Faking in Self‐Report Personality Scales: A Qualitative Analysis and Taxonomy of the Behaviors That Constitute Faking Strategies

Röhne , Jessica; Schü z, As id; Ziegle , Ma hias
A icle — Published Ve sion
Faking in Sel ‐Repo Pe sonali y Scales: A Quali a i e
Analysis and Taxonomy o he Beha io s Tha Cons i u e
Faking S a egies
In e na ional Jou nal o Selec ion and Assessmen
P o ided in Coope a ion wi h:
John Wiley & Sons
Sugges ed Ci a ion: Röhne , Jessica; Schü z, As id; Ziegle , Ma hias (2024) : Faking in Sel ‐Repo
Pe sonali y Scales: A Quali a i e Analysis and Taxonomy o he Beha io s Tha Cons i u e Faking
S a egies, In e na ional Jou nal o Selec ion and Assessmen , ISSN 1468-2389, Wiley Pe iodicals,
Inc., Hoboken, NJ, Vol. 33, Iss. 1,
h ps://doi.o g/10.1111/ijsa.12513
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/313705
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h p://c ea i ecommons.o g/licenses/by/4.0/
In e na ional Jou nal o Selec ion and Assessmen
RESEARCH ARTICLE
Faking in Sel ‐Repo Pe sonali y Scales: A Quali a i e
Analysis and Taxonomy o he Beha io s Tha Cons i u e
Faking S a egies
Jessica Röhne
1
| As id Schü z
1
| Ma hias Ziegle
2
1
Depa men o Psychology, Uni e si y o Bambe g, Bambe g, Ge many |
2
Depa men o Psychology, Humbold ‐Uni e si ä zu Be lin, Be lin, Ge many
Co espondence: Jessica Röhne ([email p o ec ed])
Recei ed: 2 Oc obe 2023 | Re ised: 14 No embe 2024 | Accep ed: 15 No embe 2024
Funding: This esea ch was pa ly unded by a g an om he Equal Oppo uni ies O ice and by a g an om he In e nal Resea ch Funding a he Uni e si y
o Bambe g.
Keywo ds: aking | p ocess o aking in SRPSs | sel ‐ epo pe sonali y scales (SRPSs) | axonomy o aking beha io s ha cons i u e aking s a egies in
SRPSs | he gene al esponse p ocess model (GRPM)
ABSTRACT
Faking in sel ‐ epo pe sonali y scales (SRPSs) is no su icien ly unde s ood. This limi s i s de ec ion and p e en ion. He e, we
in oduce a axonomy o aking beha io s ha cons i u e aking s a egies in SRPSs, e lec ing he s ages (comp ehension,
e ie al, judgmen , and esponse) o he gene al esponse p ocess model (GRPM). We eanalyzed da a om wo s udies
in es iga ing he aking o high and low sco es on Ex a e sion (E) and Need o Cogni ion (NFC) scales (Da a Se 1; N= 305) o
on an E scale (Da a Se 2; N= 251). Pa icipan s we e asked o explain exac ly wha hey did o ake, and hei esponses
(N= 533) we e examined ia a quali a i e con en analysis. The esul ing axonomy included 22 global and 13 speci ic beha io s
ha (in combina ion) cons i u e aking s a egies in SRPSs. We o ganized he beha io s in o ou clus e s along he s ages o he
GRPM. The beha io s held i espec i e o he cons uc (E o NFC), and wi h wo excep ions, also i espec i e o he da a se
(Da a Se s 1 o 2). Eigh excep ions conce ning aking di ec ion (high o low) indica e di ec ion‐speci ic di e ences in aking
beha io s. Responden s epo ed using no only di e en aking beha io s (e.g., ole‐playing, beha io s o a oid being de ec ed)
bu also mul iple combina ions he eo . The sugges ed axonomy is necessa ily limi ed o he speci ied con ex , and, hus,
addi ional aking beha io s a e possible. To ully unde s and aking, u he esea ch in o he con ex s should be conduc ed o
complemen he axonomy. S ill, he complexi y shown he e explains why adequa e de ec ion and p e en ion o aking in SRPSs
is so challenging.
Faking has been de ined as “… a esponse se aimed a p o-
iding a po ayal o he sel ha helps a pe son o achie e
pe sonal goals. Faking occu s when his esponse se is ac i-
a ed by si ua ional demands and pe son cha ac e is ics o
p oduce sys ema ic di e ences in es sco es ha a e no caused
by he a ibu e o in e es ”(Ziegle e al. 2012, p. 8). The ex en
o which esponden s ake di e s ac oss aking condi ions, such
as he aking di ec ion (e.g., aking high sco es s. aking low
sco es) and he cons uc being aked (me a‐analysis: Bi keland
e al. 2006), o ming a ce ain aking con ex . Thus, aking is a
complex phenomenon ha a ies ac oss con ex s (e.g., Bensch
e al. 2019; Röhne , Thoss, and Schü z 2022).
Subs an ial li e a u e has documen ed he nega i e e ec s o
aking (e.g., changes in mean sco es, s anda d de ia ions, ank
o de s, eliabili y, and cons uc ‐ ela ed alidi y), and e-
sea che s ag ee ha aking is a se ious p oblem in assessmen ,
ha needs o be add essed (e.g., Salgado 2016; o an o e iew,
This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s use, dis ibu ion and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly
ci ed.
© 2024 The Au ho (s). In e na ional Jou nal o Selec ion and Assessmen published by John Wiley & Sons L d.
1o 27In e na ional Jou nal o Selec ion and Assessmen , 2025; 33:e12513
h ps://doi.o g/10.1111/ijsa.12513
see Ziegle e al. 2012). Sel ‐ epo pe sonali y scales (i.e.,
SRPSs) wi h Like ‐ ype i ems a e especially p one o aking.
These measu emen p ocedu es a e used equen ly and
esponden s can easily in e hei associa ed assessmen goals,
which inc eases hei akeabili y (e.g., Fuech enhans and
B own 2022; Salgado 2016).
Se e al app oaches o he de ec ion (e.g., implemen ing scales
ha aim o measu e he endency o c ea e a desi ed imp ession:
Paulhus 2002; bu c . Lanz, Thielmann, and Ge po 2022;using
machine‐lea ning algo i hms: Calanna e al. 2020; Röhne ,
Thoss, and Schü z 2022;inspec ing eac ion imes: Holden and
Lambe 2015; bu c . Röhne and Holden 2022) and p e en ion
o aking in SRPSs (e.g., using wa nings: Lande s, Sacke , and
Tuzinski 2011; bu c . Robson, Jones, and Ab aham 2007;using
o ced‐choice o ma s: Cao and D asgow 2019; bu c . Ma ínez
and Salgado 2021;using “ ake‐ esis an ”measu es: LeB e on
e al. 2007; bu Röhne and Ewe s 2016;o using p oc o ed
es ing: Ka im, Kaminsky, and Beh end 2014; bu c . Bea y
e al. 2011) ha e been sugges ed.
Ne e heless, esea che s ha e no ye succeeded in inding
alid ways o de ec o p e en aking wi h high accu acy (e.g.,
Bill and Melche s 2022), e en hough hey ha e been a emp -
ing o do so o o e 100 yea s (Sacke e al. 2017). The e a e a
leas wo easons why hese en u es ha e ailed o succeed.
Fi s , e ec i ely de ec ing o p e en ing aking equi es a
comp ehensi e unde s anding o he mani oldness o he
beha io s in ol ed. Al hough c ucial, knowledge abou wha
exac ly esponden s do o ake emains limi ed, as we will ex-
plain below. Howe e , ailing o add ess he complexi y o hese
beha io s would h ea en e o s o de ec and p e en
hem. Second, aking depends on he con ex (e.g., Bensch
e al. 2019), and aking beha io s may di e be ween condi ions
(e.g., he aking di ec ion: aking high sco es s. aking low
sco es; Röhne , Thoss, and Schü z 2022). As a esul , me hods
ha a e e ec i e o de ec ing o p e en ing aking in one
condi ion migh ail in ano he . The e o e, a comp ehensi e
o e iew o indi idual aking beha io s ac oss condi ions in
SRPSs is essen ial.
He e, we sough o gain deepe insigh s in o he complexi y o
hese beha io s by conduc ing a quali a i e analysis o espon-
den s' epo s on he a ious aking beha io s hey s a ed o
ha e implemen ed while hey we e aking high sco es o aking
low sco es on wo SRPSs. We so ed hese beha io s along he
gene al esponse p ocess model (GRPM; see he sec ion below)
because i p o ides a use ul heo e ical amewo k o he ak-
ing p ocess in SRPSs.
1 | Posi ioning Faking in SRPSs in he GRPM
The GRPM desc ibes he gene al p ocess by which pa icipan s
espond o SRPSs i ems. On he basis o a li e a u e e iew,
K osnick (1999) sugges ed ha he GRPM comp ises ou
s ages: comp ehension, e ie al,judgmen , and mapping
(Figu e 1). Acco ding o ha model, esponden s encode an
i em and o m a men al ep esen a ion o i s con en
(comp ehension s age). They hen ecall in o ma ion ha is
ele an o he i em con en ( e ie al s age). They compa e he
in o ma ion wi h hei men al ep esen a ion o he i em
(judgmen s age). Finally, hey p ojec he espec i e esul on o
he a ing scale (mapping s age). K osnick (1999) dis inguished
be ween op imizing (i.e., when esponden s a e mo i a ed o
answe a en i ely and comply wi h he ins uc ions) and sa-
is icing (i.e., when esponden s do no answe a en i ely).
Tou angeau and Rasinski (1988) sugges ed a simila GRPM bu
assumed ha he las s age includes mo e han jus mapping
and he e o e called his s age esponse. Then hey dis inguished
be ween wo s eps in he las s age and called hese wo s eps
mapping and edi ing (Figu e 1). Whe eas mapping is in line
wi h wha K osnick (1999) desc ibed, edi ing e lec s esponse
dis o ions, such as aking in SRPSs. Acco ding o his iew,
aking would be ela ed only o he edi ing componen o his
las s age and no o any o he s age.
Howe e , Ziegle (2011) showed in hei s udy ha all ou
s ages o he GRPM a e a ec ed by aking and ha esponden s'
mo i es and pe sonali y mode a e hese aking a emp s in
SRPSs. Fo example, hey ound ha na cissism was ela ed o
aking in he las s age, bu sel ‐decep ion was ela ed o aking
in ea lie s ages. Thus, Ziegle (2011) ound ha aking is no
es ic ed o he las s age as p e iously assumed (e.g.,
Tou angeau and Rasinski 1988).
2 | Wha Do We Know Abou Faking Beha io s
in SRPSs So Fa ?
2.1 | Insigh s F om Theo e ical Models
Knowledge abou possible aking beha io s in SRPSs is pa ly
based on heo e ical models o he aking p ocess (e.g., Go in
and Boyd 2009; Te and Simone 2011). Al hough he speci ic
con en a ies ac oss aking models, a common ea u e o hese
models is ha aking includes he componen s o abili y,
mo i a ion, and oppo uni y o ake (see also Ellingson and
McFa land 2011). Ne e heless, i is s ill unclea whe he aking
beha io s in SRPSs ha ha e been sugges ed heo e ically a e
Summa y
•The ac ual beha io s unde lying aking in SRPSs a e s ill
no well unde s ood, which limi s aking de ec ion and
p e en ion.
•Conduc ing a quali a i e analysis, we in oduce a ax-
onomy o beha io s ha cons i u e aking s a egies in
SRPSs and e lec he GRPM s ages.
•Analyzing 533 esponses om pa icipan s asked o
epo on hei aking beha io s, we iden i ied 22 global
and 13 speci ic beha io s in ou clus e s.
•Highligh ing he complexi y o aking, esponden s e-
po ed ha hey no only used di e en beha io s bu
also mul iple combina ions he eo .
•The beha io s la gely held ac oss condi ions (e.g., aking
on E and NFC) wi h some di e ences ha we e p i-
ma ily associa ed wi h aking di ec ion; hus, he esul s
can be unde s ood as a map o build upon.
2o 27 In e na ional Jou nal o Selec ion and Assessmen , 2025
ac ually applied, whe he he e a e aking beha io s in SRPSs
ha a e so a no conside ed, whe he he e a e combina ions
o aking beha io s in SRPSs, o whe he aking beha io s a e
gene alizable (e.g., wi h espec o aking di ec ion, Bensch
e al. 2019; o he o‐be‐ aked cons uc , Bi keland e al. 2006).
2.2 | Indica ions F om Quan i a i e Analyses
Resea che s ha e also used quan i a i e analyses o iden i y
aking beha io s in SRPSs, ypically by compa ing esponse
pa e ns be ween aked and non aked esponses. This app oach
has o e ed se e al insigh s indica ing ha aking beha io s
may a y subs an ially. Due o space es ic ions, we only
desc ibe a ew examples he ea e . A ecen s udy e ealed ha
aking depends on aking di ec ion (e.g., Bensch e al. 2019).
O he s udies ha e ound e idence ha esponden s gi e bla an
and ex eme answe s when aking (Lande s, Sacke , and
Tuzinski 2011; Le ashina e al. 2014). Using mixed‐model i em
esponse heo y, Zicka , Gibby and Robie (2004) iden i ied
classes o esponden s who ake di e en ly (e.g., sligh aking
and ex eme aking).
Howe e , al hough compa ing esponse pa e ns be ween aked
and non aked esponses o e s aluable insigh s, he esul ing
knowledge migh be limi ed, as a ia ions in esponse pa e ns
do no necessa ily p o ide in o ma ion abou he ac ual un-
de lying beha io . S a ed di e en ly, al hough he e ec s (e.g.,
changed esponse pa e ns) migh be iden ical, he mechanisms
migh di e . Fo example, aking beha io s migh include no
only beha io s gea ed owa d aking bu also beha io s gea ed
owa d a oiding de ec ion (Fiedle and Bluemke 2005). Mo e-
o e , many o he quan i a i e app oaches model aking in
SRPSs as one a iable (e.g., one la en ac o ), hus ende ing i
di icul o iden i y pa icula aking beha io s, especially i hey
a e no used by all esponden s and i hey occu in di e en
combina ions.
2.3 | Indica ions F om Quali a i e Analyses
A mo e s aigh o wa d way o unde s and esponden s'
beha io s is o le esponden s speak o hemsel es (e.g.,
Schü z and DePaulo 1996), ask hem wha hey did o ake, and
analyze hei esponses (i.e., quali a i e da a). So a , only a ew
s udies ha e aken on he challenge o analyze aking in SRPSs
on his basis. Robie, B own, and Bea y (2007) we e he i s o
empi ically demons a e esponden s' willingness o ake in
SRPSs, bu hey did no ocus on he espec i e beha io s. Kö-
nig, Me z, and T au e (2012) epo ed esponden s' assump-
ions abou wha hey hough migh be ele an o esea che s
when in e p e ing hei SRPSs esponses (consis ency o
esponses, endo semen o middle s. ex eme esponses, and a
ce ain p o ile), bu hey did no ocus on aking beha io s. S ill,
hei inding ha esponden s ha e an idea abou wha is el-
e an o make a c edible imp ession indica es ha an ide ec ion
e o s may be an in eg al componen o aking. Ziegle (2011)
showed ha aking in SRPSs a ec s all s ages o he GRPM.
Rega ding aking beha io s, Ziegle ound ha when espon-
den s a e aking, hey e alua e each i em wi h ega d o i s
impo ance o eaching hei aking goal, and hey ake only on
he i ems hey conside o be ele an . Howe e , hei ocus was
on he p ocess in gene al and on he a iables ha po en ially
in luence such a gene al aking p ocess. Fo example, Ziegle
demons a ed ha esponden s' mo i es and pe sonali y a ec
he gene al aking p ocess in SRPSs a se e al s ages o he
GRPM. Al hough Ziegle (2011) ocused on he GRPM and
sugges ed ha indi idual di e ences migh mode a e his gen-
e al p ocess, hey did no in es iga e he mani oldness o ac ual
aking beha io s.
1
Hauens ein e al. (2017) in es iga ed u e -
ances abou se e al ypes o p ocessing (i.e., beha io al‐o ien ed
p ocessing, seman ic analysis o i em con en , ai ‐o ien ed
p ocessing, and condi ional p ocessing) unde ins uc ions o
“ ake good”in SRPSs. They e ealed inc eased seman ic p o-
cessing and ex eme esponses unde aking. Fu he mo e, hei
esul s poin ed o he ele ance o i em desi abili y and
FIGURE 1 | Gene al esponse p ocess model. S ages o he GRPM a e p esen ed in he g ay boxes in bold ype ace. In he model by Tou angeau
and Rasinski (1988), he same s ages as in he model by K osnick (1999) we e sugges ed, excep o he ac ha Tou angeau and Rasinski (1988)
sugges ed wo s eps wi hin s age IV (i.e., mapping and edi ing) and called he s age “ esponse”acco dingly.
3o 27
indica ed ha aking beha io di e s ac oss aking con ex s.
Howe e , hei ocus was no o assess aking beha io s wi h
espec o he GRPM. Las bu no leas , Fuech enhans and
B own (2022) in es iga ed aking beha io and e ealed ha
esponden s ac i a ed in o ma ion abou “a s e eo ypical good
wo ke ,” ha some esponden s e ie ed pas expe iences
whe eas o he s did no , ha esponden s selec ed i ems on he
basis o hei desi abili y, and ha hey ied o a oid looking
bad. Howe e , Fuech enhans and B own (2022) ocused on
aking in mul idimensional o ced‐choice (MFC) o ma s and
no in SRPSs.
3 | Finding Common G ound
Al hough p e ious esea ch has al eady iden i ied se e al ak-
ing beha io s in SRPSs, he complexi y o aking beha io s ha
esponden s use in hese measu es, he ela ionships be ween
hese beha io s, and he ela ionships wi h he esponse p ocess
ha unde lies aking in SRPSs a e s ill la gely uniden i ied o
se e al easons. Fi s , p e ious s udies may ha e o e looked
some aking beha io s in SRPSs. Fo example, i is likely ha
esponden s no only ake bu also y o a oid being de ec ed as
ake s. Thus, a ele an componen o aking beha io s in
SRPSs should conce n e o s aimed a an ide ec ion. In addi-
ion, al hough p e ious esea ch has ound ce ain aking
beha io s in SRPSs, esea che s ha e no in eg a ed hese
beha io s in o a espec i e axonomy, e en hough such a ax-
onomy would be use ul o esea ch and applica ions. Second,
combina ions o aking beha io s in SRPSs a e la gely unin-
es iga ed, al hough i is likely ha esponden s combine se -
e al beha io s (e.g., beha io s o inc ease hei chances o
aking success ully and beha io s o a oid ob ious aking pa -
e ns). Thi d, aking condi ions shape aking in SRPSs, and
hus, a aking beha io ha has been ound in a ce ain con-
di ion may no necessa ily be applied in ano he condi ion.
Mos s udies ha e no in es iga ed he gene alizabili y o hei
esul s conce ning aking di ec ions, cons uc s, and da a se s.
Also, he sample sizes in some s udies we e qui e small, so hey
we e somewha limi ed in mapping he aking beha io s ha
we e used less equen ly.
4 | The P esen S udy
The p esen s udy applied a quali a i e con en analysis o
in es iga e 533 esponden s' epo s o how hey aked in SRPSs
wi h Like ‐ ype i ems wi h espec o wo aking di ec ions,
wo cons uc s, and wo da a se s o aid he unde s anding
o he complexi y o aking beha io s in SRPSs in ela ion
o he unde lying GRPM (K osnick 1999; Tou angeau and
Rasinski 1988). Thus, we ex ended p e ious esea ch on aking
in he GRPM amewo k in se e al ways.
Fi s , we aimed o de i e a axonomy o aking beha io s in
SRPSs ha we e epo ed by esponden s and o so hese
beha io s along he ou s ages o he GRPM. Second, we aimed
o iden i y he combina ions o hese aking beha io s in SRPSs
and desc ibe hem. Thus, he quali a i e con en analysis
app oach is induc i e in ha we coded pa icipan s' epo s o
how hey aked in SRPSs. I is also deduc i e in ha we checked
whe he we could ela e he espec i e aking beha io s in
SRPSs o he GRPM. Fu he mo e, we aimed o es he gen-
e alizabili y o ou esul s conce ning aking di ec ions, con-
s uc s, and da a se s. To do so, we in es iga ed whe he he
axonomy o epo ed aking beha io s in SRPSs could be
applied o he aking o high sco es as well as o he aking o
low sco es ( hi d), whe he i could be applied o wo di e en
cons uc s ha had o be aked ( ou h), and whe he he esul s
could be c oss‐ alida ed wi h a second independen da a se
( i h).
Ou esea ch ques ions we e as ollows:
R1: Wha a e he aking beha io s in SRPSs epo ed by
esponden s,and how can hey be so ed along he s ages o
he GRPM?
R2: How a e hese aking beha io s in SRPSs combined?
R3: Can he esul ing axonomy o he espec i e aking
beha io s in SRPSs be applied o di e en aking di ec ions ( he
aking o high sco es s. he aking o low sco es)?
R4: Can he esul ing axonomy o he espec i e aking
beha io s in SRPSs be applied o di e en cons uc s
(Ex a e sion [E] s.Need o Cogni ion [NFC])?
R5: Can he esul ing axonomy o he espec i e aking
beha io s in SRPSs be ound in a second independen sample
(Da a Se 1 s.Da a Se 2)?
5 | Me hods
We ollowed he Jou nal A icle Repo ing S anda ds (JARS) o
ensu e maximal anspa ency o ou da a and me hods. Due o
space es ic ions, some o hem a e e e ed o in oo no es.
5.1 | Resea ch Design O e iew and Rec ui men
P ocess
5.1.1 | P ocedu es Followed in he O iginal S udies
In Da a Se 1, pa icipan s wo ked on an E scale and a NFC
scale. In Da a Se 2, pa icipan s wo ked on an E scale. In each
da a se , pa icipan s wo ked on a baseline assessmen and
a e wa ds we e andomly assigned o one o he ollowing
condi ions: aking high sco es, aking low sco es, o wo king
unde he measu es' s anda d ins uc ions (i.e., con ol condi-
ion).
2
The aking ins uc ions a e in line wi h p e ious esea ch
(e.g., McDaniel e al. 2009; Röhne and Holden 2022) and e e
o hypo he ical pe sonnel selec ion scena ios. Pa icipan s we e
p o ided wi h in o ma ion abou he cons uc s ha we e
assessed (E o NFC). They we e gi en a b ie desc ip ion o he
cons uc and some examples o how esponden s wi h high
e sus low sco es on he SRPSs would desc ibe hemsel es.
3
The de ailed aking ins uc ions, including he desc ip ion o
he cons uc s and he espec i e examples, can be ound in he
4o 27 In e na ional Jou nal o Selec ion and Assessmen , 2025

Supplemen on he OSF (h ps://os .io/2sg86/? iew_only=
841d1c66db164ac1a391ded 180e32a5). Due o space es ic-
ions, we desc ibe hem he e in an abb e ia ed o m. In he
ins uc ions o aking high sco es, pa icipan s we e asked o
imagine hey had been unemployed o 1 yea and had jus been
gi en he oppo uni y o apply o a e y a ac i e job. They
we e asked o ake high on E (o NFC) o maximize hei
chances o being o e ed he job. The ins uc ions o aking low
sco es included he desc ip ion o a e y una ac i e job. To
a oid being o e ed he job, pa icipan s we e asked o ake low
on E (o NFC). Pa icipan s we e no p o ided wi h incen i es
o ake.
A e wa ds, pa icipan s in he aking condi ions we e asked o
espond o he ques ion “Wha did you do o ake on he
ques ionnai e?”They we e asked o w i e down hei esponses
in an open‐ended esponse o ma . In Da a Se 1, hey we e
asked o espond o he ques ion sepa a ely o he E scale and
he NFC scale.
5.1.2 | To‐Be‐Faked Measu es in he O iginal S udies
5.1.2.1 | E Scale. Pa icipan s wo ked on he espec i e
scale om he NEO‐Fi e Fac o In en o y (Bo kenau and
Os endo 2008; English e sion: Cos a and McC ae 1992). This
scale consis s o 12 i ems ha a e answe ed on a 5‐poin scale
anging om 0 (s ongly disag ee) o4(s ongly ag ee). The scale
cha ac e is ics and C onbach's α eliabili y a he baseline
assessmen (Suppo ing In o ma ion S1: Table S1) we e com-
pa able o Bo kenau and Os endo 's (2008) alues o M= 28.38,
SD = 6.70, and α= 0.80. Faking led o ypical e ec s (e.g.,
Salgado 2016): The means changed in line wi h he aking
di ec ions, and he s anda d de ia ions and in e nal consis en-
cies inc eased.
5.1.2.2 | NFC Scale. Pa icipan s wo ked on he Ge man
adap a ion o he 16‐i em sho e sion o he NFC scale (Bless
e al. 1994; English e sion: Cacioppo and Pe y 1982). The scale
consis s o 16 i ems ha a e answe ed on a 7‐poin scale anging
om −3(s ongly disag ee) o+3(s ongly ag ee). The scale
cha ac e is ics and eliabili y a he baseline assessmen
(Suppo ing In o ma ion S1: Table S1) we e compa able o
Fleischhaue e al.'s (2010) alues o M= 15.28, SD = 11.14, and
α= 0.84. Faking led o ypical e ec s (e.g., Salgado 2016): The
means changed in line wi h he aking di ec ions, and he
s anda d de ia ions and in e nal consis ency sco e inc eased.
5.2 | Selec ion o Da a
We eanalyzed wo da a se s (N= 588) ha had been collec ed
wi h s uden samples in exchange o pa ial cou se c edi (Da a
Se 1: n= 318 and Da a Se 2: n= 270).
4
These da a se s we e
used o se e al easons. Fi s , we we e in e es ed in in es i-
ga ing pa icipan s' aking beha io s ha hey used nai ely.
5
Thus, we used da a se s in which pa icipan s had been asked o
ake wi hou ecommending a pa icula s a egy o hem
6
(e.g.,
Röhne and Holden 2022). Second, di e en om p e ious
esea ch (e.g., Ziegle 2011), pa icipan s we e no wa ned
abou aking de ec ion because such wa nings could change
esponden s' aking (e.g., by a enua ing aking; Li e al. 2022).
Also, wi h mos measu es, alid aking de ec ion is no ye
possible, and hus, wa ning pa icipan s ha aking can be
de ec ed (wi hou ha ing alid aking indica o s) would ha e
been p oblema ic om an e hical pe spec i e. Fo bo h easons,
i was necessa y o pa icipan s no o be wa ned. Thi d, we
we e in e es ed in in es iga ing aking on SRPSs wi h Like ‐
ype i ems, because hey a e especially suscep ible o aking
(e.g., Fuech enhans and B own 2022; Salgado 2016). Fou h, we
needed da a se s ha assessed aking on wo pe sonali y con-
s uc s (E and NFC), which was a necessa y p econdi ion o
compa ing aking beha io s ac oss SRPSs. E was included
because o i s equen use in p e ious aking esea ch
(McDaniel e al. 2009; S e ens 2004). In con as , NFC was
included because his cons uc has been only spa sely ex-
amined in aking esea ch (Röhne and Holden 2022), bu i is a
cons uc ha may be pa icula ly likely o inspi e esponden s
o answe in a manne ha may imp ess o he s (Lins de Ho-
landa Coelho, Hanel, and Wol 2020). Thus, by using hese
cons uc s, we hoped o co e a la ge span o aking beha io s.
Fi h, o demons a e he gene alizabili y o he esul s con-
ce ning aking di ec ions, cons uc s, and da a se s, we needed
da a se s in which high and low sco es we e aked, da a se s ha
included aking on di e en cons uc s, and da a se s wi h
independen samples. The desc ibed da a se s ma ched hese
c i e ia.
5.3 | Da a P epa a ion and Da a Composi ion
5.3.1 | Da a P epa a ion
We excluded da a om a o al o 45 pa icipan s who me one o
he ollowing c i e ia Suppo ing In o ma ion S1: Table S2).
Fi s , we excluded da a om pa icipan s who did no eply o
he ques ion “Wha did you do o ake on he ques ionnai e?”o
hose who answe ed ha hey did no hing o ake. Second, we
excluded da a om pa icipan s whose esponses could no be
clea ly in e p e ed (e.g., “Reading ins uc ions ca e ully”).
Thi d, we excluded da a om pa icipan s whose esponses
indica ed ha hey had no unde s ood he ins uc ions (e.g., “I
ga e he esponses associa ed wi h high Ex a e sion sco es”
when hey we e asked o ake low). Fou h, o easons o
compa abili y, o he analyses in Da a Se 1, i one o he
abo emen ioned c i e ia was me when a pa icipan was asked
o ake on E, we also excluded ha pa icipan 's da a when hey
we e asked o ake on NFC (and ice e sa).
5.3.2 | Da a Composi ion
The emaining sample size was N= 543 (n= 292 in Da a Se 1
and n= 251 in Da a Se 2), and we used hese da a o he
manipula ion checks. Da a Se 1 comp ised a o al o 292 pa -
icipan s (all s uden s; 222 women, 69 men, 1 di e se/no
esponse; a e age age: 22.41 yea s, SD = 4.11; aking high con-
di ion: 93 pa icipan s, con ol condi ion: 106 pa icipan s,
aking low condi ion: 93 pa icipan s), and Da a Se 2 comp ised
251 pa icipan s (all s uden s; 183 women, 68 men; a e age age:
5o 27
22.10 yea s, SD = 4.75; aking high condi ion: 79 pa icipan s,
con ol condi ion: 90 pa icipan s, aking low condi ion: 82
pa icipan s).
Fo he quali a i e con en analysis, pa icipan s in he con ol
condi ion o bo h Da a Se s (n= 196) we e excluded. Thus, he
emaining sample size o hese analyses was N= 360 pa ici-
pan s om he aking condi ions, and om hem, a o al o 533
pa icipan esponses we e collec ed. In Da a Se 1, 186
esponses we e ela ed o E (93 aking high s. 93 aking low)
and 186 o NFC (93 aking high s. 93 aking low). In Da a Se 2,
161 esponses we e ela ed o E (79 aking high s. 82 ak-
ing low).
5.4 | O e iew on Analyses
We
7
used quan i a i e and quali a i e da a and pe o med
quan i a i e and quali a i e da a analyses. We applied quan i-
a i e analyses ( obus ANCOVAs) o he SRPSs sco es
(pa icipan s' esponses on he E and NFC scales) o check
whe he pa icipan s in he aking condi ions we e mo i a ed
and able o ake on all he measu es (manipula ion check). To
de i e a axonomy o aking beha io s in SRPSs o ganized along
he ou s ages o he GRPM, we used pa icipan s' explana ions
o how hey aked (quali a i e da a) and analyzed hei
esponses wi h a quali a i e app oach (quali a i e con en
analysis). To de e mine whe he ou axonomy could be gen-
e alized wi h espec o aking di ec ions, cons uc s, and da a
se s, we used a bina y coding sys em (1 = aking beha io was
epo ed,0= aking beha io was no epo ed) o quan i a i ely
assess he p e alence o each iden i ied aking beha io . This
app oach was g ounded in he quali a i e da a de i ed om
pa icipan s' desc ip ions o how hey aked (quali a i e da a).
We hen applied Fishe 's exac es s (quan i a i e analyses) o
s a is ically analyze hese occu ences.
5.5 | Analy ical App oach
5.5.1 | Manipula ion Check
We compu ed obus ANCOVAs (Wilcox 2005) on each mea-
su e's sco e o check o whe he pa icipan s in he aking
condi ions we e mo i a ed and able o ake and whe he hei
sco es s ill di e ed when he baseline sco es we e con olled o
(Vicke s and Al man 2001). The signi ican di e ences be ween
immed means in all design poin s e ealed ha his equi e-
men was me (Suppo ing In o ma ion S1: Table S3).
5.5.2 | Quali a i e Con en Analysis
5.5.2.1 | Gene al P ocedu e. We conduc ed a quali a i e
con en analysis (May ing 2008; May ing and Fenzl 2014)
8
o
analyze he aking beha io s epo ed by esponden s in he
aking condi ions wi h espec o he s ages o he GRPM.
Con en analysis is de ined as “[…] any echnique o making
in e ences by sys ema ically and objec i ely iden i ying speci-
ied cha ac e is ics o messages”(Hols i 1968, p. 601). The goal
is o educe a la ge amoun o quali a i e da a in o ep esen -
a i e con en ca ego ies.
We allowed he coding ca ego ies o eme ge om he da a
a he han imposing ca ego ies a p io i, as li le heo y cu -
en ly exis s abou aking beha io s in SRPSs, and we wan ed o
a oid o e looking any aking beha io s. The uni o analysis
was he w i en desc ip ion o pa icipan s' aking beha io . To
his end, we coded pa icipan s' esponses o he ques ion:
“Wha did you do o ake on he ques ionnai e?”
A e excluding pa icipan s on he basis o he c i e ia
desc ibed abo e, we coded he 533 esponses gi en by pa i-
cipan s in he aking condi ions.
9
The au ho s independen ly
ead each esponse and iden i ied whe he he aking
beha io was p esen (i.e., coded 1) o no (i.e., coded 0).
Responses ha included mo e han one aking beha io we e
gi en a code o 1 o each aking beha io ha was epo ed.
Weme egula ly o e iew,discuss,and esol edisc epan-
cies ac oss he coded aking beha io s un il we ag eed on
how o code all o he 533 esponses. Example quo es ha
demons a e he esul ing axonomy o epo ed aking
beha io s in SRPSs can be ound in Table 1.Thecoding
sys em can be ound in Suppo ing In o ma ion (Suppo ing
In o ma ion S1: Table S4).
5.5.2.2 | Ou Te minology: Clus e s, Global Beha io s,
and Speci ic Beha io s. To o ganize he epo ed aking
beha io s, we clus e ed and labeled hem. We iden i ied se e al
global beha io s. In some cases, hese global beha io s could be
u he di e en ia ed in o speci ic beha io s. The beha io s buil
clus e s ha se ed di e en goals (Figu e 2). To gi e an ex-
ample: The global beha io “Adap ing answe s o i em con en
and di ec ion”includes wo speci ic beha io s (“Ag eeing wi h
i ems ha ma ch he desi ed imp ession”and “Disag eeing
wi h i ems ha con adic he desi ed imp ession”). The global
beha io (“Adap ing answe s o i em con en and di ec ion”)
and o he global beha io s (e.g., “Answe ing he opposi e”)
oge he buil a clus e ha we e med “Adjus men o esponse
beha io .”
Clus e s, global beha io s, and speci ic beha io s occu ed in
mo e han one s age o he GRPM (Figu e 2). Also, se e al
combina ions o hese elemen s we e ound ac oss esponden s.
Thus, esponden s' aking s a egies could be composed o
se e al global and speci ic beha io s in a ious clus e s.
5.5.3 | Fishe 's Exac Tes
The sugges ed axonomy o epo ed aking beha io s in SRPSs
includes global beha io s and speci ic beha io s ha a e
o ganized in o clus e s. We compa ed he occu ence o he
global beha io s and speci ic beha io s (i.e., whe he he
aking beha io was p esen [1] o no [0]) in he clus e s wi h
Fishe 's exac es s o analyze whe he he same global
beha io s and speci ic beha io s could be ound i espec i e o
aking di ec ion, o‐be‐ aked cons uc , and sample. The
independen a iables we e aking di ec ion (high s. low), o‐
be‐ aked cons uc (E s. NFC), and Da a Se (1 s. 2). The
6o 27 In e na ional Jou nal o Selec ion and Assessmen , 2025
TABLE 1 | Example quo es demons a ing he axonomy o aking beha io s in sel ‐ epo ed pe sonali y scales so ed along he ou s ages o he gene al esponse p ocess model.
S ages and clus e s Global beha io s Speci ic beha io s Example quo e Cons uc
Comp ehension and e ie al
Tes ‐c acking In e p e ing he cons uc Wi h i em meanings “On he i ems ha supposedly measu ed high
Ex a e sion, […], and on he ones [i ems] ha
supposedly measu ed low Ex a e sion, […].”
E
“On i ems associa ed wi h high NFC, […]. On
i ems ha we e associa ed wi h low NFC, […].”
NFC
Wi hou i em meanings “[…] sociable, ac i e […].”
“[…] likes o hink.”
E
NFC
Re ie al, judgmen , and esponse
Re ie ing in o ma ion beyond
he ins uc ions and choosing
esponses acco dingly
Role‐playing Using a p o o ype as a empla e “Imagined how a ese ed, a he shy pe son
would answe , o wha would be p e e ed by
ha pe son in he espec i e si ua ion.”
“Imagined how a pe son who is (ex emely)
a oidan in a cogni i e manne would
answe , […].”
E
NFC
Using an ideal/an undesi able
employee o he posi ion in ques ion
as a empla e
“[…], excep o hose [i ems] ha I hough
employe s would own upon i I chose ‘s ongly
ag ee’as an answe .”
“[…] especially wi h ega d o he pe sonnel
selec ion scena io (i.e., wha answe s would
s ike he ec ui e as nega i e).”
E
NFC
Using a ake as a empla e “I pu mysel in he shoes o a pe son who wan s
o sco e low o a oid ge ing o e ed his job […].”
“I pu mysel in he shoes o he pe son who was
desc ibed.”
E
NFC
Recalling and using one's own
expe iences/cha ac e is ics as a
guideline
Re e ing o one's own expe iences/
cha ac e is ics
“I ied o emembe he mos di icul momen
in my li e when I eally needed money and ime,
wi h he goal ha his memo y would mo i a e
me […].”
“I based my esponses on my own expe iences.”
E
NFC
Igno ing one's own expe iences/
cha ac e is ics
“[…] igno ed my own pe sonal a i udes/
cha ac e is ics.”
—
E
NFC
Imagining a se c i e ion and
aiming o i
—“Ticked e e y hing wi h he goal o achie ing
he desi ed sco e.”
E
(Con inues)
7o 27
TABLE 1 | (Con inued)
S ages and clus e s Global beha io s Speci ic beha io s Example quo e Cons uc
“[…] o answe he ques ions in a way ha I
conside mysel o be in he medium o low
ange in e ms o Need o Cogni ion.”
NFC
Judgmen and esponse
Adjus ing esponse beha io Adap ing answe s o i em
con en and di ec ion
Ag eeing wi h i ems ha ma ch he
desi ed imp ession
Faking high sco es: “S ongly ag eed wi h
s a emen s ha indica ed Ex a e sion.”
Faking low sco es: “[…] ag eed wi h s a emen s
ha indica ed In o e sion [low Ex a e sion].”
E
Faking high sco es: “S ongly ag eed wi h i ems
ha ob iously asked abou Need o Cogni ion.”
NFC
Disag eeing wi h i ems ha
con adic he desi ed imp ession
Faking low sco es: “Ag eed wi h i ems ha
indica ed low Need o Cogni ion.”
Faking high sco es: “S ongly disag eed wi h
s a emen s ha indica ed In o e sion.”
Faking low sco es: “Disag eed wi h he ones
[s a emen s] ha indica ed high Ex a e sion.”
E
Faking high sco es: “S ongly disag eed wi h
i ems ha e e ed o low sco es on Need o
Cogni ion.”
Faking low sco es: “I ied o espond o i ems
indica ing high Need o Cogni ion by choosing
‘disag ee’.”
NFC
Making su e o ake s ongly
enough
Choosing ex eme esponses on all
i ems
“I always chose he ex eme esponses: ‘s ongly
disag ee’o ‘s ongly ag ee’[…].”
“I always chose ex eme esponse op ions (−3
& +3).”
E
NFC
A oiding neu al esponses “[…], I ocused on ne e icking he middle, […].”
“[…] no neu al esponses […].”
E
NFC
Answe ing he opposi e —“Ticked answe s ha we e exac ly he opposi e
o Ex a e sion.”
E
“[…], I answe ed he opposi e.”NFC
Responding o i em desi abili y/
nondesi abili y
—“[…] chose op ions be ween […] wi h posi i ely
conno ed i ems.”
“Ticked nega i e i ems.”
E
NFC
(Con inues)
8o 27 In e na ional Jou nal o Selec ion and Assessmen , 2025
explain empi ical indings on wo ypes o aking in quan i a i e
da a (i.e., sligh and ex eme aking; Zicka , Gibby, and
Robie 2004; bu c . Ziegle e al. 2015).
The eigh h global beha io (“Responding andomly”; 1 u e -
ance; 0.19%) showed ha aking also in ol ed he a bi a y
selec ion o esponse op ions. We assigned his ca ego y i he
esponden s a ed ha hey had chosen esponse op ions a
andom (e.g., “T ied o mo e andomly be ween ‘SD’[s ongly
disag ee] and ‘D’[disag ee][…]”). The iden i ica ion o his
ca ego y is in line wi h esea ch om clinical psychology ha
ound ha 20% o malinge ing pa icipan s comple ed asks
quickly and ca elessly (Conne s, E ha d , and Spa ow 1998).
A i s glance, his ca ego y may seem simila o ca eless e-
sponding because one ype o ca elessness in ol es “ andom
esponding.”Howe e , he di e ence lies in mo i a ion: Ca e-
less esponde s aim o sa is ice by esponding “ andomly”(e.g.,
o e o ; Sch oede s, Schmid , and Gnambs 2022), whe eas
esponden s who ake in end o lea e a bad imp ession by
doing so.
6.1.3.2 | A oiding De ec ion. The i s global beha io
(“S a egically using ex eme esponses”; 199 u e ances;
37.34%) was he mos equen ly epo ed e o owa d “an i-
de ec ion.”The speci ic beha io s e e o how esponden s
s a egically deal wi h ex eme esponses o e ade aking
de ec ion: “Gi ing ex eme esponses on selec ed i ems only”
(assigned i he esponden s a ed ha hey had p o ided ex-
eme esponses on pa icula i ems only, e.g., acco ding o he
subjec i e impo ance o he i ems [e.g., “On i ems ha I con-
side ed o be e y cen al o he cons uc o In o e sion, I
always answe ed wi h ex emes […]”])
18
and “A oiding ex eme
esponses”(assigned i he esponden s a ed ha hey had
e ained om using ex eme esponses al oge he [e.g., “I did
no use ex eme esponses”]). The o me was mo e equen
(189 u e ances; 35.46%) han he la e (10 u e ances; 1.88%).
The iden i ica ion o hese ca ego ies is in line wi h p e ious
esea ch e ealing ha some esponden s pe cei ed ex eme
esponses on all i ems as no c edible and o he esponden s
we e conce ned ha ex eme esponding migh make hem
appea a ogan (Kuncel and Tellegen 2009).
Global beha io numbe wo (“T ying o a oid excessi e
aking”; 106 u e ances; 19.89%) shows ha some espon-
den s a emp ed o a oid ex eme aking. I was assigned i
he esponden s a ed ha hey had ied no o ake ex-
cessi ely (e.g., “[…], I made su e no o appea oo ex eme
bu o ind a good balance”). The iden i ica ion o his ca e-
go y do e ails wi h indings om quali a i e esea ch by
Ziegle (2011), who demons a ed ha excessi e aking was
ela ed o speci ic pe sonali y ai s (e.g., na cissism) and a
ake 's belie ha hey could ac ually li e up o he ole hey
we e playing.
The hi d global beha io (“Al e na ing be ween desi ed
esponses”; 96 u e ances; 18.01%) was assigned i he
esponden s a ed ha , while aking, hey had swi ched
be ween esponse op ions ha we e in line wi h he desi ed
imp ession (e.g., “Fo i ems ha we e associa ed wi h high
NFC, I icked high alues, i.e., 1–3[…]”).
19
The iden i ica ion
o his ca ego y e ealed ha , while aking, esponden s
al e na ed be ween desi ed esponse op ions p obably because
hey assumed ha his beha io would help hem a oid being
de ec ed as ake s.
Global beha io numbe ou (“Including neu al esponses”;33
u e ances; 6.19%) shows ha some esponden s included
neu al a ings o a oid aking de ec ion and i was assigned i
he esponden s a ed ha hey had done so (e.g., “Fo ques ions
abou Ex a e sion […], [I chose] mo e […] neu al a ings o
appea mo e au hen ic”). This ca ego y's iden i ica ion suppo s
p e ious quan i a i e and quali a i e esea ch ha poin ed o
he occu ence o midpoin esponding du ing aking (König,
Mu a, and Schmid 2015; Ziegle 2011) and esea ch ha
e ealed ha esponden s may conside he use o only ex-
emes o be oo ob ious (Kuncel and Tellegen 2009). Howe e ,
he p esen esea ch ex ends cu en knowledge by showing
ha esponden s may s a egically use neu al esponses when
aiming o an ide ec ion.
The i h global beha io (“S i ing o consis ency”; 7 u e -
ances; 1.31%) demons a es ha , while aking, some espon-
den s ied o p esen c edible and ma ching esponses. We
assigned he ca ego y i he esponden s a ed ha hey had
ied o lea e a consis en imp ession (e.g., “[…], ied o be
consis en be ween esponses, e.g., I don' like alking o o he
people …and I don' like being he cen e o a en ion”). The
iden i ica ion o his ca ego y is in line wi h quali a i e indings
e ealing ha pa icipan s assume ha he in e p e a ion o
hei esponses is based on hei consis ency (König, Me z, and
T au e 2012) and ha c ea ing a consis en p o ile will
dec ease he isk o being de ec ed as a ake (Fuech enhans
and B own 2022).
Global beha io numbe six (“A oiding ob ious/s able esponse
pa e ns”; i e u e ances; 0.94%) demons a es ha some pa -
icipan s esponded in a manne designed o a oid lea ing
suspicious esponse pa e ns ha migh e eal hei aking. The
ca ego y was assigned i he esponden s a ed ha hey had
ied o a oid ob ious esponse pa e ns o endencies (e.g.,
“A oided ob ious esponse pa e ns o a oid being de ec ed as a
ake ”).
The se en h global beha io (“Speci ically a ending o ‘con ol’
i ems”; i e u e ances; 0.94%) shows ha some esponden s
had expec ed con ol. We assigned he ca ego y i he espon-
den s a ed ha hey had assumed ha i ems we e included o
assess he c edibili y o hei esponses (e.g., “On an i em ha
was ambiguous, I answe ed neu ally. Could ha e been a
esponse endency i em”). The iden i ica ion o his ca ego y
do e ails wi h ea lie indings om quali a i e esea ch show-
ing ha pa icipan s belie e ha aking can be de ec ed (König,
Me z, and T au e 2012).
Global beha io numbe eigh (“Responding quickly”; h ee
u e ances; 0.56%) shows ha some esponden s ended o
speed h ough hei esponses o a oid aking de ec ion. The
ca ego y was assigned i he esponden s a ed ha hey had
esponded apidly (e.g., “T ied o come up wi h […] answe s as
quickly as possible”). We assume ha hey hough ha lying
akes ime (e.g., Sucho zki e al. 2017) and hus ied o a oid
being caugh by being quick.
20
15 o 27

The nin h global beha io (“Adjus ing aking s a egies while
aking”; wo u e ances; 0.38%) shows ha changed hei s a egy
o e he cou se o aking so ha hey we e no oo ob ious abou
hei aking. We assigned he ca ego y i he esponden s a ed
ha hey had begun aking in a ce ain manne bu hen changed
hei s a egy (e.g., “I no iced while comple ing he es ha i
migh be conspicuous o always selec […], and, i.e., why I ied o
ake less conspicuously om his poin on and also some imes
esponded […]”).
21
The iden i ica ion o his ca ego y suppo s a
aking model ha sugges s aking no as he las s ep in he
aking p ocess, bu as he s a ing poin o adap i e dynamics
be ween hose who ake and hose who e alua e he measu e
(Roulin, K ings, and Binggeli 2016). Thus, he p esen s udy
suppo s p e ious heo izing, bu also ex ends cu en knowledge
by showing ha adap a ion is no es ic ed o occu a e aking,
bu also happens du ing aking.
The en h global beha io (“T ying o con ey ce ain y”;1u e -
ance; 0.19%) shows ha a emp s we e made o con ey he
imp ession o ce ain y ega ding he aked esponses. The ca e-
go y was assigned i he esponden s a ed ha hey had aimed o
p ojec con idence (e.g., “Selec ed […] o show ha I am ce ain”).
Appa en ly, he esponden held he common‐sense assump ion
ha ce ain y would con ey heimp essiono c edibili y.
6.2 | Combina ions o Repo ed Faking Beha io s
Pa icipan s epo ed a a ie y o aking beha io s when hey aked
in SRPSs (Table 2). Some we e epo ed e y equen ly (e.g.,
“In e p e ing he cons uc ”), whe eas o he s we e a ely epo ed
(e.g., “Responding andomly”).
22
Mo eo e , he 22 global beha io s
and 13 speci ic beha io s in hei ou clus e swe ecombinedin
a ious ways o cons i u e indi idual aking s a egies. Due o space
es ic ions, we p o ide an o e iew in he Supplemen (see Sup-
po ing In o ma ion S1: Figu e S1). Al oge he , he e we e 137
combina ions. The smalles combina ion included wo aking
beha io s (e.g., “Using an ideal/an undesi able employee o he
posi ion in ques ion as a empla e”was combined wi h “T ying o
a oid excessi e aking”). The la ges combina ion in ol ed eigh
aking beha io s (e.g., pa icipan s in e p e ed he cons uc “Wi h
i em meanings”and adap ed hei answe s o he i em con en and
di ec ion by selec ing “Ag eeing wi h i ems ha ma ch he desi ed
imp ession”as well as “Disag eeing wi h i ems ha con adic he
desi ed imp ession”and “T ying o a oid excessi e aking”and
s a egically used ex eme esponses by “Gi ing ex eme esponses
on selec ed i ems only”and used “S i ing o consis ency,”
“Including neu al esponses,”and a he op “Al e na ing be ween
desi ed esponses”).
In he mos equen combina ion, pa icipan s epo ed ha
hey in e p e ed he cons uc “Wi h i em meanings”and
adap ed hei answe s o he i em con en and di ec ion by
selec ing “Ag eeing wi h i ems ha ma ch he desi ed imp es-
sion”as well as “Disag eeing wi h i ems ha con adic he
desi ed imp ession.”Eigh y‐eigh combina ions we e epo ed
wi h he leas equency (we e unique). All o hem consis ed o
wo o eigh aking beha io s. An example o such an in equen
combina ion is when pa icipan s epo ed ha hey in e p e ed
he cons uc “Wi hou i em meanings,”and hey epo ed ha
hey used “Imagining a se c i e ion and aiming o i .”
6.3 | Gene alizabili y o he Taxonomy o
Repo ed Faking Beha io s in SRPSs
Resul s o Fishe 's exac es s la gely suppo he gene al-
izabili y o he indings conce ning aking condi ions (see also
Suppo ing In o ma ion S1: Table S6–S8) wi h some excep ions
ha we e p edominan ly obse ed wi h espec o aking
di ec ion.
6.3.1 | Ac oss Faking Di ec ions
Wi h eigh excep ions, he equencies o he use o di e en
aking beha io s as epo ed by esponden s did no di e be ween
aking di ec ions (Suppo ing In o ma ion S1: Table S6). The odds
o adap ing answe s o he i em con en and di ec ion by
“Ag eeing wi h i ems ha ma ch he desi ed imp ession”we e
2.94 imes highe , he odds o making su e hey aked s ongly
enough by “Choosing ex eme esponses on all i ems”we e 2.50
imes highe , he odds o s a egically using ex eme esponses by
“Gi ing ex eme esponses on selec ed i ems only”we e 2.38 imes
highe , and he odds o in e p e ing he cons uc “Wi h i em
meanings”we e 1.86 imes highe when pa icipan s aked high
sco es han when hey aked low sco es.
23
Con e sely, he odds o
“Re e ing o one's own expe iences/cha ac e is ics”(OR = 0.18),
“Answe ing he opposi e”(OR = 0.03), and “Including neu al
esponses”(OR = 0.17) we e lowe when pa icipan s aked high
sco es han when hey aked low sco es. Thus, he axonomy was
la gely applicable o bo h aking di ec ions, bu he esul s also
indica e some di ec ion‐speci ic di e ences in aking beha io s.
24
6.3.2 | Ac oss Di e en Cons uc s
The e we e no di e ences in he occu ence o epo ed aking
beha io s be ween he wo cons uc s (Suppo ing In o ma ion
S1: Table S7). The e o e, he axonomy was applicable o bo h
cons uc s.
25
6.3.3 | Ac oss Di e en Samples
Wi h wo excep ions (i.e., “T ying o a oid excessi e aking”
and “Using a ake as a empla e”), he equencies o epo ed
aking beha io s did no di e be ween he wo da a se s
(Suppo ing In o ma ion S1: Table S8). The odds o “T ying o
a oid excessi e aking”we e 2.32 imes highe o pa icipan s
in Da a Se 2 han hose in Da a Se 1.
26
Thus, he axonomy
was la gely applicable o bo h da a se s.
27
7 | Gene al Discussion
Conduc ing a quali a i e con en analysis, we in es iga ed
esponden s' epo ed aking beha io s ha cons i u e aking
s a egies in SRPSs and we ela ed hem o he GRPM. The
p esen analyses sugges ha a wide a ie y o aking beha io s
(i.e., se e al global beha io s and e en a ia ions o hem
[speci ic beha io s]) a e in ol ed in he p ocess o aking in
SRPSs. In addi ion, hese beha io s a e ela ed o all s ages o
16 o 27 In e na ional Jou nal o Selec ion and Assessmen , 2025
TABLE 2 | F equencies o global beha io s and speci ic beha io s wi h espec o clus e s in he ou s ages o he gene al esponse p ocess model in esponding o sel ‐ epo ed pe sonali y scales when
compa ed o aking di ec ion ( aking high s. aking low), cons uc (Ex a e sion s. Need o Cogni ion), and da a se (1 s. 2).
S ages and clus e s Global beha io s Speci ic beha io s
Faking
di ec ion
F equency
Ex a e sion
(Da a Se 1)
Need o
Cogni ion
(Da a Se 1)
Ex a e sion
(Da a Se 2)
Comp ehension and e ie al
Tes ‐c acking In e p e ing he cons uc Wi h i em meanings Faking high 65 55 56
Faking low 51 43 44
Wi hou i em meanings Faking high 15 24 12
Faking low 27 20 19
Re ie al, judgmen , and esponse
Re ie ing in o ma ion
beyond he ins uc ions and
choosing esponses
acco dingly
Role‐playing Using a p o o ype as a
empla e
Faking high 5 11 7
Faking low 9 10 12
Using an ideal/an undesi able
employee o he posi ion in
ques ion as a empla e
Faking high 1 2 3
Faking low 1 4 4
Using a ake as a empla e Faking high ——1
Faking low —25
Recalling and using one's
own expe iences/
cha ac e is ics as a guideline
Re e ing o one's own
expe iences/cha ac e is ics
Faking high —12
Faking low 7 5 4
Igno ing one's own
expe iences/cha ac e is ics
Faking high ——1
Faking low ———
Imagining a se c i e ion and
aiming o i
—Faking high 1 3 1
Faking low 3 3 3
Judgmen and esponse
Adjus ing esponse beha io Adap ing answe s o i em
con en and di ec ion
Ag eeing wi h i ems ha
ma ch he desi ed imp ession
Faking high 46 37 33
Faking low 18 21 17
Disag eeing wi h i ems ha
con adic he desi ed
imp ession
Faking high 26 30 21
Faking low 21 24 28
Making su e o ake s ongly
enough
Choosing ex eme esponses
on all i ems
Faking high 7 18 5
Faking low 4 2 7
(Con inues)
17 o 27
TABLE 2 | (Con inued)
S ages and clus e s Global beha io s Speci ic beha io s
Faking
di ec ion
F equency
Ex a e sion
(Da a Se 1)
Need o
Cogni ion
(Da a Se 1)
Ex a e sion
(Da a Se 2)
A oiding neu al esponses Faking high —32
Faking low ———
Answe ing he opposi e —Faking high ——1
Faking low 12 13 5
Responding o i em
desi abili y/nondesi abili y
—Faking high 2 1 1
Faking low 2 6 4
Answe ing in an au hen ic
ashion
—Faking high ———
Faking low 2 1 1
Repea edly co ec ing one's
own esponses
—Faking high ———
Faking low —21
Inc easing alues owa d he
desi ed di ec ion
—Faking high ——1
Faking low —1—
Responding andomly —Faking high ———
Faking low 1 ——
A oiding being de ec ed as a
ake
S a egically using ex eme
esponses
Gi ing ex eme esponses on
selec ed i ems only
Faking high 44 38 38
Faking low 19 25 25
A oiding ex eme esponses Faking high 2 1 —
Faking low 1 3 3
T ying o a oid excessi e
aking
—Faking high 13 13 23
Faking low 15 18 24
Al e na ing be ween desi ed
esponses
—Faking high 17 9 17
Faking low 21 20 12
Including neu al esponses —Faking high 1 —4
Faking low 10 8 10
S i ing o consis ency —Faking high 1 —2
Faking low 2 1 1
(Con inues)
18 o 27 In e na ional Jou nal o Selec ion and Assessmen , 2025
he GRPM and in each case span mo e han one s age o he
GRPM. Mo eo e , we showed ha hey can be combined in
a ious ways, hus e ealing he complex ab ic ha o ms
indi idual aking s a egies in SRPSs. Las bu no leas , we
la gely demons a ed he gene alizabili y o ou esul s wi h
espec o wo di e en aking di ec ions, wo di e en con-
s uc s, and wo independen da a se s. Ne e heless, he esul s
also indica e some di ec ion‐speci ic di e ences in aking
beha io s. Ou axonomy is necessa ily es ic ed o he aking
condi ions used in ou s udy; u u e esea ch should add o he
( equencies o ) aking beha io s ha may po en ially occu
unde di e en condi ions.
7.1 | Theo e ical Implica ions
Faking in SRPSs is much mo e complex han equen ly
assumed. This complexi y e e s no only o he s ages o he
GRPM ha a e in ol ed in aking bu also o he a ie y o
aking beha io s, hei combina ions, and hei gene alizabili y.
7.1.1 | Faking Beha io s in SRPSs Occu in All S ages o
he GRPM
Tou angeau and Rasinski (1988) p oposed ha aking occu s
only in he las s age o he GRPM. Howe e , ou esul s expand
he unde s anding o he aking p ocess by sugges ing ha
aking beha io s in SRPSs may occu in all ou GRPM s ages.
Addi ionally, ou esul s show ha all aking beha io s span
mo e han one s age o his model. Fo example, he beha io
“Answe ing he opposi e”may belong o bo h he judgmen
s age and he esponse s age, as pa icipan s i s need o judge
wha he desi ed esponse is and hen espond in he opposi e
di ec ion. These indings emphasize he complex in e play
ac oss he GRPM s ages in he con ex o aking beha io s in
SRPSs, he eby unde sco ing he complexi y o aking as indi-
ca ed in ea lie quan i a i e esea ch (e.g., Bensch e al. 2019).
They a e also in line wi h heo ies sugges ing he possibili y o
concu en ly a ge ing GRPM s ages (Anglei ne , John, and
Löh 1986), he eby con ibu ing o a be e unde s anding o
he psychological mechanisms unde lying aking in SRPSs.
7.1.2 | Faking in SRPSs In ol es a Va ie y o Possible
Faking Beha io s
The complexi y o he aking p ocess in SRPSs was u he
highligh ed by ou iden i ica ion o ou clus e s ha include 22
global and 13 speci ic aking beha io s. The esul s o he
p esen quali a i e analysis in eg a e sugges ions om heo e -
ical models wi h insigh s om bo h quan i a i e and quali a i e
esea ch. In doing so, hey no only con i m p e iously sug-
ges ed aking beha io s bu also en ich p e ious sugges ions by
iden i ying new aking beha io s (e.g., “Using a ake as a
empla e”) in SRPSs, expanding on he iden i ica ion o aking
beha io s and hei unde s anding.
The aking beha io s iden i ied he e a e unlikely o ep esen a
comp ehensi e collec ion o all possible aking beha io s in
TABLE 2 | (Con inued)
S ages and clus e s Global beha io s Speci ic beha io s
Faking
di ec ion
F equency
Ex a e sion
(Da a Se 1)
Need o
Cogni ion
(Da a Se 1)
Ex a e sion
(Da a Se 2)
A oiding ob ious/s able
esponse pa e ns
—Faking high —2—
Faking low —21
Speci ically a ending o
“con ol”i ems
—Faking high 2 ——
Faking low ——3
Responding quickly —Faking high 1 2 —
Faking low ———
Adjus ing aking s a egies
while aking
—Faking high 1 ——
Faking low —1—
T ying o con ey ce ain y —Faking high 1 ——
Faking low ———
19 o 27
SRPSs. They a e limi ed because ou e ospec i e app oach o
asking pa icipan s abou hei aking beha io s a e hey aked
may ha e inc eased he isk o hem no emembe ing e e y
beha io hey applied. In addi ion, we es ic ed ou in es iga-
ion o a ce ain aking con ex . Addi ional aking beha io s
may occu in o he con ex s (see Limi a ions and Fu u e
Resea ch). Mos likely, he en i e y o aking beha io s in SRPSs
is he e o e e en mo e complex. Ne e heless, he esul s o he
p esen s udy ep esen a necessa y i s s ep owa d a mo e
comp ehensi e unde s anding o he a iabili y in aking
beha io s in SRPSs.
7.1.3 | Faking Beha io s in SRPSs A e Combined in
Va ious Ways
The indings show ha a ious combina ions o aking beha -
io s occu — hus ex ending p e ious insigh s in o he com-
plexi y o aking in SRPSs (e.g., Bensch e al. 2019; Röhne ,
Thoss, and Schü z 2022). Al oge he , he p esen s udy iden i-
ied 137 combina ions o aking beha io s, anging om com-
bina ions o wo o up o eigh aking beha io s. Howe e , wo
opposing biases may ha e occu ed. Fi s , ou me hod o asking
pa icipan s o e ospec i ely epo how hey aked may ha e
come wi h he d awback ha some pa icipan s migh no ha e
emembe ed e e y beha io hey used o migh no ha e been
mo i a ed o epo all beha io s. Second, we in es iga ed ak-
ing in SRPSs in a ce ain aking con ex , and he combina ions
we ound migh depend on his speci ic aking con ex . Thus,
ou esul s could o e es ima e o unde es ima e he equency
o ce ain combina ions (in o he se ings). To iden i y he exac
numbe o combina ions, u he in es iga ions wi h la ge and
a ying samples a e needed, ce ainly p o iding an a enue o
u u e esea ch.
7.1.4 | Faking Beha io s in SRPSs A e La gely
Compa able Ac oss Selec ed Condi ions
The axonomy o epo ed aking beha io s in SRPSs la gely
held ac oss di e en aking di ec ions, cons uc s, and samples,
hus indica ing a ce ain gene alizabili y o he esul s. The
di e ences obse ed occu ed p edominan ly wi h espec o
aking di ec ion, a inding ha suppo s ea lie a gumen s ha
aking high and aking low should be conside ed ela ed ye
dis inc ypes o aking (e.g., Bensch e al. 2019).
The esul s o he p esen s udy ad ance he unde s anding o
aking by iden i ying speci ic di e ences be ween aking high
and aking low. Fo example, when aking high sco es,
esponden s mo e equen ly e e ed o he ca ego y “Wi h
i em meanings” han when aking low sco es, indica ing ha
i em meaning is conside ed mo e impo an when aking high
sco es. In addi ion, he ca ego ies “Ag eeing wi h i ems ha
ma ch he desi ed imp ession,”“Choosing ex eme esponses
on all i ems,”and “Gi ing ex eme esponses on selec ed i ems
only”we e e e ed o abou wice as equen ly when aking
high han when aking low. This inding makes sense, as
“ag ee”o en e lec s desi able esponses, and including ex-
eme esponses migh be pe cei ed as pa icula ly help ul o
aking hem. Fo he same easons, wi h aking low, he ca e-
go y “Including neu al esponses”was e e ed o abou six
imes mo e o en, and he ca ego y “Re e ing o one's own
expe iences/cha ac e is ics”was e e ed o abou i e imes
mo e o en han wi h aking high. Las bu no leas , “An-
swe ing he opposi e”was almos exclusi ely epo ed when
aking low. This inding makes sense, as aking low o en
in ol es esponding in he opposi e di ec ion o wha is con-
side ed desi able.
7.2 | P ac ical Implica ions
The complexi y o aking beha io s shown he e has p ac ical
implica ions o s udying aking as well as o i s de ec ion and
p e en ion.
7.2.1 | Con adic o y Findings Rega ding he E ec s o
Faking May Be Caused by he Va ie y o Faking Beha io s
in SRPSs and Thei Combina ions
Wi h ega d o he e ec s o aking, he use o he di e en
aking beha io s in SRPSs may elici di e en e ec s (e.g.,
conce ning c i e ion‐ ela ed alidi y). Fo example, esponden s
u ilizing he beha io “Imagining a se c i e ion and aiming o
i ”migh ha e sco es ha a e comple ely made up, whe eas
hose who do no apply his aking beha io s ill ha e some
“ ue” a iance in hei sco es. Consequen ly, “ ue” a iance
may s ill p edic he c i e ion, o i migh no . Combina ions
wi h o he beha io s may educe o inc ease such e ec s. Fo
example, some pa icipan s epo ed ha hey combined
“Imagining a se c i e ion and aiming o i ”wi h “T ying o
a oid excessi e aking,”whe eas o he s epo ed ha hey
combined i wi h “Choosing ex eme esponses on all i ems.”
All in all, he indings o he p esen s udy may help explain
p e ious con adic o y esul s om quan i a i e esea ch on
whe he c i e ion‐ ela ed alidi y is impai ed by aking o no
(e.g., Ones and Viswes a an 1998; Salgado 2016).
7.2.2 | Challenges in De ec ing and P e en ing Faking
in SRPSs May Be G ounded in he Complexi y o Faking
Beha io s and Thei Combina ions
The a iabili y and di e si y in combina ions o aking beha io s
in SRPSs unde sco e how complex he de ec ion and p e en ion
o aking in SRPSs is and may explain why esea ch on he
subjec has aced se ious challenges (Bill and Melche s 2022).
Because he e a e a ious ways o ake, i migh be necessa y o
use di e en aking indices o de ec esponden s who di e in
hei aking beha io s, and he e migh be a need o use di e en
p oac i e app oaches o p e en esponden s om using hese
a ious aking beha io s. Fo example, i one ies o de ec
aking in SRPSs on he basis o bla an ex eme esponding (e.g.,
Lande s, Sacke , and Tuzinski 2011; Le ashina e al. 2014), his
app oach will mos likely wo k o esponden s who aked solely
by using he beha io “Choosing ex eme esponses on all
i ems,”bu i will no wo k o hose who used o he aking
beha io s (e.g., “A oiding ex eme esponses”). In addi ion, he
20 o 27 In e na ional Jou nal o Selec ion and Assessmen , 2025

esul s o he p esen s udy indica e he exis ence o di e en
aking s yles (e.g., conside ing he global beha io “Imagining a
se c i e ion and aiming o i ,”ei he he aked sco e s ill
somehow in ol es he ue sco e o he aked sco e is comple ely
made up) ha could no be de ec ed o p e en ed wi h only a
single me hod. Las bu no leas , he p esen esul s indica e ha
some esponden s adjus hei aking beha io s, which chal-
lenges aking indices and p e en i e app oaches e en mo e
because he suspicious beha io may change o e he cou se o
he assessmen (e en wi hin a single) measu e.
Combina ions o aking beha io s lead o u he challenges in
he de ec ion o p e en ion o aking. Fo example, i is plausible
ha de ec ing a esponden who, while aking, ies o unde -
s and he i ems, adjus s hei esponse beha io , and applies
an ide ec ion echniques may equi e a di e en app oach han
de ec ing one who engages in aking jus by ying o unde s and
he i ems and adjus ing hei esponses o achie e hei aking
goal. Al hough some (combina ions o ) aking beha io s occu
mo e equen ly han o he s in SRPSs, a e aking beha io s a e
s ill ele an , as hey a e less p edic able and a e hus mo e
challenging o de ec and p e en (Zicka and D asgow 1996).
28
Some aking beha io s (e.g., “Choosing ex eme esponses on all
i ems”) a e in line wi h ecommenda ions on how o de ec aking
(e.g., o moni o bla an ex eme esponding; Le ashina
e al. 2014). Ne e heless, esea che s should be awa e ha , gi en
he complexi y o aking beha io s and hei combina ions in
SRPSs, each app oach likely iden i ies only a ce ain ype o aking
(e.g., ex eme aking). The same is ue o he p e en ion o
aking. When keeping he complexi y o aking beha io s in SRPSs
in mind, i seems unlikely ha e e yone can be p e en ed om
aking wi h he same app oach (e.g., dec easing esponse imes).
Thus, a emp s o de ec o p e en aking in SRPSs can easily
become a complex endea o . Fu u e esea ch should he e o e
in es iga e app op ia e indices o de ec ing such complex pa -
e ns o aking and de elop app op ia e me hods o hei p e-
en ion. Mos likely, i will be necessa y o de elop de ec ion
and p e en ion me hods ha accoun o he a ie y o aking
beha io s in SRPSs and a e a leas as complex as he aking
beha io s hemsel es.
7.3 | Limi a ions and Fu u e Resea ch
The p esen s udy has a leas ou po en ial limi a ions. Fi s , o
minimize he isks o wo king memo y o e load and conse-
quen e ec s on he aking p ocess, we collec ed sel ‐ epo s on
aking beha io s a e pa icipan s had comple ed he ask.
Despi e ensu ing anonymi y and no wa ning abou aking
de ec ion o educe bias (e.g., caused by he mo i a ion o make
a good imp ession), i canno be uled ou ha biases occu ed
and limi ed he epo ed aking beha io s. Rela edly, pa ici-
pan s migh no ha e epo ed all aking beha io s (e.g., due o
a lack o memo y o mo i a ion). Thus, he esul s o he
p esen s udy likely e lec some o he possible aking beha -
io s and a e mo e o a map o build upon.
Second, esponden s we e no asked o epo hei aking
beha io a e each i em on he espec i e measu es. As Wil
and Re elle (2015) emphasized, i ems can con ain a ec i e,
beha io al, cogni i e, and desi e‐ ela ed con en . I is eason-
able o assume ha he decision on whe he an i em is ele an
and should be aked may be ela ed o i s con en . Thus, he
aking beha io s epo ed migh a y sys ema ically wi h he
i em con en . Fu u e esea ch should explo e hese associa ions.
Thi d, u u e esea ch should u he in es iga e he combina-
ions o aking beha io s ha cons i u e indi idual aking
s a egies. Despi e he need o ex ensi e da a o analyze hese
combina ions e ec i ely, he indings o he p esen s udy
indica e ha many esponden s use mul iple aking beha io s, a
inding ha wa an s u he explo a ion.
Fou h, aking has been demons a ed o depend on aking
condi ions (e.g., he o‐be aked measu e, he o‐be‐ aked con-
s uc s, and he aking di ec ions; Bi keland e al. 2006; Röhne ,
Thoss, and Schü z 2022) ha o m a ce ain aking con ex .
Thus, ou s udy's ocus on a ce ain aking con ex also implies
ha he gene alizabili y o ou axonomy is limi ed o his
con ex and, he e o e ep esen s a building block in he
de elopmen o an exhaus i e axonomy. We examined aking
in SRPSs. Faking on o he measu es (e.g., in e iews) may di e
and migh no be ully in line wi h he GRPM, which was p i-
ma ily de eloped o explain esponse p ocesses in SRPSs
(K osnick 1999). To a oid o e bu dening pa icipan s and o
a oid limi ing ou ocus o he Fi e Fac o Model, he p esen
s udy was es ic ed o wo pe sonali y ai s (E and NFC). On
he one hand, employing mo e han wo cons uc s could
inc ease cogni i e load, po en ially making aking mo e di i-
cul . On he o he hand, aking a ies wi h espec o he o‐be‐
aked cons uc (e.g., Bi keland e al. 2006). The e o e, bo h
issues may p omo e addi ional aking beha io s and s a egies
o a ec hei equencies. Mo eo e , ( equencies o ) aking
beha io s and s a egies migh di e depending on whe he he
cons uc is a ge ed o non a ge ed. Fu he mo e, o a oid
o e bu dening pa icipan s, we did no use a comple e pe -
sonali y in en o y. Howe e , in es iga ing aking in pe sonali y
in en o ies wi h i ems measu ing a ious ai s may e eal
o he ( equencies o ) aking beha io s and s a egies. In addi-
ion, in he p esen s udy aking beha io s ha cons i u e aking
s a egies in Like ‐ ype a ing scales we e in es iga ed, and
( equencies o ) aking beha io s and s a egies on o he
esponse o ma s migh di e . Mo eo e , we used ins uc ions o
ake high sco es and low sco es. Howe e , pa icipan s can be
ins uc ed o ake in di e en ways (e.g., in line wi h a ce ain
job p o ile, Geige , Bä wald , and Wilhelm 2021; making good
s. bad imp essions, Bensch e al. 2019; aking con a y o one's
ue esponses, Agos a e al. 2011), and esea ch has shown ha
aking di e s depending on such ins uc ions (e.g., Maha ,
Cologon, and Duck 1995; Ma in, Bowen, and Hun 2002). I is
plausible ha ins uc ing esponden s o answe opposi e o
hei ue esponses could inc ease he equencies o beha io s
ha we e a ely epo ed in he p esen s udy (e.g., Answe ing
he opposi e) and educe he equencies o o he s (e.g., due o
he es ic i e na u e o ins uc ions on how o ake). Ins uc-
ions ha ask esponden s o ake a job p o ile migh p oduce
addi ional aking beha io s and s a egies o a ec hei e-
quencies (e.g., conce ning an ide ec ion e o s). In addi ion,
compe i i eness impac s aking (Roulin and K ings 2016), and
we did no ewa d esponden s who aked bes . The samples in
21 o 27
he p esen s udy mainly consis ed o i s ‐semes e unde -
g adua e psychology s uden s om wo independen lab‐based
da a se s. Fu he mo e, he pa icipan s in his s udy we e in-
s uc ed o ake (o o answe hones ly). Faking in eal‐wo ld
se ings was no in es iga ed. While ins uc ing pa icipan s o
ake p o ides aluable insigh s in o he s a egies esponden s
can apply (Smi h and Ellingson 2002), i emains a simula ion
ha migh o e look sub le ies ha could (only) occu in eal‐li e
aking. Thus, in applied se ings, o he ( equencies o ) aking
beha io s and s a egies could eme ge (e.g., conce ning an i-
de ec ion e o s) because esponden s know ha aking is no
desi ed. In addi ion, he ins uc ions used in he p esen s udy
we e amed along simula ed pe sonnel selec ion scena ios, and
he equencies o se e al beha io s ha we iden i ied neces-
sa ily di e ed om hose ound in s udies whe e pa icipan s
p e ended o su e om ADHD (i.e., a di e en con ex ;
Conne s, E ha d , and Spa ow 1998). Las bu no leas ,
esponden s in he p esen s udy we e gi en in o ma ion (e.g.,
on he o‐be‐measu ed cons uc ), which could ha e a ec ed
he esul ing beha io ; howe e , such in o ma ion is also ypi-
cally accessible o e en p o ided in eal‐li e se ings (e.g.,
h ough job ad e isemen s ha clea ly desc ibe he desi ed
cha ac e is ics o he ideal candida es).
Because o hese limi a ions and o u he in es iga e he
gene alizabili y o he p esen s udy's esul s, u u e esea ch
should ocus mo e s ongly on i em‐speci ic aking, use o he
aking con ex s, and add he espec i e ( equencies o ) aking
beha io s and aking s a egies o he axonomy in oduced
he e. Fu u e esea ch should s udy aking wi h o he measu es,
in mo e and o he combina ions o SRPSs, when dis inguishing
be ween a ge ed and non a ge ed cons uc s (e.g., by using he
O*NET Wo k S yles), in es iga e aking in pe sonali y in en-
o ies and wi h o he esponse o ma s (e.g., o ced‐choice o -
ma s, see Fuech enhans and B own 2022), wi h o he ypes o
ins uc ions, wi h a ying ewa ds o aking, in o he samples
(e.g., o ensic samples o olde samples), in se ings wi h na -
u ally occu ing aking, wi h ins uc ions ela ing o o he se -
ings (e.g., aking o clinical symp oms), and unde condi ions
in which he abili y o iden i y c i e ia ele an o aking (ATIC;
Kleinmann e al. 2011) a ies.
8 | Conclusion
By asking pa icipan s o how hey aked in SRPSs, he p esen
s udy e ealed ou clus e s ha included 22 global and 13
speci ic beha io s, he eby ex ending p e ious indings. Mo e-
o e , he aking beha io s epo ed by esponden s could be
o ganized wi h espec o he s ages o he GRPM. Ou esul s
clea ly demons a e ha a lo is going on in esponden s' minds
when hey ake in SRPSs, ha usually a combina ion o aking
beha io s is used o cons i u e indi idual aking s a egies, and
ha while aking, a subs an ial numbe o esponden s also
employ beha io s o a oid de ec ion. We demons a ed a ce ain
gene alizabili y o ou esul s wi h indica ions o some
di ec ion‐speci ic di e ences in aking beha io s. Howe e , he
axonomy in oduced he e is necessa ily limi ed o he aking
con ex used in his s udy. The p esen s udy e ealed he
complex na u e o he beha io s ha o m indi idual aking
s a egies in SRPSs. In doing so, i may ad ance e o s in aking
de ec ion and p e en ion bu he esul s su ely do no ep esen
he en i e y o po en ial aking s a egies. Thus, he p esen
wo k should be complemen ed by u u e esea ch unco e ing
he ( equencies o ) aking beha io s and s a egies in o he
con ex s.
Au ho Con ibu ions
Jessica Röhne : concep ualiza ion, da a cu a ion, o mal analysis,
unding acquisi ion, in es iga ion, me hodology, p ojec adminis a ion,
isualiza ion, w i ing–o iginal d a p epa a ion, w i ing– e iew and
edi ing. As id Schü z: w i ing– e iew and edi ing, supe ision.
Ma hias Ziegle : concep ualiza ion, o mal analysis, me hodology,
w i ing– e iew and edi ing, supe ision.
Acknowledgmen s
We wan o exp ess ou g a i ude o he people who pa icipa ed in ou
s udies and o he s uden s who helped collec he da a o e he yea s:
Michael All amsede , Anna Di k, Ch is ina Doukas, Elke Hü en,
Hannah Klink, Ca men Mölle , and Anna‐Ma ie Ruda . We also wan
o exp ess ou g a i ude o Wiebke Hin ichsen, who coded 10% o he
da a and hus enabled us o es ima e he eliabili y o ou coding.
We hank Jane Zago ski o language edi ing. Las bu no leas , we a e
g a e ul o he help ul commen s we ecei ed om Oli ie Co neille
and Ronald R. Holden. This esea ch was pa ly unded by a g an om
he Equal Oppo uni ies O ice and by a g an om he In e nal
Resea ch Funding a he Uni e si y o Bambe g. The unding sou ce
had no in ol emen in he s udy design o analyses. Open Access
unding enabled and o ganized by P ojek DEAL.
E hics S a emen
All p ocedu es pe o med in s udies in ol ing human pa icipan s
we e in acco dance wi h he e hical s anda ds o he ins i u ional and/
o na ional esea ch commi ee and wi h he 1964 Helsinki Decla a-
ion and i s la e amendmen s o compa able e hical s anda ds. The
da a we eanalyzed we e collec ed in s udies ha had all been e al-
ua ed by a Uni e si y E hics Commi ee and app o al was g an ed
(app o al numbe s: V‐108‐15‐BM‐JR‐IAT‐14102015 and V‐151‐BM‐JR‐
IAT‐26072016).
Consen
All pa icipan s p o ided in o med consen o pa icipa e in he s udy
and o ha ing hei da a published in a jou nal a icle.
Con lic s o In e es
The au ho s decla e no con lic s o in e es .
Da a A ailabili y S a emen
The da a ha suppo he indings o his s udy a e openly a ailable in
OSF a h ps://os .io/2sg86/, e e ence numbe os .io/2sg86. We
desc ibe ou sampling plan, all da a exclusions, all manipula ions, all
measu es, and all analyses ha we used in he s udy. All analysis codes,
da a, a supplemen including u he analyses and also he guidelines
o code s, as well as he ou pu s a e s o ed on he OSF (h ps://os .io/
2sg86/? iew_only=841d1c66db164ac1a391ded 180e32a5). Da a we e
analyzed wi h he SPSS and he R so wa e.
Endno es
1
Mo eo e , in Ziegle (2011), pa icipan s we e wa ned ha aking
could be de ec ed. The e o e, i is no clea whe he he esul s can
be gene alized o a ypical se ing in which such wa nings do no
occu .
22 o 27 In e na ional Jou nal o Selec ion and Assessmen , 2025
2
Resea che –pa icipan ela ionship: The da a we e collec ed in he
labo a o y by s uden assis an s. None o he au ho s had di ec
con ac wi h he pa icipan s. Pa icipan s we e p edominan ly i s ‐
semes e unde g adua e s uden s. Thus, he isk o any ela ionship
o in e dependence was as low as possible, and he e we e no e hical
conside a ions ele an o p io ela ionships. Fu he mo e, pa ici-
pan s w o e down wha hey did o ake ins ead o e bally epo ing
i o an in e iewe . Thus, no only we e conce ns ela ed o social
desi abili y educed, bu in e ac ions be ween he s uden assis an s
and pa icipan s we e also kep o a minimum. Al oge he , a nega-
i e impac om he ela ionships be ween he au ho s o s uden
assis an s and he pa icipan s is unlikely.
3
A i s glance, i migh sound coun e in ui i e o desc ibe he
cons uc and gi e some examples o how esponden s wi h high
e sus low sco es on he SRPSs would desc ibe hemsel es o pa -
icipan s. Howe e , no e ha in applied se ings, job ad e isemen s
will ypically desc ibe he desi ed cha ac e is ics o he u u e job
holde , and in o ma ion on how esponden s wi h high e sus low
sco es on he SRPSs would desc ibe hemsel es can easily be
e ie ed om he In e ne .
4
The main con ibu ion o he cu en wo k is o o e an analysis o
he quali a i e da a om hese da a se s. These quali a i e da a ha e
ne e been analyzed o published be o e (see he Da a T anspa ency
Appendix and he O e iew o Va iables in Each Manusc ip in he
Suppo ing In o ma ion o de ailed in o ma ion).
5
The wo d nai ely means ha pa icipan s we e no gi en aking
s a egies bu had o use hei own s a egies (see, e.g., Röhne ,
Holden, and Schü z 2023).
6
No e ha he e a e also app oaches whe e pa icipan s we e no
naï e wi h espec o aking s a egies bu we e gi en speci ic aking
s a egies (e.g., aking opposi e o ones' ue esponses; Agos a
e al. 2011), an app oach ha p obably would ha e dec eased he
numbe o aking beha io s and s a egies ha could ha e been
iden i ied.
7
Desc ip ion o esea che s: The collabo a ion o he esea che s was
based on da a a ailabili y and expe ise in he opics o aking and
quali a i e esea ch. One o he h ee au ho s designed he s udies
and supe ised he da a collec ion. All h ee esea che s ha e been
expe s in he ield o aking and esea ch on his phenomenon o 15
o 30 yea s. The esea che s also ha e p io expe ience wi h quali-
a i e da a analyses. Thei expe ise enhanced he p ocess o ana-
lyzing he da a. Howe e , expe s may also be a isk o o e looking
aking beha io s because hey migh ocus selec i ely on a se o
assumed aking beha io s. We ackled his p oblem in wo ways.
Fi s , o a oid o e looking any aking beha io s by applying ca e-
go ies be o ehand and hus limi ing ou scope, we used quali a i e
con en analysis and induc i e coding so ha ca ego ies would
eme ge om he da a a he han being imposed a p io i (Campion
and Csillag 2022). Second, a e we de eloped he coding sys em and
coded he da a, we asked a s uden (a nonexpe ) o conduc an
independen coding. Thei coding was e y simila o ou s (see he
Reliabili y o he Coding sec ion below).
8
Speci ically, we chose he ma e ial (see also da a p epa a ion and
da a composi ion), analyzed he condi ions o he da a collec ion (see
also he p ocedu es used in he o iginal s udies), cha ac e ized he
ma e ial (pa icipan s' esponses), de e mined he di ec ion o he
analysis (p ecisely assessed he explici con en ), cla i ied ou
esea ch ques ions (see he heo y sec ion), de ined he ype o
analysis (s uc u ing; see also he coding sys em in he Supplemen ),
and de ined he uni o analysis (coding uni : aking beha io s.
con ex uni : aking s a egy) be o e we ca ied ou he analyses (see
he analy ical app oach).
9
Da a Se 1: n= 93 conce ning aking high and n= 93 conce ning
aking low [in each case mul iplied by 2: E and NFC we e coded
sepa a ely]; Da a Se 2: n= 79 conce ning aking high and n=82
conce ning aking low.
10
All analyses we e pe o med wi h R (4.1.3) using he ollowing
packages: e ec size (0.7.0) (Ben‐Shacha , Ludecke, and Makowski
2020), ggh4x (0.2.1) (Van den B and 2021), gg aph (2.0.5) (Pede sen
2021), gg ex (0.1.1) (Wilke 2020), he e (1.0.1) (Mülle 2020),
pa chwo k (1.1.1) (Pede sen 2020), Rall un 43 (Wilcox 2024), scales
(1.2.0) (Wickham and Seidel 2022), s gli e (2.1.0) (Wickham e al.
2022), idyg aph (1.2.1) (Pede sen 2022), and idy e se (1.3.1)
(Wickham, A e ick, and B yan 2019).
11
Tes ‐c acking is he p ocess o deciphe ing he unde lying mecha-
nisms o a measu e, including unde s anding wha is being mea-
su ed and iden i ying he speci ic manne in which o espond o
make a ce ain imp ession.
12
In each case, he pe cen ages e e o he p opo ions o pa icipan s
epo s ha included he beha io o clus e in bo h da a se s con-
ce ning aking high and low sco es and conce ning E and NFC (i.e.,
ou o 533— he o al numbe o epo s).
13
The occu ences o he beha io s in Ha ison, Edwa ds, and Pa ke
(2007) wi h he clinical measu e di e om he occu ences o he
beha io s in ou s udy wi h SRPSs. This inding once again poin s o
he complexi y o aking, which depends on he condi ions (e.g., on
he o‐be‐ aked measu es: SRPSs o clinical scale).
14
We did so because such a s a emen e e s o he employe 's
expec a ions.
15
We did so because his s a emen was clea ly ela ed o he
ins uc ions ha he pa icipan s should imagine applying o a job
and ha i would help i hey ied o sco e high/low—and pu ing
onesel in his posi ion is exac ly wha was explici ly s a ed in he
o he example abo e.
16
The ca ego y “Answe ing he opposi e”di e s om he ca ego y
“Disag eeing wi h i ems ha con adic he desi ed imp ession”
because “Answe ing he opposi e”means answe ing he opposi e o
a ce ain e e ence poin (in gene al) and no answe ing he opposi e
wi h espec o ce ain i em con en . Thus, he e e ence is mo e
abs ac he e. In addi ion, “Answe ing he opposi e”is no es ic ed
o an answe o “[s ongly] disag ee.”
17
Whe eas a esponse s yle e lec s a bias in esponding ha is con-
sis en ac oss ime and measu es, a esponse se is de ined as a bias in
esponding ha is ac i a ed only by con ex (e.g., Paulhus 2017).
18
This ca ego y di e s om he ca ego y “Choosing ex eme esponses
on all i ems”because, he e, he esponden s a ed he es ic ion ha
ex eme esponses we e no gi en on all i ems (e.g., “[…]alwayschose
‘s ongly ag ee’when i came o appea ing chee ul and ac i e […]”).
19
This ca ego y is dis inc om he ca ego ies o “Choosing ex eme
esponses on all i ems”and “Gi ing ex eme esponses on selec ed
i ems only”in ha “Al e na ing be ween desi ed esponses”on he
one hand includes a a ia ion in esponses wi h espec o hei
desi abili y and on he o he hand can be applied wi hou using
ex emes.
20
Indeed, esea ch has in es iga ed whe he inspec ing eac ion imes
can help de ec aking, bu esul s ha e been con adic o y (Holden
and Lambe 2015; bu c . Röhne and Holden 2022). The esul ha
some pa icipan s y o espond quickly while aking may pa ly
explain he con adic o y indings.
21
This ca ego y is di e en om he ca ego y “S a egically using ex-
eme esponses”because esponden s can s a egically use ex eme
esponses wi hou changing hei s a egy o e he cou se o hei
aking.
The ca ego y “Adjus ing aking s a egies while aking”is also di -
e en om he ca ego y “T ying o a oid excessi e aking”because
he la e can be done ei he in combina ion wi h changing he
s a egy o wi hou changing he s a egy a all.
22
One migh hink ha a e aking beha io s a e no impo an .
Howe e , Zicka and D asgow (1996) in es iga ed aking de ec ion
23 o 27
in a g oup o esponden s who had been asked o ake and compa ed
hose who had been coached on aking echniques in ad ance wi h
hose who had no been coached. They showed ha , while aking, a
maximum o 40% o esponden s wi h coaching bu only 20% o
esponden s wi hou coaching we e de ec ed, indica ing ha he
la e aked less p edic ably. The e o e, a e aking beha io s and
s a egies a e pa icula ly ele an , as hey a e less p edic able, and
hei use is ela ed o less de ec ion (and mos likely also o aking
p e en ion). Thus, we decided o epo all aking beha io s i e-
spec i e o hei epo ed equencies. Ne e heless, eplica ions a e
needed be o e he p ac ical ele ance o a e aking beha io s and
s a egies can be assessed.
23
The odds conce ning “A oiding neu al esponses”could no be
compu ed because o cells wi h ze o equencies in he c oss‐
abula ion.
24
The e a e se e al beha io s ha could indica e addi ional di ec ion‐
speci ic di e ences because esponden s ei he e e ed o his
beha io only when hey we e asked o ake high sco es (i.e.,
“Igno ing one's own expe iences/cha ac e is ics,”“Responding
quickly,”and “T ying o con ey ce ain y”) o when hey we e asked
o ake low sco es (i.e., “Answe ing in an au hen ic ashion,”
“Repea edly co ec ing one's own esponses,”and “Responding
andomly”). Howe e , due o he low base a es associa ed wi h
hese beha io s and he nonsigni ican di e ences, we s ay om
in e p e ing hese di e ences. Fu u e esea ch has o in es iga e
whe he hese di e ences a e meaning ul.
25
One may wonde whe he combining he wo aking di ec ions
impac ed he esul s o he in es iga ion o he gene alizabili y wi h
espec o he cons uc . We he e o e addi ionally compu ed he
analyses sepa a ely o each aking di ec ion. The esul s demon-
s a ed ha he e was only one signi ican di e ence (p= 0.030) in
he occu ence o aking beha io s be ween he wo cons uc s when
aking high sco es, indica ing ha , when aking high sco es, he
odds o “Choosing ex eme esponses on all i ems”we e 2.93 imes
highe when sco es we e aked on NFC han when hey we e aked
on E, and he e we e no signi ican di e ences in he occu ence o
aking beha io s be ween he wo cons uc s when aking low sco es
(Suppo ing In o ma ion S1: Tables S9 and S10). Thus, he esul s
we e la gely compa able.
26
The odds o “Using a ake as a empla e”could no be compu ed
because o cells wi h ze o equencies in he c oss‐ abula ion.
27
One may also wonde whe he combining he wo aking di ec ions
impac ed he esul s o he in es iga ion o he gene alizabili y wi h
espec o he da a se . We he e o e addi ionally compu ed he
analyses sepa a ely o each aking di ec ion. The esul s demon-
s a ed ha he e was only one signi ican di e ence (p= 0.023) in
he occu ence o aking beha io s be ween he wo da a se s when
aking high sco es, indica ing ha , when aking high sco es, he
odds o “T ying o a oid excessi e aking”we e 2.51 imes highe
when sco es we e aked in Da a Se 2 han when hey we e aked in
Da a Se 1, and he e was only one signi ican di e ence (p= 0.021)
in he occu ence o aking beha io s be ween he wo da a se s
when aking low sco es, indica ing ha , when aking low sco es, he
odds o “Using a ake as a empla e”we e highe when sco es we e
aked in Da a Se 2 han when hey we e aked in Da a Se 1
(Suppo ing In o ma ion S1: Tables S11 and S12).
28
In addi ion, i is impo an o ecognize ha psychological assess-
men s a e ypically implemen ed in con ex s whe e impo an
decisions esul (e.g., pe sonnel assessmen , o ensics). In hese
se ings, e en a ew unde ec ed ake s can cause se e e damage.
Re e ences
Agos a, S., V. Ghi a di, C. Zogmais e , U. Cas iello, and G. Sa o i. 2011.
“De ec ing Fake s o he Au obiog aphical IAT.”Applied Cogni i e
Psychology 25: 299–306. h ps://doi.o g/10.1002/acp.1691.
Anglei ne , A., O. P. John, and F.‐J. Löh . 1986. “I 's Wha You Ask and
How You Ask I : An I emme ic Analysis o Pe sonali y Ques ionnai es.”
In Pe sonali y Assessmen ia Ques ionnai es: Cu en Issues in Theo y and
Measu emen , edi ed by A. Anglei ne and J. S. Wiggins, 61–108. Be lin
Heidelbe g: Sp inge . h ps://doi.o g/10.1007/978-3-642-70751-3_5.
Bea y, J. C., C. D. Nye, M. J. Bo neman, T. M. Kan owi z, F. D asgow,
and E. G aue . 2011. “P oc o ed Ve sus Unp oc o ed In e ne Tes s: A e
Unp oc o ed Noncogni i e Tes s as P edic i e o Job Pe o mance?”
In e na ional Jou nal o Selec ion and Assessmen 19, no. 1: 1–10.
h ps://doi.o g/10.1111/j.1468-2389.2011.00529.x.
Bensch, D., U. Maaß, S. G ei , K. T. Ho s mann, and M. Ziegle . 2019.
“The Na u e o Faking: A Homogeneous and P edic able Cons uc ?”
Psychological Assessmen 31, no. 4: 532–544. h ps://doi.o g/10.1037/
pas0000619.
Ben‐Shacha , M., D. Lüdecke, and D. Makowski. 2020. “E ec size: Es-
ima ion o E ec Size Indices and S anda dized Pa ame e s.”Jou nal o
Open Sou ce So wa e 5, no. 56: 2815. h ps://doi.o g/10.1105/joss.02815.
Bill, B., and K. G. Melche s. 2022. “Thou Shal No Lie! Explo ing and
Tes ing Coun e measu es Agains Faking In en ions and Faking in
Selec ion In e iews.”In e na ional Jou nal o Selec ion and Assessmen
31: 22–44. h ps://doi.o g/10.1111/ijsa.12402.
Bi keland, S. A., T. M. Manson, J. L. Kisamo e, M. T. B annick, and
M. A. Smi h. 2006. “AMe a‐Analy ic In es iga ion o Job Applican Faking
on Pe sonali y Measu es.”In e na ional Jou nal o Selec ion and Assessmen
14, no. 4: 317–335. h ps://doi.o g/10.1111/j.1468-2389.2006.00354.x.
Bless, H., M. Wänke, G. Bohne , R. F. Fellhaue , and N. Schwa z. 1994.
“Need o Cogni ion: Eine Skala zu E assung on Engagemen und
F eude bei Denkau gaben [Need o Cogni ion: A Scale Measu ing
Engagemen and Happiness in Cogni i e Tasks.].”Zei sch i ü
Sozialpsychologie 25, no. 2: 147–154.
Boddy, C. R. 2016. “Sample Size o Quali a i e Resea ch.”Quali a i e
Ma ke Resea ch: An In e na ional Jou nal 19, no. 4: 426–432. h ps://
doi.o g/10.1108/QMR-06-2016-0053.
Bo kenau, P., and F. Os endo . 2008. NEO‐FFI Neo‐Fün ‐Fak o en In-
en a nach Cos a und McC ae –deu sche Fassung [NEO‐FFI. Neo‐Fi e‐
Fac o In en o y Acco ding o Cos a and McC ae–Ge man e sion].
Gö ingen: Hog e e.
B own, A., and U. Böckenhol . 2022. “In e mi en Faking o Pe son-
ali y P o iles in High‐S akes Assessmen s: A G ade o Membe ship
Analysis.”Psychological Me hods 27, no. 5: 895–916. h ps://doi.o g/10.
1037/me 0000295.
Cacioppo, J. T., and R. E. Pe y. 1982. “The Need o Cogni ion.”Jou nal
o Pe sonali y and Social Psychology 42, no. 1: 116–131. h ps://doi.o g/
10.1037/0022-3514.42.1.116.
Calanna, P., M. Lau iola, A. Saggino, M. Tommasi, and S. Fu lan. 2020.
“Using a Supe ised Machine Lea ning Algo i hm o De ec ing Faking
Good in a Pe sonali y Sel ‐Repo .”In e na ional Jou nal o Selec ion
and Assessmen 28, no. 2: 176–185. h ps://doi.o g/10.1111/ijsa.12279.
Campion, E. D., and B. Csillag. 2022. “Mul iple Jobholding Mo i a ions
and Expe iences: A Typology and La en P o ile Analysis.”Jou nal o
Applied Psychology 107, no. 8: 1261–1287. h ps://doi.o g/10.1037/
apl0000920.
Cao, M., and F. D asgow. 2019. “Does Fo cing Reduce Faking? A Me a‐
Analy ic Re iew o Fo ced‐Choice Pe sonali y Measu es in High‐S akes
Si ua ions.”Jou nal o Applied Psychology 104, no. 11: 1347–1368.
h ps://doi.o g/10.1037/apl0000414.
Conne s, C. K., D. E ha d , and M. A. Spa ow. 1998. Conne s' Adul
ADHD Ra ing Scales (CAARS): ADHD Ac oss he Li e Span. New Yo k:
Mul i‐Heal h Sys ems.
Cos a, P. T., and R. R. McC ae. 1992. Re ised NEO Pe sonali y In en o y
(NEO‐PI‐R) and NEO Fi e‐Fac o In en o y (NEO‐FFI) P o essional
Manual. Odessa, FL: Psychological Assessmen Resou ces.
24 o 27 In e na ional Jou nal o Selec ion and Assessmen , 2025