Personality and Social Psychology Review Copyright © 2006 by 2006, Vol. 10, No. 4, 336-353 Lawrence Erlbaum Associates, Inc.

Intergroup Threat and Outgroup Attitudes: A Meta-Analytic Review Blake M. Riek, Eric W. Mania, and Samuel L. Gaertner Department ofPsychology University ofDelaware

This article examines the relationship between intergroup threat and negative outgroup attitudes. We first qualitatively review the intergroup threat literature, de- scribing the shiftfrom competing theories toward more integrated approaches, such as the integrated threat theory (ITT; W G. Stephan & Stephan, 2000). The types of threats discussed include: realistic threat, symbolic threat, intergroup anxiety, nega- tive stereotypes, group esteem threat, and distinctiveness threat. We then conducted a quantitative meta-analysis examining the relationships between various intergroup threats and outgroup attitudes. The meta-analysis, involving 95 samples, revealed that 5 different threat types had a positive relationship with negative outgroup atti- tudes. Additionally, outgroup status moderated some ofthese relationships. Implica- tions andfuture directions are considered.

This article addresses aproposed cause ofintergroup which proposes that when two groups are in competi- bias and prejudice: intergroup threat. As a general defi- tion for scarce resources, the potential success of one nition, intergroup threat occurs when one group's ac- group threatens the well-being ofthe other, resulting in tions, beliefs, or characteristics challenge the goal at- negative outgroup attitudes (Sherif & Sherif, 1969). tainment or well-being of another group. To date, These resources may be tangible (e.g., money) or may although a number of separate studies support the im- involve issues of power or control (e.g., political portance of intergroup threat for understanding the eti- power; Bobo, 1988). Related to the competition over ology ofintergroup bias (for reviews ofspecific types of resources are issues of intergroup goals. Sherif and threats see J. Jackson, 1993; Sears & Henry, 2003) no Sherif (1969) proposed that when the goals of different systematic, quantitative review of this literature has groups are complementary, relations between the been conducted. In this article we first review the theo- groups will be positive, however, when conflicting ries linking intergroup threat to intergroup bias using a goals exist relations will deteriorate. The resulting con- qualitative approach. Second, we utilize quantitative, flict may increase ingroup solidarity, which in turn meta-analytic procedures to examine the effects of in- widens the ingroup/outgroup distinction, creating in- tergroup threat on outgroup attitudes across the litera- tergroup hostility. ture, focusing primarily ontheutilityofintegrative theo- To test RGCT, Sherif et al. (1961) divided children ries such as the integrated threat theory (ITT; W. G. in a summer camp program into two separate groups. Stephan & Stephan, 2000). This quantitative approach Once group boundaries were established, the children also permits systematic examination of potential mod- engaged in a number of zero-sum, competitive activi- erators ofthe proposed relationships between threatand ties (e.g., sporting events). Because these activities per- attitudes, such as outgroup status. Third, we discuss fu- mitted victory of only one group over the other, one ture directions for research, including potential strate- group's success precluded and thus threatened the suc- gies for reducing intergroup threat. cess ofthe other group. As competition continued, hos- tility between the groups increased. This hostility be- came quite severe and in some cases led to physical Realistic Group Conflict Theory violence. The bias between the groups decreased only when common goals that required intergroup coopera- One of the earliest theories to consider intergroup tion were introduced. In addition to the camp study, threat was realistic group conflict theory (RGCT), other researchers have shown that both incompatible goals and perceived competition between groups are We thank Lowell Gaertner, Walter Stephan, Ivan Sucharski, related to negative outgroup attitudes, lower levels of Miles Hewstone, and one anonymous reviewer for their comments support for affirmative action programs, and stereotyp- and suggestions on earlier drafts of this article. ing (Beaton & Tougas, 2001; Langford & Ponting, Correspondence should be sent to Blake Riek, Department of Psychology, University of Delaware, Newark, DE 19716. E-mail: 1992; Tougas, Brown, Beaton, & Joly, 1995; Watts, [email protected] 1996; Zarate, Garcia, Garza, & Hitlan, 2004).

Downloaded from psr.sagepub.com336 at Harvard Libraries on November 15, 2016 INTERGROUP THREATS AND ATTITUDES

Although initial studies of RGCT presented groups Symbolic racism proposes that racial bias is no longer with threats that directly impacted the individuals in- the result of perceptions that Blacks are biologically volved, it is proposed that biases between groups may inferior to Whites (i.e., traditional racism), but instead still occur when there is not a direct threat to the self is due to beliefs that Blacks violate and thereby (Bobo, 1983). In this sense, when the interests of a threaten values that are important to Whites (Sears, group as a whole are threatened, members perceive this 1988). For example, White Americans may believe as threatening even though self-interest is not directly that the value of equity prevalent in the White-domi- impacted. For example, a White male may perceive af- nated American culture is threatened by affirmative ac- firmative action as threatening the overall interests of tion programs. Many Whites may not believe that prej- his ingroup even when he is not personally affected. udice and discrimination are still prevalent and Since the initial studies on RGCT, the relationship therefore see such programs as violating the equity between conflict and intergroup bias has been shown value by giving minorities an unfair advantage. This is across a variety of situations where one group threatens a different construal ofWhites' reactions to affirmative the interest of another group. Brown, Maras, Masser, action than that proposed by RGCT, with the threat Vivian, and Hewstone (2001) examined conflict and arising from conflicting values rather than perceived intergroup bias between British ferry passengers and competition over resources. the French during a period when French fishermen Kinder and Sears (1981) compared the competing were blockading a port used by the British ferry, which theories of RGCT and symbolic racism in a study of threatened the goals of the British passengers. In line voting patterns in a mayoral race in which one candi- with the predictions of RGCT, both perceived and ac- date was Black and the other was White. The likeli- tual (objective) conflict between the groups were re- hood of an individual voting for the Black candidate lated to higher levels of aggression toward the French. was compared with reported levels of realistic and In another study, Quillian (1995) examined national symbolic threat. Realistic threat was measured by as- survey data in 12 countries and found that when eco- sessing how much individuals felt they would be per- nomic conditions were poor and the size of a racial mi- sonally affected by such issues as busing and desegre- nority or immigrant group was large in proportion to gation; the symbolic threat measure focused on general the majority group, bias toward these groups was high. perceptions of value violations by Blacks (e.g., push- Presumably, when there is a large outgroup presence, ing too hard; have gotten more than they deserve). economic difficulties are attributed to these outgroup Consistent with the symbolic racism perspective, sym- members who are seen as threatening ingroup bolic racism was a stronger predictor ofcandidate pref- well-being. erence than was realistic threat. Recently, other researchers have applied RGCT to Other research has shown that threats to important explain bias toward immigrant groups by assuming values can increase intergroup bias. Biernat, Vescio, that the perception that immigrant gains come at the and Theno (1996) found that the majority of Whites expense ofthe resident group creates negative attitudes perceived Blacks as not supporting their values and toward immigration and immigrants (Esses, Dovidio, these Whites had relatively more negative evaluations Jackson, & Armstrong, 2001; Esses, Jackson, & of Blacks than those Whites who perceived Blacks as Armstrong, 1998). For example, ifimmigrants are seen supporting their values. A related study indicated that as usurping jobs and thereby threatening the economic as the perceived gap between ingroup and outgroup well-being ofresidents, it may result in a negative view values increases, so do negative outgroup attitudes of the immigrant group. McLaren (2003) measured (Dunbar, Saiz, Stela, & Saez, 2000). Additionally, as levels of realistic threat in 17 European countries and postulated by Kinder and Sears (1981), threats to val- found that levels of perceived realistic threat were re- ues have also been shown to influence attitudes toward lated to a preference for the expulsion of immigrants. social policies aimed at helping minorities (Sawires & Peacock, 2000). Although symbolic racism was initially formulated Symbolic Threat to explain the anti-Black bias of Whites, the relation- ship between symbolic threat and intergroup bias has Although RGCT explains how conflict over re- been revealed in other intergroup settings as well, in- sources can create perceptions of threat, it does not ad- cluding body-type and sexual orientation groups dress instances of bias in the absence of such conflict. (Crandall, 1994; L. M. Jackson & Esses, 1997; Nicho- Other researchers, however, developed an alternative las & Durrheim, 1995). For example, individuals who theory involving the influence of threat on intergroup see homosexuals as violating important values (e.g., bias. Proponents of symbolic racism theory have pro- beliefthat homosexuality is amoral) tend to have nega- posed that racism results from conflicting values and tive attitudes toward homosexuals (Haddock, Zanna, & beliefs rather than from competition or conflicting Esses, 1993; Wyman & Snyder, 1997). Zanna (1994) goals (Kinder & Sears, 1981; McConahay, 1982). demonstrated that the relationship between value vio-

Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 337 RIEK, MANIA, GAERTNER lations and negative outgroup attitudes holds across a ence of outgroup members because of uncertainty number of different target outgroups. about how to behave toward them, which makes inter- Symbolic threats have also provided an alternative actions with outgroups seem threatening (W. G. explanation for biases against immigrants. Although Stephan & Stephan, 1985). Intergroup anxiety has RGCT focuses on economic threats as explanations for been demonstrated to be a predictor of outgroup atti- prejudice against immigrants, symbolic threat theory tudes and bias (Ho & Jackson, 2001; Islam & (Sears, 1988) proposes that these biases are due to per- Hewstone, 1993; Voci & Hewstone, 2003) and further- ceived threats to cultural values. Multiple studies have more, people who tend to be generally high in anxiety illustrated that perceived threats to the ingroup's values often exhibit higher levels ofprejudice (Hassan, 1978). by foreigners are related to increases in negative atti- Plant and Devine (2003) examined both antecedents tudes toward immigrants (Esses, Haddock, & Zanna, and consequences of intergroup anxiety and found that 1993; Esses, Hodson, & Dovidio, 2003). Additionally, unfavorable outgroup contact and negative expectan- in some instances, beliefs that foreigners would nega- cies both led to increases in intergroup anxiety. As in- tively impact or undermine American values is a stron- tergroup anxiety increased so did hostility and the de- ger predictor of negative attitudes toward immigrants sire to avoid contact with outgroup members. The than perceptions of realistic threats (McLaren, 2003). relationship between intergroup anxiety and negative outgroup attitudes has been observed across a variety of natural settings as well as in the laboratory (Brown Integrated Threat Theory et al., 2001). Negative stereotypes generate threat by creating Although symbolic racism and RGCT propose dif- negative expectations concerning the behavior of ferent specific causes for negative outgroup attitudes, outgroup members. Stereotypes have long been associ- each regards beliefs that the outgroup is threatening as ated with negative outgroup attitudes (Eagly & a fundamental factor in determining outgroup atti- Mladinic, 1989; Spencer-Rodgers & McGovern, tudes. Initially, realistic conflict and symbolic theories 2002). Additionally, negative expectancies influence of intergroup relations were in direct conflict with one social information processing and social judgments another as the literature of the time reflects (Bobo, (Hamilton, Sherman, & Ruvolo, 1990; Ybarra, 1983; Kinder & Sears, 1981; Sniderman & Tetlock, Schaberg, & Keiper, 1999). W. G. Stephan and Stephan 1986). More recently, researchers have begun to con- (1996) postulated that because negative stereotypes sider these two conceptions of threat as complemen- represent negative expectations about outgroups, nega- tary rather than as mutually exclusive. Studies that tive stereotypes occur in conjunction with negative have measured both realistic and symbolic threats have emotions (e.g., fear, anger) toward the outgroup, which shown that both types of threats can account for unique intensify negative outgroup attitudes. portions of the variance in attitudes toward outgroups The four threats are proposed to contribute to nega- (McLaren, 2001; Wilson, 2001). tive outgroup attitudes and robust support for the ITT Recently, W. G. Stephan and Stephan (1996, 2000) has been found across a number of intergroup contexts. proposed ITT, which classifies threats into four major All four types of threat predict attitudes toward racial types: realistic threat, symbolic threat, intergroup anxi- outgroups in both White and Black samples (W. G. ety, and negative stereotypes. Rather than conceptual- Stephan et al., 2002). The theory has also been utilized izing the relationship between threat and attitudes as to predict gender attitudes (C. W. Stephan, Stephan, stemming from either competition or value conflict, Demitrakis, Yamada, & Clason, 2000) and attitudes to- ITT proposes that both can influence outgroup atti- ward various immigrant groups (W. G. Stephan, tudes simultaneously. Within ITT, realistic threat is Ybarra, & Bachman, 1999; W. G. Stephan, Ybarra, similar to the threats considered by RGCT and in- Martinez, Schwarzwald, & Tur-Kaspa, 1998). In addi- cludes perceptions of competition, conflicting goals, tion to racial, gender, national, and ethnic groups, the and threats to physical and economic well-being of the integrated threat model has been used to predict nega- ingroup. Symbolic threat is similar to the ideas under- tive attitudes toward cancer patients and people with lying symbolic racism, where threat arises from a con- AIDS (Berrenberg, Finlay, Stephan, & Stephan, 2002). flict in values, norms, and beliefs between groups. Furthermore, a number of antecedents to intergroup Although intergroup anxiety and negative stereo- threat have been identified. High levels of negative types have not been traditionally considered to be contact between groups predict higher levels of per- forms of intergroup threat, W. G. Stephan and col- ceived threat as does high ingroup identification and leagues argued that they also reflect concerns about perceived intergroup conflict (Corenblum & Stephan, negative outcomes from intergroup relations and there- 2001; W. G. Stephan et al., 2000). Status differences fore constitute unique types ofthreat (W. G. Stephan & between groups also relate to higher levels of threat, Stephan, 1996, 2000). Intergroup anxiety involves such that as the difference in status increases the degree feelings of uneasiness and awkwardness in the pres- of perceived threat increases (W. G. Stephan et al.,

338 Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 INTERGROUP THREATS AND ATTITUDES

2002). Moreover, outgroup status differentially influ- correlational research designs. For example, ences reactions to threat. Cadinu and Reggiori (2002) Branscombe and Wann (1994) found that when highly found that group threat led to stronger devaluation of identified American participants viewed a film in low status relative to high status outgroups. which an American boxer was beaten by a Russian In terms of negative outgroup stereotypes, there has boxer it led to decreases in their collective self-esteem, been some question as to whether they are best concep- which in turn increased the likelihood of derogating tualized as an independent threat that directly predicts Russians. Similarly, Branscombe, Spears, Ellemers, outgroup attitudes or as an antecedent of the other and Doosje (2002) threatened the prestige of a group types ofthreat (i.e., realistic, symbolic, anxiety). To ex- by informing participants that members of another plore this issue, W. G. Stephan and colleagues (2002) group had negatively evaluated them, thereby threaten- measured all four types ofthreat and outgroup attitudes ing their group's esteem. Consequently, this led to among Black and White samples and compared two lower perceptions of public collective self-esteem and models, one with all four types of threat directly pre- fewer rewards allocated to the threatening outgroup. dicting outgroup attitudes and the other with negative Although group esteem threats are likely to result in stereotypes as an antecedent to the other types of negative attitudes toward the source of the threats, threat. The results suggested that the second model, us- there is also evidence that group esteem threat can re- ing negative stereotypes as an antecedent, was the su- sult in negative attitudes toward other outgroups unre- perior model. Realistic threat, symbolic threat, and in- lated to the source of the threat (Leach, Spears, tergroup anxiety all mediated the relationship between Branscombe, & Doosje, 2003). negative stereotypes and racial attitudes. Nevertheless, Identification with the ingroup moderates the rela- this mediation was not always complete and in the tionship between group esteem threats and intergroup White sample a significant direct effect of negative ste- bias. The stronger the identification with the ingroup, reotypes on racial attitudes remained even after the me- the stronger the reactions to group esteem threats diation effects were considered. Thus, although nega- (Branscombe et al., 1999). According to Branscombe tive stereotypes do predict the other types of threat, and Wann (1994), members who weakly identify with they also predict outgroup attitudes directly (i.e., inde- a group facing esteem threat may not be sufficiently pendent of other threats). motivated to maintain a positive group image by dero- gating the threatening outgroup. Ingroup identification also seems to increase commitment to the ingroup in Additional Threat Theories the face of esteem threats. For example, when the ingroup is of low status, high identifiers increase their Although the threats involved in the ITT have gar- contributions to the group significantly more than low nered considerable attention, other theories suggest ad- identifiers, most likely in an effort to increase the ditional types of intergroup threat that may relate to ingroup's status (Ouwerkerk, de Gilder, & de Vries, outgroup attitudes. 2000).

Group Esteem Threats Distinctiveness Threat Belonging to positively valued groups is important Distinctiveness threat is another type of threat re- for a member's self-image and when the actions of an lated to social identity. According to social identity outgroup potentially decrease an ingroup's esteem a theory (Tajfel & Turner, 1986), people seek member- threat is posed that may instigate intergroup bias ship in positively distinct groups and threats to ingroup (Tajfel & Turner, 1979). Branscombe, Ellemers, distinctiveness are therefore aversive. Intergroup com- Spears, and Doosje (1999) defined this circumstance parisons with a very similar outgroup threaten group as a threat to the ingroup's perceived value. To avoid uniqueness and distinctiveness (Branscombe et al., confusion with values related to symbolic threat, we 1999), therefore very similar groups may prompt more refer to this as group esteem threat. When the ingroup's competitive intergroup comparisons leading to higher image is threatened by an outgroup a person may levels of intergroup bias. In this regard, Roccas and disidentify or leave the ingroup (Tajfel & Turner, Schwartz (1993) manipulated the degree of intergroup 1986). However, Tajfel and Turner noted that when similarity and found that as similarity increased, high these reactions are not possible or desirable, another identifying group members showed an increase in in- option is to derogate the source of the threat. tergroup bias along dimensions relevant to the ingroup. Support for the relationship between group esteem Additionally, on some resource allocation tasks there is threat and intergroup bias has been found across a more discrimination against members of similar rela- number of studies. Unlike the research on the threats tive to dissimilar outgroups (Diehl, 1988). incorporated in the ITT, most of the studies on group As with group esteem threat, distinctiveness threat esteem threat have involved experimental, rather than seems to impact highly identified ingroup members

Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 339 RIEK, MANIA, GAERTNER

more than low identifiers. Jetten, Spears, and Manstead and specific predictions associated with different types (2001) found that under conditions of low intergroup of intergroup threat. distinctiveness, high identifying ingroup members dis- Initial findings have provided support for this played higher intergroup bias relative to low identifi- model. In an early study to utilize Neuberg and Cot- ers. Apparently, identification acts as a moderator of trell's (2002) model, White participants' perceptions of the relationship between distinctiveness and intergroup various threats, their general prejudice level, and emo- bias, such that only highly identified ingroup members tional reactions to a number of different outgroups are sufficiently motivated to react to low distinctive- were measured. As predicted, participants had differ- ness because the ingroup is an important part of their ent perceptions ofthreat for different outgroups as well identity. as different emotional reactions. For example, when In a recent meta-analysis, Jetten, Spears, and the outgroup was Blacks, participants reported high Postmes (2004) found that although threats to distinc- perceptions of threat to the ingroup's safety and per- tiveness did not necessarily impact intergroup judg- sonal property. Blacks were also more likely than the ments, these threats did influence behavioral reactions. other outgroups to elicit fear and anger. Whereas other When intergroup distinctiveness is low and thus dis- outgroups, such as Asian-Americans and Native tinctiveness threat is high, intergroup behavioral biases Americans, elicited different patterns of threat percep- (e.g., in resource allocation tasks) tend to be stronger tions and emotional reactions. For example, Native then when distinctiveness threat is low. However, the Americans were perceived to represent a threat to average effect size for this relationship was relatively ingroup values and evoked emotions such as pity and small (r = .046). guilt.

Biocultural Model of Threat Summary Neuberg and Cottrell (2002) examined issues of in- tergroup threat in their biocultural model of intergroup Overall, our qualitative review reveals that inter- emotions. Whereas other intergroup threat theories fo- group threats affect outgroup attitudes. Although early cus mainly on the causes of threat, Neuberg and Cot- research on intergroup threat was somewhat disparate trell's model elaborates on our reactions to threat. They and different conceptions of threat were seen as com- theorized that for adaptive reasons, different intergroup petitive, recently researchers have begun integrating threats lead to different emotional responses, which the different types of threat into more comprehensive then lead to different attitudinal and behavioral out- conceptual models (Neuberg & Cottrell, 2002; W. G. comes. Although different threats are expected to re- Stephan & Stephan, 1996). The purpose of the second sult in specific emotions, Neuberg and Cottrell sug- part of our review is to examine the relationships be- gested that anger will also accompany the majority of tween the different types of intergroup threat and intergroup threats. outgroup attitudes across the literature by conducting a Neuberg and Cottrell (2002) divided intergroup quantitative meta-analysis. Such an analysis provides a threats into two broad categories: threats to group-level measure of the magnitude of effect sizes across a large resources and threats to group integrity. The resource number of studies and we expect that the different threats include threats toward group safety, group pos- types of threat will show significant and unique rela- sessions, and economic security. These threats parallel tionships with outgroup attitudes. Additionally, a realistic threats. The integrity threats include threats to meta-analysis can reveal undetected effects for vari- group values, morality, competence, and reciprocity ables that may otherwise be overlooked when consid- relations. These threats share commonality with sym- ering each study by itself. In particular, it permits the bolic and social identity threats. exploration of potential moderating variables (e.g., the The primary advantage of Neuberg and Cottrell's social status of the threatening outgroup) on the rela- (2002) model is that it focuses on a series ofbehavioral tionships between the various types of intergroup outcomes that differ depending on the type of threat. threat and outgroup attitudes. For example, threats to the ingroup's economic secu- rity (e.g., immigrant groups "stealing" jobs) are pre- dicted to lead to and which in turn are ex- anger fear, Method pected to promote an aggressive response to maintain economic security. On the other hand, threats to the im- Literature Search age of the ingroup's competence (e.g., perception of Asians being intellectually superior to Whites) are ex- We chose to focus on the following types of inter- pected to lead to envy and anger followed by behaviors group threat in the current meta-analysis: realistic that diminishes the outgroup's accomplishments or threat, symbolic threat, intergroup anxiety, negative bolsters the ingroup's abilities. This allows for clear stereotypes, and group esteem threat. The first four are

340 Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 INTERGROUP THREATS AND ATTITUDES the threats included in W. G. Stephan and Stephan's outgroup attitudes whereas in others higher values in- (2000) ITT and group esteem threat derives from re- dicated negative outgroup attitudes. To adjust for this, search on social identity threat (Branscombe, we recoded all necessary correlations so that a positive Ellemers, Spears, & Doosje, 1999). We did not include correlation indicated that higher levels of threat were distinctiveness threat because a comprehensive related to more negative outgroup attitudes. If a rela- meta-analysis on this issue was recently conducted tionship was only reported as nonsignificant, we en- (Jetten, Spears, & Postmes, 2004). To obtain the most tered a zero for the correlation coefficient. comprehensive pool of research findings concerning the relationship between perceived threat and outgroup Effect attitudes three major strategies were implemented. Sizes First, computer searches of PsycINFO, Dissertation A mixed-effects model was used in the current anal- Abstracts International, and Sociological Inquiry data- yses, where it is assumed that there are systematic dif- bases were carried out using search terms related to the ferences in the effect sizes that can be accounted for by five threat types and outgroup attitudes (e.g., realistic moderating variables, but allows for the possibility that group conflict, value threat, symbolic racism, social other, unmeasured factors may be influencing the vari- identity threat, etc.). Second, we obtained articles or ance of the effect sizes (Raudenbush, 1994). A book chapters cited in the reference sections of any mixed-model approach was a more conservative manuscripts obtained via the computer searches. choice and less prone to Type I errors compared to the Finally, we put out a call for papers on the SPSP fixed-effects approach (Hedges & Vevea, 1998). Each Listserv and contacted researchers in the area of inter- correlation coefficient was transformed using Fisher's group threat to obtain any unpublished or in press Zr transformation (Hedges & 01kin, 1985) and was manuscripts. then weighted for differences in sample sizes. Al- though Hunter and Schmidt (1990) recommended cor- Inclusion Criteria recting each correlation for attenuation using scale reliabilities, a large number of studies did not report To be included in the meta-analysis, research re- alphas for the measures or utilized single-item mea- ports had to meet several criteria. Each study had to in- sures making correction for attenuation difficult. One clude a measure of the relationship between one or possible solution was to use the mean scale reliability more of the five types of intergroup threat and out- when no reliability statistic was reported (Rosenthal, group attitudes. The threat had to be intergroup rather 1994). However, because the measures of both threat than interindividual in nature. This meant that studies and outgroup attitudes differed widely across studies in which the threat did not arise from another group or this did not seem to be an effective solution. Therefore, as a result of group membership were not suitable for no corrections for attenuation were implemented. the current review (e.g., studies where threat was cre- Using the weighted correlations, we calculated an av- ated via negative feedback from an experimenter; see erage effect size for the relationship between each of Fein & Spencer, 1997). Additionally, studies where the the five threat types and outgroup attitudes. type of threat (e.g., symbolic or realistic) could not be distinguished were not included. Indexes of outgroup Antecedent attitudes that were acceptable included measures of Analyses outgroup evaluations, intergroup bias, prejudice, or In addition to examining the effects of the various attitudes toward programs or issues pertaining to an threat types on outgroup attitudes, we examined the outgroup (e.g., affirmative action, immigration impact of two variables on the intergroup threat types: policies). negative stereotypes and ingroup identification. Be- We decided upon correlation coefficients for our ef- cause negative stereotypes have been found to have fect size measure, so any potential study had to include both direct and indirect effects on outgroup attitudes a correlation coefficient describing the relationship be- (W. G. Stephan et al., 2002), we examined the effect of tween a type ofthreat and outgroup attitudes or provide negative stereotypes on the other types of threat in ad- enough information to calculate one (e.g., t-values, dition to their direct effect on outgroup attitudes. F-values, means and SD, etc.). We chose to use corre- Ingroup identification has also been considered an an- lation coefficients because they are commonly used as tecedent to intergroup threat (W. G. Stephan & a measure of effect size in meta-analyses and were re- Stephan, 2000), so the relationships between identifi- ported in the vast majority of studies. This required the cation and intergroup threat were examined. exclusion of any study that only reported results from multiple regression or structural equation modeling Moderator analyses, which are not useable in a meta-analysis Analyses (Lipsey & Wilson, 2001). Some studies had dependent Each study was coded for a number of potential measures where higher values indicated positive moderators of the threat-attitude relationship. Because

Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 341 RIEK, MANIA, GAERTNER

threat differentially influences evaluations of low and Table 1. Summary ofEffect Sizesfor Threat Types and high status outgroups (Cadinu & Reggiori, 2002), we Outgroup Attitudes. coded for whether the outgroup was of relatively low Fail-Safe or high social status. Social status was determined by Threat Type r k N Qw N the relative value that society generally ascribes to the Realistic 0.42* 38 37,516 794.72* 281 outgroup compared to the ingroup (e.g., Whites vs. ra- Symbolic 0.45* 53 30,004 916.73* 424 cial minorities, men vs. women, heterosexuals vs. ho- Anxiety 0.46* 34 7,964 430.84* 279 mosexuals). For example, studies where Whites were Negative 0.44* 31 9,449 478.45* 242 threatened by Blacks were coded as involving a low so- Stereotypes cial status outgroup, whereas studies with Blacks Group Esteem 0.21* 11 1,032 28.04* 35 threatened by Whites were coded as having a high so- Note: r = average Pearson correlation; k = number of samples in cial status outgroup.I We also coded for characteristics analysis; N = total number of participants; Qw = within-class homo- of each study to determine whether methodological geneity statistic. differences between the studies influenced the effect *p< .05. size of the threat-attitude relationship. First, we coded whether intergroup threat was manipulated or only measured. When threat was measured, we coded erate in magnitude (realistic threat: r = .42, symbolic whether the measures for each type of threat used a threat: r = .45, anxiety: r = .46, negative stereotypes: r = highly reliable scale (ct> .75), a low reliability scale (ca .44, group esteem threat: r = .21). Because only one of < .75), a scale with no reported reliability, or a single the studies in the group esteem threat analysis included item measure. Second, because ITT has been one ofthe an additional type ofthreat, wewereunable to determine most widely examined theories of intergroup threat, group esteem threat's relationship with the other threat we also coded whether or not each study utilized W. G. types. The meta-analyses revealed moderate to strong Stephan and Stephan's (2000) measurement scales of correlations among the remaining four threat types (re- the different intergroup threat types. Third, because alistic threat, symbolic threat, intergroup anxiety, and outgroup attitudes were measured in different ways, negative stereotypes) with one another (r = .35 to r = we coded whether the dependent variable involved ei- .59). Fail-safe analyses for each threat type were run us- ther outgroup evaluation ratings (on assorted charac- ing Orwin's (1983) approach to determine the number teristics), prejudice scales (e.g., Brigham's, 1993, Atti- offindings with effect sizes ofzero that wouldbe neces- tude Toward Blacks scale), intergroup bias (ingroup sary to reduce each threat-attitude relationship to a neg- ratings minus outgroup ratings), or attitudes toward so- ligible effect size (r = .05). Four of these relationships cial policies pertaining to the outgroup. met the criteria setby Hedges and 01kin (1985), withthe fail-safeNbeing larger than 5 times the number of stud- ies in each meta-analysis plus 10. Group esteem threat Results did not meet this criterion and although this is likely due to the relatively small number of samples available (k = Overall Effect Sizes 1 1) this suggests a potential publication bias. To determine the unique effects of each threat type, A total of 95 separate samples were used in the cur- the meta-analytically derived correlation matrix was rent analysis. All the included studies and their effect used as an input file in SPSS. This matrix included the size measures canbe seen intheAppendix (available on- average weighted correlations between four of the line at http://www.leaonline.com/doi/pdf/10.1207/ threat inter- Results ofthe types (realistic threat, symbolic threat, s15327957pspr1004.4A). meta-analyses and and for the relationships between the five threat types and group anxiety, negative stereotypes) outgroup attitudes as well as the average weighted correlations outgroup attitudes are presented in Table 1. As pre- between each of the threat types. The lack of studies dicted, all five types of threat were significantly related that included group esteem threat along with an addi- to outgroup attitudes and these relationships were mod- tional intergroup threat made it impossible to include group esteem threat in this analysis. We then ran a mul- tiple regression using the SPSS matrix function, simul- 'We would have preferred to also examine relationships involv- ing groups of equal status; however, the vast majority of studies in taneously entering the four threat types as predictors of the sample involved groups ofunequal status. We were unable to find outgroup attitudes. Although there may be some poten- enough studies where participants would perceive both the ingroup tial interpretive problems because different correla- and outgroup as having equal status. For example, in cases involving tions in the matrix were estimated from different sub- one national group evaluating another (e.g., Britain vs. France) the sets of studies, similar procedures have been applied in groups may be seen as having equal status on an objective level, but it is not clear how the actual participants of the study perceived the sta- previous meta-analyses (Judge, Heller, & Mount, tus of the outgroup relative to the ingroup. Therefore, in such in- 2002; Kuncel, Hezlett, & Ones, 2001; Parker et al., stances relative outgroup status was not discernable. 2003).

342 Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 INTERGROUP THREATS AND ATTITUDES

Table 2. Summary ofEffect Sizesfor the Antecedents ofIntergroup Threat. Fail-Safe Antecedent Variable r k N Qw N Realistic Ingroup Identification 0.16* 9 14,557 154.72* 19 Negative Stereotypes 0.35* 16 4,554 141.92* 96 Symbolic Ingroup Identification 0.29* 7 2,165 16.57* 34 Negative Stereotypes 0.37* 23 5,396 198.37* 147 Anxiety Ingroup Identification 0.22* 7 2,260 23.49* 24 Negative Stereotypes 0.36* 16 4,702 98.30* 99 Note: r = average Pearson correlation; k = number of samples in analysis; N = total number of-participants; Qw = within-class homogeneity statistic. *p < .05

Intergroup anxiety and negative stereotypes had the the threat measures, whether W. G. Stephan and strongest unique relationships with outgroup attitudes Stephan's (2000) measures ofthreat were used; and the (PB= .26, p < .01 and = .24, p < .01, respectively). type of measure used to assess outgroup attitudes each Symbolic threat and realistic threat also displayed sig- moderated some of the relationships between the inter- nificant unique influences on outgroup attitudes (fP = group threats and outgroup attitudes. .19,p < .01 and ,3 = .13,p < .01, respectively). Overall, the four threat types accounted for 36% of the variance Outgroup status. Sixty-nine of the 95 samples in outgroup attitudes (R2 = .36). Taken together, these were used in the moderator analysis for relative findings demonstrate that the various types of inter- outgroup status. In the remaining cases relative status group threats have a significant and unique impact on was not discernable. The status of the threatening outgroup attitudes. The fact that the threat types in- outgroup moderated a number of the relationships be- cluded in the ITT account for a substantial portion of tween the specific intergroup threats and outgroup atti- the variance in attitudes also suggests that ITT pro- tudes. A full description of these relationships is pre- vides a reasonable and useful, though perhaps incom- sented in Table 3.2 Unfortunately, because outgroup plete, framework for conceptualizing the relationship status was only discernable in 4 of the 11 samples uti- between intergroup threat and attitudes across the lized in the group esteem threat analysis, we were un- available literature. able to examine whether status moderated this particu- lar threat type. Relative status moderated the Antecedent Analyses relationship between anxiety and outgroup attitudes, QB(M) = 4.81, p < .05. Both ingroup identification and negative stereo- When outgroups were of relatively low status (e.g., ra- types were examined as potential antecedents of inter- cial minorities, females) anxiety had a stronger rela- group threat. Among the studies of group esteem tionship with outgroup attitudes than when outgroups threat, none included a measure of negative outgroup were of relatively high status (e.g., Whites, males). stereotypes and only two included a measure of Relative outgroup status was also a marginally signifi- ingroup identification, rendering an analysis of the ef- cant moderator of the relationship between realistic fect of identification on group esteem threat almost im- threat and outgroup attitudes, QB(M) = 3.24, p = .07, possible. Table 2 reveals that both antecedent variables and negative stereotypes and outgroup attitudes, QB(l) (i.e., ingroup identification and negative stereotypes) = 2.79, p = .09. In both cases, the relationship between had a significant impact on realistic threat, symbolic the threat type and negative attitudes was stronger threat, and intergroup anxiety, though negative stereo- when the threatening outgroup was oflower status than types showed stronger average effect sizes. when the outgroup was of higher status. Although not significant, the influence of symbolic threat on atti- Moderator Analyses 21t was not possible to examine separately each specific group di- As shown in Table 1, the relationships between the mension (e.g., race, sex) composing the low and high status five threat types and outgroup attitudes displayed het- outgroups classification because of the small number of samples in- erogeneity of effect sizes, suggesting there are poten- volving high status outgroups for each specific dimension, particu- tial moderators of these larly as they intersected with each of the four specific intergroup relationships. Moderator anal- threats. Thus, our statistical analyses required that we aggregate yses revealed that outgroup social status, whether across these specific group dimensions to consider the moderating threat was manipulated or measured; the reliability of role of relative group status.

343 Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 RIEK, MANIA, GAERTNER

Table 3. Group Status as a Moderatorfor Threat-Attitude Effect Sizes. Relative Outgroup Status QB k r Qw Realistic 2.88** Low 27 .48 58.99* High 5 .34 3.82 Symbolic 2.55 Low 39 .45 48.02 High 6 .34 4.58 Anxiety 4.83* Low 16 .52 16.59 High 10 .38 8.91 Negative Stereotypes 2.79** Low 24 .46 15.23 High 5 .29 1.21 Note: QB = between-classes effect; k = number of samples involved in analysis; r = average Pearson correlation; Qw = Within-class homogeneity. *p <.05. **p <.10. tudes showed a similar pattern. Table 3 shows that, in ety had unreliable measures or single item measures. the majority ofcases, once the outgroup status modera- The measurement reliabilities of the other threat types tor was accounted for, the Qw statistic indicated (symbolic threat, realistic threat, negative stereotypes) within-category homogeneity, suggesting that no fur- did not have a significant impact on the respective ther moderators exist within the subset of high and low threat-attitudes effect sizes. Group esteem threat was status groups. The only exception to this was a signifi- only measured in one sample, making it impossible to cant Qw for the low status outgroup studies within real- examine the effect of measurement reliability for this istic threat, indicating that further factors may be influ- threat type. encing the relationship between realistic threat and attitudes within this particular subset. Type of threat measure. The specific measure Manipulation versus measurement. We first of intergroup threat used moderated the relationship compared those studies that measured threat to those between realistic threat and outgroup attitudes, QB(l) = that involved the manipulation of threat. Only studies 7.95, p < .05, such that studies that used W. G. Stephan involving realistic threat and the symbolic threat analy- and Stephan's (2000) measures of realistic threat re- ses contained enough samples of both types to make ported larger effect sizes (r = .47) than studies that uti- meaningful comparisons. Studies that manipulated the lized other measures of realistic threat (r = .36). The presence of realistic threat obtained significantly type of threat measure did not moderate the relation- weaker relationships between realistic threat and ships between the other three threat types and outgroup outgroup attitudes (r = .14) than studies where realistic attitudes. threat was only measured as in survey studies (r = .43), QB(i) = 7.61, p < .01. However, this should be inter- preted cautiously because realistic threat was only ma- Outgroup attitude measure. The nature of the nipulated in 3 of the 38 studies involving realistic outgroup attitude measure was only a significant mod- threat. In the case of symbolic threat, there was no sig- erator of the relationship between anxiety and nificant difference in effect sizes for studies that ma- outgroup attitudes, QB(2) = 7.54, p < .05. Contrasts nipulated symbolic threat (r = .52) compared to those conducted using Quinones, Ford, and Teachout's that only measured it (r = .45), QB(1) = 0.61, p = ns. (1995) z-test method revealed that studies using inter- group bias (i.e., ingroup attitudes minus outgroup atti- Threat measure reliabilities. Whether the mea- tudes) as a dependent measure had significantly lower surement of intergroup threat was a reliable scale, an effect sizes (r = .22) than studies using only outgroup unreliable scale, a scale with no reported reliability, or evaluations (r = .46, z = 2.13, p < .05) or studies using a single item measure only moderated the relationship prejudice scales (r = .56, z = 2.74, p < .05). None of the between intergroup anxiety and outgroup attitudes, studies used in the anxiety analyses included attitudes QB(1) = 6.46,p < .05. Studies with highly reliable mea- toward programs or issues pertaining to an outgroup. sures of anxiety obtained significantly lower (r = .43) Studies using outgroup evaluations and prejudice effect sizes than studies that did not report a reliability scales as dependent measures did not significantly dif- ofthe anxiety measure (r = .65). The reason for this dif- fer in terms of the strength of the relationship between ference is unclear. None of the studies involving anxi- anxiety and outgroup attitudes.

344 Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 INTERGROUP THREATS AND ATTITUDES

Discussion cause of negative stereotypes as well as vice versa (Brewer & Alexander, 2002; Dovidio, Brigham, John- Overall, the results of the meta-analysis indicate son, & Gaertner, 1996). If true, then it seems reason- that intergroup threat has an important relationship able to assume that intergroup threats may create nega- with outgroup attitudes. As people perceive more inter- tive stereotypes (e.g., "members from Group X are group competition, more value violations, higher lev- aggressive because they are all after our jobs"). els of intergroup anxiety, more group esteem threats, Clearly, the role of negative stereotypes in future theo- and endorse more negative stereotypes, negative atti- ries of intergroup threat needs to be examined from tudes toward outgroups increase. The analysis also re- multiple perspectives. veals that four of the threat types are moderately to strongly correlated with one another, suggesting a Moderators of Threat/Bias common, underlying factor. Integrative frameworks, such as W. G. Stephan and Stephan's (2000) ITT seem Relationship to be advantageous relative to examining each threat Although the meta-analysis suggests that an inte- type separately because interrelationships between the grated approach provides a good understanding of the threat types can be examined. This suggests that future relationships between threat and outgroup attitudes, research should consider integrating as many threats as characteristics ofthe threatening outgroup may moder- possible. Although the ITT is an extremely useful ate these relationships. It appears that realistic threat, framework for conceptualizing intergroup threat, some intergroup anxiety, and negative stereotypes are stron- modifications to the model may further increase its ger predictors of outgroup attitudes toward low status effectiveness. rather than toward high status outgroups. This possibly explains why W. G. Stephan and colleagues (2002) Antecedents of Intergroup Threat found that an integrated threat model accounted for more of the variance in the attitudes of Whites toward A number of antecedents to intergroup threat have Blacks than in the attitudes of Blacks toward Whites. previously been studied (W. G. Stephan et al., 2002; W. For realistic threat, this effect may be due to high status G. Stephan & Stephan, 2000) and in the current analy- group members' perception that low status outgroups ses we were able to meta-analytically examine two of are trying to usurp their resources. This may be more them: ingroup identification and negative stereotypes. aversive than not having resources in the first place, Ingroup identification appears to have a weak, though which could explain why the realistic threats posed by statistically significant, effect on intergroup threat; low status outgroups have a larger impact on attitudes high identifiers experience higher levels of intergroup than the threats from high status outgroups. threat than low identifiers. This makes sense, because As for anxiety, members of low status groups likely the more important the ingroup is to individuals the have more experience interacting with higher status more sensitive they may be to anything that could harm outgroups than members of high status groups have in- their group. However, it is important to note that be- teracting with lower status outgroups. Thus, relatively cause the majority of studies included in our analyses low status group members may have adapted to the were correlational in nature, where the direction of anxiety associated with interacting with higher status causality is unclear, it is also possible that threats to the outgroups, which may reduce the impact of anxiety on ingroup could cause an increase in ingroup identifica- their outgroup attitudes. However, if high status group tion. Some experimental evidence has shown that members have little experience interacting with low threats increase identification (Jetten, Branscombe, status outgroup members, the anxiety they experience Schmitt, & Spears, 2001) and therefore the relation- may have a large impact on their outgroup attitudes. ship between identification and intergroup threat In the case of negative stereotypes, Fiske (1993) should be examined experimentally to determine the suggested that groups higher in power are more likely direction of causality. to rely on stereotypes than lower power groups because Negative stereotypes are positively related to the they do not need to attend as carefully to individuating other threats specified by the ITT. Furthermore, similar information about subordinates. Therefore, the greater to W. G. Stephan and colleagues' (2002) findings, neg- utilization of stereotypes by higher status group mem- ative stereotypes are shown to have both direct and in- bers may be responsible for the stronger relationship direct relationships with outgroup attitudes. Although between negative stereotypes and attitudes when the negative stereotypes have a moderate relationship with outgroup is of lower status. It is important to note that other threat types, they also maintain a direct, unique the results of the meta-analysis do not indicate that relationship with outgroup attitudes. Complicating the lower status outgroups elicit stronger perceptions ofre- issue further is the possibility that negative stereotypes alistic threat, anxiety, or negative stereotypes com- may also be a consequence of intergroup threat. Some pared to higher status outgroups, but rather that these suggestions have been made that prejudice can be the factors involve a stronger relationship with negative

Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 345 RIEK, MANIA, GAERTNER outgroup attitudes when the outgroup is of relatively 2002; Neuberg & Cottrell, 2002). Smith's (1993) inter- low status. It is also important to recognize that all of group emotions theory proposes that different the moderator analyses were exploratory in nature and intergroup contexts lead to different emotional reac- as such, any conclusions need to be carefully reconsid- tions. For example, outgroups that are seen as poten- ered in future research. tially taking away ingroup resources may elicit anger, Finally, studies using W. G. Stephan and Stephan's whereas outgroups that are perceived to block the (2000) measures obtained higher correlations between ingroup from obtaining desired resources may result in realistic threat and negative outgroup attitudes than frustration. A number of studies have supported the studies that used other measures. Although the scales idea of different intergroup contexts resulting in differ- used in some studies focused on a single aspect ofreal- ing emotional responses (Mackie, Devos, & Smith, istic threat (e.g., economic threat), the measure devel- 2000; Miller, Smith, & Mackie, 2004). Furthermore, oped by W. G. Stephan and colleagues taps a wider ar- both Neuberg and Cottrell's model and Smith's inter- ray of realistic threats and therefore may be more group emotions theory predict that different emotional effective. responses lead to different behavioral reactions. Unfor- tunately, most intergroup threat research has ignored Limitations and Future Directions the relationships between different threats, different emotions, and different behavioral reactions. Future In addition to demonstrating the impact of inter- research could benefit from a broader approach to the group threat on outgroup attitudes, our qualitative and outcomes of intergroup threat. quantitative analysis also reveals some methodological Although intergroup threat theories have become and conceptual limitations of current approaches. One more integrated (e.g., W. G. Stephan & Stephan, 2000) is the lack ofexperimental studies, especially in the do- even further theoretical integration is desirable. To our mains of realistic and symbolic threat. Although some knowledge, no study has examined issues of realistic studies have examined intergroup threat experimen- or symbolic threat in combination with group esteem tally (Branscombe & Wann, 1994; Cadinu & Reggiori, or distinctiveness threats. We postulate that group es- 2002; Grant, 1992, 1993a), the majority of studies teem threats operate at the same level and in a similar were based on correlational research designs that are fashion as realistic and symbolic threat, but it would be problematic for determining the direction of causality. interesting to examine how group esteem threat may Threat could lead to negative attitudes, but negative at- interact with and influence these other threats. For ex- titudes could also precede perceptions ofthreat. Plausi- ample, realistic threat accompanied by a loss of ble arguments exist for each causal direction, and it is ingroup prestige may result in stronger negative atti- reasonable to assume the relationship is bidirectional. tudes than either threat alone. It is unclear, however, Some experimental evidence has shown that the intro- exactly how distinctiveness threat would fit with the duction of intergroup threat leads to negative attitudes other threat types. Possibly, it operates at the same toward outgroups (Branscombe & Wann, 1994; Grant, level as realistic and symbolic threat, however, it could 1993a), however, to our knowledge no study has exam- potentially act as an antecedent to those threats. For ex- ined whether increases in negative attitudes lead to in- ample, when the ingroup's distinctiveness is threatened creases in perceived threat. Future studies may attempt by another group it could lead to efforts to emphasize to examine this issue through either a longitudinal de- group differentiation in a number of ways, including sign or controlled experimental studies. value differences or differences in material resources. Additionally, the overreliance on self-report mea- Thus, distinctiveness threat could initiate a number of sures neglects potential behavioral responses to inter- reactions, some of which may increase symbolic or re- group threat. Although all of the threats may result in alistic threat. Further research is needed to explore negative outgroup attitudes, different threats may lead these potential interrelationships among different to different behavioral outcomes. For example, anxiety threat types. may cause avoidance, whereas realistic threat may lead Another important issue involves the relationship to aggressive retaliation. Although this idea has begun between realistic and symbolic threats. Although it is to be explored by Neuberg and Cottrell's (2002) generally assumed that these two threat types operate biocultural model, most investigators have not consid- at the same level, it is possible that in certain circum- ered these possibilities. Also, there may be alternative stances one type of threat may act as an antecedent to ways of assessing perceptions and reactions to threat. the other. For example, in the cases of religious con- For example, physiological measures have been used flict, symbolic threats may initially be experienced, to examine reactions to outgroup members (Mendes, which in turn leads to conflict and thus realistic threats. Blascovich, Lickel, & Hunter, 2002) and reactions to a Conversely, realistic threats stemming from intergroup threatened group identity (Matheson & Cole, 2004). conflict may lead to symbolic threats in an attempt to Other investigators have focused on the emotional re- justify the conflict. Therefore, although we believe that actions to threat (Devos, Silver, Mackie, & Smith, realistic and symbolic threats should occupy the same

346 Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 INTERGROUP THREATS AND AYTITUDES

level in any general model of intergroup threat, the spe- (e.g., threat to a specific individual's job) or the cific characteristics of a given situation may influence ingroup as a whole (e.g., a loss of ingroup power) may this relationship. influence the strength with which these threats impact It is also unclear ifintergroup anxiety acts as a threat outgroup attitudes and behavior. However, these rela- in the same location in the casual sequence as realistic tionships would probably be moderated by the strength and symbolic threat or whether intergroup anxiety is of the individuals' ingroup identification. Bizman and the result of other intergroup threats. Perhaps realistic, Yinon (2001) have found that high identifiers react symbolic, and group esteem threats increase the likeli- more strongly to group-level realistic threats than low hood that an individual will experience intergroup identifiers, whereas low-identifiers' attitudes are more emotions, such as anxiety, which then increases nega- influenced by intergroup anxiety, which can be seen as tive outgroup attitudes and behaviors. Because anxiety an individual-level threat. This difference between in- has been described as a more individual-level phenom- dividual-level and group-level threats could also be ex- enon compared the group-level threats such as realistic plored in future integrated models ofintergroup threat. and symbolic threat (W. G. Stephan & Renfro, 2002; This meta-analysis also suggests additional direc- W. G. Stephan & Stephan, 2000), it makes sense that it tions for research on intergroup threat. Although the would be in closest proximity to the formation of nega- relationship between intergroup threat and outgroup tive outgroup attitudes. Intergroup anxiety could be re- attitudes is stronger when a high status group is threat- examined to determine its appropriate place in the ened by a lower status outgroup, than vice versa, it is causal chain involving intergroup threat and attitudes. unclear how threat will relate to attitudes when low sta- Future research could also examine more integra- tus groups are threatened by other low status groups, or tive models that include multiple sources of threat. In when high status group members are threatened by an- the proposed model presented in Figure 1, stereotypes, other high status group. For example, when racial mi- ingroup identification, and distinctiveness threats are norities perceive realistic or symbolic threat from other antecedents to realistic, symbolic, and group esteem minorities, will these threats impact attitudes to the threats. Furthermore, the relations between these same degree as when racial minorities are threatened threats with outgroup attitudes and intergroup behav- by Whites? Because group status moderates the rela- iors are mediated by intergroup emotions. This model tionship between intergroup threat and attitudes among represents one possible modification to models such as groups of unequal status, it would be useful to explore the ITT, however, there are certainly other models that these relationships among groups of equal status. could be tested and compared. Models similar to Another avenue of potential research concerns the Neuberg and Cottrell's (2002) could also be con- content of outgroup stereotypes. Although negative structed to examine ifthere are different emotional and stereotypes have been identified as predictors of per- behavioral outcomes for the different types of threat. ceived threat and outgroup attitudes, the possibility W. G. Stephan and Renfro (2002) have recently dif- that positive stereotypes may also be threatening has ferentiated between the realistic and symbolic threats been relatively unexplored. Although the negative ste- that affect an individual and those that affect the group. reotype that Blacks are violent may increase some Whether these threats directly impact an individual Whites' fear for their safety and perceived realistic

Figure 1. Potential model of intergroup threat.

Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 347 RIEK, MANIA, GAERTNER threat, the positive stereotype of Asians as highly intel- 1995; Park & Judd, 1990; Sedikides, 1997), which may ligent may increase realistic threat in job or academic decrease positive information's impact on outgroup settings. Additionally, it is not clear if all negative ste- stereotypes and attitudes. Perhaps if circumstances reotypes will increase threat and negative outgroup at- (e.g., cooperative interdependence) could motivate in- titudes. The stereotype that a group is unintelligent creased attention to outgroup information (see Erber & may not have the same impact as the stereotype that the Fiske, 1984), some of these problems could be group is aggressive. mitigated. Relatedly, Fiske's stereotype content model (Fiske, Other methods may be more applicable to the re- Xu, Cuddy, & Glick, 1999) has proposed that how duction of intergroup anxiety and the use of negative groups are stereotyped along the dimensions of stereotypes. The effects of contact on the reduction of warmth and competence determines reactions to them. intergroup biases are quite robust (see Pettigrew & Groups characterized as low in competence, but high Tropp, 2000) and it is probable that contact would have in warmth, may be perceived in a condescending fash- similar effects on intergroup threat. For example, posi- ion, whereas groups high in competence but low in tive intergroup contact (e.g., cooperation) reduces in- warmth may be viewed with envy (Fiske, Cuddy, tergroup anxiety (Hewstone, 2003; Voci & Hewstone, Glick, & Xu, 2002). Perhaps groups that differ along 2003) and there is evidence that it can also reduce other these stereotypic dimensions are perceived as posing types of intergroup threat (Brown & Hewstone, 2005). different types of intergroup threat. Outgroups that are Stereotypes can be weakened through contact with viewed as high in competence may elicit greater realis- outgroup members who are considered typical, but tic threat than outgroups viewed as low in competence. who also disconfirm the stereotype (Wolsko, Park, Outgroups that are high in warmth may evoke less in- Judd, & Bachelor, 2003). Additionally, if increasing tergroup anxiety than outgroups low in warmth. On the the salience of positive characteristics about oneself whole, there seems to be a number of potential ques- through self-affirmation reduces the impact of individ- tions concerning the role of negative stereotypes that ual-level threats on prejudice (Fein & Spencer, 1997), can be addressed in a model of intergroup threat. then a variant of this strategy may reduce group-level Other issues that could be examined include the eti- threats by affirming positive ingroup characteristics. ology and development of intergroup threat. Are cer- Because intergroup threat involves motivational and tain perceptions of intergroup threat learned and in- emotional mechanisms, targeting intergroup affect grained at an early age? To our knowledge, there has may also offer a way of reducing intergroup threat. been no longitudinal study of children to determine the Cross-group friendships are effective at reducing inter- developmental course of intergroup threat perceptions. group biases (Pettigrew, 1997), perhaps because they Also, there may be individual differences that influ- initially reduce intergroup threat as well. Perspec- ence the likelihood of perceiving a threat from another tive-taking increases empathy for outgroup members, group. For example, individuals high in Social Domi- which in turn reduces intergroup bias (Galinsky & nance Orientation (SDO; Pratto, Sidanius, Stallworth, Moskowitz, 2000; Vescio, Sechrist, & Paolucci, 2003). & Malle, 1994) react especially strongly to certain It is possible that perspective-taking reduces inter- types of threats (Quist & Resendez, 2002). Other per- group threat because the empathy induced may counter sonality characteristics (e.g., authoritarianism) may the negative emotions associated with intergroup threat also moderate the relationship between intergroup (e.g., fear, anger). For example, realistic threats often threats and outgroup attitudes. arise from fear that an outgroup will take resources away from the ingroup, but once a situation is seen from another viewpoint, this fear may be attenuated ei- Reducing Intergroup Threat ther by the empathy induced or the realization that the The results of the meta-analysis indicate that inter- initial fears were unfounded. group threat predicts negative outgroup attitudes. Another potential means of reducing intergroup However, although the majority of the studies have fo- threat is to target social categorization processes. Be- cused on how threat influences outgroup attitudes, lit- cause the categorization of individuals into ingroups tle research has addressed reducing intergroup threat. and outgroups plays a fundamental role in the initiation W. G. Stephan and Stephan (2000) have suggested that of intergroup biases and perceptions of threat, investi- cultural diversity programs may be effective at reduc- gators have explored how changing the social categori- ing symbolic and realistic threats. Because some zations people use can reduce these biases (Brewer & threats may arise out of ignorance, W. G. Stephan and Gaertner, 2001). Perhaps theses categorization strate- Stephan also propose that gaining relevant information gies could also reduce intergroup threat. about the outgroup may reduce symbolic and realistic One such strategy is the decategorization approach, threats. However, research suggests that people attend which attempts to reduce intergroup bias by dissolving to and process information about ingroup and outgroup group boundaries so that people are seen more as indi- members differently (Doosje, Spears, & Koomen, viduals rather than as group members (Brewer &

348 Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 INTERGROUP THREATS AND ATTITUDES

Miller, 1984). Because individuals are no longer cate- of the superordinate category, which results in elevated gorized as outgroup members, the biases that arise levels of bias (Waldzus, Mummendey, Wenzel, & from social categorization are reduced (Wilder, 1978). Weber, 2003). This could be due to common identity Research has indicated that groups are typically per- creating more comparisons between the included sub- ceived to be more threatening and elicit more fear than groups, resulting in members ofthe subgroups feeling a individuals (Insko, Schopler, Hoyle, Dardis, & Graetz, loss of distinctiveness. Thus, future research should 1990; Schopler et al., 1993), therefore, weakening so- seek to balance the potential gains of recategorization cial category boundaries could potentially reduce in- with the potential risks. tergroup threat. For example, in the case of symbolic These suggested methods for reducing intergroup threat, people may be more willing to accept alterna- threat are by no means comprehensive. Other models tive values or points of view from individuals rather that have been shown to be effective at reducing inter- than from a group as a whole. Additionally, Locksley, group biases should also be investigated. For example, Borgida, Brekke, and Hepburn (1980) found that peo- models such as Brown and Hewstone's (2005) inte- ple rely less on group stereotypes when interacting grated model of contact, which stresses both interper- with individuals rather than groups, which suggests sonal and intergroup contact, or the jigsaw classroom that decategorization can potentially reduce the threat approach, which stresses cooperative group work with associated with negative outgroup stereotypes. children (Aronson & Bridgeman, 1979), are two other An alternative method for reducing intergroup bias methods that could also have a large impact on percep- is recategorization, which seeks to reshape group tions of threat. boundaries rather than dissolve them (Gaertner, Dovidio, Banker, Houlette, Johnson et al., 2000). The common ingroup identity model (CIIM; Gaertner & Conclusion Dovidio, 2000) proposes that bias can be reduced by increasing the salience of a more inclusive super- The intergroup threat literature has evolved from a ordinate identity, so that a shared bond between the competition between the realistic and symbolic expla- groups is perceived. Once former outgroup members nations to an integrated view that values the importance are included within a superordinate identity, they can of these as well as additional types of threat. By potentially benefit from the same preferential treat- meta-analytically examining a large body of research, ment accorded to original ingroup members. we conclude that various intergroup threats have a sig- Although most of the research involving the CIIM nificant impact on outgroup attitudes. Realistic threat, has examined its direct effect on intergroup bias, it is symbolic threat, intergroup anxiety, negative stereo- possible that a common ingroup identity also reduces types, and group esteem threat all displayed significant perceived threat between groups. Once former out- relationships with outgroup attitudes. Additionally, al- group members are regarded as ingroup members, though current theories such as the ITT provide viable there is less reason to perceive them as threatening. frameworks for understanding these relationships, we For example, a common ingroup identity could focus encourage future researchers to develop even more members' attention on similarities rather than differ- fully integrated models and perhaps address some ques- ences between the groups and thereby decrease the sa- tions raised in this review. lience of the values and beliefs that are unshared, which should reduce symbolic threat. Intergroup anxi- ety occurs due to an apprehension of negative evalua- tions by outgroup members (W. G. Stephan & References Stephan, 1985), but with a common superordinate identity former outgroup members would be seen as References marked with an asterisk indicate studies included in the ingroup members and could elicit less anxiety, espe- meta-analysis. cially if the other types of threat are reduced. *Aberson, C. L., Hagg, S. C., Shoemaker, C. S., & Tomolillo, C. Although recategorization may reduce some types (2004, February). Contact, anxiety, perspective taking, and ste- of intergroup threat, it is possible that it may contrib- reotype endorsement as predictors of implicit and explicit bi- ute to distinctiveness threat. There is evidence that a ases. Poster session presented at the annual meeting of the Soci- ety for Personality and Social Psychology, Austin, TX. common identity can increase bias in some instances. Aronson, E., & Bridgeman, D. (1979). Jigsaw groups and the deseg- According to Mummendy and Wenzel (1999), group regated classroom: In pursuit of common goals. Personality and members project the attributes of their subgroup onto Social Psychology Bulletin, 5, 438-446. the superordinate identity, resulting in members of *Arriola, K. J., & Cole, E. R. (2001). Framing the affirmative-action each subgroup perceiving themselves as more typical debate: Attitudes toward out-group members and white identity. Journal ofApplied Social Psychology, 31, 2462-2483. of the superordinate group than members of the other *Beaton, A. M., & Tougas, F. (2001). Reactions to affirmative action: subgroup. This leads to perceiving members of the Group membership and social justice. Social Justice Research, other subgroup as norm violators and poor exemplars 14, 61-78.

Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 349 RIEK, MANIA, GAERTNER

*Beaton, A. M., Tougas, F., & Joly, S. (1996). Neosexism among *Cowan, G., Martinez, L., & Mendiola, S. (1997). Predictors of atti- male managers: Is it a matter of numbers? Journal ofApplied tudes toward illegal Latino immigrants. Hispanic Journal of Social Psychology, 26, 2189-2203. Behavioral Sciences, 19, 403-415. * Berrenberg, J. L., Finlay, K. A., Stephan, W. G., & Stephan, C. W. Crandall, C. S. (1994). Prejudice against fat people: Ideology and (2002). Prejudice toward people with cancer or AIDS: Ap- self-interest. Journal ofPersonality and Social Psychology, 66, plying the Integrated Threat Model. Journal of Applied 882-894. Biobehavioral Research, 7, 75-86. Devos, T., Silver, L. A., Mackie, D. M., & Smith, E. R. (2002). Experi- *Biernat, M., Vescio, T. K., & Theno, S. A. (1996). Violating Ameri- encing intergroup emotions. In D. Mackie & E. Smith (Eds.), can values: A "value-congruence" approach to understanding Fromprejudice to intergroup emotions: Differentiated reactions outgroup attitudes. Journal ofExperimental Social Psychology, to social groups (pp. 111-134). New York: Psychology Press. 32, 387-410. Diehl, M. (1988). Social identity and minimal groups: The effects of *Bizman, A., & Yinon, Y (2001). Intergroup and interpersonal interpersonal and intergroup attitude similarity on intergroup threats as determinants of prejudice: The moderating role of discrimination. British Journal of Social Psychology, 27, in-group identification. Basic and Applied Social Psychology, 289-300. 23, 191-196. Doosje, B., Spears, R., & Koomen, W. (1995). When bad isn't all Bobo, L. (1983). White's opposition to busing: Symbolic racism or bad: Strategic use of sample information in generalization and realistic group conflict? Journal ofPersonality and Social Psy- stereotyping. Journal ofPersonality and Social Psychology, 69, chology, 45, 1196-1210. 642-655. Bobo, L. (1988). Group conflict, prejudice, and the paradox of con- Dovidio, J. F., Brigham, J., Johnson, B. T., & Gaertner, S. L. (1996). temporary racial attitudes. In P. A. Katz & D. A. Taylor (Eds.), Stereotyping, prejudice, and discrimination: Another look. In Eliminating racism: Profiles in controversy (pp. 85-114). New N. Macrae, C. Stangor, & M. Hewstone. (Eds.), Stereotypes and York: Plenum Press. stereotyping (pp. 276-319). New York: Guilford. Branscombe, N. R., Ellemers, N., Spears, R., & Doosje, B. (1999). *Dunbar, E., Saiz, J. L., Stela, K., & Saez, R. (2000). Personality and The context and content of social identity threat. In N. Ellemers social group value determinants of out-group bias: A & R. Spears (Eds.), Social identity: Contexts, commitment, con- cross-national comparison of Gough's Pr/To Scale. Journal of tent (pp. 35-59). Oxford, England: Blackwell Science. Cross-Cultural Psychology, 31, 267-275. *Branscombe, N. R., Schmitt, M. T., & Harvey, R. D. (1999). Per- Eagly, A. H., & Mladinic, A. (1989). Gender stereotypes and atti- ceiving pervasive discrimination among African-Americans: tudes toward women and men. Personality and Social Psychol- Implications for group identification and well-being. Journal of ogy Bulletin, 15, 543-558. Personality and Social Psychology, 77, 135-149. *Eller, A. (2002). Putting Pettigrew 's reformulated model to the test: *Branscombe, N. R., Spears, R., Ellemers, N., & Doosje, B. (2002). The intergroup contact theory in transition. Unpublished doc- Intragroup and intergroup evaluation effects on group behavior. toral dissertation, University of Kent, England. Personality and Social Psychology Bulletin, 28, 744-753. *Eller, A., & Abrams, D. (2003). "Gringos" in Mexico: *Branscombe, N. R., & Wann, D. L. (1994). Collective self-esteem Cross-sectional and longitudinal effects of language consequences of outgroup derogation when a valued social school-promoted contact on intergroup bias. Group Processes identity is on trial. European Journal ofSocial Psychology, 24, and Intergroup Relations, 6(1), 55-75. 641-657. *Eller, A., & Abrams, D. (2004a). Come together: Longitudinal Brewer, M. B., & Alexander, M. G. (2002). Intergroup emotions and comparison of Pettigrew's reformulated intergroup contact images. In D. Mackie & E. Smith (Eds.), From prejudice to in- model and the common ingroup identity model in An- tergroup emotions: Differentiated reactions to social groups glo-French and Mexican-American contexts. European Jour- (pp. 209-225). New York: Psychology Press. nal ofSocial Psychology, 34, 229-256. Brewer, M. B., & Gaertner, S. L. (2001). Strategies for reducing in- *Eller, A., & Abrams, D. (2004b). [Impact of contact: Effects of in- tergroup bias. In R. J. Brown & S. L. Gaertner (Eds.), Blackwell ternational exchange on perceptions of Britain]. Unpublished handbook of social psychology: Intergroup processes (pp. data, University of Kent, England. 175-197). Oxford: Blackwell. *Eller, A., & Gomez, A. (2004). [Intergroup contact and attitudes be- Brewer, M. B., & Miller, N. (1984). Beyond the contact hypothesis: tween British university students and asylum seekers]. Unpub- theoretical perspectives on desegregation. In N. Miller & M. B. lished data, University of Kent, England. Brewer (Eds.), Groups in contact: The psychology of desegre- Erber, R., & Fiske, S. T. (1984). Outcome dependency and attention gation (pp. 281-302). New York: Academic Press. to inconsistent information. Journal ofPersonality and Social Brigham, J. C. (1993). College students' racial attitudes. Journal of Psychology, 47, 709-726. Applied Social Psychology, 23, 1933-1967. Esses, V. M., Dovidio, J. F., Jackson, L. M., & Armstrong, T. L. Brown, R., & Hewstone, M. (2005). An integrative theory of inter- (2001). The immigration dilemma: The role ofperceived group group contact. In M. Zanna (Ed.), Advances in experimental so- competition, ethnic prejudice, and national identity. Journal of cial psychology, Vol. 37. San Diego, CA: Academic Press. Social Issues, 57, 389-412. *Brown, R., Maras, P., Masser, B., Vivian, J., & Hewstone, M. Esses, V. M., Haddock, G., & Zanna, M. P. (1993). Values, stereo- (2001). Life on the ocean wave: Testing some intergroup hy- types, and emotions as determinants of intergroup attitudes. In potheses in a naturalistic setting. Group Processes and Inter- D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and group Relations, 4, 81-97. stereotyping: Interactive processes in group perception (pp. *Cadinu, M., & Cerchioni, M. (2001). Compensatory biases after 137-166). New York: Academic Press ingroup threat: "Yeah, but we have a good personality." Euro- *Esses, V. M., Hodson, G., & Dovidio, J. F. (2003). Public attitudes pean Journal ofSocial Psychology, 31, 353-367. toward immigrants and immigration. In C. M. Beach, A. G. *Cadinu, M., & Reggiori, C. (2002). Discrimination of the Green, & J. F. Reitz (Eds.), Canadian immigration policy for low-status outgroup: The role of ingroup threat. European the 21st century (pp. 507-536). Montreal, Canada: Journal ofSocial Psychology, 32, 501-515. McGill-Queen's Press. *Corenblum, B., & Stephan, W. G. (2001). White fears and Native *Esses, V. M., Jackson, L. M., & Armstrong, T. L. (1998). Intergroup apprehensions: An integrated threat theory approach to inter- competition and attitudes toward immigrants and immigration: group attitudes. Canadian Journal of Behavioral Science, 33, An instrumental model of group conflict. Journal ofSocial Is- 251-268. sues, 54, 699-724.

350 Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 INTERGROUP THREATS AND ATTITUDES

Fein, S., & Spencer, S. J. (1997). Prejudice as self-image mainte- *Hopwood, M., & Connors, J. (2002). Heterosexual attitudes to ho- nance: Affirming the self through derogating others. Journal of mosexuality: Homophobia at a rural Australian university. Personality & Social Psychology, 73, 31-44. Journal ofGay and Lesbian Social Services, 14, 79-94. Fiske, S. T. (1993). Controlling other people. American Psycholo- *Hunsberger, B., Owusu, V., & Duck, R. (1999). Religion and preju- gist, 48, 621-628. dice in Ghana and Canada: Religious fundamentalism, Fiske, S. T., Cuddy, A., Glick, P., & Xu, J. (2002). A model of (often right-wing authoritarianism, and attitudes toward homosexuals mixed) stereotype content: Competence and warmth respec- and women. The International Journal for the Psychology of tively follow from perceived status competition. Journal ofPer- Religion, 9, 181-194. sonality and Social Psychology, 82, 878-902. Hunter, J. E., & Schmidt, F. L. (1990). Methods of meta-analysis. Fiske, S. T., Xu, J., Cuddy, A., & Glick, P. (1999). (Dis)respecting London: Sage. versus (dis)liking: Status and interdependence predict ambiva- Insko, C. A., Schopler, J., Hoyle, R. H., Dardis, G. J., & Graetz, K. A. lent stereotypes of competence and warmth. Journal ofSocial (1990). Individual-group discontinuity as a function offear and Issues, 55, 473-491. greed. JournalofPersonalityandSocialPsychology, 58,68-79. Gaertner, S. L., & Dovidio, J. F. (2000). Reducing intergroup bias: Islam, M. R., & Hewstone, M. (1993). Dimensions ofcontact as pre- The common ingroup identity model. Philadelphia: Psychology dictors of intergroup anxiety, perceived out-group variability, Press. and out-group attitude: An integrative model. Personality and Gaertner, S. L., Dovidio, J. F., Banker, B. S., Houlette, M. A., John- Social Psychology Bulletin, 19, 700-710 son, K. M., & McGlynn, E. A. (2000). Reducing intergroup *Islam, M. R., & Jahjah, M. (2001). Predictors ofyoung Australians' conflict: From superordinate goals to decategorization, attitudes toward Aboriginals, Asians, and Arabs. Social Behav- recategorization, and mutual differentiation. Group Dynamics, ior and Personality, 29, 569-580. 4, 98-114. Jackson, J. (1993). Realistic group conflict theory: A review and Galinsky, A. D., & Moskowitz, G. B. (2000). Perspective-taking: evaluation of the theoretical and empirical literature. Psycho- Decreasing stereotype expression, stereotype accessibility, and logical Record, 43, 395-414. ingroup favoritism. Journal ofPersonality and Social Psychol- *Jackson,J., &Smith, E. R. (1999). Conceptualizing socialidentity: A ogy, 78, 708-724. new framework and evidence for the impact of different dimen- *Goldman, M., Stockbauer, J. W., & McAuliffe, T. G. (1977). Inter- sions. PersonalityandSocialPsychologyBulletin, 25, 120-135. group and intragroup competition and cooperation. Journal of *Jackson, J. S., Brown, K. T., Brown, T. N., & Marks, B. (2001) Con- Experimental Social Psychology, 13, 81-88. temporary immigration policy orientations among domi- *Gonnerman, M. E. (2001). Ingroup enhancement or outgroup deni- nant-group members in Western Europe. Journal ofSocial Is- gration: The effects of group identification, social identity sues, 57, 389-412. threat, and audience composition. Unpublished manuscript. *Jackson, L. M., & Esses, V. M. (1997). Of scripture and ascription: *Grant, P. R. (1992). Ethnocentrism between groups of unequal The relation between religious fundamentalism and intergroup power in response to perceived threat to social identity and val- helping. Personality and Social Psychology Bulletin, 23, ued resources. Canadian Journal of Behavioural Science, 24, 893-906. 348-370 *Jackson, L. M., & Hunsberger, B. (1999). An intergroup perspec- *Grant, P. R. (1993a). Ethnocentrism in response to a threat to social tive on religion and prejudice. Journalfor the Scientific Study of identity. Journal of Social Behavior and Personality, 8, Religion, 38, 509-523. 143-154. *Jacobson, C. K. (1985). Resistance to affirmative action: Self-inter- *Grant, P. R. (1993b). Reactions to intergroup similarity: Examina- est or racism? Journal ofConflict Resolution, 29, 306-329. tion of the similarity differentiation and the similarity attraction Jetten, J., Branscombe, N. R., Schmitt, M. T., & Spears, R. (2001). hypotheses. Canadian Journal of Behavioral Science, 25, Rebels with a cause: Group identification as a response to per- 28-44. ceived discrimination from the mainstream. Personality and *Haddock, G., Zanna, M. P., & Esses, V. M. (1993). Assessing the Social Psychology Bulletin, 27, 1204-1213. structure of prejudicial attitudes: The case of attitudes toward Jetten, J., Spears, R., & Manstead, A. (2001). Similarity as a source homosexuals. Journal of Personality and Social Psychology, of differentiation: The role of group identification. European 65,1105-1118. Journal ofSocial Psychology, 31, 621-640. *Haddock, G., Zanna, M. P., & Esses V. M. (1994). The (limited) Jetten, J., Spears, R., & Postmes, T. (2004). Intergroup distinctive- role of trait-laden stereotypes in predicating attitudes toward ness and differentiation: A meta-analytic integration. Journal of Native peoples. British Journal of Social Psychology, 33, Personality and Social Psychology, 86, 862-879. 83-106. Judge, T. A., Heller, D., & Mount, M. K. (2002). Five-factor model Hamilton, D. L., Sherman, S. J., & Ruvolo, C. M. (1990). Stereo- of personality and job satisfaction: A meta-analysis. Journal of type-based expectancies: Effects on information processing Applied Psychology, 87, 530-541. and social behavior. Journal ofSocial Issues, 46, 35-60. *Kahn, A., & Ryen, A. H. (1972). Factors influencing the bias to- Hassan, M. K. (1978). A study of ethnocentrism, prejudice, and re- wards one's own group. International Journal of Group Ten- lated personality factors in Hindu and Muslim college students. sions, 2, 33-50 Psychologia, 21, 150-154. *Kilianski, S. E. (2003). Explaining heterosexual men's attitudes to- Hedges, L. V., & Olkin, I. (1985). Statistical methods for ward woman and gay men: The theory ofexclusively masculine meta-analysis. Orlando, FL: Academic Press. identity. Psychology ofMen and Masculinity, 4, 37-56. Hedges, L. V., & Vevea, J. L. (1998). Fixed and random-effects mod- Kinder, D. R., & Sears, D. 0. (1981). Prejudice and politics: Sym- els in meta-analysis. Psychological Methods, 3, 486-504. bolic racism versus racial threats to the good life. Journal of *Hennessy, J., & West, M. A. (1999). Intergroup behavior in organi- Personality and Social Psychology, 40, 414-431. zations: A field test of social identity theory. Small Group Re- *Kleinpenning, G., & Hagendoorn, L. (1993). Forms of racism and search, 30, 361-382. the cumulative dimension of ethnic attitudes. Social Psychol- Hewstone, M. (2003). Intergroup contact: Panacea for prejudice? ogy Quarterly 56, 21-36. Psychologist, 16, 352-355. Kuncel, N. R., Hezlett, S. A., & Ones, D. S. (2001). A comprehen- *Ho, C., & Jackson, J. W. (2001). Attitudes toward Asian Ameri- sive meta-analysis of the predictive validity of the graduate re- cans: Theory and measurement. Journal ofApplied Social Psy- cord examinations: Implications for graduate student selection chology, 31, 1553-1581. and performance. Psychological Bulletin, 127, 162-181.

Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 351 RIEK, MANIA, GAERTNER

Langford, T., & Ponting, R. (1992). Canadians' responses to aborigi- *Parker, A., & Bhugra, D. (2000). Attitudes of British medical stu- nal issues: The roles ofprejudice, perceived group conflict, and dents towards male homosexuality. Sexual and Relationship economic conservatism. Canadian Review of Sociology and Therapy, 15, 141-149. Anthropology, 29, 410-166. Parker, C. P., Baltes, B. B., Young, S. A., Huff, J. W., Altmann, R. A., *Laythe, B., Finkel, D., & Kirkpatrick, L. A. (2001). Predicting prej- Lacost, H. A., etc. (2003). Relationships between psychological udice from religious fundamentalism and right-wing authoritar- climate perceptions and work outcomes: A meta-analytic re- ianism: A multiple-regression approach. Journalfor the Scien- view. Journal ofOrganizational Behavior, 24, 389-416. tific Study ofReligion, 40, 1-10. Pettigrew, T. F. (1997). Generalized intergroup contacteffects onprej- *Leach, C. W., Spears, R., Branscombe, N. R., & Doosje, B. (2003). udice. PersonalityandSocialPsychologyBulletin, 23, 173-185. Malicious pleasure: Schadenfreude at the suffering of another Pettigrew, T. F., & Tropp, L. R. (2000). Does intergroup contact re- group. Journal of Personality and Social Psychology, 84, duce prejudice: Recent meta-analytic findings. In S. Oskamp 932-943. (Ed.), Reducing prejudice and discrimination: The Claremont Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. Symposium on applied social psychology (pp. 93-114). London: Sage. Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Locksley, A., Borgida, E., Brekke, N., & Hepburn, C. (1980). Sex Plant, A. E., & Devine, P. G. (2003). The antecedents and implica- stereotypes and social judgment. Journal ofPersonality & So- tions of interracial anxiety. Personality and Social Psychology cial Psychology, 39, 821-831. Bulletin, 29, 790-801. *Maass, A. Cadinu, M., Guarnieri, G., & Grasselli, A. (2003). Sex- Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1994). So- ual harassment under social identity threat: The computer ha- cial dominance orientation: A personality variable predicting rassment paradigm. Journal ofPersonality and Social Psychol- social and political attitudes. Journal ofPersonality and Social ogy, 85, 853-870. Psychology, 67, 741-763. Mackie, D. M., Devos, T., & Smith, E. R. (2000). Intergroup emo- Quillian, L. (1995). Prejudice as a response to perceived group tions: Explaining offensive action tendencies in an intergroup threat: Population composition and anti-immigrant and racial context. Journal of Personality and Social Psychology, 79, prejudice in Europe. American Sociological Review, 60, 602-616. 586-611. *Matchinsky, M. S., & Iverson, T. G. (1996). Homophobia in hetero- Quinones, M. A., Ford, K. J., & Teachout, M. S. (1995). The rela- sexual female undergraduates. Journal of Homosexuality, 31, tionsliip between work experience andjob performance: A con- 123-128. ceptual and meta-analytic review. Personnel Psychology, 48, Matheson, K., & Cole, K. (2004). Coping with a threatened group 887-910 identity: Psychosocial and neuroendocrine responses. Journal *Quinton, W. J., Cowan, G., & Watson, B. D. (1996). Personality and ofExperimental Social Psychology, 40, 777-786. attitudinal predicators of support of Proposition 187-Califor- McConahay, J. B. (1982). Self-interest versus racial attitudes as cor- nia's anti-illegal immigrant initiative. Journal ofApplied Social relates of anti-busing attitudes in Louisville: Is it the buses or Psychology, 26, 2204-2223. the blacks? Journal ofPolitics, 44, 692-720. *Quist, R. M., & Resendez, M. G. (2002). Social dominance threat: *McLaren, L. M. (2003). Anti-immigrant prejudice in Europe: Con- Examining social dominance theory's explanation of prejudice tact, threat perception, and preferences for the expulsion of mi- as legitimizing myths. Basic and Applied Social Psychology, grants. Social Forces, 81, 909-936. 24, 287-293. Mendes, W. B., Blascovich, J., Lickel, B., & Hunter, S. (2002). Chal- Raudenbush, S. W. (1994). Random effects models. In H. Cooper & lenge and threat during social interaction with white and black L. V. Hedges (Eds.), The handbook of research synthesis (pp. men. Personality and Social Psychology Bulletin, 28, 939-962. 301-321). New York: Russell Sage Foundation. Miller, D. A., Smith, E. R., & Mackie, D. M. (2004). Effects of inter- *Renfro, C. L., & Stephan, W. G. (2004). The role ofthreats in men's group contact and political predispositions on prejudice: Role attitudes toward women. Manuscript in preparation. of intergroup emotions. Group Processes and Intergroup Rela- Roccas, S., & Schwartz, S. H. (1993). Effects of intergroup similar- tions, 7, 221-237 ity on intergroup relations. European Journal of Social Psy- *Montieth, M. J., & Spicer, C. V. (2000). Contents and correlates of chology, 23, 581-595. Whites' and Blacks' racial attitudes. Journal of Experimental Rosenthal, R. (1994). Statistically describing and combining studies. Social Psychology, 36, 125-154. In H. Cooper & L. V. Hedges (Eds.), The handbook ofresearch Mummendey, A., & Wenzel, M. (1999). Social discrimination and synthesis (pp. 231-244). New York: Russell Sage Foundation. tolerance in intergroup relations: Reactions to intergroup differ- Sawires, J. N., & Peacock, M. J. (2000). Symbolic racism and voting ence. Personality and Social Psychology Review, 3, 158-174. behavior on Proposition 209. Journal of Applied Social Psy- Neuberg, S. L., & Cottrell, C. A. (2002). Intergroup emotions: A chology, 30, 2092-2099. biocultural approach. In D. Mackie & E. Smith (Eds.), From Schopler, J., Insko, C. A., Graetz, K. A., Drigotas, S. M., Smith, V. prejudice to intergroup emotions: Differentiated reactions to A., & Dahl, K. (1993). Individual-group discontinuity: Further social groups (pp. 265-283). New York: Psychology Press. evidence for mediation by fear and greed. Personality and So- *Nicholas, L., & Durrheim, K. (1995). Religiosity, AIDS, and sexu- cial Psychology Bulletin, 19, 419-43 1. ality knowledge, attitudes, beliefs, and practices of Blacks *Schwarzwald, J., & Sabo-Manor, S. (2005). Perceptions of and atti- South-African first-year university students. Psychological Re- tudes toward foreign workers in Israel. Megamot, 44, 29-55. ports, 77, 1328-1330. *Schwarzwald, J., & Tur-Kaspa, M. (1997). Perceived threat and so- Orwin, R. G. (1983). A fail-safe N for effect size in meta-analysis. cial dominance as determinants of prejudice toward Russians Journal ofEducational Statistics, 8, 157-159. and Ethiopians. Megamot, 38, 504-527. Ouwerkerk, J. W., de Gilder, D., & de Vries, N. K. (2000). When the Sears, D. 0. (1988). Symbolic racism. In P. A. Katz & D. A. Taylor going gets tough, the tough get going: Social identification and (Eds.), Eliminating racism: Profiles in controversy (pp. 53-84). individual effort in intergroup competition. Personality and So- New York: Plenum Press. cial Psychology Bulletin, 26, 1550-1559. Sears, D. O., & Henry, P. J. (2003). The origins of symbolic racism. Park, B., & Judd, C. M. (1990). Measures and models of perceived Journal ofPersonality and Social Psychology, 85, 259-275. group variability. Journal of Personality and Social Psychol- Sedikides, C. (1997). Differential processing of ingroup and ogy, 59, 173-191. outgroup information: The role of relative group status in per-

352 Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 INTERGROUP THREATS AND ATTITUDES

meable boundary groups. European Journal ofSocial Psychol- *Stephan, W. G., Ybarra, O., Martinez, C. M., Schwarzwald, J., & ogy, 27, 121-144. Tur-Kaspa, M. (1998). Prejudice toward immigrants in Spain *Shelton, J. N. (2003). Interpersonal concerns in social encounters and Israel: An integrated threat theory analysis. Journal of between majority and minority group members. Group Pro- Cross-Cultural Psychology, 29, 559-576. cesses and Intergroup Relations, 6, 171-185. Tajfel, H., & Turner, J. C. (1979). An integrative theory ofintergroup Sherif, M., Harvey, 0. J., White, J., Hood, W., & Sherif, C. W. conflict. In W. G. Austin & S. Worchel (Eds.), The social psy- (1961). Intergroup conflict and cooperation: The robbers cave chology of intergroup relations (pp. 33-47). Monterrey, CA: experiment. Norman, OK: University Book Exchange. Brooks-Cole. Sherif, M., & Sherif, C. W. (1969). Ingroup and intergroup relations: Tajfel, H., & Turner, J. C. (1986). The social identity theory of inter- Experimental analysis. In M. Sherif & C. W. Sherif (Eds.), So- group behavior. In S. Worchel & W. G. Austin (Eds.), Psychol- cial psychology (pp. 221-266). New York: Harper & Row. ogy of intergroup relations (2nd ed., pp. 7-24). Chicago: *Skitka, L. J., Bauman, C. W., & Mullen, E. (2004). Political toler- Nelson-Hall. ance and coming to psychological closure following the Sep- *Tougas, F., Brown, R., Beaton, A. M., & Joly, S. (1995). tember 11, 2001, terrorist attacks: An integrative approach. Per- Neosexism: Plus ca change, plus c'est pareil. Personality and sonality and Social Psychology Bulletin, 30, 743-756. Social Psychology Bulletin, 21, 842-849. Smith, E. R. (1993). Social identity and social emotions: Toward *Tur-Kaspa Shimoni, M., & Schwarzwald, J. (2003). Perceived new conceptualizations of prejudice. In D. M. Mackie & D. L. threat and prejudice in three domains of inter-group rivalry in Hamilton (Eds.), Affect, cognition, and stereotyping: Interac- the Israeli Society. Megamot, 42, 549-584. tive processes in group perception (pp. 297-313). San Diego, Vescio, T. K., Sechrist, G. B., & Paolucci, M. P. (2003). Perspec- CA: Academic Press. tive-taking and prejudice reduction: The mediational role of Sniderman, P. M., & Tetlock, P. E. (1986). Symbolic racism: Prob- empathy arousal and situational attributions. European Journal lems of motive attribution in political analysis. Journal ofSo- ofSocial Psychology, 33, 455-472. cial Issues, 42, 129-150. *Voci, A., & Hewstone, M. (2003). Intergroup contact and prejudice *Spencer-Rodgers, J. (2001). Consensual and individual stereotypic toward immigrants in Italy: The mediational role ofanxiety and beliefs about international students among American host na- the moderational role of group salience. Group Processes and tionals. International Journal of Intercultural Relations, 25, Intergroup Relations, 6, 37-54. 639-657. Waldzus, S., Mumrhendey, A., Wenzel, M., & Weber, U. (2003). To- *Spencer-Rodgers, J., & McGovern, T. (2002). Attitudes toward the wards tolerance: Representations of superordinate categories culturally different: The role of intercultural communication and perceived ingroup prototypicality. Journal ofExperimental barriers, affective responses, consensual stereotypes, and per- Social Psychology, 39, 31-47. ceived threat. International Journal ofIntercultural Relations, *Watts, M. W. (1996). Political xenophobia in the transition from so- 26, 609-631. cialism: Threat, racism, and ideology among East German *Stephan, C. W., Stephan, W. G., Demitrakis, K. M., Yamada, A. M., youth. Political Psychology, 17, 97-126. & Clason, D. L. (2000). Psychology of Women Quarterly, 24, *Weigel, R. H., & Howes, P. W. (1985). Conceptions of racial preju- 63-73. dice: Symbolic racism reconsidered. Journal ofSocial Issues, *Stephan, W. G., Boniecki, K. A., Ybarra, O., Bettencourt, A., Ervin, 41,117-138. K. S., Jackson, L. A., et al. (2002). The role of threats in the ra- *Whitley, B. E. (1999). Right-wing authoritarianism, social domi- cial attitudes of Blacks and Whites. Personality and Social Psy- nance orientation, and prejudice. Journal ofPersonality andSo- chology Bulletin, 28, 1242-1254. cial Psychology, 77, 126-134. Stephan, W. G., & Renfro C. L. (2002). The role of threat in inter- Wilder, D. A. (1978). Reduction of intergroup discrimination group relations. In D. Mackie & E. Smith (Eds.), From preju- through individuation of the outgroup. Journal of Personality dice to intergroup emotions: Differentiated reactions to social and Social Psychology, 36, 1361-1374. groups (pp. 265-283). New York: Psychology Press *Wilson, T. (2001) Americans' views on immigration policy: Testing *Stephan, W. G., Renfro, C. L., Esses, V. M., Stephan, C. W., & Mar- the role of threatened groups interests. Sociological Perspec- tin, T. (2004). The effects of threat on attitudes toward immi- tives, 44, 485-501. grants. Manuscript in preparation. Wolsko, C., Park, B., Judd, C. M., & Bachelor, J. (2003). Intergroup *Stephan W. G., & Stephan, C. W. (1985). Intergroup anxiety. Jour- contact: Effects on group evaluations and perceived variability. nal ofSocial Issues, 41, 157-175. Group Processes and Intergroup Relations, 6, 93-110. *Stephan W. G., & Stephan, C. W. (1989). Antecedents ofintergroup *Wyman, M., & Snyder, M. (1997). Attitudes toward "Gays in the anxiety in Asian-Americans and Hispanic-Americans. Interna- military": A functional perspective. Journal ofApplied Social tional Journal ofIntercultural Relations, 13, 203-219. Psychology, 27, 306-329. Stephan, W. G., & Stephan, C. W. (1996). Predicting prejudice. In- Ybarra, O., Schaberg, L., & Keiper, S. (1999). Favorable and unfa- ternational Journal ofIntercultural Relations, 20, 409-426. vorable target expectancies and social information processing. Stephan,W. G., &Stephan, C. W. (2000). Anintegratedtheoryofprej- Journal ofPersonality and Social Psychology, 77, 698-709. udice. In S. Oskamp (Ed.), Reducingprejudice and discrimina- *Zanna, M. P. (1994). On the nature ofprejudice. Canadian Psychol- tion: The Claremont Symposium on applied social psychology ogy, 35, 11-23. (pp. 23-45). Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Zarate, M. A., Garcia, B., Garza, A. A., & Hitlan, R. T. (2004). Cul- *Stephan, W. G., Ybarra, O., & Bachman, G. (1999). Prejudice to- tural threat and perceived realistic group conflict as dual predic- ward immigrants. Journal of Applied Social Psychology, 29, tors of prejudice. Journal of Experimental Social Psychology, 2221-2237. 40, 99-105.

Downloaded from psr.sagepub.com at Harvard Libraries on November 15, 2016 353