Christine Dobbs, University of Wales Swansea

1
England vs. Wales Allocation-based ingroup bias
vs. evaluative ingroup bias

• Christine Dobbs, University of Wales Swansea
• Russell Spears, Cardiff University

STOP PRESS RELATIVE RESPECT
2
England vs. Wales Design

• 2 (National group English/Welsh) 3
  (Condition EU/GB/control) 2 (Target group
  ingroup/outgroup)

Great Britain
Control
Wales
England
Wales
England
Student sample Cardiff Uni, Swansea Uni, Swansea
Institute of Higher Education English sample
N 86, age M 20.29, SD 5.18, MF
1769 Welsh sample N 98, age M 23.71, SD
9.12, MF 2076
3
England vs. Wales IPM and conflicting models

• Ingroup Projection Model (Mummendey Wenzel,
  1999) If the ingroup perceives itself to be more
  relatively prototypical of the superordinate
  category than the outgroup, this may lead to a
  negative outgroup evaluation.

Great Britain
EU


England
Wales
England
Wales
4
England vs. Wales Research questions

• Four research questions
• 1. Does relative prototypicality mediate ingroup
  bias?
• Nope
• 2. Does relative power mediate ingroup bias?
• Nope
• Does relative prototypicality mediate relative
  respect?
• Nope
• Does relative power mediate relative respect?
• In one instance, yes!

5
Predictor variables 1. Trait-based relative
prototypicality 2. Proximity-based relative
prototypicality 3. Similarity-based relative
prototypicality 4. Relative power

• Outcome variables
• Evaluative ingroup bias
• Allocation-based ingroup bias
• Perceived relative ingroup respect

6
1. Trait-based relative prototypicality Ps rated
how typical 15 traits (5 x warmth, 5 x
competency, 5 x Katz Braly) were for the
ingroup, the outgroup and the superordinate
category (1 not at all to 7 very much so).
Euclidean distances between superordinate ?
ingroup and superordinate ? outgroup were
calculated. Dsup-ingroup ? Dsup-outgroup gave
measure of relative dissimilarity. Lower
scores, therefore, indicate higher trait-based
relative prototypicality.
7
2. Proximity-based relative prototypicality Ps
ranked the order in which four Venn-like diagrams
reflected the relationship between the ingroup
and the outgroup to the superordinate category (1
least best to 4 best).
Scenarios A B disregarded (no disparity between
in- and outgroup). Scenario C ? Scenario D
proximity-based relative prototypicality. -3
low to 3 high
8
3. Similarity-based relative prototypicality I
think the goals and values of the of the
following pairs are Ingroup ? Outgroup Ingroup
? Superordinate category Outgroup ? Superordinate
category 1 very dissimilar to 7 very
similar Mingroup?superordinate ?
Moutgroup?superordinate similarity-based
relative prototypicality (-6  low to 6  high).
9
Relative power Three ingroup and three outgroup
statements, e.g. England has too little impact
on matters in The EU (1 strongly disagree to 7
strongly agree). Collapsed ingroup score minus
collapsed outgroup score relative power
(-6  low to 6  high).
10
? Lower scores indicate higher relative
prototypicality Main effects and
interaction National Group F 6.07, p
lt.05 Condition n.s. 2-way interaction n.s.

11

Main effects and interaction National
Group F 126.47, p lt.001 Condition n.s. 2-way
interaction F 6.57, p lt.05


12
p .07

Main effects and interaction National Group F
30.50, p lt.001 Condition n.s. (p .07) 2-way
interaction F 8.04, p lt.01
p .08

13


Main effects and interaction National Group F
166.87, p lt.001 Condition n.s. 2-way
interaction F 9.55, p lt.01


14
(No Transcript)
15
1. Allocation-based ingroup bias Ps allocated a
grand total 15 million 15 million between the
ingroup and outgroup for three ostensible
projects, which had been organised by the
respective superordinate category, e.g. to fund
brownsite developments. Mingroup-score
Moutgroup-score delivered allocation-based
ingroup bias (-15  low to 15  high).
16
2. Evaluative ingroup bias Ps stated how
favourably they rated the ingroup, the outgroup
and the superordinate category on the Feelings
Thermometer (0  extremely unfavourable to
100  extremely favourable). Mattitude-ingroup
Mattitude-outgroup delivered the measure of
evaluative ingroup bias (-100  low to
100  high).
17
2. Evaluative ingroup bias An observation on
raw attitude scores The English show no sig.
differences in attitudes to ingroup, outgroup,
The EU and Great Britain. The Welsh rate ingroup
significantly higher than the outgroup and The
EU. Differences between the outgroup and The EU
non-significant. The Welsh rate ingroup
significantly higher than the outgroup and Great
Britain.
n.s.

EU GB
18
3. Relative respect Three ingroup and three
outgroup statements, e.g. The EU does not take
England seriously (1 strongly disagree to 7
strongly agree). Collapsed ingroup score minus
collapsed outgroup score relative respect
(-6  low to 6  high). For the purpose of
this study, perceived relative ingroup respect
measures the degree to which the ingroup feels
relatively discriminated against or not. Higher
scores indicate higher perceived respect, lower
scores, lower perceived respect.
19
Main effects and interaction National Group F
22.51, p lt.001 Condition n.s. 2-way
interaction n.s.

p .07
20
Main effects and interaction National Group F
22.12, p lt.001 Condition n.s. 2-way
interaction n.s.


21


Main effects and interaction National Group F
125.02, p lt.001 Condition n.s. 2-way
interaction F 5.58 p lt.001


22
Note a p .06.
23
England vs. Wales Prototypicality power -
ingroup bias respect

• Does relative prototypicality or relative power
• mediate ingroup bias or relative respect?
• A series of ANCOVAs performed separately for
• a) each condition (i.e. 2) and b) for each
  of the two ingroup bias measures for relative
  respect (i.e. 3 DVs) with c) one of the
  covariates (i.e. 4 possible MVs)
• DV BY National Group (IV) WITH Covariate

24
England vs. Wales Prototypicality, power and
INGROUP BIAS
EU Condition Evaluative ingroup bias No main
effect of trait-based relative prototypicality,
similarity-based relative prototypicality or
relative power (all Fs 1.21, all ps .22) Main
effect of proximity-based relative
prototypicality, F 3.27, p
.08. Allocation-based ingroup bias No main
effects of covariates (all Fs 1.09, all ps .30)
GB Condition Evaluative ingroup bias Ditto (all
Fs .67, all ps .42) Ditto, F 3.45, p
.07.Allocation-based ingroup bias Ditto (all
Fs 2.50, all ps .12)
25
England vs. Wales Prototypicality, power and
RELATIVE RESPECT
EU Condition No main effects of covariates (all
Fs 1.89, all ps .18) GB Condition No main
effect of trait-based relative prototypicality (p
.41). Main effect of Proximity-based
relative prototypicality F 3.64, p
.06 Similarity-based relative prototypicality F
9.59, p lt .01 Relative power F 37.25, p lt .001
26
England vs. Wales Prototypicality, power and
RELATIVE RESPECT
GB Condition Testing for mediation On
proximity- and similarity measures, the pathway
from IV to DV was initially significant (both ps
lt .001). After adding mediator, pathway remained
significant at .001 level. Relative power,
however, was a winner (well sort of!).
27
England vs. Wales POWER and RELATIVE RESPECT
After mediation path c -.16, n.s. (p
.17)Sobels t ?, p lt ?
After mediation path c -.19, p .07
28
England vs. Wales Summary

• Predictor variables
• Within groups, across all three prototypicality
  and relative power measures, English scores are
  higher in the GB condition than in the EU
  condition. Welsh scores are higher in The EU
  condition than in the GB condition.
• Between group differences are significant on the
  trait measure in the GB condition. Proximity,
  similarity and power measures are significantly
  different (or marginally in one instance)
  irrespective of condition.

29
England vs. Wales Summary

• Research Questions
• ? None of the prototypicality measures mediates
  ingroup bias.
• None of the prototypicality measures mediates
  relative respect.
• Relative power does not mediate ingroup bias.
• ? Relative power mediates relative respect in the
  GB condition.

30
England vs. Wales Summary

• Outcome variables
• In this sample, English participants show higher
  allocation-based ingroup bias than Welsh
  participants (EU p .07, GB p lt .05).
• Conversely, Welsh participants show significantly
  higher evaluative ingroup bias than English
  participants across both conditions.
• On perceived relative respect, both between and
  within group differences are significant. The
  English perceive highest respect and the Welsh
  lowest respect in the GB condition.

31
England vs. Wales Summary
Outcome variables The English are high in power,
are respected, and are deserving of more
resources, and this claim may be legitimate. The
Welsh are low in power, are not respected and
apparently may not claim higher allocations.
Therefore evaluative ingroup bias is a
compensatory measure. Legitimacy? How fair do
you personally find the amount of power that
Wales has as a nation? Etc. Relative fairness
32
England vs. Wales Summary
Legitimacy (Work in progress)
Main effects and interaction National Group
F 43.53, p lt.001 Condition n.s. 2-way
interaction n.s.


33
England vs. Wales Dual identification
Identification

• Ingroup identification
• Cameron (2004) measures, 1 low to 7 high.
• All three factors. Reliability good (?  .62).
  Collapsed.
• English M 4.61, SD .70
• Welsh M 5.17, SD .88 t 4.70, p lt
  .001

• Superordinate category identification
• Cameron (2004) measures, 1 low to 7 high.
• Factor centrality excluded. Reliability strong
  (?  .77). Collapsed.
• EUEnglish M 4.40, SD 1.00
• Welsh M 4.26, SD .91GBEnglish M
  5.16, SD .94
• Welsh M 4.97, SD 1.02

34
England vs. Wales Superordinate category
identification


Main effects and interaction National
Group n.s. Condition F 16.01, p lt.001
2-way interaction n.s.
However, no significant interactions between
ingroup superordinate category identification
and/or National Group Condition. Ingroup
identification therefore not a moderator of
ingroup bias or relative respect.