• Keine Ergebnisse gefunden

Generalization of Intergroup Contact Effects

N/A
N/A
Protected

Academic year: 2022

Aktie "Generalization of Intergroup Contact Effects"

Copied!
177
0
0

Wird geladen.... (Jetzt Volltext ansehen)

Volltext

(1)

Sybille Neji

Generalization of Intergroup Contact Effects

Dissertation

Psychologie

(2)

Generalization of Intergroup Contact Effects

Dissertation

zur Erlangung des Doktorgrades der Naturwissenschaften

(Dr. rer. nat.)

der Fakultät für Psychologie der FernUniversität in Hagen

vorgelegt von Sybille Neji, M. Sc.

aus Velbert

Hagen, 28.10.2020

(3)

Generalization of Intergroup Contact Effects Sybille Neji, M.Sc. Psychologie

Erstgutachter: Prof. Dr. Oliver Christ (FernUniversität in Hagen)

Zweitgutachter: Prof. Dr. Frank Asbrock (Technische Universität Chemnitz) Tag der mündlichen Prüfung: 24. März 2021

(4)

Acknowledgements

Although I had to write this work alone, so many people were involved in it and I would like to express my special thanks at this point. First and foremost, I would like to thank Oliver Christ, my supervisor. Without your tireless support, the goal of this work would not have been realized. I have never taken it for granted to get the opportunity to work on my PhD project and therefore you have my deepest gratitude. I would also like to thank Frank Asbrock, who did not hesitate for a moment to supervise my work as a second assessor. Further thanks go to all my co-authors, especially to Miles Hewstone for his always very valuable feedback and Chloe Bracegirdle for her numerous helpful comments.

Special thanks to the whole PME team, especially Anja Munder, Sarina Schäfer, and Mathias Kauff. Above all, your constant support and your friendship has helped me infinitely and I am glad to have (had) colleagues like you.

My deep gratitude goes to my incredible family: Mum, Celine & Joël, Caroline, Katharina, Uncle Ulrich, Aunt Isolde, Kristina & Karla. The love and support of my whole family means a lot to me and I’m very grateful to have (had) such wonderful people with me.

Many thanks to my amazing friends: Some of you have been with me since (my first day at) school—Anke, Alex, Ali, Bea, Bibo, Bine, Christiane, Corina, Dee, Doro, Heike, HG, Jörg, Karsten, Katja, Margha, Michi, Moris, Niki, Patrick, Reni, Roland, Sarah, Silke, Susi, Tina, Tobi, Uli, Uwe, Wencke, Werner, Wolfgang, … . Without your friendship, my life wouldn't be the same and that makes you my favorite people.

Finally, I would like to dedicate this thesis to my mother and my uncle—even though I have learned to be cautious about inferring causality I can definitely say that without them, I would not be where I am today.

(5)

Table of Content

List of Tables ... V List of Figures ... VI

Summary ... 1

Deutsche Zusammenfassung ... 2

General Introduction ... 4

Intergroup Contact ... 5

The Intergroup Contact Hypothesis ... 5

Negative Intergroup Contact ... 8

Generalization of Intergroup Contact Experiences ... 9

Primary Transfer of Intergroup Contact Effects ... 9

Moderators of Intergroup Contact Effects ... 11

Mediators of Intergroup Contact Effects ... 12

Intergroup Contact Effects for Majority and Minority Groups ... 13

Group Perception ... 16

Perceived Entitativity ... 16

The Present Research ... 19

Manuscript #1 ... 21

Supplementary Material for Manuscript #1 ... 62

Excluded studies Manuscript #2 ... 89

Manuscript #2 ... 90

Supplementary Material for Manuscript #2 ... 132

General Discussion ... 144

Limitations and Future Research ... 148

Conclusion ... 149

References ... 151

Appendix ... 167

Erklärung über den Umfang des eigenen Beitrags ... 168

Eidesstattliche Versicherung ... 170

(6)

List of Tables (Not including tables in manuscripts)

Table 1 Umfang der Beiträge der Autor_innen nach Beitragsbereichen in Prozent für Manuskript #1 ………168

Table 2 Umfang der Beiträge der Autor_innen nach Beitragsbereichen in Prozent für Manuskript #2 ………169

(7)

List of Figures (Not including figures in manuscripts)

Figure 1 A Mediated-Moderation Model for the Association between Intergroup Contact and Prejudice……….………..………20

(8)

Summary

In the field of research on intergroup relations, one of the most reliable effects is that positive outgroup contact between members of different groups is associated with lessened prejudice and more favorable outgroup attitudes. The present thesis contributes to this field by studying the generalization of intergroup contact effects as well as the difference in the strength of generalization effects for members of majority and minority groups. After introducing intergroup contact theory and reviewing intergroup contact research, I focus on the generalization of intergroup contact effects. I then consider evidence from group

perception research, more specifically research on perceived outgroup entitativity. In particular, I focus on the difference in the strength of intergroup contact effects for majority high status and minority low status groups. The implied assumption thereby is that

perceived outgroup entitativity facilitates the generalization of intergroup contact effects and is a possible explanation for the status-related difference in the generalization of intergroup contact effects. Accordingly, Manuscript #1 emphasises the generalization of intergroup contact effects, introducing perceived outgroup entitativity as a novel moderator.

In this manuscript, we found that higher perceived outgroup entitativity facilitates the generalization of positive intergroup contact and is associated with a stronger relation between positive intergroup contact and outgroup attitudes. Manuscript #2 focuses on status-related differences in intergroup contact effects, suggesting that these effects can partly be explained by differences in perceived outgroup entitativity. Overall the findings provide evidence that outgroup perception (i.e., perceived outgroup entitativity) is a critical factor influencing the generalization of intergroup contact effects. The relevance of the reported findings is discussed.

Keywords: intergroup contact effects, generalization of contact effects, perceived entitativity, difference in contact effects majority minority

(9)

Deutsche Zusammenfassung

Im Bereich der Intergruppenbeziehung ist einer der zuverlässigsten Effekte, dass positiver Kontakt zwischen Mitgliedern verschiedener Gruppen zum Abbau von Vorurteilen und zu positiveren Einstellungen zwischen den Gruppen führt. Die vorliegende Dissertation trägt zu diesem Forschungsfeld bei, indem sie sich auf die Generalisierung von

Intergruppenkontakteffekten und die unterschiedliche Stärke dieses Effekts für Mitglieder von Majoritäts- und Minoritätsgruppen konzentriert. Nach einer Einführung in die Theorie des Intergruppenkontakts und Darstellung der zugehörigen aktuellen Forschung, liegt der Schwerpunkt auf der Generalisierung von Intergruppenkontakteffekten unter Einbezug von Forschung aus dem Bereich der Wahrnehmung von Gruppen, genauer Forschung über wahrgenommene Entitativität von Fremdgruppen. Insbesondere setzte ich auch einen Schwerpunkt auf die unterschiedliche Stärke der Kontakteffekte für Status hohe Majoritäts- und Staus niedrige Minoritätsgruppen. Die Annahme ist, dass die wahrgenommene

Entitativität der Fremdgruppe die Generalisierung von Intergruppenkontakteffekten erleichtert und eine mögliche Erklärung für den statusbezogenen Unterschied der Generalisierung von Kontakteffekten ist. Entsprechend wird in Manuskript #1 der Schwerpunkt auf die Generalisierung von Intergruppenkontakteffekten gelegt und wahrgenommene Entitativität der Fremdgruppe als neuer Moderator eingeführt.

Insbesondere fanden wir heraus, dass eine höhere wahrgenommene Entitativität der Fremdgruppe die Generalisierung positiver Intergruppenkontakte erleichtert und mit einer stärkeren Beziehung zwischen positivem Intergruppenkontakt und Einstellungen gegenüber der Fremdgruppe verbunden ist. Manuskript #2 konzentriert sich auf diese statusbezogenen Unterschiede in den Intergruppenkontakteffekten und legt nahe, dass diese teilweise durch Unterschiede in der wahrgenommenen Entitativität der Fremdgruppe erklärt werden können. Insgesamt liefern die Ergebnisse Hinweise darauf, dass die Wahrnehmung der

(10)

Fremdgruppe (d.h. die wahrgenommene Entitativität der Fremdgrupppe) ein kritischer Faktor ist, der die Generalisierung der Kontakteffekte beeinflusst. Die Relevanz der Ergebnisse wird diskutiert.

Keywords: Intergruppenkontakteffekte, Generalisierung von Kontakteffekten, wahrgenommene Entitativität, Unterschiede in Kontakteffekten Majorität Minorität

(11)

General Introduction

Given ongoing globalization, opportunities for encounters between people from different groups have increased and societies have become more multicultural and diverse (e.g., in terms of ethnicity, culture, and sexual orientation). Handling this increased social diversity contains challenges to promote intergroup harmony and to decrease interethnic tensions that probably undermine intergroup relations (Blalock, 1967; Putnam, 2007). One central idea in the field of intergroup relations is that contact between individual members of different groups can improve intergroup relations (e.g., Allport, 1954). This assumption is one of the most tested hypotheses in social psychology (e.g., Pettigrew et al., 2011).

Meta-analytic integrations of the field have revealed the impact of intergroup contact on the reduction of prejudice (e.g., Pettigrew & Tropp, 2006). One central issue is whether the intergroup contact experience generalizes beyond the specific intergroup contact situation to the outgroup as a whole. Research to date has thereby identified two factors which can crucially affect the generalization of intergroup contact effects: contact effects are stronger when group memberships are salient and the contacted person is considered a typical member of their group (e.g., Brown & Hewstone, 2005; Hewstone & Brown, 1986).

However, there is evidence that the generalization of intergroup contact effects is more beneficial for members of majority high status groups compared to members of minority low status groups (for a meta-analysis see Tropp & Pettigrew, 2005)1.

By integrating theories and evidence from research focusing on intergroup contact and on group perception, the present thesis focuses, first, on perceived outgroup entitativity as a central group characteristic and a concept that might help to explain when intergroup

1 In this thesis, I define groups with fewer members as minorities and numerically larger groups as majorities and the expressions ‘majority/minority’ and ‘high status/low status groups’ are used synonymously unless otherwise specified.

(12)

contact effects are more likely to generalize and, second, on why intergroup contact effects are typically stronger for majority compared to minority groups.

In the following section, I begin with a brief introduction of intergroup contact theory, the generalization of intergroup contact effects, the difference in the strength of intergroup contact effects for majority and minority groups, and research evidence from the field of group perception (i.e., perceived entitativity). Assuming that perceived entitativity influences the generalization of a contact experience with a single outgroup member to the outgroup as a whole, Manuscript #1 focuses on perceived outgroup entitativity as an additional moderator of intergroup contact effects. Manuscript #2 elaborates on perceived outgroup entitativity as an explanation for the status-related difference in intergroup contact effects for majority and minority groups. Finally, the results of the present work are

examined in the final discussion, limitations are highlighted, and implications for further research are proposed.

Intergroup Contact

According to Turner (1982, p. 15), a social group can be defined as “two or more individuals who share a common social identification of themselves or, which is nearly the same thing, perceive themselves to be members of the same social category”. We all belong to numerous social groups, such as a family, a group of friends, or the organization we work for and we all have constant intergroup contact with members from other groups.

The Intergroup Contact Hypothesis

The original formulation of the intergroup contact hypothesis can be credited to Robin Williams (1947). On behalf of the Social Science Research Council, Williams provided an overview of previous research on intergroup relations. In his monograph The Reduction of Intergroup Tensions he stated that intergroup contact can reduce prejudices when 1) the involved groups have the same status, the same interests and share the same

(13)

tasks, 2) the situation allows personal, intimate contact, 3) the group members do not correspond to the stereotypes associated with their group and 4) the activities include both groups. Much more influential, but similar in principle, was Allport’s (1954) intergroup contact hypothesis. Building on early empirical work, Williams’ monograph, as well as on written reports from his students on their personal experiences, Allport specified that

intergroup contact can reduce prejudice when certain “optimal” conditions in the intergroup contact situation are fulfilled: 1) equal status between the groups within the intergroup contact situation, 2) common goals, 3) intergroup cooperation, and 4) support through institutions and authorities.

Despite the fact that most research on the intergroup contact hypothesis has focused on ethnic and racial groups, the effects of intergroup contact have also been tested on many other groups, including those that differ in age, gender, social status, or sexual orientation.

In their extensive meta-analysis of 515 studies and 713 independent samples, Pettigrew and Tropp (2006) showed that increased positive intergroup contact is generally associated with less prejudice and that this effect can generalize to the entire outgroup and across different situations. However, positive intergroup contact effects were present even when the optimal contact conditions specified by Allport were not met—indicating that they are not

necessary but facilitating conditions (Pettigrew & Tropp, 2006). Following these results, positive intergroup contact can also lead to more positive outgroup attitudes, for example through indirect encounters in the everyday environment, without direct interaction between members of different groups. However, intergroup contact under optimal conditions, as Allport described it, can lead to a significantly stronger improvement of outgroup attitudes.

(14)

Extensive research in the field of intergroup contact in the last decades has led to the acknowledgement of the intergroup contact hypothesis as an integrated theory (Hewstone, 2009; Hewstone & Swart, 2011). As Pettigrew and Tropp (2005) referred to it:

Moreover, Allport’s formulation specified neither the processes involved in

intergroup contact’s effects nor how these effects generalize to other situations, the entire outgroup, and other outgroups not involved in the contact (Pettigrew, 1998).

Indeed, these omissions explain why he called it an “hypothesis” and not a “theory”.

[…] Several key points illustrate our view of a reformulated approach to intergroup contact theory that builds on Allport’s formulation while it exploits the research findings of the last five decades. (p. 271)

As a consequence, not only direct but also indirect forms of intergroup contact can be distinguished. Although they are not relevant for the present thesis, they should be briefly listed for the sake of completeness. Allport (1954) himself referred to indirect

approaches and acknowledged that some strategies to implement a direct contact experience may induce threat, suggesting that vicarious experiences through films, novels, and dramas could be “more effective” (p. 488).

The term indirect intergroup contact can be seen as an umbrella term to describe intergroup contact effects that do not include the direct interaction between members of different groups. Extended contact postulates that the “knowledge that an in-group member has a close relationship with an out-group member can lead to more positive intergroup attitudes” (Wright et al., 1997, p. 74). A series of studies provided empirical evidence that people who know about or observe intergroup friendships show more positive outgroup attitudes compared to those who do not, even when controlling for direct friendship (for a review see Turner et al., 2007; for a meta-analysis see Zhou et al., 2019).

(15)

Another form of indirect contact, vicarious contact (i.e., observing ingroup members having successful cross-group contact), has also been proven as an effective tool to improve outgroup attitudes (e.g., Brown & Paterson, 2016; Mazziotta et al., 2011). As vicarious contact only requires that one observes an intergroup interaction, it is easily implemented and a more practical intervention strategy than direct forms of intergroup contact. A meta- analysis found that, in intervention studies, vicarious intergroup contact improved outgroup attitudes more effectively than other forms of contact (Lemmer & Wagner, 2015).

Additionally, Turner et al. (2007) proposed that imagined contact, that is, simply imagining a positive contact situation with an outgroup member, could improve outgroup attitudes (e.g., Miles & Crisp, 2014).

Negative Intergroup Contact

Intergroup contact experiences are not always positive and recent studies have also explored the effects of negative intergroup contact on outgroup attitudes, consistently demonstrating that positive and negative intergroup contact have opposite effects on

intergroup relations (e.g., Barlow et al., 2012; Paolini et al., 2010). However, these findings offer a less clear picture of the interplay, with some studies showing greater effects of negative intergroup contact (e.g., Barlow et al., 2012; Paolini et al., 2010) and other studies equal and, sometimes, even larger effects of positive intergroup contact (e.g., Rupar et al., 2020; Visintin et al., 2017). Thus, more research, considering valence of intergroup contact is required to gain a more comprehensive understanding of the effects of positive and negative intergroup contact on intergroup relations (for a further discussion on intergroup contact valence asymmetry, see Schäfer et al., 2020). In the present thesis I also contribute to filling this research gap by examining the generalization of positive and negative

intergroup contact effects (see Manuscript #1).

(16)

Generalization of Intergroup Contact Experiences

Intergroup contact can, and, if it is to be taken seriously as a possible intervention, it must, not only improve outgroup attitudes toward the outgroup member encountered in the contact experience. It should also generalize to the outgroup as a whole (i.e., a primary transfer effect; Pettigrew & Tropp, 2011), to other outgroups not involved in the intergroup contact situation (i.e., a secondary transfer effect; Pettigrew, 1997, 2009), and can have an influence on cognitive processes beyond intergroup relations (i.e., tertiary transfer effect;

Meleady et al., 2019).

Primary Transfer of Intergroup Contact Effects

A fundamental premise of Allport's (1954) intergroup contact hypothesis is that when individuals from different groups have positive intergroup contact experiences, this can reduce the individuals’ prejudice toward the outgroup as a whole (i.e., primary transfer effect). Indeed research has provided compelling evidence that positive intergroup contact not only improves interpersonal attitudes among the specific individuals that are present in the intergroup contact situations but, across a wide range of samples and contexts,

generalizes to the respective outgroup as a whole (for meta-analytical reviews, see Davies et al., 2011; Lemmer & Wagner, 2015; Paluck et al., 2019; Pettigrew & Tropp, 2006; for a meta-analysis on extended contact, see Zhou et al., 2019; for a meta-analysis on imagined contact, see Miles & Crisp, 2014; for a review on indirect and mass-mediated contact, see Harwood, 2017).

Although the focus of my work lies explicitly on the generalization of intergroup contact effects from the individual group member encountered during the contact

experience to the outgroup as a whole, it should be noted that the generalization of intergroup contact effects is not limited to improved outgroup attitudes toward the

immediate contact person and the entire outgroup. The secondary transfer effects describe

(17)

the process whereby contact with a member of one outgroup can generalize to other outgroups not directly involved in the contact situation (Pettigrew, 1997, 2009). In a series of studies Tausch et al. (2010) showed that secondary transfer effects also occur while controlling for direct contact with the secondary outgroups and cannot be explained in terms of socially desirable responding. Thus, positive intergroup contact can improve outgroup attitudes more generally and widely.

Additionally, recent research considered a different kind of transfer effects (i.e., tertiary) assuming that intergroup contact may have not only the potential to reduce prejudicial attitudes, but also lead to a generalized cognitive flexibility, coined cognitive liberalization (Hodson et al., 2018; Meleady et al., 2019). Research has shown that intergroup contact can trigger more systematic and complex thinking outside the box (Bowman, 2010). Cognitive liberalization following intergroup contact leads to increased productivity, higher creativity, and better problem-solving skills (Gurin et al., 2002; Lu et al., 2017; Maddux & Galinsky, 2009).

Taken together, positive intergroup contact leads not only to an improvement of attitudes towards one or multiple outgroups but can also lead to a more complex way of thinking. Thus, positive intergroup contact may become even more influential than previously assumed, thereby once again demonstrating the importance and relevance of examining factors that influence the generalization of intergroup contact effects given the extensive consequences.

A central question thereby concerns conditions under which positive changes achieved through intergroup contact between members of different groups generalize from the individual group member involved in the contact experience to the entire outgroup (e.g., Brown & Hewstone, 2005; Hewstone & Brown, 1986; see also Manuscript #1). In the

(18)

present thesis I will examine an additional and new factor, perceived outgroup entitativity, which is thought to influence the generalization of intergroup contact effects.

Moderators of Intergroup Contact Effects

Inspired by the question of when intergroup contact effects are most likely to happen, research has accumulated consistent evidence that group membership salience plays a key role in encouraging the generalization of intergroup contact effects (e.g., Brown

& Hewstone, 2005; Hewstone & Brown, 1986; see also Manuscript #1). Group

membership salience is an overarching term that can be defined and measured differently, for example through group member typicality or the extent to which group memberships are psychologically present during intergroup contact.

Brown and Hewstone (2005) argued that if group membership is cognitively accessible during intergroup contact, individuals are perceived as members of their group, and the improvement of attitudes towards the individual group member is more likely to generalize to the entire outgroup. By contrast, when group membership salience is low, the intergroup contact experience will be perceived as an interindividual interaction, thus limiting the potential to change outgroup attitudes.

According to models of perceived typicality and categorization (Rosch, 1999), improvement of attitudes will generalize to a wider social category only when the outgroup member is categorized as a typical member of his or her particular group. When the

perceived prototypicality of the outgroup member is low, the person encountered in the intergroup contact situation will be subtyped and the perceiver’s attitudes toward the outgroup as a whole would not change (Wilder, 1984).

Additionally, other factors that may facilitate or hinder the generalization of

intergroup contact effects are identified, including past intergroup contact experiences (e.g., Al Ramiah et al., 2013; Voci et al., 2017), individual dispositions (e.g., social dominance

(19)

orientation and right-wing authoritarianism, Dhont & Van Hiel, 2009; diversity beliefs, Adesokan et al., 2011; for a review see Turner et al., 2020), ingroup identification (e.g., Voci et al., 2015), neighborhood composition (e.g., Kauff et al., 2016), and group status (e.g., Tausch et al., 2009; Tropp & Pettigrew, 2005). A recent meta-analysis focused on intergroup valence as a moderator of intergroup contact effects finding stronger

generalization of intergroup contact effects for positive contact with admired outgroups and for negative contact with stigmatized outgroups (Paolini & McIntyre, 2019).

Overall, research has shown that numerous factors influence the generalization of intergroup contact effects and intergroup contact appears to be especially effective when the contact is high on both interpersonal and intergroup dimensions, as when optimal contact conditions are fulfilled in contact between cross-group friends and their respective group memberships remain salient (Brown & Hewstone, 2005). One aim of this thesis is to extend known factors explaining when generalization of intergroup contact effects is more likely (see Manuscript #1).

Mediators of Intergroup Contact Effects

A limitation of the classic intergroup contact hypothesis concerns the failure to specify the underlying processes of attitude change through intergroup contact (Pettigrew, 1998). The intergroup contact hypothesis predicts when intergroup contact leads to positive changes focusing on conditions (e.g., equal status in the intergroup contact situation) of the immediate contact situation but how these changes take place remains open—a problem that this work will also focus on (see Manuscript #2). A meta-analysis on mediators of intergroup contact (Pettigrew & Tropp, 2008) found that intergroup contact is associated with an improvement of attitudes by reducing negative affect (i.e., anxiety about intergroup contact) and by enhancing knowledge about the outgroup, and increasing empathy and perspective taking. However, knowledge about underlying processes will offer

(20)

opportunities for shaping intergroup contact situations and concrete interventions (see also Kenworthy et al., 2005). Thus, the extent to which positive changes through intergroup contact between individual members of different groups generalize to the entire outgroup (from individual to group) is crucial in determining how extensive and long-term the effects of intergroup contact are.

Intergroup Contact Effects for Majority and Minority Groups

Many, if not most, real-life intergroup contexts include groups that hold different positions, namely, either a minority or majority status position (Tajfel, 1981). Common definitions of minority or majority group membership are by group size (e.g., Brewer, 1991; Simon, 1992), power, and/or social status (e.g., Tajfel, 1981). Focusing on intergroup contact research taking group status into account, most studies showed that intergroup contact effects are generally weaker for minority compared to majority groups (e.g., Binder et al., 2009; Tropp & Pettigrew, 2005). More recent research revealed the difference in the strength of intergroup contact effects in different contexts, for example between Black and White people in South Africa, the USA, and Canada (Aboud et al., 2003; Gibson &

Classen, 2010; Swart et al., 2010; Tropp, 2007), Christians and Muslims in Indonesia (Kanas et al., 2015), Protestants and Catholics in Northern Ireland (Tausch, Hewstone et al., 2007; Tausch, Tam et al., 2007), and between Italians and non-Italians (Vezzali et al., 2010). Likewise, Lemmer and Wagner’s (2015) meta-analysis of intergroup contact interventions found that although prejudice was generally reduced for both majority and minority groups, interventions were more beneficial for the majority. Examining intergroup contact between two minority groups (Black lower status minority and Asian higher status minority students in the USA), Bikmen (2011) found a pattern similar to those typically found for minorities and majorities, showing once again that the difference in the size of

(21)

intergroup contact effects is related to the different status between groups and not limited to minority-majority group contexts.

Thus far, few possible explanations have been proposed for the commonly found difference in the strength of intergroup contact effects between majority and minority groups (see also Manuscript #2) and key findings from this research is that members of different status groups may have different intergroup contact experiences (Dovidio et al., 2017), different expectations and perceptions (e.g., Shelton et al., 2006), and different motivations and goals in intergroup relations (e.g., Bergsieker et al., 2010).

For example, Tropp and Pettigrew (2005) suggested that members of different status groups subjectively perceive intergroup contact in different ways, even when contact

conditions are objectively implemented to enhance the positive effect of intergroup contact.

The fear of discrimination and being a target of prejudice increases the salience of the group status for inferior groups and minority, compared to majority, members may be more insecure as to the extent to which optimal contact conditions are met. These differences are likely to limit the positive effects of intergroup contact independent of the positive nature of the intergroup contact experience (e.g., Hyers & Swim, 1998; Saguy et al., 2008; Tropp, 2006).

In line with this reasoning, a number of studies have confirmed that optimal contact conditions are more beneficial for majority high status, compared to minority low status, group members (e.g., Tropp & Pettigrew, 2005). However, the concept of how to

implement equal status in intergroup contact experiences between different status groups is problematic. Whereas Allport (1954) focused on equal status within the intergroup contact situation, other researchers highlight the importance and influence of preexisting status differences (e.g., Brewer & Kramer, 1985). In their critical discussion of intergroup contact research, Durrheim and Dixon (2005) noted that the optimal conditions are difficult to

(22)

achieve and that social structures hamper the implementation of conditions such as equal status (see also Hewstone & Brown, 1986).

Given that intergroup contact research has focused mainly on the improvement of outgroup attitudes and minority, compared to majority, groups have been understudied (Shelton, 2003), the understanding of factors that may potentially affect the generalization of intergroup contact effects for different status groups is limited (Tropp, 2006). There are reasons to assume that members of high and low status groups do not necessarily have the same expectations and feelings with regard to intergroup contact, or do not even perceive the intergroup contact experience in the same way (e.g., Devine & Vasquez, 1998; Dixon et al., 2010; Saguy et al., 2008; Tropp, 2003, 2006). Sigelman and Welch (1993) found that Whites and Blacks generally differ in their perception of the interracial relation and Blacks perceived more hostility whereas Whites were more likely to perceive the interracial interaction as pleasant. In line with this, Tropp (2007) found that intergroup contact was consistently associated with more interracial closeness for Whites, intergroup contact effects for Blacks were only observable for those who perceived no discrimination against their group.

Taken together, there are only a few approaches that have tried to explain the status- related difference in intergroup contact effects for majority and minority groups, and only few empirical studies examining this difference (e.g., Binder et al., 2009). Furthermore, none of the approaches mentioned above (see also Manuscript #2) found an explanation for why the generalization from individual outgroup members encountered in the intergroup contact situation to the outgroup member is typically stronger for the majority (compared to minority) group members, although this effect is not ubiquitous, and can even be reversed (e.g., Wölfer et al., 2016, Study 1; see also Manuscript #2).

(23)

This thesis proposes that the generalization process itself may be a decisive factor explaining why the strength of intergroup contact effects differs as a function of group status. There are reasons to assume that majority and minority group members not only differ in their perceptions of the contact situation, but also in their perceptions of the respective outgroup itself qua outgroup. To my knowledge there are no approaches that have sought to explain the status-related difference in the strength of intergroup contact effects by focusing on the generalization process itself; that is, potential differences

between majorities and minorities regarding the generalization of intergroup contact effects have not yet been explicitly addressed.

Group Perception

Considering that the difference in the strength of intergroup contact effects might be due to differences in the generalization process, I took research from group perception into account, arguing that over and above the well-known situational and individual

characteristics (e.g., outgroup member salience and typicality of the outgroup member), perceived characteristics of the outgroup itself may have a decisive influence (see Manuscript #1). A central question in the literature of group perception is what a group constitutes and when people are perceived as belonging together. In a classic paper, Campbell (1958) addressed this question and coined the term entitativity to refer to “the degree of having the nature of an entity, of having real existence” (p. 17).

Perceived Entitativity

Campbell (1958) defined entitativity as the degree to which a group really exists, that is, the extent to which it exists as a unified entity. Campbell derived the dimensions of entitativity from the four Gestalt principles of perceptual organization proposed by

Wertheimer (1938): proximity (i.e., elements close together are more likely to be perceived as parts of the same organization), similarity (i.e., similar elements are more likely to be

(24)

perceived as parts of the same organization), common fate (i.e., elements that move together in the same direction, or in successive temporal observation are more likely to be perceived as parts of the same organization), and Prägnanz (good continuation or good form, i.e., elements forming a part of a spatial organization or pattern, as a line or more complex form, tend to be perceived as a part of the same unit). Groups vary in the degree to which a perceiver assumes they are entitative, leading to the idea of an entitativity

continuum (e.g., Campbell, 1958; Hamilton & Sherman, 1996; Lickel et al., 2000).

The concepts of perceived entitativity and homogeneity, that is, viewing a group as composed of highly similar and interchangeable members (e.g., Yzerbyt et al., 2004), can be considered as related, but distinct aspects of group perception (e.g., Hamilton et al., 2004; Haslam et al., 2004). Although similarity (homogeneity) can increase perceptions of group entitativity—and vice versa—it is only one of several characteristics probably influencing the degree to which a group is perceived as a unified entity (e.g., Campbell, 1958; Lickel et al., 2000).

Research has shown that whereas perceiving an ingroup as highly entitative is often related to positive outcomes, like identification (e.g., Castano et al., 2002, 2003) and fulfilment of psychological needs (e.g., Crawford & Salaman, 2012), high perceived

outgroup entitativity is often associated with more negative outcomes, like higher prejudice (e.g., Agadullina & Lovakov, 2018), stereotyping (e.g., Levy et al., 2001; Rydell et al., 2007; Spencer-Rodgers et al., 2007), racism (e.g., Roets & Van Hiel, 2011), threat (Dasgupta et al., 1999), and intergroup bias (Gaertner & Schopler, 1998). Furthermore, research has shown that perceptions of entitativity influence information processing and several approaches to group perception distinguish between two different types of processing: individual- and group-level (e.g., Brewer, 1988; Fiske & Neuberg, 1990).

Hamilton and Sherman (1996) suggested that the same underlying processing system is

(25)

activated for both high and low entitative groups. When a group is perceived high in

entitativity, information processing is similar to individual-level information processing and information about low entitative groups is processed individually (Crawford et al., 2002;

Hamilton & Sherman, 1996; McConnell et al., 1997; Srull, 1981; see also Manuscript #1).

Of particular relevance, Crawford et al. (2002) found that the transfer of information from a single group member to others was stronger when the group was perceived as highly

entitative (compared to low), thereby confirming that perceived entitativity leads to integrated information processing, evaluating the group as a cohesive unit, and perceiving its members as interchangeable.

Furthermore, taking further research from group perception into account, there is ample research showing that the numerical size of a group is a critical factor influencing how entitative a group is perceived to be. It is argued that, due to its salience, minority group membership leads to higher perceived ingroup entitativity (e.g., Gerard & Hoyt, 1974; Mullen, 1991; Mullen et al., 1992). Likewise, there is research showing that group size is related to perceived outgroup entitativity (e.g., Simon, 1992; Simon & Hamilton, 1994). Thus, relative group status, defined as the relative position of groups on valued domains of comparison, such as social standing (Sachdev & Bourhis, 1987), explains the tendency of group members to perceive themselves, and to be perceived by outgroup members, either as belonging to an entitative group or as a loose association of people.

Taken together, the present thesis tests the conjecture that perceived outgroup entitativity is a decisive group dimension influencing the primary transfer of contact effects thereby explaining when generalization of intergroup contact effects is more likely (Manuscript #1) and why the generalization of intergroup contact effects differs for majority compared to minority groups (Manuscript #2).

(26)

The Present Research

The present thesis focuses on the generalization of intergroup contact effects taking group status into account. In Manuscript #1, we assumed that perceiving a group as high (compared to low) in entitativity facilitates the generalization of information and the intergroup contact experience with an individual outgroup member is thus more likely to generalize to the entire outgroup. Specifically, we examined factors that would facilitate the generalization of attitude change beyond the specific context in which the intergroup

contact occurs. We hypothesized that outgroup entitativity would play a key role in

encouraging generalization, suggesting that it moderates the primary transfer of intergroup contact effects. We tested these predictions in a random sample of the adult Northern Irish population (Study 1, N = 792), a cross-sectional online-study with German students (Study 2, N = 254), a longitudinal survey among White British majority and Asian British minority group members (Study 3, N = 460), and an online experiment with German participants (Study 4, N = 330), finding that the generalization of positive (but not negative) intergroup contact was stronger for high versus low entitative outgroups.

In Manuscript #2 we sought to better understand the differential impact of

intergroup contact effects for majority high status and minority low status group members by proposing perceived outgroup entitativity as a mediator of the relationship between ingroup status and prejudice (for a conceptual model see Figure 1). In two preregistered experiments (Ns = 347/396) we found that perceived outgroup entitativity partly explains how intergroup contact with an outgroup member improves attitudes towards the entire outgroup differently for minority low compared to majority high status groups.

(27)

Figure 1

A Mediated-Moderation Model for the Association between Intergroup Contact and Prejudice

(28)

Manuscript #1

Perceived Outgroup Entitativity as a Moderator of Intergroup Contact Effects

Manuscript submitted to: European Journal of Social Psychology

(29)

Perceived outgroup entitativity as a moderator of intergroup contact effects Sybille Neji¹, Miles Hewstone², Jared B. Kenworthy³, and Oliver Christ¹

¹FernUniversität in Hagen (University of Hagen), Germany

² University of Oxford, United Kingdom

³University of Texas at Arlington, United States of America

Corresponding Author: Sybille Neji, FernUniversität in Hagen, Faculty of Psychology, Universitätsstraße 33, 58097 Hagen, Germany, E-mail:

sybille.neji@fernuni-hagen.de

Author note. Miles Hewstone is Emeritus Professor, University of Oxford, UK.

(30)

Abstract

We introduce perceived entitativity as a novel moderator of intergroup contact effects and present results of four studies that provide supporting evidence. In Study 1 (N = 792), we predicted that higher perceived outgroup entitativity was associated with a stronger contact–

attitudes association. Our findings provide first evidence for the moderating role of entitativity (i.e., higher entitativity is associated with a stronger generalization of contact effects compared to lower entitativity). Study 2 (N = 254) replicated our findings.

Furthermore, this effect was observable even after controlling for the well-established moderators, outgroup member typicality and group membership salience. Study 3 (N = 460) revealed the moderating role of entitativity in a longitudinal survey. Finally, Study 4, an online experiment (N = 330), showed that the generalization of positive contact was stronger for highly entitative outgroups. We discuss implications for intergroup contact theory and outline avenues for future research.

Keywords: intergroup contact effects, perceived entitativity, moderator of intergroup contact, generalization of contact effects

(31)

Perceived outgroup entitativity as a moderator of intergroup contact effects One of the most enduring ideas in the study of intergroup relations is the assumption that contact between members of different groups reduces prejudice and therefore improves intergroup relations (e.g., Allport, 1954). A multitude of studies has provided evidence for the prejudice-reducing effect of intergroup contact (for meta-analytic reviews, see: Davies et al., 2011; Pettigrew & Tropp, 2006). In the intergroup contact literature one central issue of debate—with both theoretical and practical implications—has been how best to achieve generalization of attitudes from individual outgroup members encountered to the outgroup as a whole (Cook, 1978). Past research strongly suggests that positive effects of contact are more likely to generalize to the entire outgroup when the group membership of the encountered outgroup member is salient and the outgroup member is perceived to be typical rather than atypical (Brown & Hewstone, 2005). The aim of the present paper is to expand prior contact research by showing that beyond situational (i.e., group membership salience) and individual (i.e., typicality of the outgroup member) characteristics, the generalization of intergroup contact effects is facilitated by perceived attributes of the outgroup itself. As such, we introduce perceived outgroup entitativity as an additional moderator of intergroup contact effects.

Campbell (1958) defined the term entitativity as the degree to which a social aggregate is perceived as having “the nature of an entity, of having real existence” (p. 17). Research has shown that perceived entitativity influences the processing of information (e.g., McConnell et al., 1997; Welbourne, 1999) and the transfer of traits from one group member to other group members (Crawford et al., 2002). For instance, Crawford et al. (2002) showed that the transfer of information from an individual group member to other ingroup members was stronger for high than for low entitative groups or simply aggregates of individuals. We therefore suggest that perceiving a group as highly entitative facilitates the generalization of contact-derived information from individual outgroup members to the outgroup in general.

(32)

Generalization of Intergroup Contact Experiences

The premise of Allport's (1954) contact hypothesis is that contact between individuals from different groups can reduce prejudice toward the outgroup; that is, positive attitudes toward the encountered outgroup member generalize to the outgroup as a whole. Research has shown that both an outgroup member’s typicality and their outgroup category salience have important influences on the generalization of positive attitudes (e.g., Brown et al., 1999;

Brown & Hewstone, 2005; Wilder, 1984).

Outgroup member typicality reduces the likelihood of subtyping atypical individuals (Richards & Hewstone, 2001) and thus limiting their potential to change outgroup attitudes.

Wilder (1984) found that positive contact with a typical outgroup member generalizes to the whole outgroup whereas contact with an atypical outgroup member does not. A strong association between outgroup members and their corresponding group membership (i.e., typical group members) increases the likelihood that the contact experience is seen as an intergroup—rather than an interindividual—interaction making generalization of contact effects more likely.

Furthermore, Brown and Hewstone (2005) argued that the generalization of intergroup contact effects depends on relatively high category salience during the contact situation. If the interaction partner’s category membership is salient and is cognitively accessible during interactions, individuals are perceived as members of their group, and contact has the potential to change outgroup attitudes. By contrast, when salience is low, the contact

experience will be perceived primarily as an interindividual interaction and generalization of attitudes toward the entire outgroup is unlikely. Ensari and Miller (2005) showed that both outgroup member typicality and group membership salience are essential components of a contact situation that aims to reduce intergroup bias.

However, we argue that not only features of the situation (i.e., category salience) and the individual outgroup member (i.e., typicality) influence the generalization of contact

(33)

effects, but also (perceived) characteristics of the outgroup itself. Bringing together research evidence from intergroup contact and group perception literature, we hypothesize that over and above the established moderators, outgroup member typicality and category salience, perceived outgroup entitativity should plausibly play a key role in facilitating the

generalization of intergroup contact effects.

Perceived Entitativity

Stimulated by the classic work of Campbell (1958), a wave of research within the realm of group perception focused on the degree to which an aggregate of individuals is perceived as a unified entity. Campbell defined the term entitativity and focused on Gestalt principles of perceptual organization—proximity, similarity, common fate, and Prägnanz (meaning conciseness)—that lead discrete elements to be perceived as parts of an organized whole. Furthermore, Campbell proposed the idea of an entitativity continuum along which groups vary in the degree to which they are perceived as coherent units. Research has shown that perceiving a group as highly entitative often has negative consequences for intergroup relations; higher entitativity is positively associated with prejudice (Rydell et al., 2007, Study 1; Spencer-Rodgers et al., 2007) and can lead to more stereotyping and prejudice (e.g., Dasgupta et al., 1999; Newheiser et al., 2012; Rydell et al., 2007, Study 2). A recent meta- analysis (Agadullina & Lovakov, 2018) found a moderate positive effect between perceived outgroup entitativity and prejudice.

Brewer et al. (1995) argued that perceived entitativity is related to different

representation strategies and the organization of information about a group; high entitativity implies the cognitive representation of a social category “as a single, coherent unit as opposed to an associative link among individual group members” (p. 29). A fundamental question related to perceived entitativity concerns when a person is perceived as a member of a group and when they are perceived as an individual person (Crawford et al., 2002). The dual-process model (Brewer, 1988) and the continuum model (Fiske & Neuberg, 1990) both try to explain

(34)

the different types of impression formation. The two models distinguish between category- based and attribute-based impression formation, proposing that perceivers can form impressions both in stereotypic, category-based and in more individuated, attribute-based ways. While both models try to explain how information about individuals as well as groups helps perceivers to form impressions of individuals, they do not explain how impressions of groups are formed. Consequently, Hamilton and Sherman (1996) developed a model

explaining how impressions of groups are formed from information about the individual members of a group. They suggested that perceived entitativity is an explanation for the different mechanisms for processing information. When expectations of unity, consistency, and coherence are equated between individual and group targets, the information processes are very similar: information about high entitative groups is processed in an integrative manner (i.e., similarly to information processing about individuals) whereas information about low entitative groups is not (Hamilton & Sherman, 1996; Sherman et al., 2002). These different types of information processing involve the interchangeability of individual group members and the loss of their individuality when a group is perceived as highly entitative.

Of particular relevance to our argument, Crawford et al. (2002) examined how information about individual group members is integrated and generalized to other group members. In three experiments their participants viewed a series of photographs showing members of three different groups, which were described as being either high entitative, low entitative, or unrelated individuals. Additionally, the participants received behavioral

information about the respective person in the photograph. In the following task, half of the photographs were presented with a trait term that matched the behavioral information given earlier and half of the photographs were paired with a trait term that matched the behavioral information about other members of the same group. Crawford et al. found that the transfer from one group member to others was stronger for high entitative groups than for low

entitative groups or for aggregates of individuals, confirming Hamilton and Sherman’s (1996)

(35)

assumption of different kinds of information processing for high and low entitative groups.

Specifically, information about high entitative groups is processed in an integrative manner, whereas information about low entitative groups must be processed and learned individually (Crawford et al., 2002; Hamilton & Sherman, 1996; McConnell et al., 1997). Thus, perceived entitativity leads to integrated information processing, forming an impression of a group and the transfer of that group impression to all members of the group, evaluating the group as a cohesive unit, and perceiving all the members as sharing all group attributes and being interchangeable (Crawford et al., 2002).

We are thus suggesting that a positive experience with an individual group member is more likely to generalize to other group members to the degree that the group is perceived as a unified entity. That is, encountering a member from a highly entitative group and perceiving all the members as interchangeable should facilitate the generalization of the contact

experience to the outgroup as a whole. When the outgroup is seen as less entitative, by contrast, the generalization from the individual-level experiences to the group as a whole should be weaker, or perhaps not even occur.

The present research

In the present research we aimed to identify conditions which facilitate the generalization of positive attitude change generated by intergroup contact, introducing perceived entitativity as a novel potential moderator of contact effects. Based on research in the area of group perception, we assumed that entitativity is a defining characteristic in terms of which groups generally differ, and that both structural and social factors influence the extent to which we see a given group as a coherent entity or as a looser association of people (e.g., Campbell, 1958; Hamilton & Sherman, 1996; Lickel et al., 2000). Specifically, we predicted that high perceived outgroup entitativity would be associated with a stronger relation between intergroup contact and outgroup attitudes and outgroup trust, compared to low perceived outgroup entitativity (Study 1). Furthermore, we predicted that this moderating

(36)

effect would be observable independent from situational and individual features; that is, perceived outgroup entitativity would have an important moderating function over and above the well-established moderators of group membership salience and outgroup member

typicality (Study 2). Moreover, to increase confidence in and then demonstrate the causal effect of perceived outgroup entitativity on the generalization of contact effects, we examined the longitudinal effect of perceived outgroup entitativity in a two-wave survey (Study 3), and manipulated entitativity and contact valence in an online experiment (Study 4).

Study 1

In Study 1, we tested our assumptions in a context of real intergroup conflict and violence. Northern Ireland has been afflicted by decades of violence and intergroup tensions.

Although now a post-conflict society, Northern Ireland remains deeply segregated, especially in terms of housing and education (Niens et al., 2003), although one in which intergroup contact does occur, and with many positive outcomes (e.g., Voci et al., 2015). We used data from a large, pre-existing cross-sectional survey to test our main prediction, that high perceived outgroup entitativity would be associated with a stronger relation between

intergroup contact and prejudice reduction, compared to low perceived outgroup entitativity.

We included two indicators of prejudice, outgroup attitudes and outgroup trust. Whereas prejudice is typically indexed by conventional measures of outgroup attitude, outgroup trust also provides an indication of prejudice. Trust can be defined as a positive expectation about intentions and behavior of an outgroup toward the ingroup and is a necessary part of

reconciliation strategies that aim to improve intergroup relations in the aftermath of intergroup conflicts (Kenworthy et al., 2016). There is evidence showing that positive intergroup contact can help to enhance not only outgroup attitudes, but also outgroup trust (Tam et al., 2009).

Method Participants

(37)

The data collected for this study came from a random sample (N = 1,000) of the adult Northern Irish population. Professional interviewers conducted the sessions in respondents’

homes. We excluded participants who belonged to neither the Protestant nor the Catholic community (n = 11), indicated having no contact with the outgroup (n = 146), or did not know how to rate perceived outgroup entitativity (n = 51). The final sample for analysis included 792 participants (478 Protestant: 262 female [54.8 %], 216 male, age: M = 47.26, SD = 17.19, range 18-91; 314 Catholic: 175 female [55.7 %], 139 male, age: M = 44.20, SD = 16.45, range 18-89).

Measures2

All interviews were conducted in English.

Intergroup Contact. Our predictor variables were quality and quantity of contact with outgroup members (adapted from Islam & Hewstone, 1993). Quantity of contact was assessed with three items asking about the frequency of intergroup contact across different domains: “just chatting to people”, “over all social situations”, and “How many of your closest friends are from the other community?”. The first two items were assessed using 4- point scales (1 = a great deal of contact, 4 = no contact at all), and the third item using a 5- point scale (1 = none, 2 = one to five, 3 = six to ten, 4 = eleven to twenty, 5 = more than twenty), Cronbach’s α = .69. Quality of contact was measured with two items. Preceded by,

“When you meet members of the other community, in general do you find the contact…”, participants indicated on 7-point scales (1 = not at all, 7 = very much) how “pleasant”, and

“cooperative” the contact had been (r[787] = .88, p < .001). Because contact is expected to be associated with outgroup attitudes when it is cooperative and frequent (Allport, 1954), we

2 The data included in Study 1 were part of a larger survey and only measures relevant to our hypotheses are reported. Part of the data—including intergroup contact, intergroup appraisals, self- disclosure, intergroup emotions, trust, outgroup attitudes, behavioral tendencies—has already been published (see Kenworthy et al., 2016; Study 3), but the present inclusion of data, and focus, on entitativity is unique to this paper.

(38)

calculated a single index of high quality contact by multiplying the two scales (Brown et al., 2001; Voci & Hewstone, 2003)3.

Outgroup Attitudes. Respondents indicated their overall feelings toward the

outgroup along a feeling thermometer (0 = extremely unfavorable, 100 = extremely favorable;

Campbell, 1971).

Outgroup Trust. We measured general trust toward the outgroup by adapting Brehm and Rahn’s (1997) 3-item Trust Scale. The items were measured on 4-point scales, with higher numbers indicating more endorsement of each item (e.g., “Would you say that most members of the other community can be trusted or that you can't be too careful with them?”).

The items were re-coded when necessary so that higher scores indicated greater outgroup trust (α = .80).

Perceived Outgroup Entitativity. Outgroup entitativity was measured with four items, answered on 5-point scales (1 = not at all, 5 = very much), “To what degree do members of the other community share common goals with each other?”, “How similar to each other are members of the other community?”, “To what degree do members of the other community share common outcomes to daily events? In other words, how much does

something that affects one member affect others as well?” , and “To what degree are the other community divided among themselves?”. The items were re-coded when necessary so that higher scores indicated higher perceived outgroup entitativity (α = .61).

Results and Discussion

We used moderated regression analysis (PROCESS Version 3.0; Hayes, 2018; Model 1, 95% confidence intervals, 5,000 bootstrap samples, using unstandardized scores) to test for the moderating role of perceived outgroup entitativity on the association between intergroup

3 Separate analyses for contact quantity (while controlling for contact quality) and contact quality (while controlling for contact quantity) showed the same results as the single index and can be found in the supplemental online material.

(39)

contact, on the one hand, and outgroup attitudes and outgroup trust, on the other hand4. In the first step we examined whether perceived outgroup entitativity moderated the effect of contact on outgroup attitudes, F(3, 777) = 9.09, p < .001, f² = 0.03, R² = .03. Results indicated a significant interaction: entitativity b = 0.06, SE = 1.07, p = .956; contact b = -1.38, SE = 1.08, p = .200; interaction b = 0.73, SE = 0.32, p = .025, 95% CI [0.09, 1.36]. The relation between contact and outgroup attitudes was significant for high and low perceived outgroup entitativity, but, as expected, the relation was stronger for high than low entitativity, high entitativity (+1 SD) b = 1.52, SE = 0.32, t(777) = 4.76, p < .001, CI 95% [0.89, 2.15], low entitativity (-1 SD) b = 0.61, SE = 0.27, t(777) = 2.27, p = .023, CI 95% [0.08, 1.15] (see Figure 1).

Figure 1

Simple Slopes of Outgroup Attitudes on Intergroup Contact for Low and High Perceived Entitativity (+/- 1 SD), Study 1.

4Correlations among the variables, as well as their means and standard deviations, for our studies can be found in the supplemental material.

40 45 50 55 60 65 70 75 80 85 90

Low intergroup contact High intergroup contact

Outgroup Attitudes Low

perceived entitativity High perceived entitativity

(40)

A similar pattern was found for the moderating role of perceived outgroup entitativity between contact and outgroup trust, F(3, 774) = 20.34, p < .001, f² = 0.08, R² = .07. The estimates for contact and, more important, the interaction were significant: entitativity b = - 0.03, SE = 0.04, p = .360; contact b = -0.08, SE = 0.04, p = .034, 95% CI [-0.16, -0.01];

interaction b = 0.04, SE = 0.01, p < .001, 95% CI [0.02, 0.06]. The relation between contact and outgroup trust was significant for high and low perceived outgroup entitativity, but, again, the relation was stronger for high versus low entitativity, high entitativity (+1 SD) b = 0.08, SE = 0.01, t(774) = 7.27, p < .001, 95% CI [0.06, 0.10], low entitativity (-1 SD) b = 0.03, SE

= 0.01, t(774) = 3.23, p = .001 95% CI [0.01, 0.05]5. In line with our predictions, positive contact was related more strongly with outgroup attitudes and outgroup trust when the outgroup was perceived as high versus low in entitativity.

To summarize, results largely supported our prediction, that perceived outgroup entitativity would moderate the impact of contact variables on outgroup attitudes and

outgroup trust. Thus, Study 1 provides some initial evidence that the relation between contact and outgroup attitudes and trust is stronger for those who perceive the outgroup to be more entitative, compared to those who perceive it to be less entitative. Study 1 did not, however, allow us to control for the established moderators, group membership salience and outgroup member typicality; it is possible that the results of the first study could, in part, have been due to salience and typicality rather than to perceived outgroup entitativity. We therefore

conducted a second study to address this issue.

Study 2

In Study 2, we aimed to replicate the findings, and improve the research by obtaining ratings for both group membership salience and outgroup member typicality during

intergroup contact. Furthermore, while Study 1 relied on data from an existing survey, Study

5 Additional figures for simple slopes analyses are available for all studies in the supplemental online material.

(41)

2 was designed to investigate our research question, and thus relevant variables—in particular perceived entitativity—were assessed more comprehensively. We focused on Turks living in Germany as the target outgroup, because they are the largest (Statistisches Bundesamt, 2019) and most salient ethnic minority in Germany and elicit more prejudice than do other

immigrants (Asbrock et al., 2009). As in Study 1, we assumed that perceived outgroup entitativity would moderate contact effects, with higher perceived outgroup entitativity yielding stronger contact effects compared to lower perceived outgroup entitativity.

Furthermore, we expected that this moderating effect would explain differences in the generalization of contact effects over and above group membership salience and outgroup member typicality.

Method

All measures and materials were presented in German and can be found in the supplemental online material6.

Participants

Based on Study 1 we assumed a weak moderating effect of perceived outgroup entitativity on contact effects and conducted an a priori power calculation that suggested a required sample size of 268 participants (f² = 0.03, α = .05, 1-β = .80; G*Power 3.1.9.2; Faul et al., 2009).

Four hundred and thirty-five participants took part in a cross-sectional online-survey at the virtual lab of the psychology department of the University of Hagen (Germany).

Student participants were requested to pass on the link for the study. We included participants identifying as ethnic majority (i.e., German people without migration background) living in Germany (n = 283), who indicated having contact with Turks (n = 274). Furthermore, we

6 To reduce social desirability, we indicated that we were interested in general attitudes towards immigration and experiences with people of different ethnic origins. In order to make the cover story more meaningful, we first assessed attitudes towards immigration. We excluded this scale from our analyses.

(42)

excluded participants with incomplete (n = 18) and invalid (n = 2) data. These exclusions resulted in 254 participants (190 female [74.80 %], 63 male, one other; mean age = 35.59 years, SD = 10.92, range 18-65 years) in the analysis. Participants from the University of Hagen (n = 213 [83.86 %]) received 0.25 credit for their psychology course.

Measures

Intergroup Contact. Quantity and quality of intergroup contact was measured by three items from Islam and Hewstone (1993): “How often do you have contact with Turks (in the neighbourhood, at work etc.)?” (1 = never, 7 = very often), “How do you generally

experience this contact?” (1 = negative, 7 = positive), and “How many Turks are friends of yours?” (1 = none, 7 = all). We excluded the last item from the statistical analyses, because more than half of the participants (53.1 %) indicated they did not have any Turkish friends.

The predictor variable measured in this study was a composite index as in Study 1, computed as the product of the first two items, r(252) = .16, p < .01 (Brown et al., 2001; Voci &

Hewstone, 2003)7.

Group Membership Salience. Two items measured the salience of group

membership during intergroup contact (Voci & Hewstone, 2003). A stem statement, “When you met Turks ...”, was followed by two questions: “How aware were you that you belonged to different groups?”, and “Did you feel that you were two people representing their

respective membership groups?” (r[252] = .60, p < .001). Both items had 7-point response scales (1 = not at all, 7 = very).

7 Separate analyses for contact quantity (while controlling for contact quality) showed the same results as the composite index. For contact quality (while controlling for contact quantity) we found the same pattern, but the effect only approached significance (p = .093) when outgroup attitudes were measured with the feeling thermometer, and was not significant when measured with the General Evaluation Scale (see also supplemental material).

(43)

Outgroup Member Typicality. One item measured typicality (Voci & Hewstone, 2003; “When you met Turks, did you perceive the other person as a typical Turk?”). The item had a 7-point response scale ranging from 1 (not at all) to 7 (very).

Outgroup Attitudes. We measured outgroup attitudes with two scales. Respondents indicated their overall feelings toward Turks along a feeling thermometer (0 = extremely negative, 100 = extremely positive; Campbell, 1971). Furthermore, outgroup attitudes were measured using Wright et al.’s (1997) General Evaluation Scale. The question, “How do you feel about Turks in general?“ was followed by six 7-point bipolar scales (e.g., cold–warm).

The items formed a reliable scale (α= .92), with higher scores indicating more favorable attitudes.

Perceived Outgroup Entitativity. In order to assess perceived outgroup entitativity in more detail than in Study 1, we used a well-established scale and participants responded to 10 statements measuring the perceived entitativity of Turks (adapted from Castano et al., 2003). All items (e.g., “Turks have many characteristics in common”) used a 7-point response scale (1 = strongly disagree, 7 = strongly agree; α = .87).

Results and Discussion

We used moderated regression analysis (PROCESS Version 3.0; Hayes, 2018; Model 1, 95% confidence intervals, 5,000 bootstrap samples, using unstandardized scores) to test for the moderating role of group membership salience, outgroup member typicality, and

perceived outgroup entitativity, on the contact–outgroup attitudes relationship.

In the first step we confirmed the moderation effect for the established moderators, group membership salience and outgroup member typicality, on outgroup attitudes in our study (see supplemental material). Next, we considered entitativity as the moderator,

controlling for both group membership salience and outgroup member typicality, first for the feeling thermometer, and then for the General Evaluation Scale.

Referenzen

ÄHNLICHE DOKUMENTE

On the other hand, the Matrix-Tree theorem (The- orem 2.12, or [Zeilbe85, Section 4], or [Verstr12, Theorem 1]) expresses the number of spanning trees of a graph as a determinant 2..

…rms can increase their pro…ts from the preference-based price discrimination in intense competition, where consumer preferences follow an inverse U-shaped distribution..

More attention is also needed for calibration in relation to the treatment of uncertainties in model parameters and in model relations, and of errors in field data

In line with the second explanation for the variety of results yielded for the asymmetry of valenced intergroup contact effects, I suggest it is important to go beyond

The neo-classical approach may be able to explain privatisation through the idea that technological changes have changed the extent of 'natural monopoly' (telecom- munications may be

We extend the original cylinder conjecture on point sets in affine three-dimensional space to the more general framework of divisible linear codes over F q and their

τ c 2 &lt; β: first, by using the contraction mapping theorem in appropriately chosen spaces, we show a local existence result in some appropriate functional spaces, second by

The point of our concern here is that conventionally variance is visualized as the expectation of (squared) deviations of the individual variate values from the mean value of