Rösele , Lukas e al.
A icle
Need o Cogni ion, Cogni i e Load, and Fo ewa ning do no Mode a e Ancho ing E ec s. A
Replica ion S udy o Epley & Gilo ich (Jou nal o Beha io al Decision Making, 2005; Psychological
Science, 2006)
Jou nal o Commen s and Replica ions in Economics (JCRE)
Sugges ed Ci a ion: Rösele , Lukas e al. (2024) : Need o Cogni ion, Cogni i e Load, and
Fo ewa ning do no Mode a e Ancho ing E ec s. A Replica ion S udy o Epley & Gilo ich (Jou nal
o Beha io al Decision Making, 2005; Psychological Science, 2006), Jou nal o Commen s and
Replica ions in Economics (JCRE), ISSN 2749-988X, ZBW - Leibniz In o ma ion Cen e o Economics,
Kiel, Hambu g, Vol. 3, pp. 1-42,
h ps://doi.o g/10.18718/81781.38
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/304381
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 ps://c ea i ecommons.o g/licenses/by/4.0/
Jou nal o Commen s and Replica ions in Economics - JCRE
Need o Cogni ion, Cogni i e Load, and Fo ewa ning
do no Mode a e Ancho ing E ec s
A Replica ion S udy o Epley & Gilo ich (Jou nal o Beha io al Decision
Making, 2005; Psychological Science, 2006)
Lukas Rösele *1,2 Hannah L. Bögle 2Lisa Koßmann2
Sabine M. K uege 2Sab ina L. C. Bickenbach2Rica da Bühle 2
Jasmin della Gua dia2Lisa-Ma ie A. Köppel2Ja l Möh ing2
Susanne Ponade 2Kons an in Roßmaie 2Jessica Sing2
Jou nal o Commen s and Replica ions in Economics, Volume 3, 2024-8, DOI: 10.18718/81781.38
JEL: D91
Keywo ds: Ancho ing e ec , Sel -gene a ed, Expe imen e -p o ided, Replica ion, Ancho ing and
adjus men
Da a A ailabili y: The da a and R code o ep oduce he esul s o his eplica ion can be down-
loaded a JCRE’s da a a chi e (DOI: 10.15456/j1.2024270.0728512269).
Please Ci e As: Rösele , L. e al. (2024). Need o Cogni ion, Cogni i e Load, and Fo ewa ning do
no Mode a e Ancho ing E ec s. A Replica ion S udy o Epley & Gilo ich (2005, 2006). Jou nal o
Commen s and Replica ions in Economics, Vol 3(2024-8). DOI: 10.18718/81781.38 .
*Co esponding au ho . Email: lukas. oesele @uni-muens e .de
1Müns e Cen e o Open Science, Uni e si y o Müns e
2Uni e si y o Bambe g
Decla a ion: The au ho s decla e ha hey ha e no con lic s o in e es . The esea ch epo ed in his pape did no esul om
a o -pay consul ing ela ionship and ou employe has no inancial in e es in he pape ’s opic.
Recei ed Oc obe 31, 2023; Re ised May 15, 2024; Accep ed Augus 26, 2024; Published Oc obe 09, 2024.
©Au ho (s) 2024. Licensed unde he C ea i e Common License -A ibu ion 4.0 In e na ional (CC BY 4.0).
1
L. Rösele e al. – Need o Cogni ion, Cogni i e Load (Replica ion). JCRE (2024-8)
Abs ac
Ancho ing, he assimila ion o nume ical es ima es owa d p e iously conside ed numbe s, has
gene ally been sepa a ed in o ancho ing om sel -gene a ed ancho s (e.g., people i s hinking o
9 mon hs when asked o he ges a ion pe iod o an animal) and expe imen e -p o ided ancho s
(e.g., expe imen e s le ing pa icipan s spin o une wheels). Fo some ime, he wo ypes o
ancho ing we e belie ed o be explained by wo di e en heo e ical accoun s. Howe e , la e e-
sea ch showed c osso e be ween he accoun s. Wha now emains a e con adic ions be ween pas
and ecen indings, speci ically, which mode a o s a ec which ype o ancho ing. We conduc ed
h ee eplica ions (𝑁 o al =657) o seminal s udies on he dis inc ion be ween sel -gene a ed and
expe imen e -p o ided ancho ing e ec s whe e we in es iga ed he mode a o s need o cogni ion,
cogni i e load, and o ewa ning. We ound no e idence ha ei he ype o ancho ing is mode a ed
by any o he mode a o s. In line wi h ecen eplica ion e o s, we ound ha ancho ing e ec s
we e obus , bu he indings on mode a o s o ancho ing e ec s should be ea ed wi h cau ion.
2
Jou nal o Commen s and Replica ions in Economics - JCRE
1 In oduc ion
Wha is he eezing poin o odka? And a e he e mo e o ewe han nine A ican s a es in he
UN? These wo seemingly un ela ed ques ions a e examples o wo di e en kinds o ancho ing
ques ions. Tha is, bo h ha e shown ancho ing e ec s (Epley & Gilo ich, 2001; T e sky & Kahne-
man, 1974), which occu when people’s es ima es a e biased owa d p e iously conside ed ancho s
(0°C as he eezing poin o wa e ; nine A ican s a es), bu he wo ha e usually been explained
by di e en mechanisms. Ancho ing esea che s ha e hypo hesized ha mos people es ima e he
eezing poin o odka by memo izing he eezing poin o wa e (sel -gene a ed ancho ) and ad-
jus ing away om i because hey know ha odka eezes a colde empe a u es. In doing so, hey
ail o adjus a enough, and hei es ima es a e he eby biased owa d he sel -gene a ed ancho .
This has been e e ed o as insu icien adjus men model (e.g., Epley & Gilo ich, 2001). Howe e ,
when expe imen e -p o ided ancho s a e p esen , people do no adjus away om he ancho bu
ins ead engage in hypo hesis-consis en es ing, ha is, hey gene a e easons in a o o he ancho
while simul aneously p iming nume ic alues ha a e close o he ancho . This p ocess has been
e e ed o as selec i e accessibili y (e.g., Mussweile & S ack, 1999a).
2 Con adic o y Findings om Pas Resea ch
The insu icien adjus men model and he selec i e accessibili y model a e he wo mos p ominen
ancho ing models. Insu icien adjus men was he i s explana ion o ancho ing e ec s (e.g.,
T e sky & Kahneman, 1974, p. 1228), whe eas selec i e accessibili y was p oposed la e and made
p ominen by Mussweile and S ack (1999a, 1999b). No ing ha li le a en ion had been paid
o he insu icien adjus men accoun , Epley and Gilo ich (2001, 2004, 2005, 2006, 2010) a -
gued ex ensi ely ha insu icien adjus men accoun s o ancho ing e ec s ha a e caused by sel -
gene a ed ancho s, and selec i e accessibili y accoun s o e ec s ha a e caused by expe imen e -
p o ided ancho s. In hei expe imen s, Epley and Gilo ich showed ha hings such as cogni i e
load, need o cogni ion, o o ewa nings a ec adjus men om sel -gene a ed ancho s bu no
adjus men om expe imen e -p o ided ancho s. No ably, he di ec ion o adjus men was known
o sel -gene a ed ancho s bu no o expe imen e -p o ided ancho s in Epley and Gilo ich’s ex-
pe imen s (i.e., people we e awa e ha he ancho was oo high o oo low o he sel -gene a ed
ancho s). By showing ha he mo i a ion o be accu a e leads o mo e adjus men away om an-
cho s only when he di ec ion o adjus men is known bu ega dless o he ype o ancho , Simmons
e al. (2010) made a case ha insu icien adjus men accoun s o expe imen e -p o ided ancho s,
oo. And la e , Chaxel (2014) showed ha selec i e accessibili y accoun s o sel -gene a ed an-
cho s, oo. Thus, by 2014, he dis inc ion be ween he wo kinds o ancho s ha had been os e ed
by a decade o esea ch had o be d opped again. Acco ding o his logic, he mode a o s ha Epley
and Gilo ich in es iga ed should also mode a e expe imen e -p o ided ancho ing e ec s as long
as he di ec ion o adjus men is known. Despi e hei heo e ical ele ance, he mode a o s ha
Epley and Gilo ich in es iga ed ha e ecei ed li le a en ion since Simmons e al.’s (2010) indings.
Mo eo e , ecen indings ha e challenged he iew ha wo di e en heo ies a e needed o
explain he wo kinds o ancho s: Fi s , Ha is e al. (2019) ailed o eplica e a "signa u e es o
he ope a ion o selec i e accessibili y mechanisms" (Abs ac ), and Bahník (2021) showed ha an-
cho s do no ac i a e in o ma ion ha is consis en wi h he ancho (see also F ede ick & Mochon,
2012 o c i icism o he selec i e accessibili y mechanisms). Addi ionally, in Epley and Gilo ich’s
3
L. Rösele e al. – Need o Cogni ion, Cogni i e Load (Replica ion). JCRE (2024-8)
(2010) S udy 2a, suscep ibili y o sel -gene a ed ancho s was co ela ed wi h need o cogni ion
(bu suscep ibili y o expe imen e -p o ided ancho s was no ). Howe e , esea ch es ing he elia-
bili y o suscep ibili y o ancho ing sco es (e.g., Rösele , 2021; Rösele e al., 2019; Schindle e al.,
2021) has ound ex emely low alues (i.e., a e age in e i em co ela ions close o ze o). The eby,
despi e he indings epo ed by Epley and Gilo ich (2006, S udy 2a), a iables such as need o
cogni ion canno be co ela ed wi h ancho ing.
Table 1: Summa y o Findings Rega ding he Rela ionship Be ween Sel -Gene a ed and
Expe imen e -P o ided Ancho s
No di e ence be ween ancho ypes EP + SA and SG + IA
- Simmons e al. (2010): Insu icien ad-
jus men is alid o bo h ypes o ancho s as
long as he di ec ion o adjus men is known
- Chaxel (2014): Selec i e accessibil-
i y is alid o bo h ypes o ancho s
- Ha is e al. (2019) and Bahník (2021): he se-
lec i e accessibili y model is p oblema ic
- Epley and Gilo ich (2001, 2004, 2005, 2006): Sel -
gene a ed ancho s a e mode a ed by need o cogni ion,
o ewa ning, mone a y incen i es, cogni i e load, head
mo emen , a m lexion, and alcohol consump ion
No es: EP = expe imen e -p o ided, SG = sel -gene a ed, SA = selec i e accessibili y, IA = insu i-
cien adjus men .
3 Resol ing he Con adic ions
Epley and Gilo ich’s indings ha e been undamen al o he de elopmen o he insu icien ad-
jus men model o ancho ing and a e highly ci ed (e.g., 1328 ci a ions o he 2006 pape acco ding
o Google Schola in Ap il 2024). Ye , hey s ongly con adic mo e ecen wo k. The aim o
his esea ch is o b ing cla i y o he con adic ions su ounding sel -gene a ed and expe imen e -
p o ided ancho s: How can wo mechanisms be esponsible o he wo ypes o ancho ing i
he e is no e idence o one o hem (i.e., selec i e accessibili y)? Why is he e no published
e idence ha mode a o s o sel -gene a ed ancho ing e ec s also a ec expe imen e -p o ided an-
cho ing? We chose o begin by conduc ing close eplica ions o h ee seminal s udies om Epley
and Gilo ich’s esea ch in which an in e en ion a ec ed suscep ibili y o sel -gene a ed ancho s
bu no expe imen e -p o ided ancho s. We chose s udies ha could be conduc ed online as all o
hem we e conduc ed be ween June 2020 and Ma ch 2022 du ing he COVID-19 pandemic. We
epo all s udies in he o de in which hey we e conduc ed. No e ha S udies 2 and 3 began a
he same ime, bu he ec ui men o pa icipan s o S udy 3 ook longe .
• S udy 1 is a eplica ion o Epley and Gilo ich (2006, S udy 2a) and in es iga ed he mode a o
need o cogni ion.
• S udy 2 is a eplica ion o Epley and Gilo ich (2006, S udy 2c) and in es iga ed he mode a o
cogni i e load.
• S udy 3 is a eplica ion o Epley and Gilo ich (2005, S udy 2) and in es iga ed he mode a o
o ewa ning.
4
Jou nal o Commen s and Replica ions in Economics - JCRE
All a ailable s udy ma e ials and da a se s can be ound online (h ps://os .io/p wu6). We
in i e o he esea che s o eanalyze ou da a o o conduc s udies using ou ma e ials and o
eplica e he esul s. We epo how we de e mined ou sample sizes, all da a exclusions (i any),
all manipula ions, and all measu es in he s udies (Simmons e al., 2012). In all s udies, we used
SoSci Su ey (Leine , 2019) o p og am he s udies, and we used R (R Co e Team, 2023) and he
packages coco (Diedenho en & Musch, 2015), da a. able (Dowle & S ini asan, 2024), dply (Wick-
ham e al., 2023), ggplo 2 (Wickham, 2016), g idEx a (Auguie, 2017), lme Tes (Kuzne so a e
al., 2017), lub ida e (G olemund & Wickham, 2011), MBESS (Kelley, 2023), me a o (Viech baue ,
2010), psych (Re elle, 2024), pw (Champely, 2020), eshape (Wickham, 2007), and xlsx (D ag-
ulescu & A end , 2020) o analyze he da a. An o e iew o he main esul s is p o ided in Figu e 1.
S udy 3 (Fo ewa nings)
S udy 2 (Cogni i e load)
S udy 1 (Need o Cogni ion)
−0.5 0.0 0.5 1.0 1.5 2.0
Cohen's d
Type
O iginal
Replica ion
Figu e 1: O e iew o e ec s o mode a o s on adjus men om sel -gene a ed
ancho s in o iginal and eplica ion s udies
No es: E alua ion o esul s acco ding o LeBel e al.’s (2019) e minology is no signal – in-
consis en . All es s o eplica ion s udies we e p e egis e ed. Code o ep oduce his igu e:
h ps://os .io/pe5kn
4 S udy 1: Replica ion o Epley and Gilo ich, 2006, S udy 2a (Need o Cogni ion)
4.1 Me hod
Need o cogni ion is de ined as " he endency o an indi idual o engage and enjoy hinking" (Ca-
cioppo & Pe y, 1982, p. 116). In e ms o ancho ing and adjus men , mo e hinking co esponds
o mo e adjus men away om he ancho and hus less suscep ibili y o ancho s. I insu icien
adjus men occu s o sel -gene a ed ancho s bu no o expe imen e -p o ided ancho s, only he
o me should be co ela ed wi h need o cogni ion. To es whe he need o cogni ion is nega-
i ely co ela ed wi h suscep ibili y o sel -gene a ed ancho s and unco ela ed wi h suscep ibili y
5
L. Rösele e al. – Need o Cogni ion, Cogni i e Load (Replica ion). JCRE (2024-8)
o expe imen e -p o ided ancho s, we conduc ed a p e egis e ed close eplica ion o Epley and
Gilo ich (2006, S udy 2a).
We used Epley and Gilo ich’s desc ip ions o c ea e new ma e ials because he o iginal ones we e
no a ailable. The en i e s udy (hypo hesis, p ocedu e, ma e ials, analysis sc ip ) was p e egis e ed
(h ps://os .io/8k9n ) using he eplica ion ecipe (B and e al., 2014).
4.1.1 A P io i Sample Size De e mina ion
The o iginal e ec size o he di e ence in suscep ibili y o sel -gene a ed ancho s be ween peo-
ple who sco ed high e sus people who sco ed low in need o cogni ion was 𝑑=0.49, CI 95%
[0.039,0.936]. The main e ec o need o cogni ion on suscep ibili y o expe imen e -p o ided an-
cho s was no epo ed. The e ec size o he in e ac ion be ween need o cogni ion and ancho
ype was 𝜂2= .05. In he o iginal s udy, only people wi h need o cogni ion sco es in he uppe and
lowe quin iles we e used. We conduc ed a simula ion o see which co ela ions his e ec would
apply o i he da a had been collec ed om people wi h a no mal dis ibu ion o need o cogni ion
alues ins ead o he ex eme quin iles. We planned o collec da a om 𝑁=240 pa icipan s so
ha he s a is ical powe o de ec ing he co ela ion be ween sel -gene a ed ancho s and need o
cogni ion (𝑟=.16) would be 80%. We aimed o achie e 80% powe only o p ac ical easons.
When sel -gene a ed ancho ing i ems a e used in pas s udies, many pa icipan s who do no hink
o he in ended sel -gene a ed ancho mus be excluded (e.g., Epley & Gilo ich, 2006, p. 314; mo e
han 10% o all pa icipan s, Rösele e al., 2020, p. 8). The simula ion and powe analyses a e
a ailable online (h ps://os .io/n2g u).
4.1.2 Ma e ials
Need o cogni ion was measu ed wi h he alida ed Ge man e sion o he need o cogni ion
scale by Bless e al. (1994). To measu e he suscep ibili y o ancho ing, we had o de ia e om
he o iginal expe imen o some deg ee: The e we e ou sel -gene a ed ancho ing i ems and ou
expe imen e -p o ided ancho ing i ems, he la e o which we e aken om Jacowi z and Kahne-
man (1995). Un o una ely, he o iginal s udy did no disclose which ou o he 15 i ems (Jacowi z
& Kahneman, 1995, p. 1163, Table 1) hey used. Thus, we chose i ems ha we belie ed would
wo k well wi h a Ge man sample ins ead o a U.S. sample. Due o he s ic exclusion c i e ia o
sel -gene a ed ancho ing i ems (i.e., pa icipan s mus know he sel -gene a ed ancho and mus
indica e ha hey had hough o i when gi ing hei es ima e), we chose eigh i ems, a leas
ou o which had o emain a e he exclusion c i e ia we e applied. These i ems we e a com-
bina ion o i ems ha we e ansla ed o adap ed om he o iginal s udy ( ou i ems), i ems ha
we e aken om Rösele e al. (2020; h ee i ems), and a newly c ea ed i em (one i em). An
o e iew o all i ems and hei espec i e sou ce, ype, ancho , and ue alue a e a ailable online
(h ps://os .io/9dnez).
4.1.3 P ocedu e
Pa icipan s we e g ee ed and old ha he pu pose o he expe imen was o es hei gene al
knowledge. Due o di icul ies in ec ui ing pa icipan s, we added non-mone a y incen i es o
6
Jou nal o Commen s and Replica ions in Economics - JCRE
pa icipa ion o 74% o he inal sample (all s uden s we e o e ed cou se c edi bu only la e pa -
icipan s we e o e ed eedback on he co ec alues o he gene al knowledge ques ions). A e
we collec ed demog aphic da a, pa icipan s comple ed he need o cogni ion scale. Expe imen e -
p o ided ancho ing i ems we e p esen ed wi h ixed ancho s and compa a i e ques ions (e.g., A e
he e mo e o ewe han 127 A ican membe s in he UN?). Fo each i em, pa icipan s answe ed
he compa a i e ques ion (mo e o less/ ewe ), ga e hei es ima e (How many A ican membe s
a e he e in he UN?), and— o explo a o y pu poses—indica ed how su e hey we e abou hei
es ima e on a 10-poin scale wi h labeled ex emes (no a all, e y much). Fo sel -gene a ed
ancho ing i ems, pa icipan s only ga e es ima es. The ancho ing i ems we e p esen ed on wo
subsequen pages (one o each ype). Then, pa icipan s we e asked whe he hey knew he al-
ues ha hey we e supposed o use o he sel -gene a ed ancho ing ques ions (e.g., he a e age
human body empe a u e) and whe he hey hough o he alue when gi ing hei es ima e (e.g.,
he lowes empe a u e measu ed in a li ing human). Finally, o explo a o y pu poses, pa icipan s
we e asked whe he hey knew o had hea d abou ancho ing e ec s and whe he hey had hough
abou ancho ing e ec s when comple ing he s udy. An o e iew o he p ocedu e is p o ided in
Figu e 2.
4.1.4 Design
Fo he expe imen e -p o ided ancho ing i ems, we manipula ed whe he pa icipan s we e gi en
high o low ancho s. To minimize p og amming e o s o he ques ionnai e and he analyses,
pa icipan s we e gi en ancho s ha we e ei he high-high-low-low o low-low-high-high o he
ou expe imen e -p o ided ancho ing i ems. All pa icipan s ga e es ima es o all 12 ancho ing
i ems. The e we e no u he manipula ions. Need o cogni ion was no manipula ed as i is
ela i ely s able o e ime (B uinsma & C u zen, 2018).
4.1.5 S a is ical Analyses
As pe he p e egis e ed analysis sc ip , suscep ibili y o ancho ing was compu ed as he absolu e
di e ence be ween ancho and es ima e (absolu e adjus men ) o all i ems. Absolu e adjus men
was s anda dized (i.e., mean-cen e ed and scaled) by i em and agg ega ed ancho ype. Means
we e compu ed by ancho ype o ope a ionalize he suscep ibili y o ancho ing.
Pa icipan s we e excluded i (a) he need o cogni ion scale had no been answe ed o (b)
hey did no know o hink o he in ended sel -gene a ed ancho s o a leas i e ou o he eigh
sel -gene a ed ancho ing i ems. Single sco es we e excluded i (a) hey co esponded o he co ec
alue o (b) hei absolu e s anda dized (by i em) absolu e adjus men was 𝑍 > 3.
4.1.6 De ia ions F om he O iginal S udy
We decided o de ia e om he o iginal s udy in some espec s because o he COVID-19 pandemic,
insu icien epo ing in he o iginal s udy, and some o he p ac ical easons. Fo example, ou s udy
was conduc ed in Ge many, equi ing us o ansla e he ancho ing ques ions and hink o new ones.
Fo example, ew Ge man people know when Washing on was elec ed p esiden , which is an i em
used by Epley and Gilo ich. Mo eo e , we e ained om using he o iginal ancho o 30 miles/h
7
L. Rösele e al. – Need o Cogni ion, Cogni i e Load (Replica ion). JCRE (2024-8)
Figu e 2: P ocedu e Used in S udy 1
o he maximum speed o a houseca because he ac ual maximum speed o a house ca is abou
30 miles/h (new ancho : 37.28 miles/h o 60 km/h ). Ano he de ia ion was ha we used all he
need o cogni ion sco es ins ead o jus he ex eme quin iles. This choice was made possible by
ou la ge sample size and mo e di e se sample. And we used mo e sel -gene a ed ancho ing i ems
(eigh i ems o which he e had o be a leas ou es ima es ins ead o only ou i ems) due o s ic
exclusion c i e ia. Compensa ion, ype o ins uc ions (e.g., he pu pose o he expe imen ), age,
gende , and con ol ques ions we e no epo ed in he o iginal s udy, and hus, we used common
s anda ds om he ield o ancho ing esea ch. We belie e ha all o hese changes we e necessa y
o he s udy o wo k. An o e iew is p o ided in Table 2.
8
Jou nal o Commen s and Replica ions in Economics - JCRE
Figu e 4: P ocedu e Used in S udy 2
5.1.5 De ia ions F om he O iginal S udy
Due o he COVID-19 pandemic, his eplica ion was conduc ed online. As pa icipan s we e Ge -
man, we had o change some o he i ems (e.g., "Elec ion yea o Ge man chancello Sch öde "
ins ead o "Numbe o Lincoln’s p esidency") o c ea e new ones. A comp ehensi e lis o de ia ions
om he o iginal s udy is p esen ed in Table 3.
15
L. Rösele e al. – Need o Cogni ion, Cogni i e Load (Replica ion). JCRE (2024-8)
Table 3: Me hodological Di e ences Be ween he O iginal S udy and ou Replica ion S udy 2
S udy ea u e Epley & Gilo ich, 2006,
2c
This s udy Reason o change
Language o ques ionnai e English Ge man Ge man pa icipan s
Type o sample Only college s uden s College s uden s and non-
s uden s
He e ogeneous sample
should inc ease he e ec
size
Type o s udy S udy was conduc ed on
si e
Online s udy COVID-19, la ge sample
Compensa ion No epo ed Cou se c edi o none Facili a e pa icipan e-
c ui men
Type o ins uc ions No epo ed Tes o gene al knowledge O iginal ma e ials we e no
a ailable
Addi ional a iables ha
we e collec ed
No epo ed Age, sex, educa ional s a-
us, p o ession
O iginal ma e ials we e no
a ailable
Le e s ings/Cogni i e
Load
Eigh -le e s ings Memo iza ion o
expe imen e -p o ided
eigh -le e s ings
Eigh -le e s ings om
he o iginal s udy we e un-
known
O de o p esen a ion The o de o p esen a ion
o he wo ypes o ques-
ions was coun e balanced
The o de o p esen a ion
was he same o each pa -
icipan
P esen ing he ques ions in
a coun e balanced o de
would be oo complex
Expe imen e -p o ided an-
cho
Nine i ems, which we e
no speci ied any u he
( he e we e 15 i ems in he
sou ce; Jacowi z & Kahne-
man, 1995);
Choice o nine
expe imen e -p o ided
i ems om he sou ce
(Jacowi z & Kahneman,
1995)
O iginal ma e ials we e no
a ailable
Sel -gene a ed ancho Nine i ems Fou i ems om he o ig-
inal s udy plus i e ad-
di ional i ems ( wo com-
ple ely new i ems, h ee
adap ed om he o iginal
i ems)
Ge man pa icipan s ha e
di e en common knowl-
edge han Ame ican pa -
icipan s; Pa icipan s who
did no hink o he ancho
migh ha e o be excluded
Exclusion c i e ia No epo ed Pa icipan s had o hink o
he ancho ; ou lie s we e
emo ed
O he wise, he e would
be dis o ion om pa ici-
pan s who did no hink o
he ancho ; ou lie s would
ha e dis o ed e ec size
es ima es
5.1.6 De ia ions F om he P e egis a ion
We had planned o ec ui 235 pa icipan s o ou eplica ion, bu a he end o es ing on Ma ch
22nd, 2022, only 183 people had pa icipa ed. We did no de ia e om he p e egis a ion in any
o he ega d. Despi e he smalle - han-planned sample size, he s a is ical powe o he o iginal
e ec size (𝑑=0.66) was 99.75% (𝑑min >0.369), and he powe o he in e ac ion e ec (𝜂2=
.03) was 81.16%.
16
Jou nal o Commen s and Replica ions in Economics - JCRE
5.2 Resul s
5.2.1 Sample
Ou sample consis ed o 183 pa icipan s (123 women, 58 men, 2 o he ) o which 127 we e s u-
den s. Thei ages anged om 18 yea s o 65 yea s, wi h a median age o 24 yea s.
5.2.2 Da a Quali y Checks
In a one-sided es , ancho ing e ec s we e p esen o se en ou o nine expe imen e -p o ided
ancho ing i ems (see Table 5, I ems 13, 14, 15, 17, 18, 20, 21) and six ou o nine sel -gene a ed
ancho ing i ems (see Table 5, I ems 23, 24, 26-29). No e ha he es in Table 5 is mo e sophis i-
ca ed han he p e egis e ed one-sided es because he es we used also equi ed ha he mean
adjus men alues we e no in he di ec ion opposi e he ancho .
5.2.3 Hypo hesis Tes s
The i s hypo hesis was ha cogni i e load does no ha e an e ec on expe imen e -p o ided
ancho ques ions. A Welch wo-sample es e ealed ha cogni i e load did no ha e a signi -
ican e ec on expe imen e -p o ided ancho s, (180.79)=0.75,p=.451, 𝑑 =−0.112, 95% CI
[−0.402,0.178], and pa icipan s’ adjus men sco es we e simila in he wo condi ions (𝑀no cogni i e load =
0.05, 𝑆𝐷no cogni i e load =0.44, 𝑁no cogni i e load =90, 𝑀cogni i e load =0.00, 𝑆𝐷cogni i e load =0.47, 𝑁no cogni i e load =
93; see also Figu e 5). The second hypo hesis p edic ed ha cogni i e load would lead o less ad-
jus men om he ancho o sel -gene a ed ancho ing i ems. This was no he case, 𝑡(176.58)=
-1.34, 𝑝=.908,𝑑=0.198, 95% CI [−0.093,0.488]. The in e ac ion be ween o ewa ning condi ion
and ancho ype in he 2 × 2 epea ed-measu es ANOVA was also no signi ican , 𝐹(1,362)=2.18,
𝑝=.141,𝜂2=.006, 90% CI [.000, .026].
17
L. Rösele e al. – Need o Cogni ion, Cogni i e Load (Replica ion). JCRE (2024-8)
Figu e 5: E ec s o Cogni i e Load on Adjus men om Expe imen e -P o ided
and Sel -Gene a ed Ancho s
No es: Code o ep oduce his igu e: h ps://os .io/g29w8
5.3 Discussion
In ou eplica ion o Epley and Gilo ich (2006, S udy 2c), we es ed whe he adjus men om
sel -gene a ed bu no expe imen e -p o ided ancho s dec eased when pa icipan s expe ienced
cogni i e load. We de ia ed om he o iginal s udy in ha we had o c ea e new i ems o he
18
Jou nal o Commen s and Replica ions in Economics - JCRE
Ge man (ins ead o US-Ame ican) pa icipan s. Di icul ies in ec ui ing pa icipan s esul ed in a
inal sample size o 𝑁 eplica ion =183 (𝑁 a ge =235, 𝑁o iginal𝑠𝑡𝑢𝑑𝑦 =94), bu s a is ical powe was s ill
>99% o he o iginal e ec o cogni i e load on adjus men om sel -gene a ed ancho s. Acco d-
ing o ou simple es o ancho ing, ancho ing occu ed o 13 ou o 18 i ems. Mos impo an ly, we
could no eplica e he o iginal inding ha cogni i e load a ec ed adjus men om sel -gene a ed
ancho s bu no om expe imen e -p o ided ancho s.
6 S udy 3: Replica ion o Epley and Gilo ich, 2005, S udy 2 (Fo ewa ning)
6.1 Me hod
In hei S udy 2, Epley and Gilo ich (2005) ound ha wa ning pa icipan s ha hei adjus men s
om ancho s we e insu icien led o inc eases in adjus men s om sel -gene a ed ancho s bu no
om expe imen e -p o ided ancho s. We conduc ed a p e egis e ed (h ps://os .io/ 5sj8) eplica-
ion o his s udy.
6.1.1 A P io i Sample Size De e mina ion
As e ec sizes we e no epo ed in he pape , we calcula ed hem on he basis o he epo ed
esul s. The es o he sel -gene a ed ancho s yielded Cohen’s 𝑑=1.24, CI 95% [0.558,1.894],
whe eas he es o he expe imen e -p o ided ancho s yielded 𝑑=0(we assumed a null e ec
because he es was epo ed as nonsigni ican wi h < 1; Epley & Gilo ich, 2005, p. 207). The
in e ac ion e ec was 𝜂2= .17, 90% CI [.033, .319]. To de e mine he necessa y sample size, we
used he small elescopes app oach (Simonsohn, 2015). When applied o he p esen eplica ion
s udy, he app oach indica ed ha we needed 120 pa icipan s (2.5 mul iplied by he o iginal 48
pa icipan s). The inal sample size exceeded 𝑁=120, so we es ed o whe he he esul s di e ed
when hey we e based on he i s 120 pa icipan s e sus he en i e sample.
6.1.2 Ma e ials
The ancho ing i ems (six sel -gene a ed and six expe imen e -p o ided ancho ing i ems) o mea-
su ing suscep ibili y o ancho alues we e p esen ed in he eplica ion s udy. Mos o hese i ems
di e ed om he i ems used in he o iginal s udy. Two i ems, one sel -gene a ed and one expe imen e -
p o ided, we e aken om he o iginal s udy. The easons o he de ia ions we e ha we adap ed
he i ems o he Ge man cul u e and language and he ac ha some o he i ems did no show
signi ican ancho ing e ec s in ou S udy 1 epo ed abo e. I ems ha could no be adap ed we e
eplaced by newly c ea ed ones. De ails abou he i ems used in he o iginal and eplica ion s udies
and hei espec i e sou ce, ype, ancho , and ue alue a e lis ed in a sepa a e able a ailable
online (h ps://os .io/dwm82, ile name: Ancho ing_i ems; see also Table 5).
6.1.3 P ocedu e
The s udy was conduc ed online. The in oduc o y ex o he online s udy and he ec ui men
ex bo h s a ed ha gene al knowledge and es ima ion ques ions we e he subjec s o he s udy.
We ec ui ed pa icipan s by con ac ing o he uni e si ies in Ge many (Ba a ia), social media, and
p i a e acquain ances. The in oduc o y ex in he ques ionnai e explained ha he s udy would
19
L. Rösele e al. – Need o Cogni ion, Cogni i e Load (Replica ion). JCRE (2024-8)
ake abou 10 min. Cou se c edi was o e ed as a non-mone a y incen i e, and pa icipan s had he
op ion o ecei e eedback om he s udy. A e pa icipan s p o ided in o med consen , hey we e
andomly assigned o he condi ion wi h o wi hou o ewa ning. In he o ewa ning condi ion,
pa icipan s we e in o med ha esea ch had demons a ed ha judgmen s a e s ongly in luenced
by he in o ma ion ha i s comes o mind. The example om he o iginal s udy was e ained:
Fo example, eal es a e agen s’ es ima es o a house’s alue a e in luenced by he alue o he
p e iously inspec ed house. Pa icipan s we e old ha i is suspec ed ha indi iduals begin om
he alue ha i s occu s o hem and subsequen ly ail o su icien ly adjus away om ha alue.
The ins uc ions, simila o he ones used in he o iginal s udy we e: In he ollowing, you will be
asked some ques ions. Ei he ce ain alues will be gi en o you will ha e a ce ain alue in you
mind. Please y no o be in luenced by hese numbe s. Pa icipan s we e asked o a oid using any
auxilia y sou ces, and hey we e eassu ed ha i did no ma e i hey we e no su e abou he
answe . All ins uc ions and ques ions we e ansla ed in o Ge man.
The expe imen e -p o ided ancho ing i ems we e asked i s , ollowed by he sel -gene a ed an-
cho ing i ems. Fo all expe imen e -p o ided ancho ing i ems, he i s ques ion was whe he he
ue alue was abo e o below he gi en ancho ing alue (e.g., whe he he Rhine is sho e o
longe han 2,000 km). Pa icipan s could choose be ween "mo e/g ea e /longe " o "less/smalle
/ ewe /sho e ." Subsequen ly, hey we e asked o en e hei es ima ed alue in an open ield.
The sel -gene a ed ancho ing i ems we e p esen ed a e wa ds. A e all i ems has been p e-
sen ed, he pa icipan s we e asked whe he hey knew he expec ed ancho alue (e.g., whe he
hey knew he a e age body empe a u e o a human being) and whe he hey had his alue in
mind when hey ga e hei es ima e (e.g., Did you hink o app oxima ely 36/37 deg ees Celsius
when you made you es ima e?).
A he end o he ques ionnai e, ques ions abou demog aphic a iables ega ding gende , age,
and occupa ion we e asked. Finally, as an explo a o y ques ion, he pa icipan s we e asked whe he
hey knew wha ancho ing e ec s we e. An email add ess o ques ions was gi en, and pa icipan s
had he op ion o p o ide hei email add ess o pa icipa ion in u u e s udies. An o e iew o he
p ocedu e is p o ided in Figu e 6.
20
Jou nal o Commen s and Replica ions in Economics - JCRE
Figu e 6: P ocedu e Used in S udy 3
6.1.4 Design
The p esen s udy used a 2 (ancho ype: sel -gene a ed s. expe imen e -p o ided) × 2 ( o ewa n-
ing: yes s. no) design. Unlike in he o iginal s udy, pa icipan s we e no all gi en he same ancho
alues o he expe imen e -p o ided ancho ing i ems bu we e andomly assigned o di e en e -
sions o he ques ionnai e wi h high and low ancho alues. To simpli y he p og amming, ou
e sions o he ques ionnai e we e c ea ed: In he i s e sion, high ancho alues we e speci ied
o he i s h ee expe imen e -p o ided ancho ing i ems and low ancho alues o he ollow-
ing h ee expe imen e -p o ided ancho ing i ems (high-high-high-low-low-low). This e sion was
a ailable once wi h and once wi hou o ewa ning. Fo he o he wo e sions o he ques ionnai e,
he i s h ee expe imen e -p o ided ancho ing i ems we e p esen ed wi h low ancho alues, and
he ollowing h ee wi h high ancho alues (low-low-low-high-high-high). A dis inc ion was also
made be ween i ems wi h and wi hou o ewa ning o his o de o ancho ing alues. Thus, he e
we e ou e sions o he ques ionnai e in o al. The sel -gene a ed ancho ing i ems we e iden ical
o all pa icipan s. The o de o he 12 i ems was no andomized.
21
L. Rösele e al. – Need o Cogni ion, Cogni i e Load (Replica ion). JCRE (2024-8)
6.1.5 S a is ical Analyses
As o he p e ious eplica ion s udies, we i s calcula ed he absolu e di e ence be ween ancho -
ing alue and es ima e (absolu e adjus men ). Absolu e adjus men was s anda dized (i.e., mean-
cen e ed and scaled) by ancho ype. Means we e compu ed by ancho ype o ope a ionalize he
suscep ibili y o ancho ing. Es ima es we e excluded i em wise (a) i pa icipan s did no know he
in ended sel -gene a ed ancho o did no ha e he alue in mind du ing hei es ima ion, (b) i hey
speci ied he ue alue, o (c) i hei s anda dized absolu e adjus men was 𝑍 > 3. Pa icipan s
we e included in he analyses i hey answe ed a leas one expe imen e -p o ided ancho ing i em
and one sel -gene a ed ancho ing i em.
6.1.6 De ia ions F om he O iginal S udy
Due o he COVID-19 pandemic, insu icien epo ing in he o iginal s udy, and p ac ical easons,
he eplica ion s udy di e ed om he o iginal one in some espec s. Jus like he o he eplica ion
s udies desc ibed ea lie , he s udy was conduc ed in Ge many. This is why ins uc ions, ancho ing
i ems, and o he ques ions we e ansla ed in o Ge man and adap ed o he Ge man cul u e and
gene al knowledge. Likewise, uni s we e con e ed o he Ge man s anda d (e.g., miles o kilome-
e s). The o iginal s udy was ca ied ou a a Bos on ain s a ion, bu due o he pandemic, we
chose o conduc he s udy online. O he in o ma ion, such as ype o sample, compensa ion, and
ype o demog aphic da a collec ed, was no epo ed in he o iginal s udy. Addi ional de ails abou
he de ia ions a e gi en in Table 4.
22
Jou nal o Commen s and Replica ions in Economics - JCRE
Table 4: Me hodological Di e ences Be ween he O iginal S udy and Replica ion S udy 3
S udy ea u e Epley & Gilo ich, 2005,
S udy 2
This s udy Reason o change
Language o ques ionnai e English Ge man Ge man pa icipan s
Type o sample Unknown College s uden s and non-
s uden s
Su icien pa icipan s
Compensa ion Some candy Cou se c edi Su icien pa icipan s
Type o s udy S udy was conduc ed in a
Bos on ain s a ion
Online s udy COVID-19, la ge sample
Type o ins uc ions Spoken W i en o m, simila in-
s uc ions
Online s udy did no allow
o spoken wo ds
Pe sonal da a No epo ed Th ee indi idual- ela ed
pieces o da a (age, sex,
p o ession) we e collec ed
We did no know which
kinds o indi idual- ela ed
da a we e collec ed in he
o iginal s udy
Expe imen e -p o ided an-
cho ing i ems
6 i ems ( om Jacowi z &
Kahneman, 1995)
1 i em om o igi-
nal s udy, 5 di e en
i ems (see ex a ile)
- As neu al as possi-
ble, no US-speci ic
- Adap ed o Ge man cul-
u e
- Ge man s uden s ha e li -
le US-speci ic knowledge
- I ems ha we e no signi -
ican in ou p e ious epli-
ca ion s udy we e changed
Expe imen e -p o ided an-
cho s: Va ia ion o ancho
alues
E e y pa icipan ecei ed
he same ancho alues
Pa icipan s we e an-
domly gi en high o low
ancho alues
To es o whe he ancho -
ing e ec s occu ed: com-
pa ison o adjus men om
wo di e en di ec ions
Sel -gene a ed ancho s 6 i ems: 2 i ems om
p e ious esea ch om
Epley and Gilo ich; Ep-
ley and Gilo ich (2001,
2004), 4 o he i ems (sel -
gene a ed)
1 i em om o iginal
s udy used, 5 di e en
i ems (see ex a ile)
- As neu al as possi-
ble, no US-speci ic
- Adap ed o Ge man cul-
u e
- Ge man s uden s ha e li -
le US-speci ic knowledge
- I ems ha we e no signi -
ican in ou p e ious epli-
ca ion s udy we e changed
O de o sel -gene a ed
and expe imen e -
p o ided ancho ques ions
O de was coun e bal-
anced
O de was ixed ( i s
expe imen e -p o ided,
hen sel -gene a ed)
The o iginal s udy ound
no in luence o he o de o
ancho ypes on any o he
epo ed esul s
Explo a o y ques ion None Ques ion abou whe he
he phenomenon o "an-
cho ing e ec s" was known
Explo a o y
Exclusion c i e ia No de ails abou ou lie ex-
clusion
I em-based exclusion i
pa icipan s de ia ed mo e
han 3 SD om he a e age
i em es ima e
O he dis o ion o he e-
sul s by ex eme answe s
Da a analysis No speci ied Linea mixed-e ec s model
(lme4 R-package)
6.1.7 De ia ions F om he P e egis a ion
Ins ead o he planned 120 pa icipan s, ou sample size comp ised 171 pa icipan s. We included
all da a in he analysis o inc ease he s a is ical powe , bu we es ed whe he including only he
i s 120 pa icipan s would ha e yielded di e en esul s. Conce ning he analysis sc ip , we made
small adjus men s. In he sample desc ip ion pa , we added one unc ion o de e mine he ac ual
23
L. Rösele e al. – Need o Cogni ion, Cogni i e Load (Replica ion). JCRE (2024-8)
sample size. Mo eo e , we conduc ed a powe analysis o he o iginal e ec size and a sensi i i y
analysis o he gi en sample size. In he hypo hesis es ing sec ion, we added one line each o
he sel -gene a ed and expe imen e -p o ided ancho s o p o ide desc ip i e s a is ics (M,SD) e-
ga ding ancho suscep ibili y g ouped by o ewa ning condi ion. We made ano he small change in
his sec ion so ha he Cohen’s dou pu would be p esen ed o 3 o mo e decimal places. Ano he
addi ion consis ed o an e ec size calcula ion wi h CIs o he ANOVA. Coding o empe a u e es-
ima es was co ec ed as he p e egis e ed unc ions did no wo k due o he pa icipan s using a
mix o commas and poin s as decimal symbols. Finally, we ca ied ou explo a o y analyses. We
es ed whe he he hypo heses would ha e been con i med i he i ems ha did no show signi ican
ancho ing e ec s had been excluded. Ano he issue was ha some pa icipan s we e al eady awa e
o ancho ing e ec s be o e he s udy, as he explo a o y ques ion we asked a he end o he s udy
had e ealed. We he e o e examined whe he adjus men inc eased when hese indi iduals we e
included in he g oup wi h o ewa ning. A inal supplemen was a g aph showing he days o he
s udy on he x-axis and he numbe o pa icipan s on he y-axis.
6.2 Resul s
6.2.1 Sample
The s udy ook place be ween Janua y 19, 2022 and Feb ua y 3, 2022. Du ing his pe iod, 220
pa icipan s comple ed he online ques ionnai e, o which 49 had o be emo ed due o ou ex-
clusion c i e ia. Ou inal sample comp ised 171 pa icipan s, 94 o whom we e o ewa ned abou
ancho ing e ec s, and 77 who we e no . The mean age was 26.98 yea s (one missing alue), and
o all pa icipan s, 114 we e women, 56 we e men, and one was di e se. Rega ding pa icipan s’
p o essions, he la ges pa o he sample (𝑁=105) epo ed ha hey we e uni e si y s uden s,
52 we e employed, i e we e in oca ional aining, one was a high school s uden , one was unem-
ployed, and he o he se en selec ed he ca ego y "o he ."
As we exceeded ou planned sample size o 120 pa icipan s, he powe we achie ed o he es
o he sel -gene a ed ancho ing i ems was > 99.99%, and mean di e ences o d > 0.384 could be
de ec ed wi h 80% powe . Fo he in e ac ion be ween o ewa ning condi ion and ancho ing ype,
we also achie ed a powe o > 99.99%.
6.2.2 Da a Quali y Checks
All six expe imen e -p o ided ancho ing i ems p oduced signi ican ancho ing e ec s (all d ≥0.60),
ha is, he e was a signi ican di e ence be ween he es ima es gi en a e conside ing high ancho s
and hose gi en a e conside ing low ancho s. As in S udy 2, we checked o he sel -gene a ed
ancho ing i ems i he absolu e le el o adjus men was smalle han he adjus men necessa y o
es ima e he ue alue. We ound signi ican ancho ing e ec s only o h ee ou o he six sel -
gene a ed ancho ing i ems (Table 5, I ems 37, 41, and 42). In o al, 75% o all ancho ing i ems
showed signi ican ancho ing e ec s, a inding ha was lowe han he expec ed 80%.
6.2.3 Hypo hesis Tes s
One- ailed independen -samples es s e ealed ha adjus men in he o ewa ning condi ion was
no signi ican ly g ea e han in he con ol condi ion o he expe imen e -p o ided (𝑀no o ewa ning =
24
Jou nal o Commen s and Replica ions in Economics - JCRE
29 Cogni i e Load Du a ion o
Me cu y’s
o bi a ound
he sun (88
Ea h days,
SG)
168 0.534 [0.424, 0.645] 0.021 [-0.131, 0.171] -0.07 [-0.219, 0.082]
30 Cogni i e Load Yea he i s
Ch is mas
was cele-
b a ed (336,
SG)
142 2.501 [2.125, 2.877] 0.04 [-0.126, 0.203] 0.04 [-0.126, 0.203]
31 Fo ewa ning Popula ion o
Wiesbaden
(278,609, EP)
165 0.73 [0.645, 0.815] 0.009 [-0.144, 0.162] 0.062 [-0.091, 0.213]
32 Fo ewa ning Speed o a
g eyhound
(80 km/h,
EP)
148 0.783 [0.63, 0.936] 0.136 [-0.026, 0.291] 0.095 [-0.068, 0.252]
33 Fo ewa ning Leng h o he
Rhein (1233
km, EP)
163 0.449 [0.342, 0.557] 0.047 [-0.107, 0.2] -0.067 [-0.218, 0.088]
34 Fo ewa ning Daily bi h
a e in Ge -
many in 2020
(2,112, EP)
160 0.355 [0.223, 0.487] 0.004 [-0.151, 0.159] 0.031 [-0.125, 0.185]
35 Fo ewa ning Bee con-
sump ion pe
pe son 2020
(94.6 l, EP)
162 0.465 [0.123, 0.807] 0.062 [-0.093, 0.214] 0.074 [-0.081, 0.225]
36 Fo ewa ning Yea he ele-
phone was in-
en ed (1861,
EP)
168 1.376 [1.077, 1.675] 0.058 [-0.094, 0.208] 0.059 [-0.094, 0.208]
37 Fo ewa ning Lowes
eco ded
human body
empe a u e
(13.7°C, SG)
167 0.386 [0.348, 0.424] 0.044 [-0.109, 0.194] 0.044 [-0.109, 0.194]
38 Fo ewa ning Yea he
Wo ld T ade
Cen e was
opened
(1973, SG)
161 0.939 [0.828, 1.051] -0.02 [-0.174, 0.135] -0.009 [-0.163, 0.146]
39 Fo ewa ning Ge man pop-
ula ion in
1880 (45
million, SG)
165 1.29 [1.203, 1.377] 0.013 [-0.14, 0.165] 0.002 [-0.15, 0.155]
40 Fo ewa ning F eezing
poin o me -
cu y (-38.9°C,
SG)
158 0.41 [0.22, 0.599] -0.1 [-0.253, 0.057] 0.045 [-0.111, 0.2]
41 Fo ewa ning Du a ion o
Ma s’ o bi
a ound he
sun (687
ea h days,
SG)
167 0.085 [-0.049, 0.218] 0.056 [-0.097, 0.206] -0.043 [-0.193, 0.11]
42 Fo ewa ning Numbe o
s a es in he
Eu opean
Union in
1995 (15,
SG)
147 0.868 [0.779, 0.958] 0.156 [-0.007, 0.31] 0.145 [-0.018, 0.299]
31
L. Rösele e al. – Need o Cogni ion, Cogni i e Load (Replica ion). JCRE (2024-8)
No es: EP = expe imen e -p o ided ancho , SG = sel -gene a ed ancho . Values in pa en heses
ep esen co ec answe s. Values in b acke s ep esen 95% con idence in e als. 0-1 sco es a e
0 i he ue alue was es ima ed and 1 i he ancho was es ima ed. Sco es ha do no include 0
and a e be ween 0 and 1 sugges ha ancho ing occu ed and a e bold. Sco es abo e 1 indica e
o e adjus men . Sco es below 0 indica e adjus men in he w ong di ec ion. Highe co ela ions
indica e mo e adjus men and less ancho ing gi en high alues o need o cogni ion, he p esence
o cogni i e load, o he p esence o a o ewa ning. Ou lie s ±3 SD we e excluded.
8 Gene al Discussion
We conduc ed p e egis e ed eplica ions o h ee seminal indings on mode a o s o sel -gene a ed
and expe imen e -p o ided ancho ing. In cases whe e ancho s we e a ied be ween pa icipan s
(e.g., one low and one high ancho was used pe ancho ing i em), ancho ing e ec s we e la ge.
Sel -gene a ed ancho ing i ems a e cha ac e ized by he same alue consis en ly coming o pa -
icipan s’ minds, and ancho s can he eby no be manipula ed. We es ed whe he adjus men
om sel -gene a ed ancho s owa d co ec alues was insu icien and ound ha 15/21 ancho ing
i ems displayed wha we would conside ancho ing e ec s. In he o he cases, people adjus ed in
he "w ong" di ec ion (i.e., away om he ue alue ins ead o owa d he ue alue), adjus ed
un il hey a i ed a he ue alue, o adjus ed oo a ). E idence o openly a ailable da ase s using
o iginal i ems om Epley and Gilo ich sugges s ha hei i ems should no be used o in es iga e
adjus men om sel -gene a ed ancho s (e.g., mos Ame icans do no hink o he decla a ion o
independence when asked abou he yea he Bos on Tea Pa y occu ed; h ps://os .io/y2m 9).
Mos impo an ly, none o he mode a o s (need o cogni ion, cogni i e load, o ewa ning) we e
associa ed wi h adjus men om ancho s.
A p ominen de ia ion o ou s udies is ha all eplica ion s udies we e conduc ed online
whe eas he o iginal s udies we e no . The e is cu en ly no way o es whe he his has a ec ed
he quali y o he da a and we belie e ha i did no : Fi s , ancho ing e ec s we e as la ge and
he e ogeneous as usual, which hey may no ha e been i pa icipan s ac ed di e en ly om he
o iginal se ings. Gene ally, ancho ing e ec s do no di e be ween on-si e and online s udies (e.g.,
Rösele & Schü z, 2022, Table 3, p. 22). Second, i is unlikely ha all mode a o s would ha e
been a ec ed equally. Fo example, need o cogni ion is a pe sonali y ai and measu ed wi h a
alida ed scale und hus unlikely o be a ec ed by his ype o a ia ion. Mo eo e , he gene al
le el o cogni i e load could di e be ween online and labo a o y se ings (in bo h possible ways),
bu we do no see a eason why o ewa ning should wo k be e in a labo a o y han in an online
s udy. Thi d, o ac ually compa e da a be ween he o iginal and eplica ion s udies, he o iginal
s udy’s da ase s would be needed. As he o iginal s udies ha e been conduc ed app oxima ely 20
yea s ago, no da a is a ailable.
His o ically, he dis inc ion be ween he wo ypes o ancho s led o a deba e spanning 13
yea s: Epley and Gilo ich econciled he insu icien adjus men model and he selec i e acces-
sibili y model o ancho ing in 2001 by sugges ing ha one had o be applied o sel -gene a ed
ancho ing i ems and he o he o expe imen e -p o ided ancho ing i ems. A e wa d, Simmons e
al. (2010) and Chaxel (2014) e oked hese indings, sugges ing ha bo h accoun s we e alid
o bo h ypes o ancho s. No e ha ecen indings ha e sugges ed ha he selec i e accessibili y
32
Jou nal o Commen s and Replica ions in Economics - JCRE
model is in alid (Bahník, 2021; Ha is e al., 2019). Ou esul s suppo his idea as none o ou
h ee eplica ion s udies e ealed any e idence ha he e is a di e ence be ween adjus men om
expe imen e -p o ided and sel -gene a ed ancho s. Bo h ypes o ancho s p o oked ancho ing e -
ec s, bo h adjus men sco es we e un eliable, and nei he sco e was co ela ed wi h any o he
mode a o s we in es iga ed (i.e., need o cogni ion, cogni i e load, o ewa ning).
No e ha ou esul s do no sugges ha none o Epley and Gilo ich’s indings on mode a o s can
be eplica ed (e.g., p-cu e analysis indica ed o e all high powe ; h ps://os .io/c6x4q). We chose
o eplica e he indings on h ee speci ic mode a o s (i.e., need o cogni ion, cogni i e load, and
o ewa ning) because all o hem could easily be measu ed o manipula ed, and he global COVID-
19 pandemic p e en ed us om using manipula ions ha equi ed pa icipan s o be obse ed in
he lab. Addi ional mode a o s consis o nodding e sus shaking one’s head, a m mo emen ( lex-
ion s. ension), inancial incen i es, and alcohol consump ion. Al hough we canno say whe he
ou null indings can be gene alized o hese mode a o s, we belie e ha gi en ou s udies’ highe
deg ee o anspa ency and an o e all la ge sample size, epo s o di e ences be ween mode a-
o s o expe imen e -p o ided and sel -gene a ed ancho s should gene ally be aken wi h a g ain
o sal . We also do no expec mode a o s o be associa ed wi h adjus men om expe imen e -
p o ided ancho s i he di ec ion o adjus men is known as he e is o e whelming e idence agains
he hypo hesis ha hese sco es a e mo e eliable (Rösele e al., 2022). To u he cla i y he ole
o po en ial mode a o s o ancho ing e ec s, we hink ha Simmons e al.’s (2010; mone a y in-
cen i es o accu acy) and Chaxel’s (2014; p iming selec i e accessibili y) indings on mode a o s
should be eplica ed, oo. Fo example, he o me add o o he indings ha mone a y incen i es
dec ease ancho ing (e.g., Epley & Gilo ich, 2005; LeBoeu & Sha i , 2009; Meub e al., 2013) bu
a e inconsis en wi h null indings (e.g., Enke e al., 2021; Li e al., 2021; Wilson e al., 1996).
The eby, we encou age adhe ence o anspa en ly desc ibed expe imen al p ocedu es such as ou s
o ha o o he esea che s (e.g., Cheek & No em, 2022; Maye & Rebholz, 2024) o minimize
a ia ion and o bewa e o compa ibili y o incen i es (He wig & O mann, 2001). O e all, he e
is ha dly any doub ha ancho ing e ec s a e la ge and obus . Bu apa om his gene al inding,
ancho ing esea ch is no excep ion o esea ch ha has p oblems wi h eplicabili y (e.g., Bahník,
2021; Ha is e al., 2019; Rösele e al., 2020; Rösele e al., 2021; Rösele e al., 2019; Shanks e
al., 2020). We ecommend ha ancho ing esea che s pu g ea e emphasis on he eplicabili y o
p e ious indings as his could ha e sa ed us 13 yea s o esea ching and deba ing.
CRediT Au ho S a emen
• Lukas Rösele : Concep ualiza ion, Da a cu a ion, Fo mal analysis, Funding acquisi ion, In es-
iga ion, Me hodology, P ojec adminis a ion, So wa e, Supe ision, Valida ion, Visualiza ion,
W i ing – o iginal d a , W i ing, e iew and edi ing
• Hannah L. Bögle , Lisa Koßmann, Sabine M. K uege : Fo mal analysis, In es iga ion, Me hod-
ology, So wa e, W i ing – o iginal d a , W i ing – e iew and edi ing
• Sab ina L. C. Bickenbach, Rica da Bühle , Jasmin della Gua dia, Lisa-Ma ie A. Köppel, Ja l
Möh ing, Susanne Ponade , Kons an in Roßmaie , Jessica Sing: Fo mal analysis, In es iga ion,
Me hodology, So wa e, W i ing – e iew and edi ing
33
L. Rösele e al. – Need o Cogni ion, Cogni i e Load (Replica ion). JCRE (2024-8)
Acknowledgmen s
We hank Leonie Beue le, Lena Ecke , A a Fische , Maxi Gö ni z, Alexande Ji anek, Rebekka
Kelle , G e a Kick, Cla a Kö h, Rebekka K aus, Ka ha ina M. K owo sch, Lisa Lede e , Johanna
Popp, Emilia Sch amm, and Annka h in Zo bach o hei help in designing and conduc ing he
eplica ion s udies. We hank Jane Zago ski o language edi ing.
Re e ences
Auguie, B., An ono , A. & Auguie, M. B. (2017). “Package ‘g idEx a”’. Miscellaneous Func ions
o “G id” G aphics, 9. PDF: h ps://c an. -p ojec .o g/web/packages/g idEx a/g idEx a.pd .
Bahník, Š. (2021). “Ancho ing Does No Ac i a e Examples Associa ed wi h he Ancho Value”.
DOI: 10.31234/os .io/4j5wb.
Ba e , T., Dowle, M., S ini asan, A., Go ecki, J., Chi ico, M. & Hocking, T. (2024). "da a. able:
Ex ension o ‘da a. ame". URL: h ps://c an. -p ojec .o g/web/packages/da a. able/index.h ml.
Bless, H., Fellhaue , R. F., Bohne , G. & Schwa z, N. (1991). "Need o Cogni ion: eine Skala
zu E assung on Engagemen und F eude bei Denkau gaben", olume 1991/06 o ZUMA-
A bei sbe ich . Mannheim: Zen um ü Um agen, Me hoden und Analysen -ZUMA-. URL:
h ps://nbn- esol ing.o g/u n:nbn:de:0168-ssoa -68892.
B and , M. J., IJze man, H., Dijks e huis, A., Fa ach, F. J., Gelle , J., Gine -So olla, R., G ange,
J. A., Pe ugini, M., Spies, J. R. & an ’ Vee , A. (2014). “The Replica ion Recipe: Wha Makes
o a Con incing Replica ion?” Jou nal o Expe imen al Social Psychology, 50: 217–224. DOI:
10.1016/j.jesp.2013.10.005.
B uinsma, J. & C u zen, R. (2018). “A Longi udinal S udy on he S abili y o he Need o Cogni-
ion”. Pe sonali y and Indi idual Di e ences, 127: 151–161. DOI: 10.1016/j.paid.2018.02.001.
Cacioppo, J. T. & Pe y, R. E. (1982). “The Need o Cogni ion”. Jou nal o Pe sonali y and Social
Psychology, 42(1): 116–131. DOI: 10.1037/0022-3514.42.1.116.
Champely, S. (2020). "pw : Basic Func ions o Powe Analysis". URL: h ps://CRAN.R-
p ojec .o g/package=pw .
Chaxel, A.-S. (2014). “The Impac o P ocedu al P iming o Selec i e Accessibili y on Sel -
Gene a ed and Expe imen e -P o ided Ancho s”. Jou nal o Expe imen al Social Psychology, 50:
45–51. DOI: 10.1016/j.jesp.2013.09.005.
Cheek, N. N. & No em, J. K. (2022). “Indi idual Di e ences in Ancho ing Suscep ibili y: Ve bal
Reasoning, Au is ic Tendencies, and Na cissism”. Pe sonali y and Indi idual Di e ences, 184(184):
111212. DOI: 10.1016/j.paid.2021.111212.
Diedenho en, B. & Musch, J. (2015). “Coco : A Comp ehensi e Solu ion o he S a is ical Com-
pa ison o Co ela ions”. PLoS One, 10(4): e0121945. DOI: 10.1371/jou nal.pone.0121945.
D agulescu, A. & A end , C. (2020). "xlsx: Read, W i e, Fo ma Excel 2007 and Excel
97/2000/XP/2003 Files". URL: h ps://CRAN.R-p ojec .o g/package=xlsx.
34
Jou nal o Commen s and Replica ions in Economics - JCRE
Enke, B., Gneezy, U., Hall, B., Ma in, D., Nelido , V., O e man, T. & an de Ven, J. (2021).
“Cogni i e Biases: Mis akes o Missing S akes?” The Re iew o Economics and S a is ics, 105(4).
DOI: 10.1162/𝑟𝑒𝑠𝑡𝑎01093.
Epley, N. & Gilo ich, T. (2001). “Pu ing Adjus men Back in he Ancho ing and Adjus men
Heu is ic: Di e en ial P ocessing o Sel -Gene a ed and Expe imen e -P o ided Ancho s”. Psy-
chological Science, 12(5): 391–396. DOI: 10.1111/1467-9280.0037.
Epley, N. & Gilo ich, T. (2004). “A e Adjus men s Insu icien ?” Pe s Soc Psychol Bull, 30(4):
447–460. DOI: 10.1177/0146167203261889.
Epley, N. & Gilo ich, T. (2005). “When E o ul Thinking In luences Judgmen al Ancho ing: Di e -
en ial E ec s o Fo ewa ning and Incen i es on Sel -gene a ed and Ex e nally P o ided Ancho s”.
Jou nal o Beha io al Decision Making, 18(3): 199–212. DOI: 10.1002/bdm.495.
Epley, N. & Gilo ich, T. (2006). “The Ancho ing-and-adjus men Heu is ic: Why he Ad-
jus men s a e Insu icien ”. Psychological Science, 17(4): 311–318. DOI: 10.1111/j.1467-
9280.2006.01704.x.
Epley, N. & Gilo ich, T. (2010). “Ancho ing Unbound”. Jou nal o Consume Psychology, 20(1):
20–24. DOI: 10.1016/j.jcps.2009.12.005.
F ede ick, S. W. & Mochon, D. (2012). “A Scale Dis o ion Theo y o Ancho ing”. Jou nal o
Expe imen al Psychology: Gene al, 141(1): 124–133. DOI: 10.1037/a0024006.
Goh, J. X., Hall, J. A. & Rosen hal, R. (2016). “Mini Me a-analysis o You Own S udies: Some
A gumen s on Why and a P ime on How”. Social and Pe sonali y Psychology Compass, 10(10):
535–549. DOI: 10.1111/spc3.12267.
G olemund, G. & Wickham, H. (2011). “Da es and Times Made Easy wi h Lub ida e”. Jou nal o
s a is ical so wa e, 40: 1–25. DOI: 10.18637/jss. 040.i03.
Ha is, A. J. L., Blowe , F. B. N., Rodge s, S. A., Laga o , S., Page, E., Bu on, A., U lichich,
D. & Speekenb ink, M. (2019). “Failu es o Replica e a Key Resul o he Selec i e Accessibili y
Theo y o Ancho ing”. Jou nal o Expe imen al Psychology: Gene al, 148(9): e30–e50. DOI:
10.1037/xge0000644.
Hedge, C., Powell, G. & Sumne , P. (2018). “The Reliabili y Pa adox: Why Robus Cogni i e Tasks
Do No P oduce Reliable Indi idual Di e ences”. Beha io Resea ch Me hods, 50(3): 1166–1186.
DOI: 10.3758/s13428-017-0935-1.
He wig, R. & O mann, A. (2001). “Expe imen al P ac ices in Economics: a Me hodolog-
ical Challenge o Psychologis s?” Beha io al and B ain Sciences, 24(3): 383–403. DOI:
10.1017/S0140525X01004149.
Jacowi z, K. E. & Kahneman, D. (1995). “Measu es o Ancho ing in Es ima ion Tasks”. Pe sonali y
and Social Psychology Bulle in, 21(11): 1161–1166. DOI: 10.1177/01461672952111004.
Kelley, K. (2023). MBESS: The MBESS R Package. URL h ps://CRAN.R-p ojec .o g/pa
ckage=MBESS. R package e sion 4.9.3.
35
L. Rösele e al. – Need o Cogni ion, Cogni i e Load (Replica ion). JCRE (2024-8)
Kuzne so a, A., B ockho , P. B. & Ch is ensen, R. H. B. (2017). “Lme Tes Package: Tes s in Lin-
ea Mixed E ec s Models”. Jou nal o S a is ical So wa e, 82(13). DOI: 10.18637/jss. 082.i13.
LeBel, E. P., Vanpaemel, W., Cheung, I. & Campbell, L. (2019). “A B ie Guide o E alua e
Replica ions”. Me a-Psychology, 3. DOI: 10.15626/MP.2018.843.
LeBoeu , R. A. & Sha i , E. (2009). “Ancho ing on he "He e" and "Now" in Time and Dis ance
Judgmen s”. Jou nal o Expe imen al Psychology: Lea ning, Memo y, and Cogni ion, 35(1): 81–
93. DOI: 10.1037/a0013665.
Leine , D. J. (2019). “SoSci Su ey ( e sion 3.1. 06)[compu e so wa e]”. URL:
h ps://www.soscisu ey.de/.
Li, L., Maniadis, Z. & Sedikides, C. (2021). “Ancho ing in Economics: A Me a-analysis o S udies
on Willingness- o-pay and Willingness- o-accep ”. Jou nal o Beha io al and Expe imen al Eco-
nomics, 90(101629). DOI: 10.1016/j.socec.2020.101629.
Maye , M. & Rebholz, T. R. (2024). “Na iga ing Ancho Rele ance Skill ully: Expe ise Reduces
Suscep ibili y o Ancho ing E ec s”. URL: h ps://os .io/p ep in s/psya xi /69jw .
Meub, L., P oege , T. & Bize , K. (2013). “Ancho ing: A Valid Explana ion o Biased Fo ecas s
when Ra ional P edic ions a e Easily Accessible and Well Incen i ized?” cege Discussion Pape s
No 166. URL: h ps://www.econs o .eu/bi s eam/10419/78599/1/756663717.pd .
Mussweile , T. & S ack, F. (1999a). “Compa ing is Belie ing: A Selec i e Accessibili y Model
o Judgmen al Ancho ing”. Eu opean Re iew o Social Psychology, 10(1): 135–167. DOI:
10.1080/14792779943000044.
Mussweile , T. & S ack, F. (1999b). “Con amina ion & Co ec ion o Social Judgemen s: S a e-
gies o Co ec ion Re isi ed: Theo y-based Adjus men e sus Recompu a ion”.
Pea son, K. & Filon, L. N. G. (1898). “VII. Ma hema ical Con ibu ions o he Theo y o E o-
lu ion.— IV. On he P obable E o s o F equency Cons an s and on he In luence o Random
Selec ion on Va ia ion and Co ela ion”. Philosophical T ansac ions o he Royal Socie y o Lon-
don Se ies A, Con aining Pape s o a Ma hema ical o Physical Cha ac e , 191(0): 229–311. DOI:
10.1098/ s a.1898.0007.
Re elle, W. (2024). psych: P ocedu es o Psychological, Psychome ic, and Pe -
sonali y Resea ch. No hwes e n Uni e si y, E ans on, Illinois. URL: h ps://c an. -
p ojec .o g/web/packages/psych/index.h ml.
Rösele , L. (2021). Ancho ing E ec s: Resol ing he Con adic ions o Pe son-
ali y Mode a o Resea ch. Ph.D. hesis, Hochschule Ha z. PDF: h ps:// is.uni-
bambe g.de/se e /api/co e/bi s eams/6b8bc0b9-31e8-41e1-a052-4982dc18ea66/con en .
Rösele , L. & Schü z, A. (2022). “Hanging he Ancho O a New Ship: A Me a-Analysis o Ancho -
ing E ec s”. URL: h ps://os .io/p ep in s/psya xi /w 2 n.
Rösele , L., Schü z, A., Baumeis e , R. F. & S a ke , U. (2020). “Does Ego Deple ion Reduce Judg-
men Adjus men o Bo h In e nally and Ex e nally Gene a ed Ancho s?” Jou nal o Expe imen al
Social Psychology, 87(103942). DOI: 10.1016/j.jesp.2019.103942.
36
Jou nal o Commen s and Replica ions in Economics - JCRE
Rösele , L., Schü z, A. & S a ke , U. (2019). “Cogni i e Abili y Does No and Canno Co ela e
wi h Suscep ibili y o Ancho ing E ec s”. URL: h ps://os .io/p ep in s/psya xi /bnsx2.
Rösele , L., Webe , L., Helge h, K., S ich, E., Gün he , M., Wagne , F.-S. & Schü z, A. (2022).
“Measu emen s o Suscep ibili y o Ancho ing a e Un eliable: Me a-analy ic e idence om mo e
han 50,000 ancho ed es ima es”. URL: h ps://os .io/p ep in s/psya xi /b6 35.
Rösele , L., Schü z, A., Blank, P. A., Dück, M., Fels, S., Kup e , J., Scheelje, L. & Seida, C.
(2021). “E idence agains Subliminal Ancho ing: Two close, Highly Powe ed, P e egis e ed,
and Failed Replica ion A emp s”. Jou nal o Expe imen al Social Psychology, 92(104066). DOI:
10.1016/j.jesp.2020.104066.
Schindle , S., Que engässe , J., B uchmann, M., Bögemann, N. J., Moeck, R. & S aube, T.
(2021). “Bayes Fac o s Show E idence Agains Sys ema ic Rela ionships Be ween he Ancho ing
E ec and he Big Fi e Pe sonali y T ai s”. Scien i ic Repo s, 11(1): 7021. DOI: 10.1038/s41598-
021-86429-2.
Shanks, D. R., Ba bie i-He mi e, P. & Vadillo, M. A. (2020). “Do Inciden al En i onmen al
Ancho s Bias Consume s’ P ice Es ima ions?” Collab a Psychology, 6(1): 19. DOI: 10.1525/col-
lab a.310.
Simmons, J., Nelson, L. & Simonsohn, U. (2012). “A 21 Wo d Solu ion”. SSRN Elec on Jou nal.
DOI: 10.2139/ss n.2160588.
Simmons, J. P., LeBoeu , R. A. & Nelson, L. D. (2010). “The E ec o Accu acy Mo i a ion on
Ancho ing and Adjus men : Do People Adjus om P o ided Ancho s?” Jou nal o Pe sonali y
and Social Psychology, 99(6): 917–932. DOI: 10.1037/a0021540.
Simonsohn, U. (2015). “Small elescopes: De ec abili y and he E alua ion o Replica ion Resul s”.
Psychological Science, 26(5): 559–569. DOI: 10.1177/0956797614567341.
Team, R. C. & o he s" (2023). R Co e Team: R Founda ion o S a is ical Compu ing. URL:
T e sky, A. & Kahneman, D. (1974). “Judgmen unde Unce ain y: Heu is ics and Biases”. Sci-
ence, 185(4157): 1124–1131. DOI: 10.1126/science.185.4157.1124.
Viech baue , W. (2010). “Conduc ing Me a-Analyses in R wi h he me a o Package”. Jou nal o
S a is ical So wa e, 36(3). DOI: 10.18637/jss. 036.i03.
Wickham, H. (2007). “Reshaping Da a wi h he eshape Package”. Jou nal o S a is ical So wa e,
21(12). DOI: 10.18637/jss. 021.i12.
Wickham, H. (2016). ggplo 2: Elegan G aphics o Da a Analysis. Sp inge -Ve lag New Yo k.
ISBN 978-3-319-24277-4. URL: h ps://ggplo 2-book.o g/.
Wickham, H., F ançois, R., Hen y, L., Mülle , K. & Vaughan, D. (2023). dply : A G amma o
Da a Manipula ion. URL: h ps://CRAN.R-p ojec .o g/package=dply .
Wilson, T. D., Hous on, C. E., E ling, K. M. & B ekke, N. (1996). “A New Look a Ancho ing
E ec s: Basic Ancho ing and i s An eceden s”. Jou nal o Expe imen al Psychology: Gene al,
125(4): 387–402. DOI: 10.1037/0096-3445.125.4.387.
37
L. Rösele e al. – Need o Cogni ion, Cogni i e Load (Replica ion). JCRE (2024-8)
Yoon, H., Scopelli i, I. & Mo ewedge, C. K. (2021). “Decision Making Can Be Imp o ed Th ough
Obse a ional Lea ning”. O ganiza ional Beha io and Human Decision P ocesses, 162: 155–188.
DOI: 10.1016/j.obhdp.2020.10.011.
38
Jou nal o Commen s and Replica ions in Economics - JCRE
Appendix
S udy 1: Explo a o y Tes s
To allow o compa isons be ween adjus men s away om di e en ancho s o di e en i ems,
we compu ed s anda dized adjus men sco es by di iding he di e ence be ween es ima e and
ancho by he di e ence be ween ue alue and ancho . Sco es o all ques ions and hei
co ela ions wi h cogni i e load along wi h co ela ions o absolu e adjus men and cogni i e
load a e displayed in Figu e 8 and Table 5.
Mo eo e , we es ed whe he o e ing he inclusion o pa icipan s who we e ec ui ed wi h
non-mone a y incen i es o he pa icipan s a ec ed he esul s. Due o ha ing oo ew pa ic-
ipan s, we p o ided la e olun ee s wi h eedback abou hei suscep ibili y o ancho ing and
he accu acy o hei es ima es (non-mone a y incen i e). The need o cogni ion sco es in he
sample ha ecei ed incen i es was signi ican ly highe , 𝑡(153.42)=2.27,𝑝=.024 ( wo- ailed;
𝑀no incen i e =77.03,𝑆𝐷no incen i e =13.66,𝑁no incen i e =80;𝑀incen i e =81.19,𝑆𝐷incen i e =15.14,
𝑁incen i e =223). Suscep ibili y o ancho ing did no di e be ween he samples (bo h 𝑝 > .201).
As bo h o he suscep ibili y o ancho ing sco es we e un eliable, hei co ela ion was e y
low, oo, 𝑟(301)=.060,𝑝=.298 ( wo- ailed). The co ela ions be ween suscep ibili y o he
expe imen e -p o ided ancho ing i ems and need o cogni ion and suscep ibili y o he sel -
gene a ed-ancho ing i ems and need o cogni ion did no di e , 𝑧=0.78,𝑝=.435 ( wo- ailed;
Pea son & Filon, 1898). Fu he mo e, nei he ype o ancho ing was ela ed o whe he people
knew o hough abou ancho ing e ec s du ing he expe imen (bo h 𝑝 > .274). De ailed esul s
can be ob ained ia he analysis sc ip (h ps://os .io/wzhmy/).
As no exclusion c i e ia we e epo ed in he o iginal s udy, we e- an ou analyses wi hou
excluding pa icipan s. S ill, despi e le e age poin s, co ela ions be ween need o cogni ion and
adjus men om sel -gene a ed ancho s (𝑟[301]=−.100,𝑝=.042, one- ailed) and adjus men
om expe imen e -p o ided ancho s (𝑟[301]=−.011,𝑝=.846) we e close o ze o. No e ha
excluding he middle quin iles ob iously did no ende he e ec s signi ican , ei he (see
analysis sc ip o exac esul s, h ps://os .io/wzhmy/).
Due o he gende imbalance, we es ed whe he gende a ec ed he ela ionship be ween
NFC and ancho ing. Co ela ions be ween adjus men om sel -gene a ed ancho s we e
𝑟 emale(208)=.047, 95% CI [−.089, .181]and 𝑟male(91)=.056, 95% CI [−.149, .257]. This is in line
wi h me a-analy ical indings (e.g., Rösele & Schü z, 2022, Table 2, p. 21).
Finally, we di ided he di e ence be ween ancho and es ima e by he di e ence be ween an-
cho and ue alue pe pe son and pe i em. This addi ional p ocedu e leads o an al eady
s anda dized 0–1 sco e ins ead o an absolu e adjus men sco e ha has o be 𝑧- ans o med be-
o e agg ega ion (e.g., Yoon e al., 2021, p. 14). 0–1 sco es a e 0 i he ue alue was es ima ed
and 1 i he ancho was es ima ed. The esul s his p ocedu e yielded did no di e om hose
ha we ob ained wi h he p e egis e ed es s.
39
L. Rösele e al. – Need o Cogni ion, Cogni i e Load (Replica ion). JCRE (2024-8)
S udy 2: Explo a o y Tes s
Fo explo a o y pu poses, we included a ques ion abou whe he pa icipan s p e en ed hem-
sel es om expe iencing cogni i e load on he le e s ask, o example, by w i ing down he
le e s ings ins ead o memo izing hem. A o al o 11 pa icipan s indica ed ha hey had
done so. Excluding hese pa icipan s did no a ec he esul s, as he in e ac ion was s ill no
signi ican , 𝐹(1,296)=1.82,𝑝=.179.
To assess he cogni i e load manipula ion, we analyzed he memo y ask. Pa icipan s co ec ly
ep oduced 43% o all le e s ings (𝑆𝐷 =0.30%,𝑁=93). Pe o mance declined sligh ly o e
he 18 i ems, 𝑟(16)=−.484,𝑝=.042, and was eliable, 𝛼=.91. Pa icipan s’ adjus men sco es
we e no co ela ed wi h he numbe o memo ized le e s ings (𝑟sg [91]=−.176,𝑟ep [91]=.067).
To es i he e ec s we e dependen on ou exclusion c i e ia, we es ed hem again bu wi hou
applying ou exclusion c i e ia. This led o he inclusion o 3 mo e pa icipan s and a o al sample
size o 𝑁=186. Cogni i e load was s ill no associa ed wi h adjus men om expe imen e -
p o ided ancho s, 𝑑=−.105, 95% CI [−0.393,0.183], o adjus men om sel -gene a ed ancho s,
𝑑=0.060, 95% CI [−0.228,0.347].
As in S udy 1, we compu ed adjus men sco es by di iding he di e ence be ween he es ima e
and ancho by he di e ence be ween he ue alue and ancho . These 0–1 sco es o all ques-
ions and hei co ela ions wi h cogni i e load along wi h he co ela ions be ween absolu e
adjus men and cogni i e load a e p esen ed in Figu e 8 and Table 5.
S udy 3: Explo a o y Tes s
To es whe he ou analyses we e dis o ed by poo ly unc ioning ancho ing i ems, we epea ed
ou es s using only he i ems wi h signi ican ancho ing e ec s. Being o ewa ned s ill did no
signi ican ly inc ease adjus men o he sel -gene a ed ancho ing i ems (𝑀no- o ewa ning =−0.07,
𝑀 o ewa ning =0.03,𝑆𝐷no- o ewa ning =0.74,𝑆𝐷 o ewa ning =0.65,𝑡(151.13)=−0.98,𝑝=.164,
𝑑=0.155, 95% CI [−0.150,0.459]).
Some pa icipan s al eady knew abou ancho ing e ec s, which migh ha e had a simila
e ec as he o ewa ning in oduc ion in ou s udy. The e o e, we combined he people
who we e o ewa ned o al eady knew abou ancho ing e ec s and compa ed hem wi h
he o he pa icipan s. The e was s ill no signi ican di e ence in he adjus men s o he
expe imen e -p o ided (𝑀no- o ewa ning =−0.07,𝑀 o ewa ning =0.04,𝑆𝐷no- o ewa ning =0.42,
𝑆𝐷 o ewa ning =0.53,𝑁no- o ewa ning =54,𝑁 o ewa ning =117,𝑡(126.16)=−1.409,𝑝=.081,𝑑=0.153,
95% CI [−0.149,0.455]) o o he sel -gene a ed ancho ing i ems (𝑀no- o ewa ning =−0.04,
𝑀 o ewa ning =−0.01,𝑆𝐷no- o ewa ning =0.50,𝑆𝐷 o ewa ning =0.53,𝑁no- o ewa ning =54,
𝑁 o ewa ning =117,𝑡(108.28)=−0.280,𝑝=.390,𝑑=−0.120, 95% CI [−0.182,0.421]).
Because ou sample size exceeded ou a ge ed sample o 120 pa icipan s e en a e we
emo ed pa icipan s on he basis o he exclusion c i e ia, we e an he s a is ical analyses
wi h only he i s 120 pa icipan s o compa ison. The smalle sample yielded no ema kable
40
