lecture9.prejudice

advertisement
Lecture Outline
Extra Credit experiment
Stereotypes defined
Diagnostic ratio revisited
Origins of stereotypes
Models of stereotype change/maintenance
Prejudice defined
1
Ashmore & Del Boca (1981)
A stereotypes is…...
“A set of beliefs about the
personal attributes of a
group of people”
2
Ashmore & Del Boca (1981)
Limitation:
Many attributes are perceived as
typical of a group and yet are not
part of people’s stereotypes
3
Stereotypes include attributes that are
perceived as:
TYPICAL
and
DISTINGUISH
BETWEEN GROUPS
4
Diagnostic Ratio
DR = % of group (with attribute)
% of reference (with attribute)
5
Diagnostic Ratio
When DR = 1 (or close to 1),
attribute does not distinguish
between groups
attribute not part of stereotype
6
Diagnostic Ratio
When DR substantially > than 1
attribute distinguishes between
groups
attribute is stereotypic of group
7
Diagnostic Ratio
When DR substantially < than 1
attribute distinguishes between
groups
attribute is counterstereotypic of
group
8
McCauley & Stitt (1978)
Purpose:
1. Show utility of DR
2. Measure (in)accuracy of
stereotype about African Americans
9
McCauley & Stitt (1978)
Sampled five groups
Created DR’s based on perceptions
of African Americans and Americans
Created criterion DR’s based on
census information
10
Results: McCauley & Stitt (1978)
Attribute
Criteria HS College Union
HS
.65
(.68)
(.73)
(.67)
Illegitimate 3.10 (1.80) (1.70) (2.10)
Unemployed 1.90 (1.90) (1.60) (1.80)
Victims
1.50
.83 (1.80)
(2.00)
Welfare
4.60 (2.30) (1.90) (1.60)
Kids
1.90 (1.60) (1.40)
1.60
Female head 2.80 (1.70) (1.90)
(1.70)
Choir
(.68)
(1.90)
(2.60)
1.50
(1.80)
(1.30)
(1.50)
SW
(.60)
(2.30)
(2.30)
2.30
1.40
(1.30)
(1.70)
(Green) DR’s: different from 1 (p < .05); n = 30
Black DR’s not different from 1 (p > .05); n = 5
Underlined DR’s = different from criteria (p < .05); n = 16 11 76
Results: McCauley & Stitt (1978)
Attribute
Criteria HS College Union Choir SW
HS
.65
(.68) (.73)
(.67)
(.68) (.60)
Illegitimate 3.10 (1.80) (1.70)
(2.10) (1.90) (2.30)
Unemployed 1.90 (1.90) (1.60)
(1.80) (2.60) (2.30)
Victims
1.50
.83 (1.80) (2.00) 1.50 2.30
Welfare
4.60 (2.30) (1.90) (1.60) (1.80) 1.40
Kids
1.90 (1.60) (1.40)
1.60
(1.30) (1.30)
Female head 2.80 (1.70) (1.90) (1.70) (1.50) (1.70)
Most DR’s different from one (green):
People held stereotype of African Americans
12 77
Results: McCauley & Stitt (1978)
Attribute
Criteria HS College Union Choir SW
HS
.65
(.68) (.73)
(.67)
(.68) (.60)
Illegitimate 3.10 (1.80) (1.70)
(2.10) (1.90) (2.30)
Unemployed 1.90 (1.90) (1.60)
(1.80) (2.60) (2.30)
Victims
1.50
.83 (1.80) (2.00) 1.50 2.30
Welfare
4.60 (2.30) (1.90) (1.60) (1.80) 1.40
Kids
1.90 (1.60) (1.40)
1.60
(1.30) (1.30)
Female head 2.80 (1.70) (1.90) (1.70) (1.50) (1.70)
Some DR’s different from criteria [underlined]. Other DR’s
not different from criteria [not underlined]: People’s
stereotypes were both inaccurate [underlined] and accurate
13
[not underlined]
78
Results: McCauley & Stitt (1978)
Attribute
Criteria HS College Union Choir SW
HS
.65
(.68) (.73)
(.67)
(.68) (.60)
Illegitimate 3.10 (1.80) (1.70)
(2.10) (1.90) (2.30)
Unemployed 1.90 (1.90) (1.60)
(1.80) (2.60) (2.30)
Victims
1.50
.83 (1.80) (2.00) 1.50 2.30
Welfare
4.60 (2.30) (1.90) (1.60) (1.80) 1.40
Kids
1.90 (1.60) (1.40)
1.60
(1.30) (1.30)
Female head 2.80 (1.70) (1.90) (1.70) (1.50) (1.70)
When DR’s indicated inaccurate stereotype [underlined],
difference was smaller than criteria: People’s stereotypes
underestimated real differences. They did not exaggerate
real differences
14
79
McCauley & Stitt (1978): Summary
People endorsed a stereotype of AA
most DR’s different than 1
AA stereotype was accurate & inaccurate
some DR’s different from criteria (inaccurate)
other DR’s not different from criteria (accurate)
AA stereotype underestimated real difference
when DR different from criteria, it was
smaller
15
Origin of Stereotypes:
Where do they come from?
Socio-cultural perspective
Kernel of Truth hypothesis
Illusory correlations
16
Socio-Cultural Perspective
Premise: Individuals are
socialized into a particular
culture
(e.g., media or significant others)
17
Socio-Cultural Perspective
1. People are born into a culture
2. People are rewarded/punished for their
beliefs, values, behaviors
3. People act in accord with norms
4. People internalize norms
5. Internalization perpetuates the norms
18
Socio-Cultural Perspective
Two versions of socio-cultural view
Structuralist-Functionalist
Conflict theory
19
Structuralist-Functionalist
Version
A single culture accepted throughout
a society
i.e., individuals in a society are similar
in their beliefs, values and behaviors
20
Structuralist-Functionalist
Version
Function of stereotypes:
stereotypes communicate expected
behavior
stereotypes communicate how
different people should be treated
21
Structuralist-Functionalist
Version
More evident in more homogeneous
and collectivist societies
22
Conflict Theory Version
Multiple subcultures within society
People accept norms of their subculture
23
Conflict Theory Version
People within a subculture are similar in
their beliefs, values, behaviors
People in different subcultures are
different in their beliefs, values,
behaviors
The more different two subcultures, the
greater the conflict in their beliefs,
24
values, behaviors
Conflict Theory Version
Function of stereotypes:
stereotypes justify prejudice
incompetence justifies lower pay
laziness justifies poverty
25
Conflict Theory Version
More evident in more
heterogeneous societies
26
Kernel of Truth Hypothesis
Premise: Stereotypes are
exaggerations that exist in
some measure in a group
27
Kernel of Truth Hypothesis
1. The larger a real difference between
groups, the more likely the attribute will
be in the stereotype
Example: Circumscribing and noncircumscribing tribes
28
Kernel of Truth Hypothesis
2. Stereotypes become more accurate as
contact between groups increases
Example: women/men v.s. African
Americans/Whites
29
Kernel of Truth Hypothesis
3. Behaviors punished in one group, but
not in another, tend to be in a stereotype
Example: nudity and bathroom practices
30
Kernel of Truth Hypothesis
4. Similar behaviors that groups perform in
different situations tend to be in
stereotypes, but connote different
valences.
Example……...
31
Kernel of Truth Hypothesis
We are loyal.
They are clannish.
We are brave and
progressive.
They are aggressive
and expansionistic.
We are thrifty.
They are cheap.
32
Kernel of Truth
Cautionary Statements
Perceived differences are not veridical
Perceived differences are exaggerated
Perceived differences reflect social
factors, not genetic differences
33
Illusory Correlations
Definition: People overestimate
how strongly two things are
related
(e.g., arthritis pain and changes in the weather)
34
Illusory correlations & stereotype formation
People associate a group with an
attribute (African Americans & crime)
Cognitive biases “corroborate” the
perceived association
confirmation biases in hypothesis testing
remember consistent information better
35
Illusory Correlation
People most susceptible to
illusory correlations when:
group is relatively small
attribute is rare in population
36
Illusory Correlation
Example
African Americans are a minority in
the US. Whites are the majority
Being a media superstar is rare
Illusory correlation likely……
 More AA (small group) superstars (rare event)
than White (large group) superstars (rare
event)
37
Illusory Correlation
Negative behavior more rare than positive
behavior
Implication:
Negative behavior by minority more
memorable and salient than same
behavior by majority
Negative behavior becomes part of
stereotype of minority
38
Stereotype Change
Consensual stereotypes change over
time, across individuals.
Very little known about stereotype change
over time, within individuals (see Weber
& Crocker, 1983, for an exception)
39
Models of Stereotype Change
Bookkeeping Model
Conversion Model
Subtyping Model
Focus on stereotype-inconsistent information
40
Bookkeeping Model
Stereotype change is incremental
Each instance of inconsistent
information modifies the stereotype
Single instance = small change
Accumulation = large change
41
Bookkeeping Model
Implication:
Stereotype change will be similar
regardless of whether inconsistent
information is concentrated or
dispersed. Amount (not dispersion)
matters.
42
Conversion Model
Stereotype change is dramatic
Stereotypes change in response to
large and salient inconsistent info.
Stereotypes remain unchanged by
minor inconsistent information
43
Conversion Model
Implication:
Stereotype change will be greater
when inconsistent information is
concentrated v.s. dispersed
44
Subtyping Model
Stereotypes hierarchically structured
Rare, inconsistent instances lead to
creation of subtypes.
Instances regarded as “exceptions”
Stereotype protected from change
Common, inconsistent instances result
45
in stereotype change
Subtyping Model
Implication:
Stereotype change will be greater
when inconsistent information is
dispersed v.s. concentrated
46
Weber & Crocker (1983)
Purpose:
Tested the three models of
stereotype change
47
Weber & Crocker (1983)
Procedure:
Given information about corporate
lawyers
Rated each lawyer on stereotypic
traits
48
Weber & Crocker (1983)
Manipulations:
Dispersion of Inconsistent info:
Dispersed across all members
Concentrated in 1/3 of members
Group size:
6 members v.s. 30 members
Amount of inconsistent info higher in
larger group
49
Weber & Crocker (1983)
Predictions
Dispersion has no effect on stereotype change,
but amount does (bookkeeping)
Stereotype change greater when inconsistent
info concentrated (conversion)
Stereotype change greater when when
inconsistent info dispersed (subtyping)
50
Weber & Crocker (1983)
Operationalization of Stereotype Change
More change = lower stereotypic judgments
Less change = higher stereotypic judgments
51
Weber & Crocker (1983)
Stereotypic Judgments
(lower = more change)
Effect of Dispersion
6
5
Dispersed
4
3
Dispersed
Concentrated
Which stereotype change model
does this result support?
52
Subtyping Model
Stereotypic Judgments
(lower = more change)
Weber & Crocker (1983)
6
Effect of Group Size
5
judge
4
3
Small group
Large group
Which stereotype change model
does this result support?
53
Bookkeeping Model
Weber & Crocker (1983)
Supported subtyping model:
stereotype change > dispersed
Supported bookkeeping model:
stereotype change > large group
54
Stereotype Maintenance
Subtyping Model
Subtypes help to maintain stereotype
Cognitive Biases
Better memory for stereotypeconsistent information
Confirmation biases in hypothesis
testing
55
Cognitive Biases
Cognitive biases maintain stereotype
by increasing confidence in the
stereotype’s accuracy
56
Cohen (1981)
96 participants watched video of a
librarian or waitress and her
husband
Some attributes fit stereotype of
librarian or waitresses (see next slide for
examples), others did not
Recalled as many of the woman’s
attributes as they could
57
Example of woman’s attributes
Half fit stereotype of librarians
wore glasses
ate roast beef
Half fit stereotype of waitresses
affectionate with husband
ate hamburger
58
% recalled correctly
Cohen (1981)
90.00%
85.00%
80.00%
Immediate
recall
75.00%
70.00%
65.00%
60.00%
Consistent
Information
Inconsistent
Information
Conclusion: Better recall for stereotypeconsistent information
59
Confirmation Biases in
Hypothesis Testing
Definition: Search for
information that confirms one’s
expectations (stereotype)
60
Snyder and colleagues
Through series of studies
showed that people engage in
this bias
Example…...
61
Snyder and colleagues
Told participants they would interview
another individual
Told to figure out if other person was
introverted or extroverted (initial hypothesis)
Given suggested questions to ask
1/2 introverted; 1/2 extroverted……..
62
Example questions
Introverted:
“What factors make it hard for you to
really open up to people?”
Extroverted:
“What kind of situations do you seek
out if you want to meet new people?”
63
Snyder & Colleagues
Results
Participants preferentially chose to ask
questions that would confirm their
initial hypothesis
64
Prejudice
Definition of Prejudice
A positive or negative attitude,
judgment or feeling about a person
that is generalized from attitudes or
beliefs held about the group to which
the person belongs.
65
Prejudice
Negative forms of prejudice studied
more because has greatest potential
to create social problems
Cautionary statement: preferential
treatment (positive prejudice) can also
cause problems
66
Zanna (1994)
Purpose:
Demonstrate that prejudice is made up
of different components
Correlated prejudice scores with three
proposed components of prejudice
67
Zanna (1994)
Components of prejudice:
Stereotypic beliefs: typical attributes
Symbolic beliefs: values, traditions, customs
Emotions: affective reactions (e.g., disgust)
68
Zanna (1994)
Procedure
1) Participants indicated their stereotypic
beliefs, symbolic beliefs, and emotions
about these social groups:
 English Canadian (ingroup)
 French Canadian
 Native Indian
 Pakistani
 Homosexual
69
Zanna (1994)
Procedure continued
2) Participants rated how favorable each
group was (i.e., prejudice)
70
Zanna (1994)
Results
1) On average, prejudice correlated positively
with each component (all p’s < .05)
2) But, correlations varied by target group…….
71
Zanna
Zanna(1994)
(1994)
0.6
Correlation
Correlationbetween
betweenprejudice
prejudiceand
and
components
of prejudice
components
of prejudice
by group
0.5
0.4
0.3
0.2
0.1
0
EC
FC
Stereotypic beliefs
NI
P
Symbolic beliefs
H
Emotion
72 72
Result 1: weakest correlation b/t prejudice
and components for English Canadians
overall
0.6
0.5
0.4
0.3
0.2
0.1
0
EC
FC
Stereotypic beliefs
NI
P
Symbolic beliefs
H
Emotion
73
73
Result 2: strongest correlation b/t prejudice
and components for French Canadians
overall
0.6
0.5
0.4
0.3
0.2
0.1
0
EC
FC
Stereotypic beliefs
NI
P
Symbolic beliefs
H
Emotion
74
74
Result 3: prejudice correlated with
stereotypic beliefs most strongly for French
Canadian and Homosexual
0.6
0.5
0.4
0.3
0.2
0.1
0
EC
FC
Stereotypic beliefs
NI
P
Symbolic beliefs
H
Emotion
75
75
Result 4: prejudice correlated with
symbolic beliefs most strongly for French
Canadian
0.6
0.5
0.4
0.3
0.2
0.1
0
EC
FC
Stereotypic beliefs
NI
P
Symbolic beliefs
H
Emotion
76 76
Result 5: prejudice correlated with
emotion most strongly for Pakistani
0.6
0.5
0.4
0.3
0.2
0.1
0
EC
FC
Stereotypic beliefs
NI
P
Symbolic beliefs
H
Emotion
77
77
Zanna (1994)
Conclusions:
Prejudice consists of at least three
components
stereotypic beliefs
symbolic beliefs
emotion
The components most central to
prejudice varies across groups
78
Download