Based on “OSF p e- egis a ion” o m
Ti le: The E ec s o S abili y o Inequali y and Ing oup Iden i ica ion on he P inciple-
Implemen a ion Gap among Ad an aged G oups.
Desc ip ion:
This is an anonymous online expe imen al s udy among Whi e ci izens in he UK (ad an aged
g oup), ha will be conduc ed ia P oli ic as a ollow-up s udy o he OSF p e- egis e ed s udy
p ojec en i led: “The F agili y o P i ilege: How Pe cei ed (In)s abili y o Inequali y and he
Demands o he Disad an aged In luence Suppo Among he Ad an aged” (DOI:
h ps://doi.o g/10.17605/OSF.IO/UYV3K).
In his s udy we examine how s abili y o acial heal h ca e inequali y (s able s. uns able) a ec s
in en ions o suppo measu es o add ess hese inequali ies. Ou main amewo k is he
P inciple-Implemen a ion Gap (PI Gap) e e ing o he disc epancy be ween abs ac suppo o
equali y and esis ance o conc e e measu es aimed a achie ing i among he p i ileged, (Bobo,
1988; Jackman & C ane, 1986; Dixon e al., 2017). We also ake in o accoun he ole o Ing oup
Iden i ica ion as a mode a o based on explo a o y indings om ou a i s p e- egis e ed s udy
and esul s om p e ious esea ch showing ha eac ions om ad an aged g oups o inequali y
depend on he in e play be ween con ex ual ea u es a ound inequali y and ing oup iden i ica ion
(Teixei a e al.; 2023, Teixei a e al., 2022). Speci ically, we aim o explo e he impac o s abili y
o inequali y (Teixei a e al., 2023;Knigh & Meh a, 2017; Scheepe s e al., 2015) and ing oup
iden i ica ion on he wid h o he PI Gap among ad an aged g oup membe s
Pa icipan s will be whi e UK ci izens om P oli ic Academic. We will manipula e s abili y o
acial inequali y in he UK be ween-pa icipan s. Be o e he manipula ion we will measu e
ing oup iden i ica ion. Ou main dependen a iables will be a ious scales ope a ionalizing
mo e o less abs ac suppo o inequali y educ ion. These measu es a e aimed a imp o ing
measu emen o he P-I gap, due o sho comings (linked o in a iance and ceiling e ec s)
obse ed in ou i s s udy. We will also measu e some o he concep s in an explo a o y manne
(c . sec ion on dependen a iables).
Pa icipan s will be exposed o:
1) Desc ip ions o s able s uns able acial heal h inequali y (i.e., a ying mo e o less in
ime and space, depending on he condi ion).
2) Fic ious campaign ad oca ing o an ac o be app o ed by he UK go e nmen
p omo ing a se ies o policies aimed a dis ibu ing heal h ca e esou ces mo e equi ably
be ween whi e and black UK ci izens.
Hypo heses:
We expec an in e ac ion be ween s abili y o inequali y and ing oup iden i ica ion on he PI Gap.
Based on ou p e ious esul s and p e ious esea ch, we expec he e ec s o iden i ica ion on
he P-I gap o be s onge unde s able inequali y. In ou p e ious esea ch we had measu emen
issues (c . abo e) ega ding he P-I gap. Fu he mo e, p e ious esea ch on ad an aged g oup
membe s some imes shows e ec s o be s onge amongs high (Teixei a e al. 2020) and
some imes amongs low iden i ie s (Teixei a e al. 2023). Wi h high-iden i ie s some imes
showing highe le els o suppo o inequali y educ ion (Shuman e ., 2020) and some imes
lowe (Teixei a e al., 2020; compa ed o low-iden i ie s). The e o e, he p edic i e di ec ion o
he esul s emains en a i e.
Design Plan
S udy ype
Expe imen
Blinding
No blinding is in ol ed in his s udy.
S udy design
This is a be ween-subjec s design wi h one ac o including 2 le els:
S abili y o Inequali y: S able s. Uns able
Randomiza ion
We used he embedded da a ac ion on Qual ics, o andomize pa icipan s in one o 2 condi ions
(S able s. Uns able).
Sampling Plan
Regis a ion p io o c ea ion o da a
Da a collec ion p ocedu es
We will ocus on Whi e B i ish ci izens li ing in he UK ec ui ed h ough P oli ic Academic.
Pa icipan s will be paid £1,5 gi en hey ha e comple ed he su ey and all a en ion checks
co ec ly. I hey ail he comp ehension checks, hey will ge pa ial compensa ion (acco ding o
P oli ic guidelines). Those who ail he a en ion checks will be asked o e u n hei submissions
wi hou being compensa ed. Pa icipan s who comple e he en i e su ey bu do no gi e consen
a e he deb ie ing will be ully compensa ed bu hei da a will no be s o ed. Pa icipan s mus
be a leas 18 yea s o age. Da a collec ion will ake place on Oc obe 2025.
Sample size
Ou a ge sample size is 200 pa icipan s.
Sample size a ionale
We used he R package "Supe powe and online Shiny apps", de eloped by Ladkens & Cadwell,
which enables us o pe o m simula ion-based powe analysis o ANOVA designs o up o h ee
wi hin- o be ween-subjec ac o s. We based ou calcula ion on he co ela ions ound in he i s
s udy be ween suppo (implemen a ion) and ing oup iden i ica ion unde s able inequali y ( = -
.484) and uns able inequali y ( = -.054) unde edis ibu i e demands as hese we e he
condi ions o in e es . The analyses sugges ed 194 pa icipan s o achie e a powe o .90. We will
op o 200 pa icipan s o accoun o po en ial exclusions.
S opping ule
Once he 200 quo a o comple e esponses on P oli ic is eached, da a collec ion will be
au oma ically s opped.
Va iables
Manipula ed a iables
We will manipula e s abili y o inequali y using ic i ious ideos and he assis ance o a
con ede a e. Pa icipan s will wa ch a ideo highligh ing heal hca e dispa i ies be ween Whi e
and Black B i ish ci izens, wi h inequali ies p esen ed as ei he s able (consis en dispa i ies
be ween Black and Whi e B i ish ci izens ac oss ime and space) o uns able ( luc ua ion
dispa i ies be ween Black and Whi e B i ish ci izens ac oss ime and space).
A e he manipula ion, pa icipan s will wa ch a second ideo ( he same ac oss bo h condi ions)
showcasing in o ma ion abou a ic i ious bill. This bill p oposes edis ibu ions on heal hca e
budge s be ween Whi e and Black B i ish ci izens o achie e equali y, p esen ed by a con ede a e
ac ing as he ep esen a i e o a ic i ious o ganiza ion.
Measu ed a iables
Demog aphic Ques ions (gende , e hnici y, age)
Comp ehension checks: Two ques ions ega ding s abili y o inequali y (T ue/False) and wo
ques ions on he bill (asking o he name o he bill and i s pu pose) will be used.
Open Ques ion: (se ing as a quali y check) asking pa icipan s o sha e some o hei hough s
on bo h ideos hey (a summa y o he in o ma ion p esen ed and hei opinions abou i )
Manipula ion Checks:
1) Legi imacy o Inequali y: (“In you opinion, o wha ex en a e he heal hca e inequali ies
desc ibed in he ideos a e…jus i ied, 1= no a all, 7 = e y much)
2) S abili y o Inequali y: To wha ex en do you ag ee wi h he ollowing s a emen s (e.g. “The
heal hca e inequali ies be ween Whi e and Black people in he UK a e s able.”, -3 = disag ee,
3 = ag ee)
3) Redis ibu ions: A e he ollowing s a emen s abou he Redis ibu i e Heal hca e Ac ,
ue? (e.g. “The Redis ibu i e Heal hca e Ac sugges s mo ing NHS unds om whi e o
black communi ies., T ue o False)
Main ou come measu es:
1) P inciple suppo : Measu ing suppo on he gene al goals o he bill (e.g. “C ea ing a
Heal hca e sys em ha is ai and does no disc imina e agains Black indi iduals, 1= No a
p io i y, 7 = Top p io i y)
2) Suppo i e A i udes: measu ing gene al a i udes owa ds he bill (e.g. “I suppo he
Redis ibu i e Heal hca e Ac .”, -3 = disag ee, 3 = Disag ee)
2) Implemen a ion suppo : Measu ing suppo on speci ic policies add essed in he bill (e.g.
P io i ize he opening o new GP posi ions in municipali ies mainly composed o Black
esiden s.-3= disag ee, 3 = ag ee),
3) Suppo h ough Ac ion: measu ing willingness o ake ac ion in suppo o he bill (e.g. “Sign
a pe i ion suppo ing he p oposed Redis ibu i e Heal hca e Ac bill.”, -3 = no willing a all o
pa icipa e; -3 = e y much willing o pa icipa e).
O he measu es:
1) Ing oup Iden i ica ion scale (Leach e al.,2008) (e.g. “I eel commi ed o membe s o my
e hnic g oup”, 1= No a all, 7= Ve y much)
2) Poli ical o ien a ion: Using a scale om 1 (= le ) o 11(= igh )
3) Pe cei ed blame by he disad an aged will be measu ed wi h one i em (Teixei a e al., 2020)
(e.g. To wha ex en do you hink ha he ad oca es o he RHA blame Whi e people o he
acial inequali y in heal hca e?”, 1= No a all, 7 = Ve y much).
4) Backlash (Adap ed om Teixei a e al., 2020) measu ing willingness o ake ac ions agains
he bill ( e.g. “Join a demons a ion agains he “Redis ibu i e Heal hca e Ac ” campaign -3 =
no willing a all o pa icipa e; -3 = e y much willing o pa icipa e)
5) Resou ce Th ea (Adap ed om Teixei a e al., 2020): o measu e o wha ex en do
pa icipan s belie e ha , i app o ed, he Ac will ha e nega i e consequences o hei ing oup’s
esou ces (e.g. “Dec ease Whi e B i ish ci izens' chances o quali y heal hca e?” 1= no ; 7=
likely)
6) Mo al Image Th ea (Adap ed om Teixei a e al., 2020): To measu e o wha ex en
pa icipan s belie e (e.g. “The p oposed bill will make Whi e B i ish ci izens seem un ai o he
es o he wo ld”, 1= no likely, 7= likely) ha he campaign will damage hei ing oup’s image.
Indices
We will i s conduc explo a o y ac o analyses on he di e en suppo scales. These esul s
will guide he c ea ion o mean-base indices ep esen ing he P-I gap.
Fo he o he dependen a iables, we will check eliabili y using C onbach’s alpha and c ea e
mean-sco es o he i ems measu ing he same concep .
Analysis Plan
S a is ical models
We will use Gene al Linea Model Analysis (GLM) o es he in e ac ion be ween s abili y o
inequali y and ing oup iden i ica ion on he PI-Gap.
I ound, in e ac ions will be decomposed looking a bo h 1) e ec s o iden i ica ion wi hin each
condi ion (using eg ession, see below o ans o ma ions) and 2) looking a e ec s o s abili y
a +1SD and -1SD o iden i ica ion.
No iles selec ed
T ans o ma ions
T ans o ma ions
The s abili y a iable will be coded as -1 and 1.
Ing oup iden i ica ion will be cen e ed.
I we use eg ession models o es explo a o y hypo heses, he P-I gap will be calcula ed by sub-
ac ing “implemen a ion” om “p inciple” suppo .
In e ence c i e ia
We will use he s anda d p <. 05 c i e ia ( wo- ailed) o de e mining analyses sugges ha he
esul s a e signi ican ly di e en om hose expec ed i he null hypo hesis we e co ec .
Da a exclusion
Ou lie s p esen ing s uden ized esiduals > = o 3 in ou main dependen a iables will be
excluded.
Missing da a
Pa icipan s who did no comple e any o he suppo scales, will no be included in he analysis.
Explo a o y analysis
I poli ical o ien a ion is co ela ed wi h ing oup iden i ica ion, we will e- un he main analyses
con olling o poli ical o ien a ion based on he s eps sugges ed by Yze by e al., 2005. We will
also explo e he e ec s on pe cei ed blame and h ea s using he same p edic o s as in he main
analyses.
