relationship bet"een optimally "eighted aggregated sets o+
criterion and predictor measures can be obtained by e@amining the
canonical correlation bet"een the predictor and criterion vectors(
Canonical components +or personality only /eight predictors and eight
criterion measures4 and personality and ability /nine predictors and eight
criterion measures4 +or the same set o+ cases accounted +or 82(2$N o+ the
variance +or personality only /0 (;34 and <%(99N o+ the variance +or
personality and ability combined /0 (;;4( !lthough these +igures
provide an estimate o+ the validity obtainable "ith aggregated multiple
criteria, the +ollo"ing should be noted)
These results are based on the uncorrected correlation matrices
and as such are underestimates o+ the covariance bet"een predictor and
criterion sets(
8 Canonical correlation, li-e multiple regression, capitali5es on
chance and provides us "ith a best +itting solution
+or this data set( Thus, the covariance accounted +or in ne" data sets by
the canonical e?uations developed +rom this data set "ould sho"
shrin-age(
The e@tent to "hich this level o+ correlation is obtainable in practice can
be estimated by assuming that ratings o+ :,P provide an indication o+ the
relative "eights an organi5ation places on the eight criterion
competencies( By regressing the :,P measures, "hen these are available,
onto the Great 'ight criterion competency ratings, it "ill be possible to
create aggregated criterion competency scores using the regression beta
values as "eights( 6ultiple regression can then be used to see ho" the
predictor competency potential scores relate to this aggregate +or each
study(
+redi!ting /.+
Data on :,P ratings "ere available +or < studies( :+ these, only % also
had ratings on all eight o+ the criterion competencies( The current analysis
is restricted to those % studies /N ,2934( =egression analyses "ere
carried out on the ra" data /"ith no corrections +or arti+acts4 +or each
study( Hndividual study beta "eights and multiple correlations, and
sample-"eighted average beta "eights are sho"n in Table 9(
:+ the eight competency ratings, those most strongly related to :,P
ratings, in order o+ importance /see Table 94, are !naly5ing 0
Hnterpreting, :rgani5ing 0 '@ecuting, 'nterprising 0 Per+orming,
Leading 0 Deciding, and Creating 0 Conceptuali5ing( The more
conte@tual competencies /Supporting 0 Cooperating, !dapting 0 Coping,
Hnteracting 0 Presenting4 are less strongly related( This suggests that :,P
ratings are primarily in+luenced by tas- per+ormance competencies(
Ho"ever, the data do sho" considerable di++erences in patterns bet"een
studies, "ith the standard deviation o+ the beta "eights across studies
averaging about %(;( This suggests that the competency +actors that
in+luence .udgments o+ :,P do vary +rom situation to situation(
In+ortunately, there is insu++icient in+ormation available about the
individual studies to establish "hether these variations are meaning+ul in
terms o+ di++erences in .ob content or organi5ational culture(
The model underlying the Great 'ight competencies /#ur5 0 Bartram,
8%%84 is consistent "ith Spencer and SpencerCs /11<4 causal +lo" model,
"hich postulates that personal characteristics predict :,P achievement
through competencies( The regression "eights sho"n in Table 9 "ere
used to construct a single"eighted aggregate competency criterion +or
each study( The composite competency criterion variables "ere then
regressed on the eight predictor competencies /"eighted composites o+
personality and ability +or eight o+ the studies and personality only +or
t"o4( The average multiple correlation across the % studies "as (<;
/un"eighted4 or (8$ /"eighted by sample si5e> Table $4( !s a cross-
validation, the sample-"eighted average beta "eights +rom Table 9 "ere
used to create composite criterion scores +or 3 additional studies /N
33%4, "hen there "ere data on all eight o+ the criterion competencies( The
same process "as then used to regress these composite criterion variables
on the eight predictors( Bor these 3 studies, mean 0 (<; /both
un"eighted and "eighted by sample si5e4( !cross all 3 studies, the
sample-"eighted mean 0 (82, and the average un"eighted 0 (<;(
Given that the impact o+ correcting +or arti+acts /range-restriction and
criterion reliability4 +or these data is to increase obtained coe++icients by
Table 9
a
Total mean %D %(< %(;(
b
Total sample si5e ,293(
Study b
Competency potential
predictors 8 < 3 ; 2 % < 3
( %
D
LeadingPDeciding %(%8 %(% %(%3 %( %(8; %(%2 %(% %(%; %(< %(8$ %(%$
%(
<
SupportingPCooperating %(%8 %(89 %(1 %(8% %(%% %(%% %(%9 %(%% %(; %(%; %(%9 %(
%
HnteractingPPresenting %(%; %(8% %(%; %(%2 %(% %(%; %(< %(%1 %(% %(1 %(%8 %(
8
!naly5ingPHnterpreting %(%8 %(<< %(<1 %(82 %(%3 %(% %(8; %(%$ %(%; %($ %(%3 %(8

CreatingPConceptuali5ing %(%$ %(<9 %(<8 %(81 %(<9 %(%% %(83 %(9 %(%< %(<< %(%; %(8
$
:rgani5ingP'@ecuting %( %(3 %(8 %(1 %(1 %(%3 %(%8 %(%9 %(2 %(% %(%2 %(
9
!daptingPCoping %(%< %(<8 %(%2 %(%< %(81 %(%9 %(%2 %(%8 %(% %(3 %(%8 %(
<
'nterprisingPPer+ormi
ng
%(8 %(%9 %(%; %(%8 %(88 %(%8 %(%8 %(%8 %(%< %(%% %(%8 %(
%1
0
%(2 %(;$ %(<% %(<3 %(;; %( %(<; %(83 %(81 %(;$
N 3 <; 1 <8 9; <$2 29 ;%< <98 12
,eta 3eights and 0 2al&es for 0egression of /.+ on Line-(anager Cometen!y
0atings
Study no(
a
Line-manager competency ratings 8 < 3 ; 2 % < 3 ( %D
LeadingPDeciding %(< %(3; %(9 %(<$ %(% %( %(<% %(88 %(%8 %($ %(9 %(; SupportingPCooperating %(%1 %( %(%% %(% %(82 %(% %(<< %( %(%8 %(<8
%(%9 %(2 HnteractingPPresenting %(%8 %(9 %(%; %(%% %( %(8% %( %(%$ %(% %(38 %(%3 %($ !naly5ingPHnterpreting %(8 %(; %(83 %(%1 %($ %(2 %(<8 %(81
%(;% %(%% %(89 %(3 CreatingPConceptuali5ing %(8% %(%; %(%1 %(1 %(8 %( %(%9 %(1 %(8 %(%2 %(8 %( :rgani5ingP'@ecuting %(<9 %(33 %(3% %(8; %(<% %(2
%(% %(88 %(%9 %(% %(1 %(; !daptingPCoping %(%2 %(3% %(8< %(% %(8 %(; %(% %(%1 %(% %(%< %(%< %(2 'nterprisingPPer+orming %(%< %( %(%%
%(% %(1 %( %(8; %(; %(3< %(81 %(2 %(3
0 %(9< %(2; %($8 %(9< %(9$ %($$ %($2 %(23 %(1< %(92
N
b
3 <; 1 <8 9; <$2 29 ;%< <98 12
Note" 6ean beta values are sample "eighted( !ll correlations are uncorrected +or e++ects o+ arti+acts( :,P overall .ob per+ormance(
bet"een 3$N and 9%N /average increase ;N) see Tables <A;4, one
can estimate that operational validities +or aggregated predictor and
aggregated criterion measures "ould be in the range o+ %(38 to %(;<(
Binally, the predictor composite scores "ere correlated "ith the :,P
ratings /Table 24( The average /uncorrected4 correlation "as (8$
/un"eighted4 or (88 /"eighted by sample si5e4(
The pattern o+ results +or the prediction o+ the composite competency
criterion +rom personality and ability-based competency predictors /see
Table $4 is "hat one "ould e@pect +rom previous research) The strongest
predictor is :rgani5ing 0 '@ecuting /"hich is related to Big Bive
Conscientiousness in the predictor domain4( The other main predictors are
Leading 0 Deciding, Creating 0 Conceptuali5ing, !naly5ing 0
=eporting, and Hnteracting 0 Presenting, "hich are most strongly related
to general mental ability and '@traversion in the predictor domain( Beta
"eights +or Supporting 0 Cooperating are negative or near 5ero(
Table $
(eta-1nalysis of 0elationshis ,etween
Com)ined +ersonality and 1)ility
+redi!tors and /.+
6eta-analysis o+ the eight competency potential scores as direct predictors
o+ :,P "as carried out +or the % studies e@amined previously( Bor this
analysis, the reliability o+ the :,P criterion "as set at %($% +or the seven
studies using the si@-item :,P rating scale and at %(;8 /as recommended
by Hermelin 0 =obertson, 8%%4 +or the remaining three studies, "hich
used single-item ratings o+ :,P( &o correction +or range restriction "as
carried out( The results /Table 14 sho" a similar pattern to that reported
previously) Leading 0 Deciding and :rgani5ing 0 '@ecuting have the
strongest average relationships but have lo" generali5ability( !naly5ing 0
Hnterpreting, in contrast, although having a lo"er level o+ average
correlation, sho"s high generali5ability /"ith arti+act corrections actually
overcorrecting +or bet"een-study variance4(
,eta 3eights and 0 2al&es for 0egression of 1ggregated Great Eight Line-(anager Cometen!y 0atings on Great
Eight Cometen!y +otential +redi!tors
a
Note" !ll correlations are uncorrected +or e++ects o+ arti+acts( 6ean beta values are sample "eighted(
a
Personality and !bility composites, e@cept +or Study 2 and Study 3, "hich had personality data only(
b
In"eighted total mean %(<;> sample-"eighted total mean %(8$(
Table 2
7ariable &o( studies
N r

%D
Nvar %N CrH
LeadingPDeciding % 293 %(% %(3
%(3
<$(%N %(%3
SupportingPCooperating % 293 %( %(9 %(< <$(;%N %(<<
HnteractingPPresenting % 293 %(%; %(%$ %(8 3<(9N %(%1
!naly5ingPHnterpreting % 293 %(%9 %(%1 %(%% %%(%%N %(%1
CreatingPConceptuali5ing % 293 %(%9 %(%2 %(%1 ;9(1%N %(%3
:rgani5ingP'@ecuting % 293 %(%1 %(8 %(%1 ;1(8%N %(%
!daptingPCoping % 293 %(%% %(% %(%% %%(%%N %(%
'nterprisingPPer+ormi
ng
% 293 %(%9 %(%1 %(< <2(1;N %(%2
!ggPred !ggCrit !ggPred
Study "P!ggCrit "P:,P "P:,P
N
%(2 %(9< %(3 3
8 %(;$ %(2; %(<; <;
< %(<% %($8 %(81 1
3 %(<3 %(9< %(2 <8
; %(;; %(9$ %(<9 9;
2 %( %($$ %(9 <$2
%(<; %($2 %(8 29
8 %(83 %(23 %(9 ;%<
3 %(81 %(1< %(89 <98
; %(;$ %(92 %(;1 12
Sample "eighted mean %(8$ %($1 %(88
In"eighted mean %(<; %($; %(8$
Note" !ll correlations are uncorrected +or e++ects o+ arti+acts( Total sam-
ple ,293(
!s noted in the earlier analyses, there is a negative correlation bet"een
:,P and Supporting 0 Cooperating /"hich is mainly Big Bive
!greeableness in the predictor domain4, suggesting that people "ho are
high on Big Bive !greeableness are .udged less +avorably on their :,P
than those "ho are lo"er on this attribute( :ther studies have noted small
negative correlations bet"een agreeableness and .ob per+ormance /e(g(,
Hunthausen, Tru@illo, Bauer, 0 Hammer, 8%%<4, but the reported e++ects
are usually less than %((
%&mmary
The results support the validity o+ point-to-point relationships bet"een
Great 'ight competencies and their predictors( The obtained correlations
"ere consistently higher than those bet"een unmatched pairs o+ predictors
and criteria( !lthough personalitybased predictors sho"ed moderate to
high validities +or all o+ the Great 'ight, ability tests only added to the
prediction o+ criteria +or +our o+ the eight competencies( !s hypothesi5ed,
ability is most
Table 1
strongly predictive o+ competencies in the areas o+ !naly5ing 0
Hnterpreting and Creating 0 Conceptuali5ing( The correlation bet"een
aggregated multiple predictors and aggregated multiple criteria "as, as
one "ould e@pect, substantially higher than the relationships bet"een the
predictors and :,P "ould suggest(
Discussion
The current results sho" that "hen there is a strong rationale de+ining the
predictors and the criterion, and "hen these can then be matched on a one-
to-one basis /rather than the traditional many predictors to one or many to
+e" criteria4, a clear pattern o+ results is +ound, "hich is consistent "ith
the hypotheses presented early in this article( The results con+irm the
hypothesi5ed Great 'ight pair"ise predictorAcriterion competency
relationships( &ot only are the relationships bet"een matched predictorA
competency pairs substantially higher than those bet"een unmatched
pairs, but it "as also sho"n that personality and ability together and in
isolation predict competency per+ormance ratings in a meaning+ul manner(
Speci+ically, ability tests predict +our o+ the Great 'ight> the strongest
relationship bet"een ability and competencies occurred +or the !naly5ing
0 Hnterpreting competencies( Personality provides a +ar broader coverage
o+ the competency domain than ability, but ability data add to the level o+
prediction one obtains +rom personality measures on their o"n in those
areas "here ability is relevant(
The results sho" that "hen Big Bive measures are used as predictors, they
also provide good coverage o+ the Great 'ight criterion competency
model( Ho"ever, the evidence suggests that applying the Great 'ight
competency model to the predictor domain provides a clearer and stronger
pattern o+ relationships than using a mi@ed model "ith the Big Bive as
predictors and the Great 'ight as criteria( 6apping the predictor domain to
the Great 'ight de+initions rather than the Big Bive accounts +or more o+
the criterion variance and also provides a stronger practitioner +ocus by
concentrating on "hat is being predicted rather than "hat is doing the
prediction(
The main advantage o+ the Great 'ight model is that it provides
/a4 a +rame"or- +or integrating measures in the predictor domain, such as
ability, personality, and motivation scales, and /b4 a clear set o+ a priori
hypotheses regarding the e@pected eight one-to-one
Correlations 5/erational 2alidities6 ,etween Com)ined +ersonality-and
1)ility-,ased +redi!tions of Cometen!ies and 'ndeendent 0atings of
/verall .o) +erforman!e
Correlations 1mong Great Eight 1ggregated
Cometen!y +otential +redi!tors 51gg+red6: Great
Eight 1ggregated Line (anager90ated
Cometen!ies 51ggCrit6: and /verall .o)
+erforman!e 5/.+6
Note" r sample-"eighted average o+ uncorrected correlations> sample-"eighted average o+
correlations corrected +or arti+acts> %D standard deviation o+ corrected correlations> %N CrH lo"er
%N boundary o+ 2%N Credibility interval> Nvar percentage variance accounted +or by arti+act corrections
+or corrected correlation distribution(
predictor
A
criterion relationships( The contribution o+ the Great 'ight model +or
understanding o+ .ob per+ormance is clear( 'ach o+ the eight predictors "as
sho"n to predict a di++erent area o+ .ob per+ormance consistently across
.obs, measurement instruments, and cultural conte@ts(
The correlation o+ these predictors "ith :,P "as lo"er than one "ould
e@pect +rom the combination o+ the eight pair"ise predictorAcriterion
relationships) The average uncorrected correlation across studies bet"een
aggregated predictor competencies and :,P "as bet"een (88 /"eighted4
and (8; /un"eighted4, "hereas the average uncorrected correlation
bet"een aggregated predictor and criterion competencies "as bet"een (8$
/"eighted4 and (<; /un"eighted4( The average uncorrected correlation be-
t"een aggregated criterion competency ratings and :,P, ho"ever, "as
bet"een ($; /"eighted4 and ($1 /un"eighted4( These results are consistent
"ith the causal +lo" model /Spencer 0 Spencer, 11<4) Personality and
ability act to predict the related behaviors as rated by line managers as
competencies, and these ratings o+ competencies are, in turn, related to
:,P ratings( Considering .ob per+ormance in an undi++erentiated manner
/as ratings o+ :,P do4 hides the pattern o+ relationships bet"een predictors
and more speci+ic competency +actors( This suggests that the current meta-
analysis literature may be underestimating the capacity +or personality
measures and ability test data to add value to assessment procedures by
enhancing their overall predictive po"er and providing more detailed
diagnostic in+ormation on per+ormance(
The levels o+ correlation obtained in the current meta-analysis are ?uite
high +or personality-based predictors in comparison "ith other meta-
analysis studies /and clearly substantive in practical terms4( These results
have been obtained using personality instruments that adopt a clear "or--
related +rame o+ re+erence /B:=4( !s Hunthausen et al( /8%%<4 have
sho"n, instruments that have a "orld-o+-"or- B:= do yield higher
validities than those that are more general( :nes and 7is"esvaran /8%%4
have revie"ed the use o+ Dcriterion-+ocused occupational personality
scales /C:PS4G in selection and have also noted the higher validities
associated "ith scales that directly address issues o+ relevance in the
"or-place compared "ith more general personality assessment
instruments( Ht should be noted that the :PJ instruments "ere developed
as "or--related measures o+ personality, and the item content and scales
"ere developed through "or-ing "ith people in industry( Ht "as as a
conse?uence o+ this that di++erent instruments "ere developed to cover
general graduate and managerial use /:PJ4, customer service roles
/CCSJ4, and blue-collar .obs /ESJ4( Be-cause the development process
"as centered on covering all aspects o+ personality that are considered to
be o+ relevance in the "or-place, the :PJ inventories cover a "ider range
o+ personal attributes than instruments developed +rom a personality
theory +ocus, such as the +ive-+actor model( Hn particular, aspects o+
motivation are covered( This greater breadth becomes important "hen one
is attempting to cover the +ull range o+ personal attributes assessed by line
managers in their competency ratings(
The results +or :,P sho" that this is not predicted e?ually by all eight o+
the personality-based predictors, by ability, or by personalityAability
combinations( :,P is predicted mainly by :rgani5ing 0 '@ecuting,
Leading 0 Deciding, and !naly5ing 0 Hnterpreting, "ith a negative
association "ith Supporting 0 Cooperating competencies( This may have
more to say about "hat +actors drive managersC general ratings o+ .ob
per+ormance than
anything else( Ht suggests a pattern "hereby managers +avor people "ho
are dependable, high achieving, and +ocused on the tas- rather than those
"ho display the prosocial behaviors o+ helping and supporting others(
Burther "or- is needed to determine "hether actual productivity or other
outcome measures are more strongly related to aggregated multiple
criteria or to single :,P ratings( Bor the current data, at least, the
relationship bet"een the eight predictors and eight criteria sho"s ho"
much stronger validities can be obtained by aggregation o+ multiple
criteria than by the use o+ single overall rating measures(
Conte7t&al and Tas- +erforman!e
The analyses o+ the Great 'ight predictors and criteria /see Table 14
indicate a more general level o+ description that ties this "or- in closely
"ith the literature on conte@tual versus tas--related per+ormance
constructs, revie"ed earlier here( Bor the criterion measures, the +irst
principal component is loaded by competencies that are closely tied to tas-
per+ormance and that are best predicted by motivation, general ability,
conscientiousness, and openness to ne" e@perience( The second principal
component is related to competencies associated "ith supporting and
cooperating "ith others and coping and adapting to change( The
distinction bet"een these is very similar to that described by =( Hogan
/12<4 and ,( Hogan and Holland /8%%<4 as dynamism, or Dgetting ahead,G
and social propriety, or Dgetting along(G
The personality-based Great 'ight predictors provide some more
di++erentiation, in that the Dgetting aheadG competencies are divided into
t"o principal components in terms o+ potential( The +irst re+lects
motivation, e@traversion, and openness to ne" e@perience, "hereas the
second represents the con+iguration o+ variables that one sees emerging as
consistent predictors o+ per+ormance across a "ide range o+ studies)
conscientiousness and related aspects o+ thin-ing styles /e(g(, Barric- 0
6ount, 11> Salgado, 11$, 112> Schmidt 0 Hunter, 1124(
!pproaches to the criterion domain re+lect the same issues that "e have
seen in the personality +ield in terms o+ ho" di++erentiated the domain
needs to be( Le@ical analysis studies and research on sel+-report
personality measures have provided evidence +or a number o+ "ays o+
loo-ing at the domain space( !t the most general level, there is single
personality +actor, "hich di++erentiates bet"een desirable and undesirable
attributes /Boies, Lee, !shton, Pascal, 0 &icol, 8%%> Goldberg 0 Somer,
8%%%> Saucier, 11$4( This is analogous in the criterion domain to ratings
o+ :,P, "hich typically re+lect broad evaluations o+ ho" good or bad a
person is( !lso "ell established is the t"o-+actor solution /Boies et al(,
8%%> Digman, 11$> Paulhus 0 ,ohn, 112> Saucier, 11$4, in "hich the
+irst +actor relates to positive dynamic attributes and individual
ascendancy and the second to social propriety and community cohesion(
Ba-an /1994 described these +actors as agency and communion( They
also clearly relate to tas- and conte@tual per+ormance +actors in the
criterion domain(
The Big Bive level o+ analysis is the one +ocused on here, in that the Great
'ight competencies are designed to represent a level o+ generality
comparable to that represented by the Big Bive in the personality domain(
Bactor analyses o+ the :PJ produces +ive+actor solutions that map onto
the Big Bive( Si@-and seven-+actor solutions can also be +ound that provide
a di++erentiation similar to that o+ the HPH and that di++erentiate
achievement +rom depend
ability /these tend to be combined as Conscientiousness in most Big Bive
models4 and that di++erentiates sociability +rom impul
sivity /o+ten combined as '@traversion4( Ho"ever, the argument as to
"hether there are +ive, si@, or seven +actors is primarily one about the
scope o+ the domain rather than the level o+ aggregation( Ehichever
solution one adopts, it is then possible to disaggregate these +actors into
+acets or more speci+ic subscales /.ust as the :PJ<8 resolves them into <8
scales, and the &':-PH-= into <%-+acet scales4( Ho"ever, the main thrust
o+ the current argument is that it is more use+ul to map predictor
instruments onto criterion models +or the purposes o+ validation rather
than to map the criterion models onto predictor models( To do this, the
level o+ aggregation o+ the predictor scales should match that o+ the
criterion(
H+ su++icient data are available, it "ould be better to operate at a more
detailed level o+ description /e(g(, in terms o+ the 8% competencies
presented in the !ppendi@4 "ith comparable, more speci+ic composites o+
personality and ability tests as predictors( This "as not possible +or the
current research, because insu++icient in+ormation "as available on the
criterion competencies to allo" mapping to a more detailed level than the
Great 'ight(
Bor the current research, the ability test data that "ere available /verbal
reasoning or numerical reasoning or both4 "ere treated as providing an
estimate o+ general mental ability( !t the Great 'ight level o+ description,
it is probably not appropriate to consider the di++erential impact o+ speci+ic
abilities( '@amination o+ the component level o+ !naly5ing 0 Hnterpreting
/see !ppendi@4 sho"s, ho"ever, that one "ould e@pect di++erent patterns
o+ validity +or more speci+ic ability tests at more detailed levels o+
description "ithin the +rame"or-( Bor e@ample, Component 3(8($ relates
to Ddemonstrating spatial a"arenessG and Component 3(8(; to Ddem-
onstrating physical and manual s-ills(G :ne "ould e@pect to +ind tests o+
spatial ability having higher validities than other ability tests +or .obs in
"hich this component "as relevant( Ee "ould also e@pect tests o+ creative
thin-ing to increase levels o+ prediction +or Creating 0 Conceptuali5ing,
"hereas tests o+ oral comprehension and e@pression should relate to
Hnteracting 0 Presenting( Bor the current set o+ studies, ho"ever, the
emphasis on service and managerial positions and the +ocus on the general
Great 'ight level o+ aggregation entails that general mental ability, as
assessed by verbal and numerical reasoning, is li-ely to account +or most
o+ the ability-related variance( Buture research needs to consider "hether
the levels o+ prediction +ound here "ould be increased by the use o+ more
di++erentiated competency models and the use o+ a "ider range o+ more
speci+ic ability tests(
Limitations
The studies reported here "ere collected +rom corporate archives( Ht "as
not possible to e@ercise control, in retrospect, over data collection,
supervisor rating procedures, and other design +actors( Hn many cases,
study design details "ere not available( The total sample o+ studies,
ho"ever, included +ive I(#( samples /Studies A; in Table <4 "here there
"as good independent control over the design o+ the studies and the data
collection( These supervised studies /previously reported in &y+ield et al(,
11;4 provide a benchmar- against "hich the ?uality o+ the data +rom the
other studies can be evaluated( =emoval o+ these supervised studies +rom
the data set generally resulted in a lo"ering o+ average
validities +or the remaining studies( Ho"ever, the overall pattern o+ results
"as not a++ected(
:n the issue o+ statistical po"er, 6urphy /11$4 suggested that caution
should be used in interpreting tests o+ situational speci+icity "hen the
average sample si5e is less than %% to 8%% or "hen the number o+ studies
/-6 is less than ; to 8%( The current data set has a median sample si5e o+
8;, "ith - generally in the 8% range( :n that basis, this collection o+
studies seems to have su++icient po"er to detect situational speci+icities(
The samples e@amined here are predominantly +rom management or are
graduates in technical or sales positions or management training( !lthough
some o+ the data came +rom blue-collar positions, their in+luence on the
overall results "ill be ?uite small( Ht is ?uite possible that the pattern o+
these results "ill vary +or di++erent types and levels o+ .ob( Hn particular,
"e might e@pect to +ind greater emphasis on the competency area most
closely related to .ob -no"ledge and s-ills /!naly5ing 0 Hnterpreting4 in
lo"er level .obs( The language used to de+ine the competencies "ithin the
current +rame"or- +its most easily "ith the "ay these behaviors are
described +or management positions( Ho"ever, the +rame"or- is intended
to be generic( Buture research needs to e@plore the degree to "hich it is
possible to produce better operational de+initions o+ the competencies in
the +rame"or- +or di++erent .ob levels(
Binally, all the data presented here have used predictor instruments +rom
one publisher /although they have included a variety o+ instruments "ith a
variety o+ response +ormats4 "ith competency measurement tools that
"ere either +rom the same source or +rom client-constructed measures o+
un-no"n psychometric ?uality( !lthough all the personality instruments
and much o+ the competency ratings data "ere collected using instruments
+rom a single publisher, there is no direct content or construct overlap
bet"een the models underlying the personality instruments and the
competency instruments( Hndeed, one reason +or developing a generic
competency +rame"or- "as to be able to map bet"een competency
models li-e H6C, CCCH, and P6C, "hich are other"ise very di++erent in
structure and content as "ell as in item type and +ormat(
Buture "or- on the Great 'ight "ould bene+it +rom research that mapped
other predictor instruments and other criterion measurement procedures
onto this structure to test its generali5ability and robustness(
Conclusions
Perhaps "e have been preoccupied +or too long "ith the "onder+ul
personality ?uestionnaires and ability tests "e have constructed to
measure all sorts o+ aspects o+ human potential( Hn so doing, "e may have
lost sight o+ "hy it is important to be able to measure these characteristics(
!s a conse?uence, practitioners have o+ten had di++iculty e@plaining to
their clients the value o+ "hat "e have to o++er( Ee need to reali5e that this
inability may be due in no small part to our +ailure to address the issues
that actually concern clients) per+ormance at "or- and the outcomes o+
that per+ormance( The Big Bive and other classi+ications o+ personality
+actors, +or e@ample, are classi+ications +ocused on the predictor domain(
Hn this article, "e have endeavored to sho" the value o+ changing this
+ocus to the criterion domain "hile still providing the same level o+
di++erentiation "ithin that domain(
!lthough the Great 'ight provides an analogue in that domain to
D
g,
G
motivation measures, and the Big Bive o+ the predictor do
main, it has the advantage o+ addressing directly the issues that are o+
prime practical importance in selection testing) "hat it is that is being
predicted(
By di++erentiating per+ormance in the criterion domain in this "ay, "e can
better articulate the value o+ "hat "e can provide as predictors o+ "or--
related behaviors +rom a practice point o+ vie" and better understand "hy
particular patterns o+ predictorA criterion relationships occur( To +acilitate
this, practitioners need to encourage their clients to adopt more
di++erentiated appraisal tools( This "ould be not only o+ scienti+ic value in
improving the ?uality o+ validity studies but also o+ value to the client in
providing more reliable and more valid measures o+ peopleCs per+ormance(
By comparing measures o+ actual per+ormance on the Great 'ight or
similar criterion classi+ication models "ith mea-sures o+ potential on the
same constructs /using personality and ability tests as predictors4, clients
"ould be better able to identi+y those areas in "hich people "ould bene+it
most +rom learning opportunities and developmental e@periences(
=e+erences
Ba-an, D( /1994( The d&ality of h&man e7isten!e: 'solation and
!omm&nion in 3estern man" Boston) Beacon Boo-s(
Barric-, 6( =(, 0 6ount, 6( #( /114( The Big Bive personality dimen-
sions and .ob per+ormance) ! meta-analysis( +ersonnel +sy!hology:
;;: A89(
Bartram, D( /8%%4( +redi!ting !ometen!y dimensions from
!omonents: 1 validation of the two-ste ro!ess" Thames
Ditton, I#) SHL Group(
Bartram, D(, =obertson, H( T(, 0 Callinan, 6( /8%%84( Hntroduction) !
+rame"or- +or e@amining organi5ational e++ectiveness( Hn H( T( =obertson,
6( Callinan, 0 D( Bartram /'ds(4, /rgani*ational effe!tiveness: The
role of sy!hology /pp( A%4( Chichester, I#) Eiley(
Boies, #(, Lee, #(, !shton, 6( C(, Pascal, S(, 0 &icol, !( !( 6( /8%%4(
The structure o+ the Brench personality le@icon( E&roean .o&rnal of
+ersonality: #<: 8$$A81;(
Borman, E( C(, Buc-, D( '(, Hanson, 6( !(, 6oto"idlo, S( ,(, Star-, S(, 0
Drasgo", B( /8%%4( !n e@amination o+ the comparative reliability,
validity, and accuracy o+ per+ormance ratings made using computeri5ed
adaptive rating scales( .o&rnal of 1lied +sy!hology: =>: 19;A1$<(
Borman, E( C(, 0 6oto"idlo, D( ,( /11<4( '@panding the criterion
domain to include elements o+ conte@tual per+ormance( Hn &( Schmitt 0
E( C( Borman /'ds(4, +ersonnel sele!tion in organi*ations /pp(
$A124( San Brancisco) ,ossey-Bass(
Borman, E( C(, Penner, L( !(, !llen, T( D(, 0 6oto"idlo, S( ,( /8%%4(
Personality predictors o+ citi5enship per+ormance( 'nternational
.o&rnal of %ele!tion and 1ssessment: ?: ;8A91(
Campbell, ,( P( /11%4( 6odeling the per+ormance prediction problem in
industrial and organi5ational psychology( Hn 6( D( Dunnette 0 L( 6(
Hough /'ds(4, Hand)oo- of ind&strial and organi*ational
sy!hology /8nd ed(, 7ol( , pp( 92$A$<84( Palo !lto) Consulting
Psychologists Press(
Campbell, ,( P(, 6cCloy, =( !(, :ppler, S( H(, 0 Sager, C( '( /11<4( !
theory o+ per+ormance( Hn &( Schmitt 0 E( C( Borman /'ds(4, +ersonnel
sele!tion in organi*ations /pp( <;A$%4( San Brancisco) ,ossey-Bass(
Campbell, ,( P(, 6cHenry, ,( ,(, 0 Eise, L( L( /11%4( 6odeling .ob
per+ormance in a population o+ .obs( +ersonnel +sy!hology: ;@: <<A
<<<(
Con"ay, ,( 6( /1194( !dditional construct validity evidence +or the
tas-Pconte@tual per+ormance distinction( H&man +erforman!e: ?:
<%1A <81(
Digman, ,( 6( /11$4( Higher order +actors o+ the Big Bive( .o&rnal of
+ersonality and %o!ial +sy!hology: A@: 839A8;9(
Goldberg, L( =(, 0 Somer, :( /8%%%4( The hierarchical structure o+ com-
mon Tur-ish person-descriptive ad.ectives( E&roean .o&rnal of +er-
sonality: #;: 31$A;<(
Gotoh, !( /1114( The eval&ation of !ometen!ies redi!tors
a!ross nine st&dies in five !o&ntries" Inpublished masterCs thesis,
Goldsmith College, London(
Hermelin, '(, 0 =obertson, H( T( /8%%4( ! criti?ue and standardi5ation o+
meta-analytic validity coe++icients in personnel selection( .o&rnal of
/!!&ational and /rgani*ational +sy!hology: A;: 8;<A8$$(
Hogan, ,(, 0 Holland, B( /8%%<4( Ising theory to evaluate personality and
.ob-per+ormance relations) ! socioanalytic perspective( .o&rnal of 1-
lied +sy!hology: ==: %%A8(
Hogan, =( /12<4( ! socioanalytic theory o+ personality( Hn 6( 6( Page
/'d(4, Ne)ras-a %ymosi&m on (otivation /pp( <<9A<;;4( Lincoln)
Iniversity o+ &ebras-a Press(
Hough, L( 6( /1184( The Big Bive personality variablesQconstruct
con+usion) Description versus prediction( H&man +erforman!e: <:
<1A ;;(
Hough, L( 6(, :nes, D( S(, 0 7is"esvaran, C( /112, !pril4(
+ersonality !orrelates of managerial erforman!e !onstr&!ts"
Paper presented at the annual con+erence o+ the Society +or Hndustrial and
:rgani5ational Psychology, Dallas(
Hunter, ,( '(, 0 Hunter, =( B( /1234( 7alidity and utility o+ alternative
predictors o+ .ob per+ormance( +sy!hologi!al ,&lletin: ?>: $8A12(
Hunter, ,( '(, 0 Schmidt, B( L( /11%4( (ethods of meta-analysis"
&e"bury Par-, C!) Sage(
Hunthausen, ,( 6(, Tru@illo, D( 6(, Bauer, T( &(, 0 Hammer, L( B( /8%%<4(
! +ield study o+ +rame-o+-re+erence e++ects in personality test validity(
.o&rnal of 1lied +sy!hology: ==: ;3;A;;(
Hurt5, G( 6(, 0 Donovan, ,( ,( /8%%%4( Personality and .ob per+ormance)
The Big Bive revisited( .o&rnal of 1lied +sy!hology: =<: 291A
2$1( ,ohnson, ,( E( /8%%4( The relative importance o+ tas- and conte@tual
per+ormance dimensions to supervisor .udgements o+ overall per+ormance(
.o&rnal of 1lied +sy!hology: =>: 123A119(
,udge, T( !(, Bono, ,( '(, Hlies, =(, 0 Gerhardt, 6( E( /8%%84( Personality
and leadership) ! ?ualitative and ?uantitative revie"( .o&rnal of
1lied +sy!hology: =A: $9;A$2%(
#ur5, =( /1114( !utomated prediction o+ managerial competencies +rom
personality and ability variables( Hn +ro!eedings of the ,+% Test
4ser Conferen!e /pp( 19A%4( Leicester, I#) British Psychological
Society(
#ur5, =(, 0 Bartram, D( /8%%84( Competency and individual per+ormance)
6odeling the "orld o+ "or-( Hn H( T( =obertson, 6( Callinan, 0 D(
Bartram /'ds(4, /rgani*ational effe!tiveness: The role of
sy!hology /pp( 88$A8;;4( Chichester) Eiley(
#ur5, =(, Bartram, D(, 0 Baron, H( /8%%34( !ssessing potential and
per+ormance at "or-) The Great 'ight competencies( Hn +ro!eedings of
the ,ritish +sy!hologi!al %o!iety /!!&ational Conferen!e /pp(
1A1;4( Leicester, I#) British Psychological Society(
6oto"idlo, S( ,(, Borman, E( C(, 0 Schmit, 6( ,( /11$4( ! theory o+
individual di++erences in tas- and conte@tual per+ormance( H&man +er-
forman!e: #B: $A2<(
6urphy, #( /11$4( 6eta-analysis and validity generali5ation( Hn &(
!nderson 0 P( Herriot /'ds(4, 'nternational hand)oo- of sele!tion
and assessment /pp( <8<A<384( Chichester, I#) Eiley(
&y+ield, G(, Gibbons, P( ,(, Baron, H(, 0 =obertson, H( /11;, 6ay4( The
!ross !&lt&ral validity of management assessment methods"
Paper presented at the %th !nnual Con+erence o+ the Society +or
Hndustrial and :rgani5ational Psychology, :rlando(
:nes, D( S(, 0 7is"esveran, C( /8%%4( Hntegrity tests and other criterion-
+ocused occupational personality scales /C:PS4 used in personnel se-
lection( 'nternational .o&rnal of %ele!tion and 1ssessment: ?:
<A<1(
Paulhus, D( L(, 0 ,ohn, :( P( /1124( 'goistic and moralistic biases in
sel+-perception) The interplay o+ sel+-descriptive styles "ith basic traits
and motives( ,ourna
l
o
f
Personality
:
99
:
%8;
A
%9%
(
Leading and Deciding 8(8 !dhering to Principles and 7alues
( Deciding 0 Hnitiating !ction 8(8( Ipholding 'thics and 7alues
(( 6a-ing Decisions 8(8(8 !cting "ith Hntegrity
((8 Ta-ing =esponsibility 8(8(< Itili5ing Diversity
((< !cting "ith Con+idence 8(8(3 Sho"ing Social and 'nvironmental =esponsibility
((3 !cting on :"n Hnitiative < Hnteracting and Presenting
((; Ta-ing !ction <( =elating 0 &et"or-ing
((9 Ta-ing Calculated =is-s <(( Building =apport
(8 Leading and Supervising <((8 &et"or-ing
(8( Providing Direction and Coordinating !ction <((< =elating !cross Levels
(8(8 Supervising and 6onitoring Behavior <((3 6anaging Con+lict
(8(< Coaching <((; Ising Humor
(8(3 Delegating <(8 Persuading and Hn+luencing
(8(; 'mpo"ering Sta++ <(8( 6a-ing an Hmpact
(8(9 6otivating :thers <(8(8 Shaping Conversations
(8($ Developing Sta++ <(8(< !ppealing to 'motions
(8(2 Hdenti+ying and =ecruiting Talent <(8(3 Promoting Hdeas
8 Supporting and Cooperating <(8(; &egotiating
8( Eor-ing "ith People <(8(9 Gaining !greement
8(( Inderstanding :thers <(8($ Dealing "ith Political Hssues
8((8 !dapting to the Team <(< Presenting and Communicating Hn+ormation
8((< Building Team Spirit <(<( Spea-ing Bluently
8((3 =ecogni5ing and =e"arding Contributions <(<(8 '@plaining Concepts and :pinions
8((; Listening <(<(< !rticulating #ey Points o+ an !rgument
8((9 Consulting :thers <(<(3 Presenting and Public Spea-ing
8(($ Communicating Proactively <(<(; Pro.ecting Credibility
8((2 Sho"ing Tolerance and Consideration <(<(9 =esponding to an !udience
8((1 Sho"ing 'mpathy 3 !naly5ing and Hnterpreting
8((% Supporting :thers 3( Eriting and =eporting
8(( Caring +or :thers 3(( Eriting Correctly
8((8 Developing and Communicating Sel+--no"ledge and Hnsight 3((8 Eriting Clearly and Bluently
=obertson, H( T(, Baron, H(, Gibbons, P( ,(, 6acHver, =(, 0 &y+ield, G(
/8%%%4( Conscientiousness and managerial per+ormance( .o&rnal of /!-
!&ational and /rgani*ational +sy!hology: A@: $A2%(
=obertson, H( T(, 0 #inder, !( /11<4( Personality and .ob competencies)
The criterion-related validity o+ some personality variables( .o&rnal of
/!!&ational and /rgani*ational +sy!hology: >>: 88;A833(
=otundo, 6(, 0 Sac-ett, P( =( /8%%84( The relative importance o+ tas-,
citi5enship, and counterproductive per+ormance to global ratings o+ .ob
per+ormance) ! policy-capturing approach( .o&rnal of 1lied
+sy!hology: =A: 99A2%(
Salgado, ,( B( /11$4( The +ive +actor model o+ personality and .ob per+or-
mance in the 'uropean Community( .o&rnal of 1lied +sy!hology:
=$: <%A3<(
Salgado, ,( B( /1124( Big Bive personality dimensions and .ob per+or-
mance in army and civil occupations) ! 'uropean perspective( H&man
+erforman!e: ##: 8$A822(
Saucier, G( /11$4( '++ects o+ variable selection on the +actor structure o+
person descriptors( .o&rnal of +ersonality and %o!ial +sy!hology:
A@: 819A<8(
Schmidt, B( L(, 0 Hunter, ,( '( /1124( The validity and utility o+ selection
methods in personnel psychology) Practical and theoretical implications o+
2; years o+ research +indings( +sy!hologi!al ,&lletin: #$;: 898A8$3(
Scullen, S( '(, 6ount, 6( #(, 0 ,udge, T( !( /8%%<4( 'vidence o+ the
construct validity o+ developmental ratings o+ managerial per+ormance(
.o&rnal of 1lied +sy!hology: ==: ;%A99(
SHL Group( /11<a4( 'nventory of (anagement Cometen!ies:
(an&al and &serCs g&ide" Thames Ditton, Inited #ingdom) !uthor(
SHL Group( /11<b4( /+D !on!et model: (an&al and &serCs
g&ide" Thames Ditton, Inited #ingdom) !uthor(
SHL Group( /1134( +erse!tives on managerial !ometen!ies:
4serCs man&al" Thames Ditton, Inited #ingdom) !uthor(
SHL Group( /11$4( C&stomer !onta!t: (an&al and &serCs g&ide"
Thames Ditton, Inited #ingdom) !uthor(
SHL Group( /111a4( /+D@$: (an&al and &serCs g&ide" Thames
Ditton, Inited #ingdom) !uthor(
SHL Group( /111b4( 3or- %tyles D&estionnaire: 2ersion n:
(an&al and &serCs g&ide" Thames Ditton, Inited #ingdom) !uthor(
Spencer, L( 6(, 0 Spencer, S( 6( /11<4( Cometen!e at wor-:
(odels for s&erior erforman!e" &e" Oor-) Eiley(
Tett, =( P(, Guterman, H( !(, Bleier, !(, 0 6urphy, P( ,( /8%%%4( Devel-
opment and content validation o+ a DhyperdimensionalG ta@onomy o+
managerial competence( H&man +erforman!e: #@: 8%;A8;(
Tett, =( P(, ,ac-son, D( &(, 0 =othstein, 6( /114( Personality measures
as predictors o+ .ob per+ormance) ! meta-analytic revie"( +ersonnel
+sy!hology: ;;: $%3A$38(
!ppendi@
Great 'ight, 8% Competency Dimension and 8 Competency Component titles +rom the SHL
Iniversal Competency Brame"or-
3((< Eriting in an '@pressive and 'ngaging Style
3((3 Targeting Communication
3(8 !pplying '@pertise and Technology
3(8( !pplying Technical '@pertise
3(8(8 Building Technical '@pertise
3(8(< Sharing '@pertise
3(8(3 Ising Technology =esources
3(8(; Demonstrating Physical and 6anual S-ills
3(8(9 Demonstrating Cross Bunctional !"areness
3(8($ Demonstrating Spatial !"areness
3(< !naly5ing
3(<( !naly5ing and 'valuating Hn+ormation
3(<(8 Testing !ssumptions and Hnvestigating
3(<(< Producing Solutions
3(<(3 6a-ing ,udgments
3(<(; Demonstrating Systems Thin-ing ;
Creating and Conceptuali5ing
;( Learning and =esearching
;(( Learning Juic-ly
;((8 Gathering Hn+ormation
;((< Thin-ing Juic-ly
;((3 'ncouraging and Supporting :rgani5ational Learning
;((; 6anaging #no"ledge
;(8 Creating and Hnnovating
;(8( Hnnovating
;(8(8 See-ing and Hntroducing Change
;(< Bormulating Strategies and Concepts
;(<( Thin-ing Broadly
;(<(8 !pproaching Eor- Strategically
;(<(< Setting and Developing Strategy
;(<(3 7isioning 9
:rgani5ing and '@ecuting
9( Planning and :rgani5ing
9(( Setting :b.ectives
9((8 Planning
9((< 6anaging Time
9((3 6anaging =esources
9((; 6onitoring Progress
9(8 Delivering =esults and 6eeting Customer '@pectations
9(8( Bocusing on Customer &eeds and Satis+action
9(8(8 Setting High Standards +or Juality
9(8(< 6onitoring and 6aintaining Juality
9(8(3 Eor-ing Systematically
9(8(; 6aintaining Juality Processes
9(8(9 6aintaining Productivity Levels
9(8($ Driving Pro.ects to =esults
9(< Bollo"ing Hnstructions and Procedures
9(<( Bollo"ing Directions
9(<(8 Bollo"ing Procedures
9(<(< Time #eeping and !ttending
9(<(3 Demonstrating Commitment
9(<(; Sho"ing !"areness o+ Sa+ety Hssues
9(<(9 Complying "ith Legal :bligations $
!dapting and Coping
$( !dapting and =esponding to Change
$(( !dapting
$((8 !ccepting &e" Hdeas
$((< !dapting Hnterpersonal Style
$((3 Sho"ing Cross-cultural !"areness
$((; Dealing "ith !mbiguity
$(8 Coping "ith Pressure and Setbac-s
$(8( Coping "ith Pressure
$(8(8 Sho"ing 'motional Sel+-control
$(8(< Balancing Eor- and Personal Li+e
$(8(3 6aintaining a Positive :utloo-
$(8(; Handling Criticism 2
'nterprising and Per+orming
2( !chieving Personal Eor- Goals and :b.ectives
2(( !chieving :b.ectives
2((8 Eor-ing 'nergetically and 'nthusiastically
2((< Pursuing Sel+-development
2((3 Demonstrating !mbition
2(8 'ntrepreneurial and Commercial Thin-ing
2(8( 6onitoring 6ar-ets and Competitors
2(8(8 Hdenti+ying Business :pportunities
2(8(< Demonstrating Binancial !"areness
2(8(3 Controlling Costs
2(8(; #eeping !"are o+ :rgani5ational Hssues &ote that each component is
+urther de+ined "ithin the +rame"or- in terms o+ negative and positive
behavioral indicators(
The competency titles in this !ppendi@ are ta-en +rom the SHL Iniversal
Competency Brame"or-K Pro+iler and Designer cards /copyright L 8%%3
by SHL Group plc, reproduced "ith permission o+ the copyright holder4(
These titles may be +reely used +or research purposes sub.ect to due
ac-no"ledgement o+ the copyright holder(
=eceived September <, 8%%< =evision
received &ovember , 8%%3 !ccepted
December 8%, 8%%3