Intergroup Contact as an Agent of Cognitive Liberalization

Basic contact as group member or representative

In operational terms, intergroup contact represents the actual or symbolic interaction between representatives of different social groups. The central premise underlying contact theory is that, as a representative of one’s own group, contact with an out-group member has the potential (over time) to reduce prejudice toward the out-group as a whole (Allport, 1954; Hewstone, 2009; Pettigrew & Tropp, 2011; Williams, 1947). According to Allport (1954), this potential is better realized when the group members are considered relatively equal in status, working toward common goals, cooperating (rather than competing), and supported by institutional norms and rules. Moreover, Allport argued that contact functions best when it is relatively more intimate. Superficial or artificial contact, in contrast, is too shallow to be psychologically effective. Indeed, such contact can backfire and worsen relations, given the tendency to look for confirmatory stereotypes and the continuation of thinking about out-groups on what Allport (1954) called an “autistic level” (p. 264).

In his influential Annual Review article, Pettigrew (1998) emphasized cognitive factors, suggesting that representations of contact should initially be decategorized (i.e., interacting as individuals), followed by salient categorization (i.e., as members of different groups), followed by recategorization (i.e., highlighting common and overlapping group memberships). The degree to which people are represented as individuals or group members (out-group or shared group) matters critically to the potency of contact. In their reformulation, Brown and Hewstone (2005) emphasized the importance of categorization as group members. Without thinking of the out-group partner as a representative of the out-group, and a prototypical one at that, the experience fails to be intergroup in nature. As a consequence, the potential to generalize across situations, to the out-group, or to other out-groups, is lost.

Despite the fact that intergroup interactions are highly emotional events (MacInnis & Page-Gould, 2015), it is also important to highlight the cognitive nature of the contact experience, recognizing that the categorization of oneself (as an in-group member) and the other (as an out-group member) plays a central role. Without categorization, the interaction is merely between two people and is essentially irrelevant to intergroup relations, regardless of any emotional impact. After social categorization, the interaction between group members then plays out in the mind as a symbolic interaction between the social groups. In the absence of categorization, contact with a member cannot generalize to the group as a whole (see Brown & Hewstone, 2005), presumably because it was not mentally represented and encoded as intergroup contact. Therefore, although the effects of contact operate more through affective (rather than cognitive) processes (Pettigrew & Tropp, 2008), cognitive processes such as categorization are critical in characterizing the contact as intergroup in nature.

There is by now little doubt that basic contact improves intergroup attitudes, and we refer the interested reader to the many reviews and meta-analyses detailing this point (e.g., Beelmann & Heinemann, 2014; Brown & Hewstone, 2005; Davies et al., 2011; Hodson & Hewstone, 2013b; Lemmer & Wagner, 2015; Pettigrew & Tropp, 2006, 2011; Vezzali, Hewstone, Capozza, Giovannini, & Wölfer, 2014). Instead, in the subsequent sections concerning multifaceted outcomes, we focus on the potential for contact to change the way that people think about the world and solve problems more generally.

Cross-group friendship

Pettigrew (1998) is widely credited with the first clear articulation of the potential of cross-group friendships. He argued that positive emotions (e.g., empathy) are fostered through friendship. Thus, friendships across group lines are theoretically strong vehicles for improving attitudes toward the out-group as a whole. Accordingly, cross-group friendship represents the “ultimate” or “ideal” potential for contact. In a test of a model based on more than 3,800 survey responses from across Europe, Pettigrew confirmed the proposed relation: People with more cross-group friends held significantly less prejudicial attitudes toward that friend’s group. Cross-group friends are now deemed critical to contact-based prejudice reduction, either as an extra condition (facilitator) of contact (Pettigrew, 1998), or even as the only essential condition (e.g., Davies, Wright, Aron, & Comeau, 2013).

Cross-group friendships have indeed proven effective. In a compelling study, Levin, van Laar, and Sidanius (2003) tracked approximately 2,000 students through university, monitoring their prejudices and friendship formations. Those with more friends from one particular out-group (e.g., Blacks) also reported more cross-group friends in other out-group categories (e.g., Asians, Latinos). This suggests to us that cross-group friendships exert a liberalizing effect, opening one up to the potential for friendships and intimate relations more broadly. As one might expect, those initially more biased in favor of their own in-groups, or those higher in intergroup anxiety, ended up with fewer out-group friends. However, those with more out-group friends halfway through their studies ended up exhibiting less in-group bias at graduation. This held even after controlling for friendships made before going to college, conservatism, gender, and religion.

Theorists have speculated about the special advantages of cross-group friendships over basic contact. For instance, Aron et al. (2004) proposed that an inclusion of other in the self concept plays an important role in explaining why contact (and especially friendship) reduces out-group prejudice (see also Davies et al., 2013). Cross-group friendships also foster self-disclosure and builds intimacy, which not only deepens bonds (Davies et al., 2013) but also affords insights into the inner lives of out-group members (decentralizing one’s in-group focus). Consistent with our central premise, contact and friendship make one more open to, and react positively toward, diversity and novel experiences.

Overall, cross-group friendships have proven consistent and reliable predictors of lower prejudice. Davies et al. (2011) reported a meta-analytic effect size comparable with that of basic contact, and the effect was again stronger for affective (vs. cognitive) measures (see Tropp & Pettigrew, 2005a). Moreover, the effects of cross-group friendship are stronger when more time is spent with the friend, and when self-disclosure is involved (Davies et al., 2011). As argued by Davis and colleagues, time spent and disclosure are evidence of actual engagement with the other. To us, this suggests that the deeper and more psychologically involving the contact experience, the greater the payoff and potential for personal change and growth. And, as discussed in a later section, cross-group friendships indeed promote creativity and innovation in the workplace (Lu et al., 2017), consistent with our liberalization premise.

Indirect contact

Although intimate forms of contact (e.g., friendship) are highly effective at reducing prejudice, they work only when social groups are afforded the opportunity to engage in contact. Unfortunately, many societies remain highly segregated, and it is difficult to bring groups together when they live, work, and attend school separately. Even in mixed social environments, interracial communication is often fleeting or superficial (e.g., Dixon & Durrheim, 2003; Taylor & Moghaddam, 1994). To enhance the applicability of intergroup contact, recent developments have focused on how contact may be implemented indirectly when opportunities for direct face-to-face contact are scarce.

Wright, Aron, McLaughlin-Volpe, and Ropp’s (1997) extended-contact hypothesis argues that the knowledge of intergroup friendships between others may indirectly reduce bias toward the out-group. Initial experimental studies using artificially created groups confirmed that individuals who learned about a positive experience between an in-group member and an out-group member subsequently developed more positive out-group attitudes (Wright, Aron, McLaughlin-Volpe, & Ropp, 1997, Studies 3 and 4). Subsequent cross-sectional research demonstrates that individuals who know in-group members with out-group friends tend to hold more positive out-group attitudes. This conclusion is supported across a range of intergroup contexts, including between Catholics and Protestants in Northern Ireland (Paolini, Hewstone, Cairns, & Voci, 2004), between Whites and South Asians in England (Turner, Hewstone, & Voci, 2007; Turner, Hewstone, Voci, & Vonofakou, 2008), and between Whites and African Americans in the United States (e.g., Schofield, Hausmann, Ye, & Woods, 2010). These associations hold when controlling for direct contact experience (see Zhou, Page-Gould, Aron, Moyer, & Hewstone, 2018), meaning that cognitive knowledge plays an independent and critical role.

Research has identified several mediators of extended-contact effects (for review, see Vezzali et al., 2014). One particularly important mechanism underlying extended-contact effects is perceived norms. Knowing that the in-group member is engaged in a friendship with an out-group member spreads the perception that there are positive in-group norms regarding the out-group (a liberalizing effect). Likewise, learning of positive cross-group interactions may also provide information about out-group norms, conveying out-group interest in positive intergroup relations. Positive in-group and out-group contact norms, in turn, exert positive influence on the perceiver’s own out-group attitudes and desire for future contact (e.g., Pettigrew, Tropp, Wagner, & Christ, 2011; Turner et al., 2008; Vezzali, Stathi, Giovannini, Capozza, & Visintin, 2015).

Whereas extended contact involves the knowledge that an in-group member has cross-group friendships, vicarious contact involves the observation of an in-group member interacting with an out-group member (Dovidio, Eller, & Hewstone, 2011; Vezzali et al., 2014). Despite an extensive literature on the effect of media depictions of out-group members on audience’s attitudes (see Mastro, 2009; Mutz & Goldman, 2010), the implications for intergroup-contact theory are underrecognized. Yet Schiappa, Gregg, and Hewes (2005) proposed the parasocial-contact hypothesis: If people acquire information about out-group members by viewing televised characters, this parasocial contact positively affects attitudes toward these social groups. In support, those reporting greater consumption of TV with a prominent gay character (Schiappa, Gregg, & Hewes, 2006) or trans character (Hoffarth & Hodson, 2018) also report less prejudice toward that character’s group, findings that have been corroborated in experiments (Schiappa et al., 2005).

In related research, Mazziotta, Mummendey, and Wright (2011) showed that viewing successful video-based intergroup contact improved attitudes toward the out-group, increased willingness to engage in direct intergroup contact through the acquisition of behavioral knowledge, and enhanced efficacy beliefs in one’s ability to navigate an imagined future contact situation. Researchers have even begun to explore whether watching intergroup-relevant TV coverage in the presence of an out-group coviewer can reduce out-group stereotypes and prejudice, and preliminary results look promising (see Tal-Or & Tsfati, 2016). Finally, a year-long extensive field test Paluck (2009; see also Paluck & Green, 2009) examined the effect of a radio soap opera on Rwandan citizens. The program featured a fictitious story of two Rwandan communities struggling to overcome prejudice and violence, including stories of positive intergroup relations and cooperation. The radio program affected social beliefs, empathy, and behaviors with respect to intermarriage, dissent and cooperation, and trauma healing.

Simulated contact

Whereas extended and vicarious contact involve the knowledge or observation of other in-group members interacting with out-group members, simulated contact involves the self more directly. Imagined contact involves the mental simulation of one’s own social interaction with a member or members of an out-group (Crisp & Turner, 2009, 2013). The broader literature on mental simulation demonstrates that imagery can elicit emotional, motivational and neurological responses similar to those of real experiences (for review, see Crisp, Birtel, & Meleady, 2011). Thus, when individuals imagine themselves engaging in a social interaction with a member of another group, they should engage in conscious processes that parallel those involved in actual contact. This should result in more positive perceptions of out-groups, similar to the effects of face-to-face contact, albeit in a situation that is safer, more secure, and less stressful.

This proposition has now received considerable empirical support (for a meta-analysis of more than 70 studies, see Miles & Crisp, 2014). Early enquiries demonstrated that imagined contact improves intergroup attitudes (Turner, Crisp, & Lambert, 2007), increases perceptions of out-group variability (Stathi & Crisp, 2008), and enhances the projection of positive traits to the out-group (Stathi & Crisp, 2008). Subsequent studies demonstrated that imagined contact can also reduce out-group dehumanization (Vezzali, Capozza, Stathi, & Giovannini, 2012), reduce negative stereotypes (Brambilla, Ravenna, & Hewstone, 2012; Cameron, Rutland, Turner, Holman-Nicolas, & Powell, 2011; Stathi, Tsantila, & Crisp, 2012), and reduce stereotype threat (Abrams et al., 2008; Crisp & Abrams, 2008). There is also evidence that, beyond self-reported outcomes, imagined contact improves verbal and nonverbal behavior toward out-group members (Birtel & Crisp, 2012a; Meleady & Seger, 2017; Turner & West, 2012).

It is noteworthy that the primary benefit of imagined contact may be that it offers a way of psychologically preparing (or orienting) people for future contact (Crisp & Turner, 2009), in much the same way that a liberal education prepares students for future challenges generally. Imagined contact may make people more likely to seek out and seize opportunities for contact. It may also improve the quality of direct contact by preparing them to engage in these interactions with a positive and open mind. Husnu and colleagues provide evidence in the context of prolonged conflicts in Cyprus (Crisp, Husnu, Meleady, Stathi, & Turner, 2010; Husnu & Crisp, 2010). Turkish Cypriot and Greek Cypriot participants who imagined out-group contact subsequently expressed greater intentions to engage positively with out-group members in the future. Subsequent research replicated this effect of imagined contact on intentions to engage in future intergroup contact (e.g., Asbrock, Gutenbrunner, & Wagner, 2013; Birtel & Crisp, 2012b; Stathi, Crisp, & Hogg, 2011; Vezzali et al., 2012) and provides evidence that these intentions translate into actual contact-seeking behavior (Vezzali, Crisp, Stathi, & Giovannini, 2015). Consistent with our overall contact-liberalization premise, imagined contact increases contact self-efficacy in interacting with out-groups. Across three studies, Stathi et al. (2011) demonstrated that after participants imagined positive contact with a single out-group member, they felt more confident in their ability to interact effectively with the out-group in general.

Theoretical models suggest that there are two main routes—affective and cognitive—through which imagined contact exerts a positive effect on contact-related attitudes and intentions (Crisp et al., 2010). Imagined contact, like direct contact, reduces intergroup anxiety, which in turn predicts more positive intergroup attitudes (e.g., Turner et al., 2007; West, Holmes, & Hewstone, 2011). The cognitive route is via the formation of a contact script—when imagining a scenario, a behavioral script will be formed and stored in memory. This script subsequently provides a cognitively available source of diagnostic information that is used to make judgments about one’s expectations and intentions (Crisp & Husnu, 2011; Husnu & Crisp, 2010, 2011).

Much of the theorizing and work on imagined contact transfers well to the use of immersive media to create opportunities for virtual out-group contact. Whereas playing a violent game against out-group characters (e.g., Arabs) increases prejudice toward that group (Saleem & Anderson, 2013), playing a video game cooperatively with an out-group member, one in which players form a team and engage in contact to work on a common goal, can improve out-group attitudes (Adachi, Hodson, Willoughby, & Zanette, 2015; Adachi, Hodson, & Hoffarth, 2015; Velez, Mahood, Ewoldsen, & Moyer-Gusé, 2014). Adachi, Hodson, Willoughby, Blank, and Ha (2016) demonstrated that these immersive, media-based, intergroup-cooperative-contact effects could be explained through cognitive recategorization (changing the intergroup mental representation from “us and them” to a more inclusive and flexible “us”), consistent with a liberalization effect of contact. Other research demonstrates that becoming immersed in fiction, as when reading a book, serves an important mental-simulation role. For instance, narratives about gay or Black out-group members, particularly if such membership is disclosed later in the narrative, can induce greater “experience taking” and also reduce stereotypes and prejudice (Kaufman & Libby, 2012). ² In many ways, this “letting go of the self” mirrors the deprovincialization outcomes of contact (see forthcoming section) in which less inward focus broadens the mind.

Target focused

“Target focused” refers to effects that are directly relevant to the contact group in question. That is, does contact with Group X lead to measurable outcomes (e.g., lower prejudice, greater trust) with regard to Group X? This research, summarized in our introduction, constitutes the vast majority of contact research, and we direct the interested reader elsewhere for detailed reviews and integration (Hodson & Hewstone, 2013b; Pettigrew & Tropp, 2006, 2011). Aside from these oft-discussed effects on attitudes or evaluations, however, there is evidence that increased frequency of contact predicts greater perceived uniqueness (or heterogeneity) among out-group members, which is consistent with the idea of contact as liberalizing agent (Islam & Hewstone, 1993; Voci & Hewstone, 2003, Study 1). Contact, therefore, does not simply influence liking of the out-group; it also influences the manner in which the other group is cognitively represented (as homogeneous or more heterogeneous). This reflects a cognitive liberalization effect of contact.

Less clear are the effects of contact on implicit attitudes, or the more subtle, often unconscious evaluations and/or associations linked to out-groups. There is, to the best of our knowledge, no meta-analytic integration for this question. Yet there is evidence that contact shapes implicit attitudes. For instance, among university students, more positive quality of contact with the elderly has been associated with more positive explicit attitudes toward them, whereas greater quantity of contact with the elderly predicted more positive implicit attitudes (Tam, Hewstone, Harwood, Voci, & Kenworthy, 2006). However, among White university students, contact quantity and quality have been negatively associated with explicit (but not implicit) anti-Black attitudes (Aberson & Haag, 2007). Particularly illustrative are the meta-analytic results from studies on imagined contact (Miles & Crisp, 2014) and cross-group friendships (Davies et al., 2011), both of which report significant benefits of contact on implicit attitudes. Although there is a clear need for meta-analytic integration of basic contact effects on implicit attitudes, the available evidence suggests something rather compelling and central to our main tenet—intergroup contact shapes the way the human brain forms associations between out-groups and evaluations. That this occurs at a less conscious, more implicit level suggests that contact effects reach deep to change the way people mentally represent the social world.

Secondary transfer

Thus far, we have primarily emphasized how contact with a member of Group X generalizes to that out-group as a whole. Consistent with our liberalizing premise, however, is the notion that contact with a member or members of Group X can translate into more favorable attitudes toward other groups (i.e., Groups Y and Z) not involved in the contact experience. Pettigrew (1997) initially discussed these as generalized intergroup-contact effects, now more widely referred to as secondary-transfer effects (Pettigrew, 2009). This process concerns the “spread” of positive group evaluations and a reduction in xenophobia. Using nationally representative data from four European countries, Pettigrew (1997) found that more immigrant contact (and especially friendship) predicted not only improved attitudes toward immigrants but also support for more lenient immigration policies and more favorable attitudes toward other groups that were generally not present in the host society (and hence were not contact-relevant groups). In two nationally representative German data sets, Pettigrew (2009) subsequently observed that contact with foreigners predicts positive attitudes toward foreigners, which in turn predicts positive attitudes toward gay and homeless people. These effects held when controlling for initial attitudes, collected a year previously, toward these noncontact groups. Supporting the notion of secondary transfer, therefore, contact with Group X is associated with positivity toward Group X, that itself predicts more positive attitudes toward Group Y.

Other studies have corroborated such findings cross-sectionally (e.g., Schmid, Hewstone, Küpper, Zick, & Wagner, 2012) and longitudinally (e.g., Eller & Abrams, 2004), including a test confirming the attitude-generalization effect (Tausch et al., 2010). Research has also confirmed the occurrence of secondary transfer even when prejudice-prone ideologies (e.g., authoritarianism) are controlled for statistically (e.g., Hodson, Costello, & MacInnis, 2013; Hoffarth & Hodson, 2018). Such findings bolster the idea that positive benefits of contact activate change in the mind and not merely among those initially more open to contact. Rather, even people who are otherwise resistant to engaging in contact can show benefits in terms of secondary transfer (for an exception, see Schmid et al., 2012).

Further evidence of secondary transfer effects comes from experiments in which contact is manipulated and participants are randomly exposed to contact conditions. For instance, in a large field experiment involving almost 2,000 university students (White, Black, Asian, Latino) at the University of California, Los Angeles, students were tracked annually across five waves (Van Laar, Levin, Sinclair, & Sidanius, 2005). The researchers were able to compare students who selected their roommates and those who were randomly assigned. Roommate racial heterogeneity was associated with more favorable attitudes toward the majority of out-groups. Furthermore, having a roommate from one out-group (e.g., Black) was associated with more positive attitudes toward others (e.g., Latino, Asian). ³ These patterns were found among students who were randomly assigned roommates and those who self-selected; the findings also held after controlling for a range of background variables. Shook, Hopkins, and Koech (2016) also examined university students who were randomly assigned same-race versus different-race roommates over their first term. The researchers found that having a roommate of a different race (e.g., a Black roommate for White participants) predicted more positive attitudes toward a noncontact group (e.g., Muslims). Here, rather than examining attitude generalization (e.g., whether more positive attitudes toward Blacks predicted more positive attitudes toward Muslims), the authors tested whether lower levels of social-dominance orientation (i.e., the endorsement of intergroup hierarchies and inequality) mediated or explained the effect. The results confirmed the proposed mediation: Cross-racial roommate assignment predicted lower social-dominance orientation later, which spread into more favorable attitudes toward noncontact groups. Contact, therefore, fostered a less hierarchical and more egalitarian ideology or worldview.

Secondary transfer effects reveal that contact works at a deep level to change how people view the social world. As noted above, this is evident in more positive attitudes toward other groups and toward policies (not just evaluations) regarding the contact group. And additional evidence lends further support to the liberalization hypothesis. Vezzali and Giovannini (2012) examined contact with immigrants among Italian teens in a school context. Greater contact was associated with lower social distance from immigrants, which generalized to lower social distance toward disabled people and toward gay people. Compellingly, these relations were mediated or explained by a transfer of perspective taking (cognitive) and lower intergroup anxiety (affective). Here, instead of attitudes generalizing across groups, what generalized to a noncontact group was the tendency to put oneself mentally into the shoes of an out-group member and the tendency to feel less awkward and anxious around an out-group member. This again speaks to the liberalizing power of contact and sheds light on the spread of inclusivity across groups. We refer the interested reader to more in-depth reviews of individual-to-group generalization (see McIntyre, Paolini, & Hewstone, 2016) and group-to-group (i.e., secondary) generalization or transfer (see Lolliot et al., 2013).

Deprovincialization

Arguably one of the more promising, yet least studied, aspects of contact involves what Pettigrew (1997, 1998) refers to as deprovincialization. He argues that contact can help people to learn that “ingroup norms and customs turn out not to be the only ways to manage the social world,” which “can reshape your view of your ingroup and lead to a less provincial view of outgroups in general [emphasis added]” (Pettigrew, 1998, p. 72). This process humanizes the out-group and puts psychological distance between oneself and the in-group (Pettigrew, 1997). Put simply, contact shakes up one’s perspective and encourages novel ways of thinking about how the world works. Indeed, Brewer (2008) conceptualizes deprovincialization as the propensity to perceive and appreciate the complexity of social life and intergroup relations. This operationalization is consistent with the notion of cognitive liberalization.

There is empirical support for the relevance of contact to the process of deprovincialization. In the most widely discussed study, Verkuyten, Thijs, and Bekhuis (2010) examined contact among Dutch participants in three samples—preadolescents (Study 1) and teens (Studies 2 and 3)—regarding their contact with Turks, Moroccans, or Muslims. The results of Study 1 indicated that more opportunity for contact predicted less in-group positivity and especially lower in-group identification, through an endorsement of multiculturalism (i.e., deprovincialization). Study 2 found similar results but with regard to self-reported contact (not opportunity for contact) with the out-group. Out-group contact, therefore, boosts endorsement of equality and multiculturalism and lowers the sense of in-group centrality and importance. This represents what Pettigrew (1998) referred to as a reappraisal of the in-group. Others have also observed that increased contact is associated with lower in-group identification (e.g., Cakal, Hewstone, Schwär, & Heath, 2011; Hodson, Harry, & Mitchell, 2009), although in cross-sectional data sets, the direction of causality is open to question. Nonetheless, there is an association between greater contact and more distance from the in-group.

Others have treated contact as a covariate or nuisance variable when examining the effects of deprovincialization (e.g., Mepham & Martinovic, 2018; Verkuyten, Martinovic, Smeekest, & Krost, 2016). This suggests an implicit recognition among researchers that contact is a deprovincializing (and hence liberalizing) agent that would need to be statistically controlled when the researcher is interested in other effects. And indeed, there is some evidence, at least for minorities, that those with more contact hold more deprovincialized views (Verkuyten et al., 2016, Footnote 7). Although the notion of deprovincialization following contact is intriguing, some have characterized the research evidence as “mixed” (Lolliot et al., 2013). Indeed, in their investigation of secondary transfer effects (described above), Tausch et al. (2010) found that deprovincialization mediated (or explained) contact effects in some of their studies but not in others. And a brief 2-week longitudinal examination of Americans studying in Mexico found no impact on in-group reappraisals (Eller & Abrams, 2003).

We believe that the potential power of deprovincialization is remarkably understudied and underappreciated, particularly when considering related constructs, such as openness to experience. For instance, among heterosexuals, contact with gay men is associated with greater openness (Barron, Struckman-Johnson, Quevillon, & Banka, 2008). Moreover, in a reanalysis of Jackson and Poulsen (2005), greater openness predicted lower anti-Black and anti-Asian prejudice through more positive quality contact with the group in question, even after controlling for the amount of contact (Hodson, Turner, & Choma, 2017). Although researchers have concentrated theoretically on the notion that openness predicts greater contact, there is reason to believe that the path would be reciprocal (i.e., that contact would also facilitate greater openness), consistent with the notion of deprovincialization. In a sample of almost 800 Dutch adults (Mepham & Martinovic, 2018), self-reported cognitive flexibility (i.e., being open and creative when approaching general problems) and deprovincialization were substantially correlated (r = .51). Such findings highlight the value of intergroup contact in broadening and liberalizing the mind, not only regarding social issues but also regarding problems in general. In the words of Mepham and Martinovic (2018), “deprovincialization implies relativizing one’s culture” (p. 54), seeing the greater context, in which one’s group is recognized as only one of many. Implicit with such thinking is the notion that other groups may have equally valid (if not more valid) ways of thinking about the world. Coming into contact with these other groups is arguably a central way to foster such thinking and thereby broaden the mind.

Indeed, some recent experimental evidence is encouraging in this regard. In their initial correlational study among American undergraduates, Sparkman, Eidelman, and Blanchar (2016) found that those reporting more multicultural experiences (e.g., travel, contact with foreigners) expressed less ethnic prejudice, and the effect was partially explained by greater openness to experience. In an experimental follow-up, participants were randomly exposed to a multicultural experience in which they learned about a group’s geopolitics, eating practices, and social norms. In the control condition, they learned about the out-group but only in direct comparison with the in-group (i.e., perceptual differences were heightened). The multicultural experience led to lower levels of prejudice toward the out-group, an effect partially explained by enhanced openness to experience. Finally, in an impressive and exceptionally rare experimental study, Dhont, Roets, and Van Hiel (2011) were able to track the attitudes of Belgian high school students; some students had been randomly assigned to partake in a 1-week trip to Morocco, and other students were not sent on the trip. The authors found that contact with Moroccans most improved attitudes toward Moroccans among those relatively higher (vs. lower) in need for cognitive closure. That is, contact was more effective at opening up (and “liberalizing”) students among those who were initially the most closed-minded and predisposed to preferring simple answers and cognitive structure. It is not clear whether such contact itself lowers one’s need for closure, but it is clear that contact makes those with epistemic needs for simple structure more receptive to a foreign out-group.

Ideology and worldview beliefs

Of course, contact with other groups has relevance for people’s ideologies and worldviews. Ideologies link moral and political attitudes, organize people’s values and belief systems, and guide political behavior (Jost, 2006). In this way, ideologies serve psychological functions, offering “a sense of certainty, predictability, and control; a sense of safety, security, and reassurance; and a sense of identity, belongingness, and shared reality” (Jost, 2017, p. 168) reflecting epistemic, existential, and relational motivations, respectively. It is easy to appreciate, therefore, how contact with other groups can draw into question one’s own set of beliefs and provide impetus for mental change and the integration of new ideas.

Of most concern to contact researchers have been ideologies represented by right-wing authoritarianism (RWA; Altemeyer, 1996) and social dominance orientation (SDO; Sidanius & Pratto, 1999). The former characterizes individuals with a propensity to conform to tradition, submit to authority, and aggress against norm violators. The latter represents the belief that, in general, groups should be hierarchically arranged and that inequality is relatively inevitable and acceptable. The most important differences explaining the ideological space differentiating people revolve around resisting social change and supporting or accepting inequality (Jost, 2006); the former is well-captured by RWA, and the latter by SDO. And, in keeping with the dual-process model, evidence supports the contention that RWA (and hence resistance to change) is associated with the cultural worldview that the world is dangerous, whereas SDO (and acceptance of inequality) is associated with the worldview that the world is competitive (Duckitt & Sibley, 2009). ⁴

Researchers in the contact field initially ignored or downplayed the importance of individual differences and ideology (for a review, see Hodson et al., 2013). But there is now considerable evidence that, all else being equal, people with more right-leaning ideologies (e.g., RWA and SDO) are more avoidant of contact (e.g., Dhont & Van Hiel, 2009; Hodson, 2008; Hodson et al., 2009; Pettigrew, 2017; Pettigrew & Tropp, 2011, pp. 208–213). ⁵ But most contact research on individual differences has focused on the degree to which contact reduces prejudice as a function of ideology (i.e., with ideology as a moderator). Given that those higher in RWA or SDO generally dislike out-groups, and typically avoid contact with out-groups, one can appreciate pessimism about successful contact among such people. But contact generally works well, and sometimes better, among those endorsing such ideologies (see Dhont et al., 2011; Dhont & Van Hiel, 2009; Hodson, 2008; Hodson et al., 2009, 2013, 2017), especially RWA (Asbrock, Christ, Duckitt, & Sibley, 2012).

In perhaps the most comprehensive test of the ideological-contact question, in a reasonably large sample with full measures of constructs, greater contact quality was associated with less racism for White Americans with higher (vs. lower) SDO, RWA, need for closure, or ethnic identification (Kteily, Hodson, Dhont, & Ho, 2017). Thus contact “works,” even among those predisposed to prejudice (for reviews, see Hodson, 2011; Hodson et al., 2013, 2017^). Moreover, there is evidence that, among those with higher scores on measures of these right-leaning ideologies, contact predicts lower prejudice through greater empathy (Hodson, 2008), greater inclusion of the other in the self, and lowered perceived threat (Hodson et al., 2009). These are liberalizing processes that operate even among those on the ideological right (i.e., those endorsing ideologies concerned with resisting change and/or accepting inequality). Such findings corroborate the liberalizing effect of contact. Although people with higher scores in measures of right-leaning ideologies might dislike racial or sexual minority out-groups, the more contact they have with these groups, the more tolerant and open-minded their attitudes about the groups become.

But might contact not only improve attitudes among those with right-leaning ideologies but also actually influence ideology itself? Some researchers (e.g., Hodson et al., 2013) were understandably skeptical about this potential. Contact, for instance, might not change ideas about “how the world should operate” (i.e., ideologies). But evidence now suggests that contact can affect worldviews, particularly in cases of meaningful contact that unfolds over time. For instance, in their longitudinal study of American university students over the course of their degree programs, Van Laar et al. (2005) found that having an interracial roommate exerted little or no impact on SDO levels at the end of the students’ first year. But when measured several years later, factors such as roommate heterogeneity (i.e., proportion of out-group roommates) predicted significantly lower SDO in subsequent years, even after controlling for initial group attitudes. ⁶ Dhont et al. (2014, Study 2) similarly used a longitudinal design in a sample of Belgian adults (n = 363) over the span of 3 months. In addition to revealing considerable stability in immigrant contact and SDO across time, contact at Time 1 nonetheless predicted lower levels of SDO at Time 2 (whereas SDO at Time 1 did not predict contact at Time 2). This study demonstrates that links between contact and lower SDO, an ideology supporting group hierarchies and inequality, operate in the direction at the heart of our central premise: Contact serves as a liberalizing agent over time. ⁷

Experimental evidence, and studies in which contact experiences are not self-selected, further support these ideas. For instance, Dhont et al. (2014, Study 1) followed a group of high school students in Belgium as they travelled to Morocco on a 1-week trip, where they interacted with Moroccan students in learning and athletic activities. The teachers had chosen who would participate in the trip, removing concerns about self-selection. In a pretest-posttest design, not only was prejudice reduced after the contact experience, but so were levels of SDO. After intimate cross-group contact, therefore, students were more rejecting of ideologies promoting intergroup dominance and hierarchy. Recall also the study by Shook et al. (2016) discussed earlier, in which first-year university students at an American university were randomly assigned a roommate of another race (rather than one of their own race). Not only were secondary out-group attitudes more positive after the first term (see the Secondary Transfer section), but also those with cross-race roommates showed a significant decrease in SDO. ⁸ In keeping with our proposition that contact can have a liberalizing effect independent of formal education, this finding was not solely attributable to the “liberalizing” effects of being at university per se, given that university students assigned to the same-race roommate condition exhibited no such pattern. Collectively, these studies on the effects of contact on ideological variables such as SDO are challenging former sceptics (e.g., Hodson et al., 2013). Having meaningful contact with an out-group, it appears, encourages the adaptation of a more liberalized and tolerant mind-set. As in the study by Shook et al. (2016), having an inter-racial roommate led to more positive attitudes toward multiple other groups (i.e., not just the group to which the roommate belonged) and also made participants less accepting of dominance and hierarchy as a general outlook on life.

Problem solving

Pushing the contact-liberalization idea even further, research suggests that intergroup contact may not only influence what people think (their attitudes, beliefs, and ideologies) but also how they think (their cognitive skills and tendencies). Because diversity experiences require people to look beyond existing knowledge and perspectives, such experiences should not only countervail the cognitive processes that promote prejudice but also generally reduce reliance on simplistic heuristic thinking. This reasoning is rooted in recent models of diversity-driven social cognition (Crisp & Meleady, 2012; Crisp & Turner, 2011). We know that although people tend to use category-based information as a default to form impressions of others, experiences that challenge existing stereotypes force perceivers to cognitively “shift gears” from heuristic to more individuated and systematic modes of information processing (e.g., Fiske & Neuberg, 1990; see also Crisp, Hewstone, & Rubin, 2001). This change in information processing is characterized by a reduced tendency to assign traits to individuals based on their group membership and a greater focus on using new, emergent attributes in impression formation (Hutter & Crisp, 2006). We also know that just as with physical exercise, repeatedly “working out” the brain improves its processing power (Jaeggi, Buschkuehl, Jonides, & Shah, 2011; Muraven, Baumeister, & Tice, 1999). It follows that contexts characterized by diversity and intergroup contact should train a more active and complex cognitive style through the repeated experience of switching from a heuristic (stereotype-based) mode of thinking to a more flexible, multicultural construal of out-groups (Crisp & Turner, 2011).

If this is the case, intercultural contact experiences, and the mind-set they trigger, should carry over to other decision domains that are not obviously related to stereotyping and impression formation. Evidence stems from research on the outcomes of counterstereotypic exposure. Although this literature is not typically deemed directly relevant for contact theory, we argue that it represents a rich seam of evidence for integration given that intergroup contact—in which people must engage with the different perspectives espoused by different cultures—necessarily involves a challenging of stereotypic preconceptions. Prati, Vasiljevic, Crisp, and Rubini (2015), for instance, asked participants to consider a counterstereotypic gender combination (e.g., female mechanic, male midwife) compared with control and to list all the words that came to mind to describe this target. Participants were subsequently given 10 reasoning problems drawing on measures used by Tversky and Kahneman (1974). They were instructed to solve as many of the problems as they could within a time frame of 10 min. An example was: “A bat and a ball cost £1.10 in total. The bat costs £1 more than the ball. How much does the ball cost?” The correct answer is 5 pence, but usually people tend to say 10 pence. Results demonstrated that participants who formed an impression of a counterstereotypic target subsequently answered more of Tversky and Kahneman’s conundrums correctly, indicating they had switched out of a heuristic way of thinking and utilized a more systematic approach to problem solving.

Research has also considered whether people who have long-term exposure to counterstereotypical experiences (by virtue of being counterstereotypical in the environments in which they find themselves) may also demonstrate a lesser reliance on heuristic thinking. For their experiment, Di Bella and Crisp (2016) recruited a sample of female undergraduates studying science, technology, engineering, and math (STEM) subjects. This group is persistently confronted with gender stereotypes, which can actually impair performance (known as the stereotype-threat effect; Steele & Aronson, 1995). After imagining or recollecting their experience of being a woman in a male-dominated field, women from STEM fields performed better on a test of quantitative judgment skills compared with women from non-STEM fields. In a similar vein, Crisp, Bache, and Maitner (2009) showed that although female psychology majors showed a typical stereotype-threat effect (lower performance on a math test when informed of the comparison with men), female engineers—a group with experience in successfully negotiating counterstereotypical domains—showed no performance detriment when exposed to the same comparison. The implications for contact are clear—increased contact exposes one to more instances of counterstereotypes. Such exposure, therefore, should enhance problem-solving abilities among intergroup interactants.

Flexible thinking

Further evidence consistent with this heuristic-switching hypothesis was provided by Vasiljevic and Crisp (2013). Participants were asked to generate five counterstereotypic (e.g., “female mechanic”) or five stereotypic (e.g., “female nurse”) social-category combinations. Those in the counterstereotype condition subsequently performed better in a test of lateral thinking (i.e., the ability to consider a problem from multiple perspectives and discard traditional modes of thinking). For instance, they were more likely to correctly answer the question, “A police officer saw a truck driver clearly going the wrong way down a one-way street, but did not try to stop him. Why not?” (Answer: Because the truck driver was walking). In another experiment, participants in the counterstereotype condition also performed better on a Stroop task, which similarly requires increased executive control and the inhibition of dominant responses. Research using a similar paradigm has shown that contesting gender stereotypes can counter an established bias in favor of “safe” organizational leaders, who are closer to established group norm, and encourage people to embrace new, innovative leaders (Leicht, Randsley de Moura, & Crisp, 2014).

Evidence of the benefits of interactions with racially diverse others for cognitive development is also provided by research within the educational literature. Developmental theorists emphasize that cognitive growth is fostered by discontinuity and discrepancy (Erikson, 1946, 1959; Piaget, 1971). Ethnic diversity represents a source of multiple and different perspectives and thus may be expected to improve academic achievement by triggering more active and complex thinking. Although it is well-established that cross-racial interactions during college are associated with improvements in intergroup outcomes (e.g., Levin et al., 2003; Shook et al., 2016; Sidanius, Levin, Van Laar, & Sears, 2008; Van Laar et al., 2005), research also demonstrates that cross-racial interactions are associated with a range of cognitive skills and tendencies, including improved attributional complexity (Hurtado, 2005), analytic and problem-solving skills (Hurtado, 2001), and critical thinking (Nelson Laird, 2005; Pascarella, Palmer, Moye, & Pierson, 2001). Accordingly, interactions with racially diverse peers have been associated with higher graduation rates (Chang, 1999), higher levels of educational engagement (e.g., drive to achieve, postgraduate aspirations; Gurin, Dey, & Hurtado, 2002), and more positive academic self-concepts (Chang, 1999). In their study of White students from less diverse contexts who were subsequently studying at the relatively diverse University of Hawaii, Pauker, Carpinella, Meyers, Young, and Sanchez (2017) found that greater time spent in the diverse setting was associated with significantly higher cognitive flexibility on an ability task. The researchers also found that, over time, these students reported less essentialized thinking about racial out-groups, which mediated the effect of time on cognitive flexibility.

Note that meta-analytic evidence (Bowman, 2010) demonstrates that intergroup interactions are stronger predictors of cognitive development than are mere measures of structural diversity (i.e., the percentage of minority students within a student body). Many of these researchers argue that the human element of face-to-face interaction is most likely to trigger cognitive disequilibrium and effortful thinking; thus, although structural diversity may increase the probability that students will encounter diverse others, its value depends on whether it leads to intergroup interaction. Further evidence for the notion that contact can stimulate cognitive growth comes from research on mixed-ethnicity cooperative-learning groups. Cooperative-learning groups were designed as a technique to improve intergroup attitudes in elementary schools by fulfilling Allport’s conditions of optimal intergroup contact (Aronson, Blaney, Stephan, Sikes, & Snapp, 1978). Evidence suggests that such techniques can result in improvements not only in intergroup outcomes (e.g., see Aronson & Patnoe, 2011), but also in student achievement (Carroll, 1986; Crone & Portillo, 2013; Perkins & Saris, 2001; Walker & Crogan, 1998).

Creativity

Creative performance is also strongly tied to flexible thinking (De Dreu, Baas, & Nijstad, 2008; Galinsky, Magee, Gruenfeld, Whitson, & Liljenquist, 2008; Maddux, Adam, & Galinsky, 2010; Maddux & Galinsky, 2009). Flexible thinking entails a capacity to “break set,” go beyond established and mentally accessible ways of thinking, and think differently from what is habitual. This type of information processing has been repeatedly shown to increase performance on various tests of creativity (De Dreu et al., 2008; Nijstad & Kaps, 2008; Sligte, de Dreu, & Nijstad, 2011). The notion that diversity experiences can activate a cognitively flexible mind-set is therefore consistent with research finding that diversity experiences can boost creative performance. In one experiment participants were asked to form impressions of an individual whose group membership was consistent or inconsistent with stereotypic expectations (e.g., a male or female mechanic). On a second, unrelated task, participants were asked to produce original names for a new brand of pasta and were given existing pasta names as examples. Results showed that participants asked to form impressions of a counterstereotype were more flexible: they relied less on schematic knowledge embedded in the task instruction (Gocłowska, Crisp, & Labuschagne, 2013, Experiment 1). In a second experiment, the task of generating various counterstereotypes led participants to develop more creative ideas for theme night at the university nightclub (Gocłowska et al., 2013, Experiment 2). Again, these findings have direct relevance to contact, given that contact increases exposure to counterstereotypes and the need to process and deal with such information.

Research suggests that diversity is most likely to increase creativity when people feel comfortable abandoning stereotypic thinking. Gocłowska and Crisp (2014) demonstrated that thinking of diverse individuals (e.g., gender counterstereotypes) led to higher performance on a divergent creativity task, but only for individuals with lower scores in measures of personal need for structure (PNS; Neuberg & Newson, 1993). Participants were asked to think of multiple uses for a plastic bottle. When participants thought about a counterstereotypic individual, they generated a higher number of ideas and their ideas were rated by a third party as more novel and original. Likewise, Gocłowska, Baas, Crisp, and De Dreu (2014) found that thinking about counterstereotypes increased creative performance among people with lower PNS scores, but decreased performance among people with higher PNS scores. Most recently, Vezzali, Gocłowska, Crisp, and Stathi (2016) found evidence of the diversity-creativity link within children, but this effect was present only when a communal mind-set was active but not when a divisional mind-set (i.e., one that emphasized group distinctions) was active.

Research also demonstrates that living in diverse social environments can heighten creative cognition. For instance, Maddux and Galinsky (2009) examined the performance of people who have lived abroad for a long term on a range of creativity tasks. In two correlational studies, the amount of experience living abroad (but not of merely traveling abroad) predicted enhanced performance on two different creativity tasks. Further experimental studies found that priming cultural diversity temporarily increased creativity among participants who had previously lived abroad, consistent with the idea that out-group contact creates new ways of thinking that can be reinstigated in the face of a contact-relevant cue. The relationship between amount of time lived aboard and creativity was mediated by adaptation to the foreign culture: Participants who had lived longer abroad tended to adapt more to the new culture, incorporating new modes of thinking and behaving, and this adaptation predicted increased creativity.

Longitudinal work has also demonstrated that across time, adapting to and learning about new cultures increases integrative complexity, which in turn is predictive of greater career success (Maddux, Bivolaru, Hafenbrack, Tadmor, & Galinsky, 2014). It appears that this type of immersive intergroup contact makes individuals aware of multiple perspectives, thus increasing the ability to think outside the box. Similar findings have been observed for close cross-group friendships and romantic relationships. Lu et al. (2017) used many methods and samples to test these ideas in laboratory and applied work settings. Dating across cultural lines predicted, over time, more creativity on divergent and convergent tasks, suggesting flexibility along with creativity. As with Maddux and Galinsky (2009), making past cross-group intimacy mentally salient via experimental manipulation boosted creativity among those with such past histories, consistent with the idea that contact changes how people think in deep ways that can be reactivated after contact.

In sum, this impressive body of work suggests that when people engage with someone from a different background or culture, when they make a mental leap beyond their comfort zone, it triggers some powerful psychological processes. These processes involve taking other groups’ perspectives, seeing things from different cultural standpoints, making compromises, and putting aside one’s existing stereotypes and prejudices. Note that these leaps literally give the mind a workout. The consequence is that, just like athletes training their legs so they can run faster, the mind will subsequently work better, faster and more efficiently. And when people use any muscle over and over again, it gets stronger, fitter, and more efficient. The studies outlined above support just this proposition: Creativity, flexibility, and problem-solving skills are enhanced after contact experiences. Contact liberalizes cognition and brings flexibility, fresh perspectives, and the propensity to see things differently.