Secondary Transfer Effects of Intergroup Contact: A Cross-national Comparison in Europe
Katharina Schmid1, Miles Hewstone1, Beate Küpper2, Andreas Zick3, and Ulrich Wagner4
1University of Oxford, UK
2University of Applied Sciences Niederrhein, Germany
3University of Bielefeld, Germany
4University of Marburg, Germany
Social Psychology Quarterly, 75, 28-51.
Department of Experimental Psychology, University of Oxford, South Parks Road, Oxford OX1 3UD, UK
This paper examines so-called ‘secondary transfer effects’ of intergroup contact, a phenomenon whereby positive intergroup contact experiences can influence attitudes not only towards encountered (primary) outgroups, but also towards other (secondary) outgroups that were not initially involved in the intergroup encounter. This paper relies on a unique cross-sectional sample of the general population drawn from eight European countries (N=7042), to examine the relationship between intergroup contact with immigrants, and attitudes towards primary (immigrants) and secondary (homosexuals and Jews) outgroups. Results showed that intergroup contact was not only directly associated with primary outgroup attitudes, but also indirectly associated with secondary outgroup attitudes, via attitude generalization. These relationships occurred primarily for individuals low in social dominance orientation. Findings are discussed in terms of their contribution to understanding secondary transfer effects of contact as well as the role of social dominance orientation as a moderator of such effects.
intergroup contact, secondary transfer effects, intergroup attitudes, social dominance orientation, intergroup relations
The reduction of prejudice has long been at the heart of social psychological and social scientific enquiry. The ‘contact hypothesis’ (Allport 1954) argues that direct, face-to-face contact with members of other groups may reduce negative, and foster positive, intergroup attitudes, when the following conditions are met: contact between group members is cooperative; they meet under equal status conditions; there is some pursuit of common goals; and contact is in some form institutionally supported. A meta-analysis of over 500 studies (Pettigrew and Tropp 2006) suggests, however, that these conditions are facilitating, rather than necessary. It is now widely agreed that the quality of contact, and not only the frequency with which it occurs, is crucial to effecting positive attitude change. Cross-group friendship is thus a particularly powerful type of high-quality contact and extended, or indirect, contact, is also typically associated with positive intergroup effects. Extended contact refers to the (mere) knowledge of a fellow ingroup member’s contact with outgroup members, and often involves family members or friends.
Does contact also have the potential to reduce prejudice more generally - not only prejudice towards groups the individual has been in contact with, but also prejudice towards other, even unrelated, groups? These ‘secondary transfer effects of intergroup contact’ (Pettigrew 2009) describe the generalization of contact effects from an encountered (primary) outgroup to other (secondary) outgroups. This paper examines secondary transfer effects of intergroup contact using a unique, large-scale cross-national comparison of eight European countries. This research examines the extent to which intergroup contact with immigrants is not only associated with more positive attitudes towards immigrants, but also the extent to which intergroup contact is indirectly associated with more positive attitudes towards two other groups: homosexuals and Jews. We thus examine the extent to which secondary transfer effects of contact may occur due to a process of attitude generalization, whereby intergroup contact effects on secondary outgroup attitudes are mediated by positive attitude change towards the primary outgroup. As well as addressing this question of how the secondary transfer effects happen, we also consider a factor that has the potential to affect when the secondary transfer effects of intergroup contact occur. This factor is a central social psychological predictor of prejudice, that of social dominance orientation (Sidanius and Pratto 1999). Exploring this moderating factor contributes to our understanding of when secondary transfer effects of intergroup contact are more, or less, pronounced.
SECONDARY TRANSFER EFFECTS OF INTERGROUP CONTACT
Intergroup contact exerts positive effects on outgroup attitudes via generalization of positive attitudes from the encountered individual to the wider outgroup. Indeed such generalization of intergroup contact effects from the individual outgroup member to the collective outgroup, but also from the specific intergroup situation to other situations, is crucial for intergroup contact to be an effective strategy for promoting positive intergroup attitudes (e.g., Brown and Hewstone 2005). We consider an additional type of generalization in the present research: the generalization of contact effects from members of one group to other groups, so-called secondary transfer effects of intergroup contact.
Secondary transfer effects describe the generalization of contact effects from an encountered, primary outgroup to other, secondary outgroups, that may or may not have been previously encountered. Thus, these effects hold far-reaching potential consequences for intergroup contact as a means for prejudice reduction.
In a study conducted in Germany, Weigert (1976) found that Black soldiers’ contact experiences with White US soldiers were related not only with positive attitudes towards White US soldiers, but also with positive attitudes towards Germans, even when ideological orientation, demographics, and contact with Germans were controlled for in the analysis. Similar results were witnessed among White, non-Jewish Americans’ attitudes towards Jewish, Latino and Asian Americans attitudes, in a study examining positive intergroup contact experiences with Black Americans (Wilson 1996). And in his analysis of over 3800 majority group respondents in France, Germany, Great Britain and the Netherlands, Pettigrew (1997) found that cross-group friendship with minority group members present in each country tended to be associated with reduced prejudice both towards the minority groups, as well as towards secondary outgroups that were not present in the country, providing further evidence for secondary transfer effects.
Using both cross-sectional and longitudinal investigations of Germans’ prejudiced attitudes towards a range of groups, Pettigrew (2009) showed that contact with foreigners exerted positive effects on attitudes towards the homeless and homosexuals. And a longitudinal study by Van Laar et al. (2005), involving approximately 2000 White, African American, Asian American, and Latino students at a large U.S. college, showed that contact with roommates who belonged to a respective outgroup predicted attitudes not only towards the primary outgroup, but also towards other ethnic groups, even when contact with, as well as pre-existing attitudes towards, the secondary outgroups were controlled for in analyses.
The most comprehensive test of secondary transfer effects carried out to date, however, is based on a series of cross-sectional and longitudinal studies carried out by Tausch et al. (2010). These studies, conducted in different contexts (Cyprus, Northern Ireland, and the US), provided consistent support for secondary transfer effects, even when controlling for contact with the secondary outgroups, and socially desirable responding. This research also identifies the psychological mechanisms that help explain how such secondary transfer effects come about. Two key mediators of secondary transfer effects have been proposed to date: “deprovincialization” (or ingroup reappraisal; see Pettigrew ) and attitude generalization (see e.g., Tausch et al. 2010). We focus on the latter in the present study.
Attitude generalization as a mediator of secondary transfer effects
Conceptually, attitude generalization defines a process by which attitudes towards a particular attitude object generalize to other, related attitude objects. In the theoretical realm of secondary transfer effects, attitude generalization refers to a generalization of attitudes held towards a primary outgroup to other, secondary outgroups. Attitude generalization processes may thus explain how contact may indirectly affect attitudes towards secondary outgroups, via changes in attitudes towards a primary outgroup. In other words, according to the attitude generalization hypothesis, the attitude towards the encountered primary outgroup mediates the effects of contact with the primary outgroup on attitudes towards secondary outgroups. In principle, attitude generalization need not be restricted to groups that can be classified as similar (e.g., both being ethnic or religious groups). For example, attitudes towards a particular ethnic group are thought to generalize towards different religious groups, or other social categories. Attitude generalization is, however, likely to be stronger if the secondary outgroup is more similar than dissimilar to the primary outgroup.
The little research that has investigated mediators of secondary transfer effects to date has yielded unequivocal evidence supporting the attitude generalization hypothesis. Pettigrew (2009), for example, reported that the relationship between contact with foreigners and attitudes towards homosexuals and the homeless was partially mediated by attitudes towards the primary, encountered outgroup, foreigners, i.e. via a process of attitude generalization.
Social dominance orientation as a moderator of intergroup contact
In this paper we consider an individual difference variable known to be strongly linked with prejudice: social dominance orientation, an individual difference variable that taps individuals’ preference for unequal and hierarchically structured group relations between salient groups. Individuals with high social dominance orientation have a strong desire to promote group-based hierarchies, and extensive prior research has found that social dominance orientation is a consistently strong predictor of prejudice. People higher in social dominance orientation are more sexist, racist, prejudiced toward immigrants, lesbians, gay men, feminists, housewives, and physically disabled people (Son Hing and Zanna 2010). In this paper we argue that the effectiveness of contact may, in fact, be limited for individuals who believe strongly in group based hierarchies, i.e. who are high in social dominance orientation. Since individuals high in social dominance hold a competitive motivation to enforce or enhance social hierarchy and dominance (typically achieved through prejudicial attitudes towards the lower-status and/or competitive outgroups), they should be particularly reluctant to change their attitudes. We therefore argue that individuals high in social dominance orientation may not display positive attitude change as a result of positive intergroup contact to the same extent as individuals low in social dominance orientation.
THE PRESENT RESEARCH
The present research explicitly examined secondary transfer effects of intergroup contact in the context of intergroup relations between members of the host nation and immigrants, as well as other social groups, in eight European countries (France, Germany, Hungary, Italy, the Netherlands, Poland, Portugal, and the UK). The study was carried out within the realm of a cross-national research program, the ‘Group-Focused Enmity Europe Project’ (Zick, Küpper, and Hövermann 2011). Our test of secondary transfer effects involved examining the relationship between intergroup contact with members of the immigrant population and attitudes towards immigrants (the primary outgroup), as well as attitudes towards homosexuals and Jews (the secondary outgroups). Our explicit aim was to test the extent to which the relationship between intergroup contact and attitudes towards the two secondary outgroups is mediated by attitudes towards the primary outgroup, immigrants (i.e. to test the attitude generalization hypothesis). Specifically, we derived and tested the following hypotheses:
Hypothesis 1: Secondary transfer effects
Hypothesis 1a. There will be a direct, positive relationship between positive intergroup contact with immigrants and positive attitudes toward immigrants (the primary outgroup).
Hypothesis 1b. There will be a direct, positive relationship between negative attitudes toward immigrants and negative attitudes toward Jews and homosexuals (the secondary outgroups).
Hypothesis 1c. The relationship between intergroup contact and attitudes towards Jews and homosexuals will be mediated by primary outgroup attitudes. This indirect relationship will reflect secondary transfer effects of intergroup contact via attitude generalization.
Our research further sought to examine the relative impact of intergroup contact on attitudes towards the primary and the secondary outgroups alongside a social dominance orientation, as well as the standard demographic variables that should be controlled for in such research (i.e., age, gender, education, income and political orientation). By testing secondary transfer effects of intergroup contact while examining potential interaction effects of contact with social dominance orientation, our test of secondary transfer effects is a particularly stringent test of the wider effectiveness of intergroup contact. We thus examine the relationship between contact with the primary outgroup and attitudes towards secondary outgroups while taking into consideration individual differences in social dominance orientation that may qualify the effectiveness of secondary transfer effects. We thus derived and tested the following additional hypotheses:
Hypothesis 2: Moderation by social dominance orientation
Hypothesis 2a. Social dominance orientation will be positively associated with prejudiced attitudes toward all three outgroups (the primary and the two secondary outgroups).
Hypothesis 2b. The direct relationship between intergroup contact and attitudes toward immigrants will be moderated by social dominance orientation. When social dominance orientation is low, the direct, positive relationship between positive intergroup contact and positive attitudes will be stronger than when social dominance orientation is high.
Hypothesis 2c. The indirect relationship between intergroup contact with the primary outgroup and attitudes towards Jews and homosexuals will be moderated by social dominance orientation. When social dominance orientation is low, the indirect relationship between primary outgroup contact and secondary outgroup attitudes will be stronger than when social dominance orientation is high. This will reflect a moderated mediation of secondary transfer effects of intergroup contact.
Participants, Procedure and Measures
Participants were 7042 adults, aged 16 and over, drawn from nationally representative samples in eight European countries. Data were collected using computer administered telephone interviewing. Respondents completed the following measures (all response scales ranged from 1– strongly disagree to 4 – strongly agree unless otherwise noted):
Intergroup contact. Two items measured intergroup contact: ‘How many of your friends are immigrants?’, and ‘How many of your [INGROUP] friends have friends who are immigrants?’. Both items were measured on a scale ranging from 1 = none, to 4 = many. Although typically treated as separate conceptual constructs of intergroup contact, the two items were consistently and significantly correlated in each country (range of correlation: r = .42, p r = .68, p
Social dominance orientation. Two items measured social dominance orientation: ‘Inferior groups should stay in their place’, and ‘It is probably a good thing that certain groups are at the top, while others are at the bottom’. Both items were significantly correlated in each sample (range of correlation: r = .36, p r = .43, p
Anti-immigrant attitudes. Four items measured anti-immigrant attitudes: ‘Immigrants enrich our culture (reverse coded)’, ‘Because of the number of immigrants, I sometimes feel like a stranger in [country]’, ‘There are too many immigrants in [country]’, ‘When jobs are scarce, [country natives] should have more rights to a job than immigrants’. The four items formed a reliable index in each country (range of Cronbach’s α = .65 to .81).
Attitudes towards homosexuals. Two items measured attitudes towards homosexuals: ‘It is a good thing to allow marriages between two men or two women’, and ‘There is nothing immoral about homosexuality’. The two items were reverse coded, and were significantly correlated in each country (range of correlation: r = .52, p r = .76, p
Anti-Semitism. Three items measured anti-Semitism: ‘Jews in general do not care about anything or anyone but their own kind’, ‘Jews have too much influence in [country]’, and ‘Jews try to take advantage of having been victims during the Nazi era’. The three items formed reliable indexes in each country (range of Cronbach’s α = .69 to .78, with the exception of Portugal which yielded a low reliability score, α = .51).
When estimating our model for the full sample (i.e., for the total sample aggregated across the eight countries), our results generally supported our first hypothesis (see Figure 1). In line with predictions (Hypothesis 1a), the direct relationship between intergroup contact and anti-immigrant attitudes was significant. More positive intergroup contact was thus associated with less negative attitudes towards the primary outgroup. Also in line with predictions (Hypothesis 1b), attitudes towards the primary outgroup (immigrants) were directly associated with attitudes towards homosexuals and Jews, such that more negative attitudes towards the primary outgroup were associated with more negative attitudes towards both secondary outgroups.
In support of the critical part of our first hypothesis, which predicted secondary transfer effects via attitude generalization (Hypothesis 1c), we also obtained a statistically significant indirect relationship between intergroup contact with the primary outgroup and attitudes towards homosexuals. The indirect relationship between primary outgroup contact and anti-Semitism also reached significance. As predicted, our estimated model therefore yielded secondary transfer effects of intergroup contact on secondary outgroup attitudes, via attitude generalization.
We further tested our second hypothesis. In line with predictions (Hypothesis 2a), social dominance orientation was consistently negatively associated with attitudes towards both primary and secondary outgroups. Social dominance orientation was directly associated with more anti-immigrant attitudes, more negative attitudes towards homosexuals, and with greater anti-Semitism. Importantly, the interaction term between contact and social dominance orientation also reached statistical significance for primary outgroup attitudes. Supporting predictions (Hypothesis 2b), results revealed that the relationship between contact and anti-immigrant attitudes was statistically significant and stronger for individuals low in social dominance orientation than for individuals high in social dominance orientation.
Since we obtained a significant moderation of the relationship between contact and anti-immigrant attitudes we proceeded to test the predicted moderated-mediation hypothesis (Hypothesis 2c). We thus estimated the indirect relationships between intergroup contact with the primary outgroup and attitudes towards the two secondary outgroups, homosexuals and Jews, separately for respondents at low and high levels of social dominance orientation. These moderated mediation results revealed significant indirect relationships between contact with immigrants and attitudes towards both secondary outgroups (homosexuals and Jews) for individuals at low and high levels of social dominance orientation. Although these moderated-mediation tests were significant in both cases, indicating the presence of secondary transfer effects of contact regardless of respondents’ level of social dominance orientation, the indirect relationships were stronger for individuals low in social dominance orientation than for individuals with relatively higher levels of social dominance orientation.
We followed up our full sample results by testing our predictions for each country separately. These country-specific analyses by and large confirmed the results obtained from the full sample, with a number of exceptions. Notably, in Portugal neither of our two hypotheses could be confirmed. And although secondary transfer effects of intergroup contact on both secondary outgroup attitudes were obtained in France and Germany (supporting Hypothesis 1), no moderated-mediation effects by social dominance orientation emerged in these contexts. Finally, in the Netherlands and Poland the secondary transfer effects hypothesis as well as the moderated secondary transfer effects hypothesis was only confirmed for one of the two secondary outgroups (Jews).
This paper adds to and extends a growing body of research on secondary transfer effects of intergroup contact. First, we demonstrated the effect in seven out of the eight national samples studied, and we did so while controlling for standard demographic variables, including political orientation. Second, we found that secondary transfer effects were, by and large, moderated by the extent to which individuals supported an ideology of inequality, such that we found the relationships to be stronger for individuals low in social dominance orientation. We discuss our findings with regard to (1) secondary transfer effects of intergroup contact and the attitude generalization hypothesis, (2) the moderating role of social dominance orientation in qualifying secondary transfer effects, and (3) methodological and conceptual contributions and limitations of the study, as well as future research directions.
Secondary transfer effects of intergroup contact and the attitude generalization hypothesis
Our research demonstrated that intergroup contact with immigrants was by and large associated with less negative attitudes towards two secondary outgroups, homosexuals and Jews. Both sets of relationships between contact and the secondary outgroups were typically mediated by attitudes towards immigrants, the primary outgroup attitude, such that contact with immigrants was associated with less anti-immigrant attitudes which mediated the relationship between contact and attitudes towards the secondary outgroups. These results are thus generally in line with the attitude generalization hypothesis, confirming, with large-scale data sets from multiple European countries, recent research by Tausch et al. (2010). Our findings thus also align with previous research which has found that people who hold a negative view of one outgroup often tend to also think negatively about other outgroups. This suggests that if and when positive change in attitudes towards one outgroup can be achieved by means of positive intergroup contact, it may then transfer to and manifest itself in more positive attitudes towards other outgroups also.
The attitude generalization hypothesis was, however, not universally confirmed in each context and for all outgroups. Thus in Poland no indirect relationship between contact and attitudes towards homosexuals emerged. Interesting in this context, however, is that intergroup contact with immigrants was directly related with less negative attitudes towards homosexuals. Moreover, in the Netherlands intergroup contact with immigrants was neither directly nor indirectly associated with attitudes towards homosexuals, nor was the primary outgroup attitude related with attitudes towards this particular secondary outgroup. One potential reason for this finding could be that in the Netherlands homosexuals may not be perceived as a highly salient outgroup. A report by the Netherlands Institute for Social Research suggests that in comparison to other European countries, the Netherlands has one of the highest levels of tolerance and acceptance of homosexuals (Keuzenkamp 2010). Hence it may be that homosexuals are not perceived as on a par with immigrants, thereby impeding generalization to this group. Indeed, the mean score of negative attitudes towards homosexuals was considerably lower in the Netherlands than in the other seven countries, whereas we detected no such pronounced difference between the anti-immigrant attitudes of our Dutch respondents in comparison to the other countries.
But what are the wider implications of these country-specific findings for the attitude generalization hypotheses? It has been argued that attitude generalization should occur even for unrelated outgroups (albeit more strongly so for outgroups that are more similar to one another). Some of our country-specific findings could call this claim into question, suggesting rather that the outgroups under consideration, as well as the potential perceived similarity between them, may matter, and may hold the potential to give rise to context-specific effects. Goffman (1963), for example, argues that one can distinguish between three different types of social stigma: tribal stigma (e.g., devalued ethnic, racial or religious groups), abominations of the body (e.g., physical handicapped groups) and blemishes of individual character (e.g., homosexuals, homeless, drug-addicts). According to this distinction, our research showed generalization from attitudes towards one tribal stigma group (immigrants) to another tribal stigma group (Jewish people), as well as to a different stigma group (homosexuals) only in some contexts; in other contexts (e.g., the Netherlands) this generalization did not occur. Yet where generalization did occur we cannot be entirely certain that respondents regarded the secondary outgroups under focus in this study as entirely unrelated. Although our outgroups, objectively considered, reflect unrelated social groups, it is of course possible that respondents subjectively perceived them to be part of a shared common-minority outgroup category, e.g., as outgroups of similar status, rather than perceiving them as separate outgroup categories. And it may be this that explains the mechanisms underlying attitude generalization. Indeed Pettigrew (2009) argues that secondary transfer effects may only be witnessed with regard to outgroups for which similar stereotypes, stigma or status prevail.
The role of social dominance orientation in secondary transfer effects
Our research considered the previously unexplored possibility that secondary transfer effects of contact may depend on individual difference variables, in this case on the extent to which individuals support an ideology of inequality. A key finding of our research is that secondary transfer effects did not occur universally but may depend on individuals’ level of social dominance orientation. Our findings thus showed that social dominance orientation not only moderated the relationship between contact with and attitudes towards the primary outgroup attitude (immigrants) (with the exception of France, Germany, and Portugal), but also by and large moderated the indirect relationships between contact with immigrants and anti-Semitism, as well as the indirect relationship between contact with immigrants and attitudes towards homosexuals (with the exception of Poland where we only witnessed a direct relationship between contact with immigrants and attitudes towards homosexuals, and the Netherlands where the primary outgroup attitude was of course unrelated to attitudes towards homosexuals). In all cases where a significant moderated mediation relationship emerged, findings were such that more positive contact was associated more strongly, or primarily, with the secondary outgroup attitudes for those individuals who were low in social dominance orientation, i.e. who were less inclined to support an ideology of inequality.
No moderation effect emerged in France or in Germany, where the attitude generalization hypothesis was upheld regardless of individuals’ levels of social dominance orientation. Thus in these two contexts, secondary transfer effects via attitude generalization occurred for all individuals, whether they supported a hierarchy of social dominance or not. This is interesting since it could suggest that intergroup contact may be particularly effective in these two contexts, or perhaps that social dominance orientation is a less important predictor of intergroup attitudes in these two contexts. It is noteworthy, for example, that levels of social dominance orientation in France were relatively lower than in the other countries considered, while also being unrelated to both secondary outgroups. Again, however, no definite conclusions about country-specific differences are warranted.
To conclude, the results of this large-scale cross-national study confirm the wider, and potentially the most far-reaching, effects of intergroup contact in fostering positive intergroup relations, thus providing a powerful testament to the broad effectiveness of intergroup contact in contributing to positive intergroup relations more generally. The potential implications of this research, and research on secondary transfer effects more generally, are thus far- reaching, given the growing need for effective strategies to promote intergroup harmony in diverse societies, including groups varying in ethnicity, religion, social status, sexual orientation and multiple other forms of stigmatized identity. Intergroup contact is one such strategy (see Hewstone 2009), yet intergroup contact with a wide range of groups is, of course, not a realistic possibility for everyone. Understanding when and how positive effects of encountering only one outgroup may generalize to, i.e., exert secondary transfer effects towards, secondary unrelated outgroups thus holds particular merit for the promotion of intergroup harmony. Nonetheless, our research also highlights that such secondary transfer effects may not be universal. Firstly, they may depend on individual differences, that is, on individuals’ predisposition to support an ideology of group-based inequality as captured in social dominance orientation. Secondly, they may depend on contextual differences across countries, and the differential status of the minority groups under investigation in each context. Taking into consideration these additional complexities is thus crucial when considering intergroup contact as a means for achieving positive social change.
Acknowledgments and funding
This research was initiated by Wilhelm Heitmeyer, University of Bielefeld, and funded by a grant on ‘Group Focused Enmity in Europe’ from the Compagnia di San Paolo, the Volkswagen foundation, the Freudenberg and Groeben foundation, and two additional private foundations, and was financially supported by the Institute of interdisciplinary Research on Conflict and Violence at the University of Bielefeld. We also gratefully acknowledge support from the Leverhulme Trust that facilitated the writing of this paper.
Allport, Gordon W. 1954. The nature of prejudice. Reading, MA: Addison-Wesley.
Goffman, Erving. 1963. Stigma: Notes on the Management of Spoiled Identity. Englewood Cliffs NJ: Prentice Hall.
Hewstone, Miles. 2009. Living apart, living together? The role of intergroup contact in social integration. Proceedings of the British Academy 162:243–300.
Keuzenkamp, Saskia. 2010. Monitoring acceptance of homosexuality in the Netherlands. The Hague: Netherlands Institute for Social Research/SCP.
Pettigrew, Thomas F.1997. “Generalized intergroup contact effects on prejudice”. Personality and Social Psychology Bulletin 23: 173–85.
Pettigrew, Thomas F. 2009. “Secondary transfer effect of contact: Do intergroup contact effects spread to noncontacted outgroups?”. Social Psychology 40:55–65.
Pettigrew, Thomas F., and Linda R. Tropp. 2006. “A meta-analytic test of intergroup contact theory”. Journal of Personality and Social Psychology 90:751–83.
Sidanius, James and Felicia Pratto. 1999. Social dominance. An intergroup theory of social hierarchy and oppression. New York: Cambridge University Press.
Son-Hing, Leanne S., and Mark P. Zanna. 2010. Individual differences in prejudice. In The Sage handbook of stereotyping, prejudice, and discrimination (pp. 163–178). Ed. J. Dovidio, M. Hewstone, & V. Esses. London: Sage.
Tausch, Nicole, Miles Hewstone, Jared B. Kenworthy, Charis Psaltis, Katharina Schmid, Jason R. Popan, Ed Cairns, and Joanne Hughes. 2010. “ ‘Secondary transfer’ effects of intergroup contact: Alternative accounts and underlying processes”. Journal of Personality and Social Psychology 99:282–302.
Van Laar, Colette, Levin, Shana, Stacey Sinclair, and Jim Sidanius. 2005. “The effect of college roommate contact on ethnic attitudes and behaviors”. Journal of Experimental Social Psychology 41:329–45.
Weigert, K. M. 1976. “Intergroup contact and attitudes about a third group: A survey of Black soldiers’ perceptions”. International Journal of Group Tensions 6:110–24.
Wilson, Thomas. C. 1996. “Prejudice reduction or self-selection? A test of the contact hypothesis”. Sociological Spectrum 16:43–60.
Zick, Andreas, Beate Küpper, and Andreas Hövermann (2011). Intolerance, prejudice and discrimination. A European report. Berlin: Friedrich Ebert Foundation.
Figure 1. Estimated model of secondary transfer effects of intergroup contact in full sample. Unstandardized regression coefficients are shown for statistically significant regression coefficients only. Note: SDO = social dominance orientation, Contact X SDO refers to the latent interaction between intergroup contact and SDO; OG = outgroup; *p p p