Overlapping Attributes and Stereotype Construal

1
Overlapping Attributes and Stereotype Construal
Dora Bernardes, Tom Postmes, Louise Pendry
Green peas in a pod may be very similar to each
other in relevant ways and very different ways
from (say) snow peas. Unlike members of a human
social groups though, the peas do not become more
similar and different, and stereotypes need to be
able to capture this dynamism. (McGarty et al.,
2002, p. 6)
2
Thanks

• Tom Postmes
• Michelle Ryan Greg Maio
• Russell Spears
• FCT ESRC

3
Stereotypes share attributes

• Both Canadians and Chinese are polite Physicists
  and Historians are boring
• Research Question Which situational factors
  influence the construal of stereotypes in terms
  of their shared (unique) attributes?
• Hypothesis The perceived relationship among
  groups is a key determinant of what aspects of
  stereotypes are used in the construals of the
  groups in question (overlapping/differentiating
  traits)

4
Stereotypes as products/means of differentiation

• Definition Stereotypes as an instantiation of
  categorization processes (Tajfel, 1969 1981
  Mackie, Hamilton, Susskind, Rosseli, 1996)
  Stereotyping is categorization at the intergroup
  level (Oakes, 1996)
• We can only tell an impression of a group by
  telling the difference between this group and
  some other group (e.g., meta-contrast)
• Categorical accentuation effect (Tajfel Wilkes,
  1963)
• Categorization is the process by which
  regularities among stimuli are crystallized so
  they can be recognized, remembered and responded
  to
• Stereotypes are generalizations about a social
  group that distinguish them from other groups
  (McCauley, Stitt, Segal, 1980)

5
Stereotyping _at_ intra-group level

• Function schemas (Fiske Taylor, 1991) that
  serve as heuristics reduce the need to elaborate
  individuated information (Fiske, 1993 Macrae,
  Hewstone, Grifiths, 1993 Macrae, Milne,
  Bodenhausen, 1994)
• Stereotype activation vs. application (Hilton
  von Hippel, 1996)

6
Stereotyping _at_ intra-group level

• Stereotypical knowledge is context-dependent (for
  review, see Blair, 2002)
• Preserve self-image (Fein Spencer, 1997
  Sinclair Kunda, 2001 Spencer, Fein, Wolfe,
  Fong, Dunn, 1998)
• Social motives (Lowery, Hardin, Sinclair, 2001
  Richeson Ambady, 2001 Sechrist Stangor,
  2001)
• Strategies to counter-stereotype (Galinsky
  Moskowitz, 2000 Gollwitzer Schall, 1998
  Kawakami, Moll, Hermsen, Russin, 2000)
• The availability of exemplars (Dasgaputa
  Greenwald, 2001)
• Visual contexts (Wittenbrink, Judd, Park,
  2001)
• Racial attitudes and social roles (Barden,
  Maddux, Brewer, 2004)

7
Stereotyping _at_ inter-group level

• Stereotypes as explanations (for review, see
  McGarty, Yzerbyt, Spears, 2002)
• Self-categorization theory (Oakes, Haslam,
  Turner, 1994 Turner et al., 1987)
• Stereotypes reflect social roles (Eagly
  Steffen, 1984 Eagly, 1987)
• Stereotype content model (Fiske, Cuddy, Glick,
  Xu, 2002)
• Stereotypes serve to maintain relations within
  societies by providing justifications for
  socio-political systems (Jost Banaji, 1994
  Kay, Jost, Young, 2005 Sidanius Pratto,
  1999)
• Stereotypes make sense for the co-occurrence of
  set of attributes (Wittenbrink, Gist, Hilton,
  1997 Wittenbrink, Hilton, Gist, 1998)
• Peoples naïve beliefs influence stereotype
  formation
• Estimates of the weather (Krueger Clement
  (1994)
• Entity vs. incremental theories (Chiu, Hong
  Dweck, 1997)
• Entitativity and group essences (Rogier
  Yzerbyt, 1999 Yzerbyt, Corneille, Estrada,
  2001)

8
Stereotyping _at_ inter-group level

• Nature of inter-group relations (Asch, 1952
  Sherif, 1967)
• SCT Social categories form to maximise meta-
  contrast (Turner, 1987)
• But, stereotypical content depends on the frame
  of reference (comparative context)
• Category prototypicality (e.g., Haslam, McGarty,
  Oakes, Turner, Onorato, 1999 Hogg, 1992
  Oakes, Haslam, Reynolds, 1999 Oakes, Haslam,
  Turner, 1998)
• Content What is typical of a category
  (Americans Haslam, Turner, Oakes, McGarty,
  Hayes, 1992 or Psychology students Doosje,
  Haslam, Spears, Oakes, Koomen, 1998 Scottish
  Hopkins, Regan, Abell, 1997) Self-stereotypes
  (Spears Manstead, 1989 Onorato Turner, 1996
  Turner Onorato, 1999)
• Perceived similarity of others with the self
  (Haslam Turner, 1992 1995)

9
Stereotyping _at_ inter-group level

• Inter-group context
• Inter-group context and perceptions of
  inter-group relationships (Alexander, Brewer,
  Livingston, 2005)
• Social judgment (Bless, Igou, Schwarz, Wankë,
  2000 Bless, Schwarz, Bodenhausen, Thiel, 2001)

10
So far

• Stereotypes accentuate between group-differences
  and within-group similarities
• Both intra and inter-group approaches to
  stereotyping have shown that stereotypical
  knowledge is not static (i.e., context-specific)
• The nature of stereotypes (McGarty, 1999)
• Relative enduring system of inter-related
  concepts that inform perceptions of certain
  groups (stereotypical knowledge p. 23)
• A specific representation of a particular group
  at a particular time (stereotypical depictions
  p. 23) stereotype construction (Spears, 2002, p.
  145)
• Stereotypes as relational constructs that allow
  us to differentiate groups
• Neglected the role of overlapping traits in
  shaping stereotype construal

11
Why the neglect of overlap?

• What makes a Stereotype? Two perspectives on
  perceptions of trait typicality
• Attribution hypothesis (Olson Fazio, 2001
  Staats Staats, 1958) (a group characteristic
  is) a characteristic which is present in the
  majority of the members of the group (51)
• Categorization hypothesis (e.g., Oakes, Haslam,
  Turner, 1994) a group characteristic which
  makes possible a distinction between two groups
  (Group A 40 and Goup B 20) (Zawadski, 1948,
  p. 135-136)
• Differentiating traits are more diagnostic
  (Murphy Medin, 1985) than overlapping traits

12
Why the neglect of overlap?

• Issues of measurement of stereotypes
• Explicit measures
• Checklist (Katz Braly, 1933, 1935) people had
  to assign traits from a list of 84 to 10
  different national and ethnic groups and then
  select the 5 traits that were most typical of
  those groups
• Percentages Brigham (1971) assumed that group
  stereotypes are represented by the percentage of
  group members who are thought to possess a given
  attribute
• Likert-type scales measuring the amount of a
  given attribute possessed by individual members
  of a group (e.g., Sigall Page, 1971)
• Differentiation scores/Diagnostic ratios
  (Distinction with other groups McCauley, Stitt,
  Segal, 1980 Yzerbyt Rocher, 2002 The
  difference between the sociability and rating for
  group A and B served as an index of
  differentiating on social competence p. 61)
• Self-report (Attributes that come to mind Ford
  Stangor, 1992 Judd Park, 1993)

13
Current research

• Previous research
• Stereotypes can be used to reflect and create
  inter-group relationships
• Stereotypes content is construed depending on
  the comparative context
• Neglected the overlap between stereotypes
• Current research
• Keeps the comparative context constant
• Still expects stereotyping content to be
  construed in terms of overlap or in terms of
  differences depending on nature of group
  relations
• Studies 1 2
• manipulate the relationship directly (two groups
  either in competition or cooperation)
• Study 3
• manipulates the relationship indirectly by
  manipulating group goals that imply positive or
  negative interdependence
• Dependent measures
• implicit and explicit measures attempting to
  capture overlap

14
Studies 1 2 Ad hoc groups

• Cover story
• Human evolution What factors made some
  societies fail and others do well?"
• Yanomami tribe River people and Forest people
  either in cooperation or in competition
• Procedure (Study 1)
• Article with manipulation of group relations
• Impression formation (10 behavioural
  descriptions for each group, of which 3 implied a
  unique attribute 5 implied attributes shared by
  both tribes and 2 consisted of characteristics of
  the tribes that were irrelevant for the
  stereotype)
• Stereotype content (implicit measure)
  Word-search task
• Stereotype content (explicit measure) Free
  recall (Study 1)
• Stereotype content (explicit measure) Venn
  diagram (Study 2)
• Relationship check
• Entitativity check

15
Studies 1 2 Ad hoc groups

• Cover story
• Human evolution What factors made some
  societies fail and others do well?"
• Yanomami tribe River people and Forest people
  either in cooperation or in competition
• Procedure (Study 2)
• Impression formation (10 behavioural
  descriptions for each group, of which 3 implied a
  unique attribute 5 implied attributes shared by
  both tribes and 2 consisted of characteristics of
  the tribes that were irrelevant for the
  stereotype)
• Article with manipulation of group relations
• Stereotype content (implicit measure)
  Word-search task
• Stereotype content (explicit measure) Free
  recall (Study 1)
• Stereotype content (explicit measure) Venn
  diagram (Study 2)
• Relationship check
• Entitativity check

16
Impression formation task
Uses complex reasoning strategies in adapting to
social context (Intelligent) Works hard in each
and every activity (Industrious)
Forest people
Is not disposed to friendliness towards
others (Unfriendly) Shows hard work and great
perseverance (Industrious)
River people
17
Stereotype content (implicit)Word-search task
18
Stereotype content implicitWord-search task
19
Stereotype content (explicit)Free Recall

• Please write down the maximum amount of
  information they could recall about each tribes
  profile.

20
Stereotype content (explicit)Venn Diagram
Participants are presented with a list of 24
characteristics (8 overlapping 16
differentiating 9 stereotypical of Forest people
and 7 stereotypical of River people) and asked to
place them in the Venn Diagram
River people Both
Forest people
Neither
21
Relationship check

• Rate the extent to which they perceived the
  relationship of the two tribes on four semantic
  differential scales (1-7), (Study 1 a 0.72
  Study 2 a 0.80)
• antagonistic/ complementary (recoded)
• related / unrelated
• competitive / cooperative (recoded)
• convergent / divergent

22
Entitativity check

• How do you perceive River and Forest people
  within the Yanomami?

• Study 1
• F(1, 47) .13, p .72 (Competition M 4.14,
  SD 1.32 Cooperation M 4.30, SD 1.68)
• Study 2
• F(1, 33) .15, p .69 (Competition M 4.31,
  SD 1.00 Cooperation M 4.19, SD .91)

23
Dependent measures and analysis

• Ratio
• Proportion of identification of differentiating
  traits number of differentiating traits
  identified/total differentiating traits
• Proportion of identification of overlapping
  traits number of overlapping traits
  identified/total overlapping traits
• Analysis
• 2 Groups relationship (competition vs.
  cooperation) X Attribute-type (overlapping vs.
  differentiating) mixed-model ANOVA with repeated
  measures on the last factor

24
Stereotype content implicit (Study 1
Word-search task)
Groups relationship X Attribute-type
interaction, F(1, 48) 4.99, p .03, ?2 .09.
25
Stereotype content implicit (Study 2
Word-search task)
Groups relationship X Attribute-type
interaction, F(1, 48) 4.99, p .03, ?2 .09.
26
Stereotype content explicit (Study 1 Free
Recall)
Groups relationship X Attribute-type
interaction, F(1, 48) 11.71, p .001, ?2
.20.
27
Stereotype content explicit (Study 2 Venn
Diagram)
Groups relationship X Attribute-type
interaction, F(1, 34) 3.89, p .057, ?2 .10.
28
Discussion studies 1 2

• Perceived group relations lead to different
  stereotypical construals (implicit and explicit
  levels)
• But, different patterns in terms of the measures
  used
• Free recall more differences in the competition
  condition
• ( differentiating traits)
• Venn diagram more differences in the cooperation
  condition ( overlapping traits)

29
Study 3 Real groups (History Physics
students)

• Pilot study
• Differentiating traits (History literary,
  assertive, relaxed Physics mathematical,
  scientific, geeky)
• Overlapping traits (committed, clever,
  determined)
• Manipulation of group relations The purpose of
  the superordinate category (University) as being
  concerned with
• Research (in which these two groups have
  different goals and thus are somewhat negatively
  interdependent) or
• Teaching (in which groups have similar goals and
  are somewhat positively interdependent).
• Dependent variables
• Stereotype content (Word-search, Venn diagram)
• Group bias (i.e., resource allocation task)
  expect no overall differences in the amount of
  resources but variation in terms of conditions

30
Stereotype content explicit (Study 3 Venn
Diagram)
Groups relationship X Attribute-type
interaction, F(1, 56) 5.41, p .024, ?2 .09
31
Group biases

• 2 Items
• Considering that the University X has 19
  bursaries for the present academic year. How do
  you think that these should be allocated between
  History and Physics?
• Considering that University X has 320000 extra
  to invest. How do you think that this money
  should be allocated between Physics and History?
• 2 Scores
• BIAS the amount of resources donated to each
  group/total resources available, -1 1, -1
  all given to Physics 1 all given to History
• DIFFERENTIATION previous in absolute terms, 0,
  1 0 equal resource allocations 1 unequal
  resource allocations
• Predictions
• No overall differences in terms of BIAS across
  conditions
• But expected systematic differences in terms of
  DIFFERENTIATION (individual differences in
  participants theories about each groups needs)

32
Group biases

• Score of Bias
• Allocations between groups were equal in both
  conditions (M for Negative interdependence
  -.03, SD 0.30 M for Positive interdependence
  .04, SD 0.10), F(1, 53) 1.30, ns.
• There was no between-condition difference for
  each donation separately, either, Fs lt 1.27, ns.
  
• Score of Differentiation
• A significant difference in group discrimination
  emerged between conditions F(1, 52) 5.82, p
  .02, ?2 .10.
• Participants discriminated more (unequal resource
  allocations) in the Negative interdependence
  condition than in the Positive interdependent
  condition (M .18, SD .24, M .08, SD .07,
  respectively).

33
Discussion study 3

• Again, the perceived nature of group relations
  affected the construals of stereotypes of real
  groups (at least in explicit construal)
• Consequences for group biases in resource
  allocation task
• Although no overall differences in terms of the
  total resources given to each group
• Participants differentiated between groups when
  allocating resources
• We can speculate that participants had different
  theories about which resources should be given to
  each group in each condition (e.g., compensate
  for weaknesses vs. maximize strengths)

34
Conclusions

• People construe different stereotypes of groups
  depending on the perceived relationship between
  them
• Differentiating and overlapping attributes in
  stereotypes both are used flexibly to construe
  stereotypes
• Pattern that occurred both explicitly and
  implicitly and using new measures of stereotype
  construal (Word-search task, Venn diagram)
• Construals of stereotypes resulted in group
  biases that reflected differentiation but not
  overall bias

35
Implications

• Conceptualizations of stereotypes stereotypical
  knowledge and stereotypical depictions
• Stereotypes as flexible and dynamic construals
  depending on group relations (differentiation and
  overlap among groups)
• Stereotyping can have positive outcomes, when
  overlapping attributes are made salient
• Measurement of stereotypes (from focus on
  differences to capture similarities)
• From third-parties to self-categories
• Reduction of inter-group bias
• In-group love vs. out-group hate