Assignment: Attribution Errors

Reeb79

SocialDominanceOrientationAttenuatessterotypeChangeintheFaceofDisconfirmingInformation.pdf

Home >Psychology homework help >Assignment: Attribution Errors

Original Article

Social Dominance Orientation Attenuates Stereotype Change in the

Face of Disconfirming Information Nicole Tausch1 and Miles Hewstone2

1Cardiff University, UK, 2University of Oxford, UK

Abstract. This study examined whether social dominance orientation (SDO) affects the malleability of group stereotypes in the face of disconfirmation. Data were collected at two time points: At Time 1, baseline stereotypes and SDO were assessed, and at Time 2, either moderately or extremely stereotype-inconsistent information was presented and stereotyping measures were repeated. Consistent with previous research, exposure to moderately inconsistent information resulted in greater stereotype change than exposure to extremely inconsistent information. As expected, SDO was negatively related to stereotype change, in particular after presentation of moderately inconsistent information. The judged typicality of the target exemplar mediated the effects of the manipulation but did not mediate the effects of SDO. Implications for future research and interventions to reduce stereotyping are discussed.

Keywords: stereotype change; disconfirming information; social dominance orientation

The present research examined whether the malleability of group stereotypes in the face of disconfirming information is affected by characteristics of the social perceiver. Research on the role of individual differences in social perception pro- cesses has a long tradition in social psychology (e.g., Adorno, Frenkel-Brunswick, Levinson, & Sanford, 1950), and it has long been assumed that the “prejudiced personality” “does not change his mental set easily, but persists in old ways of reasoning” (Allport, 1954, p. 171). Although more recent re- search has extensively studied individual differences as pre- dictors of social information processing strategies and has successfully demonstrated the link between personality and stereotyping (see Duckitt, 2005, for a review), little is still known about whether individual differences predispose so- cial perceivers toward the maintenance of prejudiced beliefs even in the face of disconfirmation. The present research aims to fill this gap in the literature by examining whether individual differences in ideological orientation, specifically social dominance orientation (SDO; Pratto, Sidanius, Stall- worth, & Malle, 1994), affect the amount of stereotype change in response to disconfirming information.

Stereotype Change in the Face of Disconfirming Information

A large number of empirical studies have shown that peo- ple often resist stereotype change when presented with in-

consistent information (see Hewstone, 1994, for a review). This stability is achieved through cognitively isolating ste- reotype-inconsistent group members from the group repre- sentation; that is, stereotype disconfirmers are judged as atypical and not considered when judging the group in gen- eral (e.g., Kunda & Oleson, 1997; Weber & Crocker, 1983). There are both cognitive and motivational accounts of this phenomenon. Accounts based on models of exemplar-cat- egory dynamics suggest that the goodness of fit of the ex- emplar to the category determines the likelihood of a new group exemplar becoming linked to the category and de- termining the attributes associated with it (see Rothbart, 2001). Consistent with this idea are findings that exposure to moderately inconsistent group members leads to greater stereotype change than exposure to extremely inconsistent exemplars, and that the judged typicality of an inconsistent exemplar mediates the effect of extent of inconsistency on stereotype change (e.g., Johnston & Hewstone, 1992; Kun- da & Oleson, 1997).

Motivational accounts, on the other hand, suggest that because stereotypes serve important needs such as simpli- fying the social world or defending the social order (see Snyder & Miene, 1994), disregarding inconsistent exem- plars may result from a constructive attempt to preserve the stereotype (Kunda & Oleson, 1995, 1997; Yzerbyt, Coull, & Rocher, 1999). Although the role of motivational factors in stereotype change has been less extensively researched, there is some evidence underlining the importance of mo- tivational factors from research showing that perceivers use

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

neutral, pseudorelevant information to discount a stereo- type-disconfirming exemplar and maintain their stereotype (Kunda & Oleson, 1995) and that this is an active process that requires ample cognitive resources (Yzerbyt et al., 1999).

The Present Research

Rather than viewing these accounts of stereotype changes as competing models, we suggest that they are complemen- tary, and that both cognitive and motivational factors need to be taken into account if we are to fully understand how stereotypes change and are maintained. In the present re- search we thus integrate these approaches by simultaneous- ly examining (a) the category fit of the disconfirming ex- emplar as a cognitive variable, (b) SDO (Pratto et al., 1994) as an ideological orientation that is likely to motivate social perceivers to maintain the stereotype, as well as (c) the in- terplay of these variables.

Social dominance theory (SDT, Sidanius & Pratto, 1999) posits that societies minimize conflict by creating consensus on ideologies that promote the superiority of one group over others. Group stereotypes serve as hier- archy-enhancing legitimizing myths that justify status differences and discriminatory policies against members of certain social groups. According to SDT, individuals vary according to their desire for hierarchical vs. egali- tarian intergroup relations. The SDO scale developed by Pratto et al. (1994) captures such individual differences. Pratto et al. demonstrated that people high in SDO, com- pared to people low in SDO, tend to (1) more strongly hold beliefs that legitimize the unequal distribution of so- cial values such as jobs and healthcare (e.g., meritocracy, racism, sexism, ethnic prejudice), and (2) agree more with discriminatory policies and disagree more with pol- icies that promote equality. We propose that, if group ste- reotypes can serve to legitimize hierarchy, high SDO may predispose social perceivers toward maintaining stereo- types that justify the lower status of certain groups, even when confronted with stereotype-inconsistent informa- tion.

The present study was conducted in the context of young adults’ stereotypes of the elderly. Discrimination against old adults in society is one of the most legiti- mized, socially accepted and institutionalized forms of bias and is manifested in many areas of public life, such as the workplace and the health system (see Nelson, 2002). Moreover, ageist behavior and institutional prac- tices are legitimized by a stereotype of the elderly as in- tellectually incompetent, forgetful, senile, frail, slow, and dependent on others (e.g., Diekman & Hirnisey, 2007; Hummert, 1990; Perdue & Gurtman, 1990). In this study we examined whether the pretested elderly stereotype changes through exposure to a moderately vs. extremely stereotype-inconsistent exemplar and whether the

amount of change depends on individual differences in SDO.

Consistent with previous research (e.g., Johnston & Hewstone, 1992; Kunda & Oleson, 1997) we expected that stereotype change would be greater after exposure to moderately as opposed to extremely inconsistent infor- mation (Hypothesis 1), that is, we expected postmanipu- lation endorsement of stereotypical traits to be lower (Hypothesis 1a) and endorsement of counterstereotypical traits to be higher (Hypothesis 1b) after exposure to mod- erately as opposed to extremely inconsistent information. Based on previous research we further predicted that these effects would be mediated by differences in the judged typicality of the target person for elderly people in general (Hypothesis 2). Moreover, because stereotypes help to legitimize hierarchy and justify the low status of certain outgroups, we predicted that SDO would predict reduced stereotype change (Hypothesis 3). Specifically, SDO would be positively related to postmanipulation en- dorsement of stereotypical traits (Hypothesis 3a) and negatively to postmanipulation endorsement of counter- stereotypical traits (Hypothesis 3b), while controlling for baseline stereotyping.

We further examined whether the effects of SDO would depend on the extent of stereotype inconsistency. As noted earlier, people generally seem to view extreme- ly unrepresentative group members as uninformative and tend not to generalize their attributes to the group as a whole (Rothbart, 2001). Thus, motivational factors may have little impact when stereotype-inconsistent exem- plars are highly untypical. How perceivers view exem- plars which only moderately disconfirm the stereotype may, however, be more likely to be subject to motivation- al goals. This idea is consistent with the literature on Per- sonality × Situation interactions, which suggests that the impact of personality on behavior is greater in “weak” compared to “strong” situations (Mischel, 1977). Thus, we examined the possibility that SDO would have a greater impact on postmanipulation stereotyping when moderately (compared to extremely) inconsistent infor- mation is presented (Hypothesis 4).

Finally, we explored how SDO might affect stereotype maintenance. Kunda and Oleson (1995) suggested that social perceivers search for subjectively satisfactory rea- sons not to revise their stereotypes in the light of discon- firmation. One way of constructing such a justification is to discount a stereotype-inconsistent exemplar as atypi- cal of the group in general (see Kunda & Oleson, 1997). This is likely to be an active process that requires cogni- tive resources (Yzerbyt et al., 1999). If judging stereo- type-inconsistent exemplars as “exceptions to the rule” provides a means by which the stereotype can be main- tained, then SDO should be negatively related to the judged typicality of the stereotype-inconsistent target person, and the effects of SDO on postmanipulation ste- reotyping should be mediated by judged typicality (Hy- pothesis 5).

170 N. Tausch & M. Hewstone: Social Dominance Orientation and Stereotype Change

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

Method

Participants and Design

Participants were psychology undergraduates at a British university. They were recruited through a research partici- pation scheme (RPS) and received course credit for their participation. Data for this study were collected at two time points. Seventy (52 female, 18 male; mean age 19.06 years) participants completed the baseline assessment (Time 1) and 40 (29 female, 11 male; mean age 18.85 years) returned on average 10.90 (SD = 5.20) days later to participate in an “impression formation” experiment (Time 2). In the exper- imental session at Time 2, we used a 2 (Stereotype incon- sistency: moderate vs. extreme) × 2 (Sex of target person: male vs. female) between-subjects experimental design. The second factor was varied for control purposes only. Ten participants were randomly assigned to each experimental condition.

Procedure and Measures

Time 1

Participants signed up for a study on “personality and social attitudes.” They received a questionnaire, which, among a number of other measures, contained Pratto et al.’s (1994) SDO scale (16 items; e.g., “Sometimes other groups must be kept in their place”; α = .90), to which participants re- sponded using scales ranging from 1 (strongly disagree) to 7 (strongly agree), as well as four stereotyping measures: the perceived central tendency of the elderly on stereotyp- ical and counterstereotypical traits and measures of confi- dence that the elderly possess stereotypical and counterste- reotypical traits. To rate perceived central tendency, partic- ipants marked an “X” on a 100 mm scale with the endpoints not at all and very, according to where they thought the elderly on average fell on five pretested stereotypical traits (ailing, forgetful, frail, slow, tired) and five pretested count- erstereotypical traits (energetic, fit, healthy, sturdy, quick). The traits were presented in one of two different random orders. We averaged ratings (in mm) to create indices of central tendency on stereotypical (α = .83) and counterste- reotypical (α = .72) traits. Participants then rated how con- fident they were that elderly people, on average, possessed each of these traits on 100 mm scales ranging from not at all confident to extremely confident. Two indices, for con- fidence in stereotypical (α = .69) and in counterstereotyp- ical (α = .74) traits, were created. Upon completion, par- ticipants were thanked for their participation and partially debriefed.

Time 2

Participants signed up for an apparently unrelated study on “impression formation,” which was advertised 1 week la- ter. Participation in the “personality and attitudes” study (i.e., Time 1) was, however, set as a precondition for par- ticipation. Upon arrival in the laboratory, participants were randomly assigned to one of the four experimental condi- tions and were then presented with information about a tar- get person via a computer screen. First, participants read a name (John or Marie) and an introductory paragraph that stated the person’s age (76), that he or she lived in Dorches- ter, England, and that he or she used to work for a local company until retiring 7 years ago. Participants were then presented with 11 allegedly self-reported statements de- scribing activities of the target person that were either mod- erately stereotype inconsistent (“I get up early and get the household chores done quickly, so I can take my dog for a walk”; “I joined the local dancing club 2 years ago; it’s a great way to meet new people”) or extremely stereotype inconsistent (e.g., “Every day I get up very early in the morning and go on a fast run with my two dogs”; “I joined the local squash club 2 years ago; it’s a great way to meet new people”). The sentences were presented one at a time upon key press and appeared in a randomized order. The statements and overall profiles were extensively pretested to ensure that they represented moderately and extremely stereotype-inconsistent information, respectively1.

Next, participants rated the target person on their typi- cality using 7-point scales ranging from 1 (not at all) to 7 (extremely) (5 items; e.g., “How typical is John/Marie of elderly people in general?”; α = .97) as well on a number of filler items. They then completed the stereotyping mea- sures, which were identical to those used at Time 1. Again we computed indices of central tendency on stereotypical (α = .90) and counterstereotypical (α = .77) traits, and con- fidence in stereotypical (α = .84) and counterstereotypical (α = .78) traits. Finally, as a manipulation check, partici- pants rated the extent to which they thought the target per- son possessed each of the stereotype-consistent and stereo- type-inconsistent traits on scales ranging from 1 (not at all) to 7 (extremely). We created indices of the extent to which the target person was judged to possess stereotypical (α = .60) and counterstereotypical (α = .93) traits. Upon com- pletion, participants were thanked for their participation, probed for suspicions about the hypotheses of the study, and fully debriefed.

Results

Responses from the two measurement points were matched using participants’ RPS numbers. There were no signifi-

N. Tausch & M. Hewstone: Social Dominance Orientation and Stereotype Change 171

� Further information on pretesting procedures and the impression formation task can be obtained from the first author.

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

cant differences between those participants who returned at Time 2 and those who did not in terms of SDO scores or stereotyping measures. There was no effect of gender of target person on any of the dependent variables, nor were there interactions between gender of target and gender of participant; we therefore collapsed the two counterbal- anced versions.

Preliminary Analyses

Table 1 presents the means and standard deviations for all dependent variables as a function of time of measurement and experimental condition. Our main analyses assessed whether the amount of stereotype change depended on the extent of inconsistency of the target person and individual differences in SDO.

First, we checked whether our manipulation of extent of stereotype inconsistency was successful. The extremely stereotype-inconsistent target was rated significantly lower on stereotypical traits (M = 1.59, SD = .50) than the mod- erately inconsistent target (M = 1.91, SD = .41), F(1, 38) = 4.93, p = .032, and significantly higher on stereotype-in- consistent traits (M = 6.36, SD = .37) than the moderately inconsistent target (M = 5.79, SD = .63), F(1, 38) = 12.05, p = .001, demonstrating a successful manipulation of ex- tent of stereotype inconsistency.

Second, we examined whether there were significant changes in stereotyping following the presentation of ste-

reotype-inconsistent information. Overall, there was a (marginally) significant reduction in the rated central ten- dency of stereotypical traits (MT1 = 71.09, MT2 = 68.23; t(39) = 1.95, p = .058). There was also a significant reduc- tion in confidence that elderly people possessed stereotyp- ical traits (MT1 = 68.50, MT2 = 65.03; t(39) = 2.36, p = .023) and increased confidence that elderly people possessed counterstereotypical traits (MT1 = 33.20, MT2 = 36.06; t(39) = –2.34, p = .025). There was no overall change, how- ever, in the rated central tendency of counterstereotypical traits (MT1 = 33.50, MT2 = 34.19; t(39) = –.54, p = .590).

The Impact of Target Inconsistency and Social Dominance Orientation on Stereotype Change

First, we performed a series of hierarchical regressions to examine the overall effects and interaction of target incon- sistency and SDO on stereotype change (see Aiken & West, 1991). In Step 1 we regressed postmanipulation stereotyp- ing scores on baseline stereotyping, the experimental ma- nipulation (effects-coded –1 for the moderately inconsis- tent condition and +1 for the extremely inconsistent condi- tion), and centered SDO scores. As expected, and consistent with previous research, stereotype change was reduced when the disconfirming exemplar was extremely stereotype-inconsistent than when it was only moderately inconsistent. Specifically, the manipulation was positively

Table 1. Means and standard deviations of dependent variables as a function of condition

Dependent variable Moderately inconsistent Extremely inconsistent

M (SD) M (SD)

Central tendency

Stereotypical

Time 1 71.10 (12.34) 71.07 (8.85)

Time 2 63.85 (13.93) 72.61 (9.22)

Change (Time 2 – Time 1) –7.25 (10.50) +1.54 (4.99)

Counterstereotypical

Time 1 34.89 (10.69) 32.11 (10.19)

Time 2 36.78 (11.09) 31.59 (9.24)

Change (Time 2 – Time 1) +1.89 (9.08) –0.52 (6.71)

Confidence

Stereotypical

Time 1 67.78 (13.16) 69.22 (7.47)

Time 2 62.83 (15.84) 67.22 (9.31)

Change (Time 2 – Time 1) –4.95 (10.90) –2.00 (7.38)

Counterstereotypical

Time 1 34.66 (11.65) 31.74 (9.18)

Time 2 37.90 (13.23) 34.21 (7.87)

Change (Time 2 – Time 1) +3.24 (8.93) +2.47 (8.01)

Target Typicality (Time 2) 3.86 (1.20) 2.03 (.54)

Note. Ratings of central tendency and confidence could range from 0 to 100, and target typicality from 1 to 7.

172 N. Tausch & M. Hewstone: Social Dominance Orientation and Stereotype Change

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

related to the central tendency of stereotypical traits at Time 2 (β = .40, p = .001), negatively related to the central tendency of counterstereotypical traits at Time 2 (β = –.23, p = .048), and positively related to confidence in stereotyp- ical traits at Time 2 (β = .24, p = .025). Although the ma- nipulation was, as expected, negatively related to confi- dence in counterstereotypical traits at Time 2, this relation- ship was not significant (β = –.14, p = .227).

Consistent with our hypotheses, SDO was positively re- lated to confidence in stereotypical traits at Time 2 (β = .24, p = .025), and negatively related to central tendency of (β = –.26, p = .027) and confidence in (β = –.24, p = .032) counterstereotypical traits at Time 2. Although SDO was positively related to central tendency of stereotypical traits at Time 2 as we had predicted, this relationship did not reach statistical significance (β = .15, p = .186). In Step 2, we added the product term of the manipulation and cen- tered SDO scores to examine the interaction between the effects of SDO and the stereotype-consistency manipula- tion. Consistent with our expectations, SDO only predicted postmanipulation stereotyping in the moderately (and not in the extremely) inconsistent condition (for central tenden- cy of counterstereotypical traits β = –.34, p = .053, and β = –.13, p = .419, respectively; for confidence in stereotypical traits β = .48, p = .001, and β = .22, p = .252, respectively; and for confidence in counterstereotypical traits (β = –.34, p = .041, and β = –.05, p = .795, respectively). However, the interaction terms for these tests did not reach statistical significance (β = .10, p = .389; β = –.17, p = .090; and β = .14, p = .195, respectively; see Table 2 for a summary of results).

Mediation by Judged Target Typicality

We examined whether the effects of the stereotype-consis- tency manipulation and SDO on stereotype change were mediated by judged target typicality. The manipulation of stereotype consistency had the expected negative effect on target typicality over and above SDO scores (β = –.77, p < .001), indicating that the target person was judged to be less

typical of elderly people in the extremely inconsistent com- pared to the moderately inconsistent condition. To test for mediation, we regressed Time 2 stereotyping measures on baseline scores, SDO, the consistency manipulation, and typicality ratings. Typicality emerged as a significant pre- dictor of Time 2 central tendency of stereotypical traits (β = –.30, p = .050) and the effect of the manipulation ceased to be significant (β = .17, p = .276), suggesting that judged target typicality fully mediated the effect of the manipula- tion (Sobel Test: z = 1.94, p = .052).

Similarly, typicality was a significant predictor of con- fidence in stereotypical traits at Time 2 (β = –.43, p = .002) and the effect of the manipulation ceased to be significant with typicality in the equation (β = –.09, p = .499), indicat- ing full mediation (Sobel Test: z = 2.98, p = .003). Further- more, although the manipulation did not have a direct effect on confidence in counterstereotypical traits, typicality emerged as a significant predictor of postmanipulation scores (β = .54, p = .001), consistent with an indirect effect of the manipulation via typicality (Sobel Test: z = –3.06, p = .002). For central tendency of counterstereotypical traits, however, typicality did not emerge as a significant predictor (β = .19, p = .285) and was thus not a mediator.

Next we examined whether judged target typicality me- diated SDO effects. SDO had only a marginally (although in the predicted direction) significant effect on typicality ratings (β = –.20, p = .100). There was, however, a signif- icant interaction between SDO and the stereotype-consis- tency manipulation (β = .23, p = .043), which indicated that SDO had an effect on typicality ratings only in the moderately (β = –.45, p = .049) and not the extremely (β = .20, p = .370) inconsistent condition. To assess whether typicality ratings mediated SDO effects on the dependent variables in the moderately inconsistent condition, we re- gressed stereotyping scores on baseline scores, SDO, and judged target typicality. For central tendency ratings of counterstereotypical traits, typicality did not emerge as a significant predictor (β = .04, p = .833) and was therefore not a mediator. For confidence in consistent traits, both typ- icality (β = –.31, p = .013) and SDO (β = .34, p = .007) were predictors but the mediated effect did not reach sig-

Table 2. Results of hierarchical regression analyses

Criterion variable Central tendency Confidence Target

Stereotypical Counterstereotypical Stereotypical Counterstereotypical typicality

Step 1

Time 1 score .66*** .69*** .69*** .74*** –

Target inconsistency condition .40** –.23* .24* –.14 –.77***

SDO .15 –.26* .40*** –.24* –.20+

R2 .62*** .59*** .67*** .61*** .54***

Step 2

SDO × Target inconsistency .00 .10 –.17+ .14 .23*

R2 change .00 .01 .03+ .02 .05*

Note. Table shows standardized regression weights; N = 60; *p < .05, **p < .01, ***p < .001, + p = .10. Target inconsistency was coded as –1 = moderate; +1 = extreme.

N. Tausch & M. Hewstone: Social Dominance Orientation and Stereotype Change 173

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

nificance (Sobel Test: z = 1.69, p = .091). Finally, although SDO ceased to be a significant predictor of confidence in inconsistent traits (β = –.15, p = .315) when typicality (β = .38, p = .041) was entered into the equation, the size of the mediated effect was not significant (Sobel Test: z = 1.53, p = .125) Thus, overall, there was little evidence that typi- cality ratings mediated SDO effects.

Discussion

The present research replicated the well-established find- ing that moderately inconsistent information is more effec- tive than extremely inconsistent information in reducing stereotyping (e.g., Johnston & Hewstone, 1992; Kunda & Oleson, 1997). The main objective of this study, however, was to extend previous research by investigating the impact of perceivers’ ideological orientation on stereotype change in response to disconfirming information. Motivational ac- counts of stereotype maintenance suggest that perceivers are motivated to preserve their stereotypes because stereo- types serve important needs, such as defending the social order (see Kunda & Oleson, 1995; Snyder & Miene, 1994). Based on this idea we predicted that, because stereotypes can serve as hierarchy-enhancing legitimizing myths (Sida- nius & Pratto, 1999), people high in SDO, an ideological orientation related to the desire for hierarchical vs. egali- tarian intergroup relations, should be more motivated to maintain their stereotypes compared to people low in SDO, and SDO should therefore be negatively related to stereo- type change. Consistent with this idea, we presented initial evidence that SDO attenuates stereotype change in re- sponse to disconfirming information. Specifically, SDO was (over and above baseline stereotyping scores and the stereotype-consistency manipulation) significantly nega- tively related to the perceived central tendency of and con- fidence in counterstereotypical traits and positively related to confidence in stereotypical traits. There was also a pos- itive (albeit nonsignificant) relationship between SDO and the central tendency of stereotypical traits.

Furthermore, based on the idea that individual differenc- es have implications in particular when other constraints (such as the extreme atypicality of an exemplar that would generally inhibit generalization) are absent (Mischel, 1977), we expected that SDO would have an impact on stereotyping in particular when moderately rather than ex- tremely inconsistent information is presented. In line with this idea, SDO was related to stereotyping only in the mod- erately inconsistent condition. Nonetheless, the interaction terms for these tests did not reach statistical significance. It should be noted, however, that interaction effects are no- toriously difficult to detect in multiple regression because of low power (see also McClelland & Judd, 1993). Thus, the present results can only be seen as initial evidence and need to be replicated in further studies, which could, for example, simultaneously manipulate motivational goals

and exemplar goodness of fit. Such future work could have important implications for further theoretical integration of the prototype (Rothbart, 2001) and motivational (Kunda & Oleson, 1995) accounts of stereotype change. We would expect that motivational factors interact with the goodness of fit of stereotype-inconsistent exemplars and operate in particular when the counterexample is relatively typical of the group.

In addition to examining the impact of extremity of ste- reotype-inconsistent information and individual differenc- es in SDO, as well as their interaction, on stereotype change, we also aimed to illuminate the psychological pro- cess by which these variables may lead to stereotype main- tenance vs. change. Previous research on stereotype change has indicated that the judged typicality of a disconfirming exemplar is central to the process of stereotype change and mediates the effect of the extent of disconfirmation on ste- reotyping (Johnston & Hewstone, 1992). Replicating this work, judged typicality of the target person fully mediated the effects of the manipulation on stereotype change in the present study. Kunda and Oleson (1995, 1997) further sug- gested that motivated social perceivers search for subjec- tively satisfactory reasons not to revise their stereotypes, and that judging a stereotype-disconfirmer as atypical may provide a successful means of constructing such a justifi- cation. We thus explored the possibility that SDO would exert its effect on stereotype change by affecting judgments of the target person’s representativeness. Although, consis- tent with such an account, SDO was negatively related to judged typicality in the moderately inconsistent condition, there was overall very little evidence that judged typicality mediated the effects of SDO on stereotype change. This pattern is intriguing as it suggests that perceivers who are motivated by an ideology that supports group-based hier- archy do not necessarily use target atypicality (see Kunda & Oleson, 1997) to maintain their stereotypes. Future re- search should examine this idea further, and also explore the potential importance of other subjective justifications for stereotype maintenance in the face of disconfirmation.

Implications

The present research presents initial evidence showing that individual differences in ideological orientation attenuate stereotype change in response to disconfirmation. Our work underlines the importance of including both cognitive and motivational factors, as well as their interplay, in anal- yses of the malleability of social stereotypes. Nonetheless, the present study can only be seen as a first step in that direction. Future research should aim to replicate and ex- tend the present findings and examine additional individual difference variables that may further bias perceivers toward stereotype maintenance, such as right-wing authoritarian- ism (RWA, Altemeyer, 1988) and need for structure (Neu- berg & Newsom, 1993), which represent different motiva- tional goals (defending the social order and simplifying the

174 N. Tausch & M. Hewstone: Social Dominance Orientation and Stereotype Change

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

social world). A potentially fruitful avenue for future re- search could also be to examine whether certain individual difference variables differentially predict the maintenance of stereotypes of specific target outgroups. Duckitt (2006), for example, demonstrated that while RWA predicts atti- tudes toward groups seen as threatening social control and order (e.g., drug dealers, rock stars), SDO predicts attitudes toward socially subordinate groups with lower status and power (housewives, physically disabled persons, and un- employment beneficiaries) that may activate competitive- ness over relative dominance. Thus, future research may examine whether RWA and SDO differentially motivate the maintenance of stereotypes of different types of out- groups.

To conclude, we emphasize that, rather than implying that stereotyping and prejudice are inevitable for “preju- diced personalities,” as suggested by early theorists (All- port, 1954), our findings give insight into how interven- tions to reduce stereotyping could be customized to be maximally effective. Levy (1999) suggested that although individual differences that affect stereotyping and preju- dice reflect relatively stable tendencies, they can be mal- leable if targeted by customized intervention programs (e.g., Schaller, Asp, Rossel, & Heim, 1996). For example, Pratto, Tatar, and Conway-Lanz (1999) demonstrated that SDO can be reduced when participants take on nondomi- nating roles in role-plays. Thus, a fruitful avenue for future research would be to examine whether training programs preceding interventions to reduce stereotyping through ex- posure to stereotype-inconsistent information would in- crease the effectiveness of such interventions. We hope that the present research will stimulate further work on this im- portant issue.

Acknowledgments

The research reported in this article was carried out as part of the first author’s doctoral thesis, supervised by the sec- ond author. It was funded by a doctoral studentship awarded by the University of Oxford. This manuscript was prepared while Nicole Tausch was a British Academy Post- doctoral Fellow.

References

Adorno, T. W., Frenkel-Brunswick, E., Levinson, D. J., & San- ford, R. N. (1950). The authoritarian personality. New York: Harper.

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interaction. New York: Sage.

Allport, G. W. (1954). The nature of prejudice. Reading, MA: Ad- dison-Wesley.

Altemeyer, R. A. (1988). Enemies of freedom: Understanding right-wing authoritarianism. San Francisco: Jossey-Bass.

Diekman, A. B., & Hirnisey, L. (2007). The effect of context on the silver ceiling: A role congruity perspective on prejudiced responses. Personality and Social Psychology Bulletin, 33, 1353–1366.

Duckitt, J. (2005). Personality and prejudice. In J. Dovidio, P. Glick, & L. A.Rudman (Eds.), On the nature of prejudice: Fifty years after Allport (pp. 395–412). Oxford: Blackwell.

Duckitt, J. (2006). Differential effects of Right Wing Authoritarian- ism and Social Dominance Orientation on outgroup attitudes and their mediation by threat from competitiveness to outgroups. Personality and Social Psychology Bulletin, 32, 684–696.

Hewstone, M. (1994). Revision and change of stereotypic beliefs: In search of the elusive subtyping model. In W. Stroebe & M. Hewstone (Eds.), European review of social psychology (Vol. 5, pp. 69–109). Chichester, UK Wiley.

Hummert, M. L. (1990). Multiple stereotypes of elderly and young adults: A comparison of structure and evaluations. Psy- chology and Aging, 5, 182–193.

Johnston, L., & Hewstone, M. (1992). Cognitive models of ste- reotype change (3): Subtyping and the perceived typicality of disconfirming group members. Journal of Experimental Social Psychology, 28, 360–386.

Kunda, Z., & Oleson, K. C. (1995). Maintaining stereotypes in the face of disconfirmation: Constructing grounds for subtyp- ing deviants. Journal of Personality and Social Psychology, 68, 565–579.

Kunda, Z., & Oleson, K. C. (1997). When exceptions prove the rule: How extremity of deviance determines the impact of de- viant examples on stereotypes. Journal of Personality and So- cial Psychology, 72, 965–979.

Levy, S. R. (1999). Reducing prejudice: Lessons from social-cog- nitive factors underlying perceiver differences in prejudice. Journal of Social Issues, 55, 745–765.

McClelland, G. H., & Judd, C. M. (1993). Statiscal differences of detecting interactions and moderator effects. Psychological Bulletin, 114, 376–390.

Mischel, W. (1977). The interaction between person and situation. In D. Magnusson & N. S. Endler (Eds.), Personality at the crossroads: Current issues in interactional psychology. Hills- dale, NJ: Erlbaum.

Nelson, T. D. (Ed.). (2002). Ageism: Stereotyping and prejudice against older persons. Cambridge, MA: MIT Press.

Neuberg, S. L., & Newsom, J. T. (1993). Personal need for struc- ture: Individual differences in the desire for simple structure. Journal of Personality and Social Psychology, 65, 113–131.

Perdue, C. W., & Gurtman, M. B. (1990). Evidence for the autom- atism of ageism. Journal of Experimental Social Psychology, 26, 199–216.

Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1994). Social dominance orientation: A personality variable predict- ing social and political attitudes. Journal of Personality and Social Psychology, 67, 741–763.

Pratto, F., Tatar, D. G., & Conway-Lanz, S. (1999). Who gets what and why: Determinants of social allocations. Political Psychol- ogy, 20, 127–150.

Rothbart, M. (2001). Category dynamics and the modification of outgroup stereotypes. In R. Brown & S. Gaertner (Eds.), Blackwell handbook of social psychology: Intergroup process- es. Oxford: Blackwell.

Schaller, M., Asp, C., Rosell, M. C., & Heim, S. J. (1996). Train- ing in statistical reasoning inhibits the formation of erroneous

N. Tausch & M. Hewstone: Social Dominance Orientation and Stereotype Change 175

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

group stereotypes. Personality and Social Psychology Bulle- tin, 22, 829–844.

Sidanius, J., & Pratto, F. (1999). Social dominance: An intergroup theory of social hierarchy and oppression. Cambridge: Uni- versity Press.

Snyder, M., & Miene, P. (1994). On the functions of stereotypes and prejudice. In M. P. Zanna & J. M. Olson (Eds.), The psy- chology of prejudice: The Ontario symposium (pp. 33–54). Hillsdale, NJ: Erlbaum.

Weber, R., & Crocker, J. (1983). Cognitive processes in the revi- sion of stereotypic beliefs. Journal of Personality and Social Psychology, 45, 961–977.

Yzerbyt, V. Y., Coull, A., & Rocher, S. J. (1999). Fencing off the deviant: The role of cognitive resources in the maintenance of

stereotypes. Journal of Personality and Social Psychology, 77, 449–462.

Nicole Tausch

School of Psychology Cardiff University Tower Building, Park Place Cardiff, CF10 3AT UK E-mail [email protected]

176 N. Tausch & M. Hewstone: Social Dominance Orientation and Stereotype Change

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .