Stigma Controllability: A Fresh Look at Attribution Theory (RES

REFERENCE No. RES-000-22-0288
Stigma Controllability: A Fresh Look at Attribution Theory
(RES-000-22-0288)
End of Grant Research Report
Investigator: Peter Hegarty (University of Surrey)
Background
Stigmata are defined as deeply discrediting traits that can reduce a ‘whole
and usual person to a tainted discredited one.’ (Goffman, 1963, p. 3). Since
Goffman’s early work, social psychologists have questioned whether some
stigmatized groups and individuals are treated particularly harshly because they are
held personally responsible for their fate. Weiner, Perry and Magnusson (1988)
argued that beliefs about the controllability of the onset and expression of a
stigmatized trait determine emotional, attitudinal, and behavioural attitudes toward
stigmatized individuals and groups. According to their attributional theory of stigma,
uncontrollable stigmas evoke pity, sympathy, and helping behaviour, and controllable
stigmas evoke anger and a refusal to extend aid.
Attribution theory has been extended to argue that people who believe stigmatized
traits to be non-controllable will, as a result, take on more tolerant attitudes
towards stigmatized groups. The present research examined this claim. Only two
experiments have directly shown that biological determinist information improves
people’s attitudes towards stigmatized groups such as the obese (Crandall, 1994;
Study 4) and lesbians and gay men (Piskur and Delegman, 1992). However, in
neither study were pre-manipulation attitudes measured. In one case, the
biological determinist text was 19% longer than the text read by control
participants, no evidence was presented that beliefs had been successfully
manipulated, and effects were observed only for men (Piskur and Delegman,
1992). This suggested the need for further better-conducted tests of this
hypothesis.
Rather, most attribution theory research only supports the hypothesis that
attributions affect attitudes in an indirect way. In vignette studies, participants
have been shown to have more positive emotional reactions and greater
intentions to extend aid to stigmatized individuals whose status is determined by
uncontrollable factors (e.g., biology) rather than controllable factors (e.g., their
own behavior patterns). Studies of this type have been conducted in such
domains as HIV/AIDS, mental and physical illness, poverty, sexual orientation,
and obesity (Armesto & Weisman, 2001; Cobb & Chabert, 2002; Corrigan et al.,
2001; Crandall & Moriarty, 1995; DeJong, 1980; Dooley, 1995; Graham, Weiner,
Giuliano, & Williams, 1993;Menec & Perry, 1998; Rush, 1998; Steins & Weiner,
1999; Weiner et al., 1988; Study 2; Zucker & Weiner, 1993). A second line of
evidence points to robust correlations between global negative attitudes toward
stigmatized groups and beliefs in the controllability of stigmatized traits. This
finding has been reported most often in regard to sexual orientation (Aguero,
Bloch & Byrne, 1984; Ernulf, Innala, & Whitam, 1989; Haslam, Rotschild, & Ernst,
1
REFERENCE No. RES-000-22-0288
2000, 2002; Hegarty, 2002, Study 1; Hegarty & Pratto, 2001; Sakalli, 2002;
Whitley, 1990) and occasionally with regard to mental illness (Corrigan, 2000) and
obesity (Crandall, 1994).
Neither line of evidence directly supports the claim that attributions affect attitudes
toward stigmatized groups. Vignette studies about individuals say little about
changes in attitudes to stigmatized groups. Information about a person only
affects beliefs about that individual’s group when that person is seen as typical of
their group (Wilder, Smith, & Fielding, 1996). The targets presented in vignette
studies might be easily subtyped (see Kunda & Oleson, 1995). For example,
positive reactions to an obese woman with a thyroid condition (DeJong, 1980)
may do little to change global beliefs or attitudes towards fat people. Of course,
correlations between attitudes and beliefs do not support causal arguments either.
Yet, major literature reviews cite Weiner et al. (1986) to argue that such
correlations are evidence of the direct effects of attributional beliefs on global
attitudes (e.g., Anderson, Krull, & Weiner, 1996; Crocker, Major, & Steele, 1998).
Some researchers have even recommended the teaching of biological determinist
information to mitigate prejudice (Finell, 2002; Sakalli, 2002, p. 267-268; Weiner,
1995, p. 15; Whitley, 1990, p. 375).
Beyond the issue of the scant evidence base, there are three principal reasons to
doubt that attributions of non-controllability directly affect attitudes. First, attitudes
towards stigmatized groups could shape causal thinking as much as the reverse.
Second, attributions of non-controllability may engender essentialist thinking.
Third, attributions of non-controllability may be consistent with entity theorizing
about personality (Dweck, 19xx) and as such engender stereotyping (Levi et al.,
19xx).
First consider the claim that attributions might be consequences rather than
causes of stigmatizing attitudes. While seemingly counter to attribution theory, this
idea is implicit within the theory itself. Weiner et al. (1988, p. 739) note that
‘negative events or effects in particular initiate attributional search’ such that ‘a
search to determine the origin of the stigma is presumed to be undertaken by the
stigmatized person as well as by observers’ (emphasis added). As nonstigmatized persons’ attitudes vary, the attributional theory implies that those with
more negative attitudes will initiate an attributional search about the cause of a
stigma more readily.
The hypothesis that beliefs about controllability are consequences rather than
causes of negative attitudes is further warranted by Hegarty’s (2002) research. In
two studies with heterosexual participants, correlations between the belief that
sexual orientation was non-controllable and positive attitudes towards lesbians
and gay men were observed; but only among participants who judged immutability
beliefs to express tolerance. No such correlation was observed among
participants who judged beliefs about immutability to express mixed views about
lesbians and gay men. These results do not suggest direct effects of attributions
on attitudes. Rather they suggest that attributional beliefs might be taken up by
heterosexuals who wish to present themselves as tolerant of lesbians and gay
men to varying degrees.
Second, attributions of non-controllability often refer to a biological ‘essence’ such
as a heritable predisposition to explain the presence of the stigmatized trait. Such
explanations are then ‘essentialist’ and as such may engender prejudice in some
cases (Haslam, Rotschild, & Ernst, 2002) or provide a ground from which to
stereotype stigmatized groups (Yzerbyt, Schadron, & Rocher, 1997). Thus,
persons who attribute stigmatized traits to non-controllable biological causes such
as genetics might hold more extreme stereotypes about stigmatized groups. This
2
REFERENCE No. RES-000-22-0288
hypothesis is also warranted by findings that individuals who consider human
personality to be fixed rather than fluid have also been shown to stereotype outgroups more readily (Levi). The claim that stigmatized traits are fixed and noncontrollable may similarly encourage the stereotyping of out-groups.
Objectives
The original aim of the research was to test four hypotheses. The first study
tested competing predictions derived from attribution theory (Weiner et al., 1986) and
Hegarty’s (2002) work on symbolic beliefs. Its results supported attribution theory
only weakly and did not support the symbolic beliefs position at all. As a result, a
second study was conducted to check that these negative findings were not
attributable to methodological weaknesses (see Study 2 below). As null results were
observed across a range of measures in Study 2 it was concluded that the results of
Study 1 were reliable and that correlations between attributions and attitudes were
weaker in this population than in previous populations studied. The original grant
proposal also described a study which experimentally tested attribution theory’s claim
that changes in attributions about stigmatized traits would lead to changes in
attitudes toward stigmatized groups. This study was the most successful of all and
provided consistent evidence that attribution theory’s claims are incorrect (in Study
3). A third study had been proposed to test the claim that people with different
attitudes engage in different kinds of causal thinking about stigmatized traits. As
some evidence in support of this hypothesis had already been gathered (see Study 3
below) and an unanticipated study had been required (Study 2) this study was not
conducted. The final study proposed had been a vignette study centring on
attributions and beliefs about stigmatized individuals. However, further reading of the
literature, and the new data gathered so far suggested that attributions about
stigmatized individuals and about stigmatized groups might be psychologically
distinct. A new fourth study was designed, and it examined the contribution of
implicit personality theory to causal thinking about stigmatized groups (see Study 4).
Completed Studies
Studies 1-3 all examined links between stigmatizing attitudes and attributions
about the origins of stigmatized traits. Previous studies in this area have relied on
standardized scales to measure attitudes (e.g., Crandall, 1994; Crandall & Martinez,
1996; Crandall et al., 2001; Hegarty, 2002; Hegarty & Pratto, 2001; Sakalli, 2002;
Whitley, 1990). One shortcoming of such measures is that they sometimes conflate
positive attitudes with ontological claims that are relevant to controllability. For
example, Hegarty (2002; Hegarty & Pratto, 2001) used Herek’s (1984) ATLG scale to
measure heterosexist attitudes. This scale contains the item ‘Just as in other
species, male homosexuality is a natural expression of sexuality in human men’
which could imply that homosexuality is biologically determined. Standardized
measures of controllability can also connote negative attitudes toward the
stigmatized groups rather than simply measure ontological beliefs. For example,
Crandall’s (1994) measure assessing the degree to which bodyweight is under
personal control includes the item ‘people who weigh too much could lose at least
some part of their weight through a little exercise’ (emphasis added). This item might
assess beliefs about whether body weight can be controlled, but it also assumes that
fat bodies lie beyond some unspecified norm for body shape, and this assumption
may carry negative connotations. For these reasons we used ‘contentless’ measures
to assess correlations in Study 1, compared contentless and standardized measures
directly in Study 2, and used both contentless and standardized measures in Study 3.
The contentless measures were thermometer-like visual analogue scales (Campbell,
3
REFERENCE No. RES-000-22-0288
1971) and personal stereotype measures (Eagly, Mladinic, & Otto, 1991; Esses,
Haddock & Zanna, 1994).
Study 1: A Test of Attribution Theory and Symbolic Attitudes Theory with
Correlational Data
Method. Study 1 tested competing hypotheses derived from attribution theory
and symbolic attitudes theory. 66 undergraduates completed two waves of a
questionnaire that were four weeks. During Wave 1, participants attitudes and beliefs
towards seven stigmatized groups were measured (i.e., alcoholics, gay men, people
with learning disabilities, obese persons, people who are mentally ill, drug addicts,
and lesbians). During Wave 2 we measured participants judgments about the
attitudes that those beliefs expressed.
During Wave 1, attitudes were assessed using thermometer measures and
personal stereotype measures. In the first case, the name of each stigmatized group
was presented with a 10 cm unmarked horizontal scale which ranged from 0
(favourable) to 100 (favourable). Participants were instructed to mark the scale at
the point that best represented their attitude. Personal stereotypes were next
assessed. The name of each of the seven stigmatized groups was presented and
participants were asked to write down up to three terms that characterized that
group, to describe the percentage of that group that were described by each term
and to rate the valence of each term on a five point scale from –2 ‘extremely
negative’ to +2 ‘extremely positive’. Personal stereotypes for each participant for
each stigma were computed according to the formula ∑ (p x v)/n, where p = the
percentage of the group judged to share the trait (0 to 100), v = the valence of the
trait (-2 to 2) and n = the total number of traits described.1 Beliefs about genetics,
personal control and personal experience as causes of the stigmatised traits were
assessed using single item measures. For example, beliefs about genetics were
assessed by asking whether the participant agreed or disagreed ‘with the idea that
genes determine who becomes a member of that group.’ Similar items were used to
assess beliefs in experiential and controllable factors. The names of all seven
groups were presented followed by 7-point Likert items that ranged from 1 (disagree)
to 7 (agree). Similar instructions elicited beliefs that membership in the stigmatised
groups was caused by personal control and personal experience.
During Wave 2, participants judged the attitudes that these attributional
beliefs expressed. Participants were presented with 21 statements describing each
of the seven stigmas as determined by genetics, personal experience or personal
control. They were asked to rate who would be most likely to express each of these
21 beliefs using 7-point Likert items anchored at 1 (a tolerant person) and 7 (a
prejudiced person).
Results. The data from Wave 1 were used to test attribution theory.
Correlations between the two attitude measures (thermometers and personal
stereotypes) and belief in genetics, personal control, and personal experience are
reported in Table 1 below. Of the 42 correlations computed, only four were
significant, and all of these involved the thermometer measure. This result fails to
replicate previous attribution theory findings. However, 35 of the 42 correlations
were in the expected direction, such that participants with more negative attitudes
1 In the original grant proposal, it was proposed that in all studies personal stereotype measures would also
be used as measures of stereotype extremity to test the hypothesis that non-controllable attributions lead to
more extreme stereotyping. Steroetype extremity was measured by averaging the percentage ratings
accorded to each trait mentioned regardless of their valence. This hypothesis was tested for Studies 1, 2, &
3 but no patterns of significant findings were observed. This hypothesis is not discussed further.
4
REFERENCE No. RES-000-22-0288
attributed the stigma to personal control and to personal experience but not to
genetics (Sign test, p <.01). These results support attribution theory, but only
weakly.
Wave 2 allowed a test of the symbolic attitudes theory, which predicts that
correlations between attitudes and beliefs depend upon the symbolic values those
beliefs are judged to express. Three composite measures of participants’ judgments
about each of the three attributional dimensions were constructed, Cronbach’s α =
.84, .78, .74 for genetics, personal control, and personal experience beliefs
respectively. We performed a median split on each composite measure.
Participants who judged each set of beliefs to express relatively tolerant attitudes
were categorized as tolerant-expressives. Participants who considered the beliefs to
express condemning attitudes were categorized as prejudice-expressives.
Correlations between each belief item and each attitude measures from Wave 1 were
calculated separately for these two groups, and Z-scores were calculated to examine
differences between the attitude-belief correlations among the tolerant-expressive,
and prejudice-expressive groups. As Table 1 shows, trends in the expected direction
were observed in only 27 out of 42 cases. Significant moderation was observed in
only two cases, and significant moderation in the direction opposite to predictions
was observed in one case, Z = 2.24, p < .05. These results do not support symbolic
attitudes theory.
Study 2: Attributions, Symbolic Theory and Standardized Measures.
The results of Study 1 were surprising in that the basic attitude-belief
correlations presumed by both attribution theory and the symbolic attitudes position
were much weaker than expected. There are at least three possible explanations of
these results.
First, the use of contentless measures may have suppressed attitude-belief
correlations. As noted above, standardised measures often conflate attitudes and
beliefs, and this may have artificially inflated correlations in past studies relative to
the present studies. Second, participants’ attitudes to some groups were positive in
Study 1. For example, attitudes toward gay men and lesbians were significantly more
positive than the midpoint of both the thermometer (Ms = 67.2, 64.9) and the
personal stereotype measures (Ms = 52.5, 30.4). Attribution theory implies that the
perception of a stigmatized trait as negative initiates a causal search. However, this
explanation cannot account for all failures to replicate past results; attitudes towards
obese people were neutral on the thermometer measure (M = 49.3) and negative on
the personal stereotype measure (M = -87.3). Finally, such correlations may reflect
genuine differences between the present sample and American undergraduates who
are normatively studied in such research. Hegarty (2002) found stronger correlations
between attitudes and beliefs about sexual orientation among American than British
samples, although identical items were used to study both groups.
5
REFERENCE No. RES-000-22-0288
Table1: Correlations between Belief and Attitude Measures Among Entire Sample,
Tolerant-Expressives and Prejudice-Expressives (Study 1).
Stigma
Alcohol
Gay
Learn
Obese
M.Ill
Drug
Lesbian
Thermometer Measure
Belief
Genetics
All
.094
Prj (30)-.304
Tol (35).472**
Experience
All
-.066
Prj (33)-.288
Tol (33) .042
Control
All
-.388**
Prj (35) -.312†
Tol (31) -.459**
.040
-.202
.031
-.088
.226†
.203
-.003
.163
.335**
.238
.077
-.078
.142
.064
.200
-.120
.439**
.134
-.010
.068
-.207†
-.124
-.252*
-.201
.119
-.160
-.134
.249
-.206
-.137
-.001
-.124
-.397*
-.135
-.142
-.274
- .036
-.209
.088
-.073
-.065
.102
- .233†
-.299
-.190
-.077
.107
-.108
-.369**
-.439**
-.310†
-.155
-.146
-.108
Personal Stereotype Measure
Genetics
All
-.083
Prj (30)-.111
Tol (35)-.057
Experience
All
.177
Prj (33).046
Tol (33).310
Control
All
-.183
Prj (35) -.176
Tol (31)-.202
.183
.098
.231
.064
-.184
.277
.027
-.029
-.008
-.047
-.071
.001
.019
.186
-.081
.082
-.143
.117
-.115
.179
-.406*
-.188
.044
-.381*
-.057
-.048
.073
-.232†
-.282
-.137
-.102
.022
-.253
-.047
-.074
-.058
.165
.104
.184
-.210
-.136
-.264
-.205
-.284
-.072
.001
-.044
.034
-.097
-.060
-.153
-.200
.069
.380†
†
p<.10, *p<.05, ** p<.01, correlations differ from 0.
Note: Learn= People with Learning Disabilities, M.Ill = Mentally Ill People, Drug =
Drug Addicts, Prj = Prejudice-Expressives, Tol = Tolerance-Expressives.
Study 2 examined attitude-belief correlations using both contentless and
standardized measures of attitudes towards lesbians/gay men and obese people.
These groups were chosen because previous studies that support attributional theory
have used standardized measures in these domains (e.g., Crandall, 1994; Whitley,
1990). Also, participants in Study 1 varied in their attitudes towards these groups;
they viewed lesbians and gay men more positively than obese people did. Thus,
Study 2 aimed to assess if the low correlations observed in Study 1 were due to the
6
REFERENCE No. RES-000-22-0288
exclusive use of contentless measures. Study 2 also included belief items that
presented stigmatized traits as either abnormal or normal. We aimed to see if belief
items that presented stigmatized groups as abnormal would be endorsed to a greater
degree, particularly among more prejudiced participants.
Method. 76 undergraduates participated in an experiment with a 2x2 design.
They completed attitude and belief items about either sexual orientation or
bodyweight. The belief items that they completed either connoted that the relevant
stigmatized trait was normal or abnormal. Participants completed thermometer and
personal stereotype measures as in Study 1, and standardized measures of
attitudes; Herek’s (1984) ATLG in the sexual orientation conditions and Factor 1 of
Crandall’s (1994) AFA in the bodyweight conditions.
The abnormalizing bodyweight items were drawn from Factor 3 of Crandall’s
(1994) AFA scale. For example, one item read as follows;
People who weigh too much could lose at least some part of their weight through
a little exercise
The corresponding normalizing bodyweight items read as follows;
It is possible for people to change their body size by exercising
The abnormalizing sexual orientation items were loosely based on items described
by Hegarty and Pratto (2001). For example one item read as follows;
Homosexuality is caused by biological aberrations
The equivalent normalizing item read as follows;
An individual’s sexual orientation is caused by natural variation in human
biology.
Participants were randomly assigned to condition.
Results. Participants’ attitudes were first examined using 2x2 ANOVAs with
stigmatized group (bodyweight vs. sexual orientation) and belief (normalizing vs.
abnormalizing) as independent factors. As predicted, and as Study 1, participants
evinced more positive attitudes toward lesbians and gay men than obese people on
the thermometer measure, F (1, 75) = 25.42, p < .001 (Ms = 62.5, 42.4 respectively),
and on the personal stereotype measure, F (1, 75) 57.91, p <.001, (Ms = 50.54, 60.60). The main effects of belief type were not significant and did not interact with
the effects of stigmatized group, all F <1 (see Table 2).
Mean differences in beliefs were examined separately for the sexual
orientation and bodyweight conditions as different items had been used in each. In
the sexual orientation conditions, responses to the first item did not vary by condition,
t<1. However, participants endorsed the second and third items more when
gay/lesbian identity was normalized rather than abnormalized, t (37) = 4.87, p <.001,
t (37) = 2.39 p <.05 respectively. In the bodyweight conditions, responses to the first
and third items did not vary by condition, both t <1. However, participants supported
the second item to a marginally greater extent when it was framed as an
abnormalizing belief, t (35) = 1.96, p<.06. In other words, participants endorsed
beliefs about the mutability of a liked stigmatized group to a greater degree when
7
REFERENCE No. RES-000-22-0288
those items presented the group’s identity as normal. However, normalizing the
identity of a disliked group produced trends to endorse the belief items less (see
Table 2).
Finally, we examined correlations between beliefs and each of the three
attitude measures; thermometers, personal stereotypes, and standardized measures.
As in Study 1, few of the individual correlations were significant. Thus these results
suggest that the choice of contentless measures in Study 1 did not lead to a
suppression of attitude-belief correlations.
Table 2: Mean Endorsement of Belief Item and Correlation with Attitude Measures by
Condition (Study 2).
Stigma
Belief
Sexual Orientation
Bodyweight
Abnormal
Normal
Abnormal
Normal
(n = 18)
(n = 21)
(n = 19)
(n = 18)
Item 1
Mean
r.Thermometer
r.Personal Stereotype
r.Standard Measure
3.48
.05
-.06
.17
3.10
-.11
-.09
-.32
5.74
-.28
.01
.09
6.05
-.27
-.21
.24
Item 2
Mean
r.Thermometer
r.Personal Stereotype
r.Standard Measure
3.10
-.21
.04
.26
5.50
-.03
.02
-.04
4.89
-.26
.28
.51*
4.00
-.16
-.26
.10
Item 3
Mean
r.Thermometer
r.Personal Stereotype
r.Standard Measure
3.57
.18
.26
.09
4.70
.40
.40
-.29
3.68
-.18
.28
.41
3.75
-.83*
-.33
.54*
Note: r. = correlation between belief item and named attitude measure.
Study 3: An Experimental Test of Attribution Theory’s Causal Hypothesis.
Studies 1 and 2 found only weak correlations between attributions about
stigmatized traits and attitudes toward stigmatized groups to be weaker than
previously reported findings. Study 3 tested two competing explanations of why
such correlations are observed. First, we tested the claim that manipulations of
attributions about stigmatized groups lead to changes in attitudes toward those
groups. Second, we tested the claim that more prejudiced persons engage in more
spontaneous causal thinking about stigmatized targets.
Method. 166 undergraduates participated in an experiment with a 2 x 4
design. Participants read about one of four stigmatized traits; homosexual
orientation, obesity, depression, or alcoholism. Each of these was presented as
either controllable or uncontrollable in the study vignettes.
8
REFERENCE No. RES-000-22-0288
Participants first completed a questionnaire consisting of 24 thermometer
items like those used in Study 1. These assessed attitudes towards several social
groups including the four stigmatized groups of interest.
Vignettes were next presented to manipulate beliefs. Each vignette was three
paragraphs long and described the stigmatized trait as either controllable or
uncontrollable by presenting different results to bogus behavioural genetics and
biological studies.
Participants were next instructed to list between five and twelve free thoughts
about the vignette. These were used to assess spontaneous causal thinking.
Several measures followed that assessed participants’ comprehension and
evaluation of the vignette.
Post-manipulation attitudes were assessed using a thermometer measure, a
personal stereotype item and a standardized measure. The standardized measure
in the sexual orientation conditions was the ATLG, and in the obesity conditions
was the ‘dislike’ subscale of the AFA as in Study 2. In the alcoholism and
depression conditions five semantic differential items selected from Crisp et al.
(2000) were used. High scores on all standardized scales indicate higher levels of
prejudice.
Results. The 168 participants produced 1049 thoughts in all, which were
coded independently by two coders in two stages. Participants thoughts about the
vignettes were first categorised as either attributional (35.5%) or non-attributional
(64.5%). Attributional thoughts were further subdivided into those that imply
controllability (52.5%), were ambiguous (10.5%), or that imply non-controllability
(37%). A 4x2x2 ANOVA with stigma group (homosexuality, obesity, depression,
alcoholism) and manipulation (controllable vs. uncontrollable) as between-subjects
factors and type of attributional thought (controllability implied vs. uncontrollability
implied) as a within-subjects factor was conducted to assess effects of the vignettes
on participants’ causal thinking. Participants produced more thoughts that implied
controllability rather than uncontrollability overall, F (1, 158) = 6.80, p =.01 (Ms =
1.18, .83). This effect was moderated by a significant interaction between the
controllability manipulation and type of thought produced, F(1, 158) = 20.49, p<.001.
A significantly greater number of thoughts that implied controllability were produced
by those who read that that stigmas were controllable rather than uncontrollable (Ms
= 1.45, .92 respectively). However, a significantly greater number of thoughts that
implied uncontrollability were produced by those who read that stigmas were
uncontrollable rather than controllable (Ms = 1.19, .46 respectively). In other words
the vignettes effectively manipulated participants’ causal thinking.
We next tested the hypothesis that more prejudiced persons engaged in more
spontaneous causal thinking. Participants’ pre-manipulation attitudes were
uncorrelated with the total number of thoughts produced r (165) =.013, but were
negatively correlated with the number of attributional thoughts produced r (165) = .182, p <.02. Attitudes were unrelated to the production of thoughts that implied
non-controllability r (165) =-.029, but were negatively correlated with the number of
attributional thoughts produced that implied controllability. This was true in the
experiment as a whole, r (165) = -.255, p =.001, and when beliefs had been
manipulated in both the controllable and non-controllable directions, r (82) = -.259, .264 respectively, both p<.05. In other words, more prejudiced persons
spontaneously expressed more thoughts that implied the controllability of the
stigmatised traits, regardless of which vignette they had read.
Where causal thinking has been effectively manipulated, attribution theory
predicts that attitudes should be more negative where stigmatized traits are
perceived as controllable rather than non-controllable. However, no such effects of
the controllability manipulation were observed with the thermometer measures, the
9
REFERENCE No. RES-000-22-0288
personal stereotype measures or the standardised measures (see Table 3). Nor
were the effects consistently in the direction predicted by attribution theory.
Table 3: Attitudes towards Stigmatized Groups by Experimental Condition and
Attitude Measure (Study 3).
.
Stigmatized Group
Obese
Men/Lesbians
Thermometer
Controllable
47.3
Non-controllable 53.5
Personal Stereotype
Controllable
62.1
Non-controllable 68.0
Standardized Measure
Controllable
3.41
Non-controllable 2.61
Depressed
Alcoholics
Gay
53.9
56.5
37.3
38.1
61.7
50.4
70.5
72.9
69.6
69.6
64.4
56.3
4.10
3.82
3.60
3.92
2.48
3.11
.
Study 4: Implicit Personality Theories and Attributions about Stigmatized
Traits.
The final study examined implicit personality theories and beliefs about
controllability. On cognitive consistency grounds, it might be predicted that people
who believe human personality to be fixed rather than fluid would also believe that
stigmatized traits are fixed and non-controllable. However, ‘entity theorists’ who
believe personality to be fixed also stereotypes out-groups more quickly and express
other kinds of prejudice more quickly than do ‘incremental theorists’ who hold
personality to be malleable (Levi et al., ). Thus beliefs about the fixity of human
personality and about the fixity of stigmatized traits might be cognitively consistent,
but may be affectively dissonant. Study 4 explored the empirical relationship
between these constructs.
Method. Eighty-three undergraduate participants completed questionnaires
that contained Dweck’s four-item measure of implicit personality theory. All items
were presented as six point scales. Participants also rated the degree to which the
onset and expression of twelve stigmatized traits was under personal control. The
relevant stigmas were alcoholism, AIDS, depression, drug addiction, homelessness,
lesbianism, lung cancer, male homosexuality, obesity, poverty, schizophrenia and
transsexuality.
Results. The four items on the implicit personality scale formed a reliable
measure (α= 0.85), and were averaged to form an overall measure of implicit
personality. Following Dweck, participants whose score on this measure was less
than 3 were categorised as entity theorists (n = 23) and those whose score was
greater than 4 were categorised as incremental theorists (n = 36). The remainder
were considered ambiguous cases (n = 24).
T-tests were conducted to test the hypothesis that entity theorists endorsed
the controllability of stigmatized traits more than did incremental theorists (see Table
4). In 17 of the 24 cases, trends were consistent with this prediction (p<.05, onetailed sign test). Moreover, in five cases results reached statistical significance, and
all five cases involved sexual and gender minorities. Thus, implicit personality
theories appear to be cognitively inconsistent with beliefs about the controllability of
stigmatized traits.
10
REFERENCE No. RES-000-22-0288
Controllability
Personality Theory
Alcoholism
AIDS
Depression
Drug Addiction
Homelessness
Lesbianism
Lung cancer
Male homosexuality
Obesity
Poverty
Schizophrenia
Onset
Entity
4.70
4.52
2.52
5.13
2.00
4.43
1.74
4.61
3.47
1.78
Inc.
5.14
3.97
2.83
5.27
4.13
2.89
4.05
2.75
4.67
3.18
2.03
t
1.19
-1.22
.92
.37
3.97 -.46
1.65†
.96
2.05*
.16
-.84
.72
Transexuality
2.08 3.75 3.18**
†
p <.10, *p <.05, **p<.01, ***p<.001.
Note: Inc = Incremental Theorists.
Expression
Entity. Inc.
3.65 4.53
3.39 3.50
2.78 2.94
3.83 4.50
3.13
3.00 4.81
3.26 3.14
3.04 4.61
3.57 3.22
3.27 2.97
2.13 2.36
t
1.91†
.19
.43
1.90†
2.92 -.50
3.42***
-.25
2.88**
-.64
-.94
.52
3.04
3.30**
4.69
Activities
The original aim was to present the findings of this research at two national
conferences. A paper was presented based on the results of Studies 1 and 3
above to the British Psychological Society’s Annual Social Psychology Section
Conference in Liverpool in September 2004 by Peter Hegarty. A submission was
made to the British Psychological Society’s Annual Division of Health Psychology
Conference by Anne Golden but was rejected on the grounds that the content was
not sufficiently relevant to Health Psychology.
Outputs
Thus far, a paper has been submitted for publication to The British Journal of
Social Psychology based on the results of Studies 1 and 3. A copy of the paper is
attached. A shorter summary of these studies has been disseminated to the
several public interest groups listed in Section 2B: ‘Dissemination’.
Impacts
The submission to The British Journal of Social Psychology and the summary
of research findings for non-academic users were prepared and mailed out in
November 2004. Thus it is too early to expect evidence of the use of this research in
either academic or non-academic settings.
11
REFERENCE No. RES-000-22-0288
Future Research Priorities
The most important finding to emerge from this research is that changes in
attributions do not necessarily lead to changes in attitudes as attribution theory
suggests. Rather, more prejudiced persons are more likely to spontaneously
consider the controllability of stigmatized traits (Study 3). This suggests that
explanations of the origins of stigmatized traits might be expressions of prejudice
rather than causal antecedents to prejudice.
This finding not only challenges a major theory within social psychology, it
also has clear applied relevance by showing that public opinion campaigns aimed at
reducing stigma are unlikely to be effective by targeting beliefs about the causes of
stigmatized traits. Future research might examine such educational campaigns using
experimental and quasi-experimental designs to determine their effectiveness.
References
Aguero, J.E., Block, L., & Byrne, D. (1984). The relationships among sexual beliefs,
attitudes, experience, and homophobia. Journal of Homosexuality, 10, 95107.
Anderson, C.A., Krull, D.S., & Weiner, B. (1996). Explanations: processes and
consequences. In E.T. Higgins & A. W. Kruglanski (Eds.), Social psychology:
Handbook of basic principles (pp. 271-296). New York: Guildford.
Armesto, J.C., & Weisman, A.G. (2001). Attributions and emotional reactions to
the
identity disclosure (“coming out”) of a homosexual child. Family Process, 40,
145-161.
Campbell, A. (1971). White Attitudes Toward Black People. Ann Arbor, MI:
Institute
for Social Research.
Cobb, M. & deChabert, J.T. (2002). HIV/AIDS and care provider attributions: Who’s
to blame? AIDS Care, 14, 545-548.
Corrigan, P.W. (2000). Mental health stigma as social attribution: Implications for
research methods and attitude change. Clinical Psychology: Research and
Practice, 48-67.
Corrigan, P.W., River, L.P., Lundin, R.K., Penn, D.L., Uphoff-Wasowski, K.,
Campion, J., Mathisen, J., Gagnon, C., Bergman, M., Goldstein, H., & Kubiak,
M.A. (2001). Three strategies for changing attributions about severe mental
illness, Schizophrenia Bulletin, 27, 187-195.
Crandall, C.S. (1994). Prejudice against fat people: ideology and self-interest.
Journal of Personality and Social Psychology, 66, 882-894.
Crandall, C.S., & Moriarty, D. (1995). Physical illness stigma and social
rejection.
British Journal of Social Psychology, 34, 67-83.
12
REFERENCE No. RES-000-22-0288
Crocker, J., & Major, B. (1989). Social stigma and self-esteem: the selfprotective
properties of stigma. Psychological Review, 96, 608-630.
Crocker, J., Major, B., & Steele, C. (1998). Social stigma. In D.T. Gilbert, S.T.
Fiske, & G. Lindzay (Eds.) The handbook of social psychology: Volume 2 (pp.
504-553). New York: Oxford University Press.
DeJong, W. (1980). The stigma of obesity: The consequences of naïve
assumptionsconcerning the causes of physical deviance. Journal of Health
and Social Behavior, 21, 75-87.
Dooley, P.A. (1995). Perceptions of the onset controllability of AIDS and
helping
judgments: An attributional analysis. Journal of Applied Social
Psychology, 25, 858-869.
Eagly, A. H., Mladinic, A., & Otto, S. (1991). Are women evaluated more favorably
than men? An analysis of attitudes, beliefs, and emotions. Psychology of
Women Quarterly, 15, 203-216.
Ernulf, K.E., Innala, S.M., & Whitam, F. (1989). Biological explanation,
psychological explanation, and tolerance of homosexuals: A cross-national
analysis of beliefs and attitudes. Psychological Reports, 65, 1003-1010.
Esses, V.M., Haddock, G., & Zanna, M.P. (1994). The role of mood in the expression
of intergroup stereotypes. The psychology of prejudice: The Ontario
Symposium, Vol. 7, 77-101.
Finell, D.S. (2002). The case for teaching patients about the neurobiological basis of
addictions. Journal of Addictions Nursing, 12, 149-158.
Goffman, E. (1963). Stigma: Notes on the management of spoiled identity. Prentice
Hall: Englewood Cliffs, NJ.
Graham, S., Weiner, B., Guiliano, T., & Williams, E. (1993). An attributional
analysis of reactions to Magic Johnson. Journal of Applied Social
Psychology, 23, 96-1010.
Haddock, G. & Zanna, M. P. (1998). On the use of open-ended measures to assess
attitude components. British Journal of Social Psychology, 37, 129-149.
Haslam, N., Rotschild, L., & Ernst, D. (2002). Are essentialist beliefs associated with
prejudice? British Journal of Social Psychology, 41, 87-100.
Hegarty, P. (2002). ‘It’s not a choice, it’s the way we’re built’: Symbolic beliefs
about sexual orientation in the US and Britain. Journal of Community and
Applied Social Psychology, 12, 153-166.
Hegarty, P., & Pratto, F. (2001). Sexual orientation beliefs: Their relationship to antigay attitudes and biological determinist attitudes. Journal of Homosexuality,
41, 121-135.
Herek, G.M. (1984). Attitudes towards lesbians and gay men: A factor analytic study.
Journal of Homosexuality, 10, 1-21.
Kunda, Z., & Oleson, K.C. (1995). Maintaining stereotypes in the face of
disconfirmation: Constructing grounds for subtyping deviants. Journal of
Personality and Social Psychology, 68, 565-579.
Levy, S.R., Stroessner, S.J., & Dweck, C. S. (1997). Stereotype formation and
endorsement: The role of implicit theories. Journal of Personality and Social
Psychology, 74, 1421-1436.
Menec, V.H., & Perry, R.P. (1998). Reactions to stigmas among Canadian students:
Testing an attribution-affect-help judgment model. Journal of Social
Psychology, 138, 443-454.
Menec, V.H., & Weiner, B. (2000). Observers’ reactions to genetic testing: The role
13
REFERENCE No. RES-000-22-0288
of hindsight bias and judgments of responsibility. Journal of Applied Social
Psychology, 30, 1670-1690.
Piskur, J., & Degelman, D. (1992). Effect of reading a summary of research about
biological bases of homosexual orientation on attitudes toward homosexuals.
Psychological Reports, 71, 1219-1225.
Rush, L.J. (1998). Affective reactions to multiple social stigmas. Journal of Social
Psychology, 138, 421-430.
Sakalli, N. (2002). Application of the attribution-value model of prejudice to
homosexuality. Journal of Social Psychology, 142, 264-271.
Steins, G., & Weiner, B. (1999). The influence of the perceived responsibility and
personality characteristics on the emotional and behavioral reactions to
people with AIDS. Journal of Social Psychology, 139, 487-495.
Weiner, B. (1995). Inferences of responsibility and social motivation. Advances in
Experimental Social Psychology, 27, 1-47.
Weiner, B., Perry, R.P., & Magnusson, J. (1988). An attributional analysis of
reactions to stigmas. Journal of Personality and Social Psychology, 55, 738748.
Whitley, B.E. Jr (1990). The relationship of heterosexuals’ attributions for the
causes
of homosexuality to attitudes toward lesbians and gay men.
Personality and Social Psychology Bulletin, 16, 369-377.
Wilder, D.A., Simon, A.F., & Faith, M. (1996). Enhancing the impact of
counterstereotypic information: Dispositional attributions for deviance.
Journal of Personality and Social Psychology, 71, 276-287.
Zucker, G.S., & Weiner, B. (1993). Conservatism and perceptions of poverty: An
attributional analysis. Journal of Applied Social Psychology, 23, 925-943.
14