Intimate Intergroup Contact Across the Lifespan. Journal of Social Issues

Intergroup contact can be as casual as members of different groups walking past one another on the street or as intimate as developing cross-group friendships or romantic relationships. To date, however, the majority of inter-group contact research has focused on examining the effects of contact through self-report measures of interactions and friendships. While this research has made a substantial contribution to scientific understanding, less is known about how different forms of contact (casual vs. intimate) influence each other and are associated with outcomes across the lifespan. The present article focuses on intimate contact; a close and meaningful relationship or interaction with either an ingroup or outgroup member. We critically review the nature and consequences (good and bad) of intimate contact for children, youth and adults and for both majority and minority group members, focusing primarily on intimate intergroup contact. We also consider how intimate contact might be best measured in future research. Implications for research design, policy, and practice are considered.


INTIMATE INTERGROUP CONTACT ACROSS THE LIFESPAN
Until recently, research on intergroup contact primarily utilized self-report surveys of contact quality and quantity or interventions in which minority and majority strangers interact briefly with one another prior to reporting their prejudice toward the opposing group (Paolini et al., in press).The contact observed and reported in such studies, however, is not necessarily representative of the constellation of cross-group interactions that exist in the real world, which differ in duration, engagement, and meaningfulness.Cross-group interactions can be as superficial as an exchange in a shop or as intimate as a marriage.Contact research has traditionally concentrated at the shorter, less meaningful end of the spectrum, in direct opposition to Allport's (1954; see also Pettigrew, 1997) proposition that deeper engagement would be more beneficial than casual contact (p.263).More recently, there has been a proliferation of research on intimate forms of contact, with these studies often revealing a stronger relationship between deeper forms of interaction resulting in cross-group friendships and reduced prejudice compared to less intimate forms (e.g., Pettigrew, 1997;Pettigrew & Tropp, 2006).Nonetheless, we know relatively little about how to facilitate intimate intergroup relationships.
Intimate contact can be understood as a close and meaningful relationship or interaction with either an ingroup or outgroup member.It is likely to involve repeated contact and be characterized by reciprocal self-disclosure and trust.The nature of intimate contact may differ for children, adolescents, and adults, varying from close friendships to romantic relationships.In the present article, we discuss the nature and consequences of intimate contact across the lifespan, focusing particularly but not exclusively on intimate intergroup contact.We critically appraise the ways in which intimate interactions within and between groups can be associated (or not) with a range of benefits and how intimate contact might be experienced differently for majority and minority group members.We also review and offer suggestions for the measurement of intimate contact and consider the implications for interventions, and for policy.

THE NATURE AND CONSEQUENCES OF INTIMATE INTERACTIONS
Throughout the lifespan, individuals engage in interactions across a range of social situations and develop relationships of varying levels of intimacy.During childhood, children connect mostly with family members and with peers in schools and neighborhoods.In adolescence and young adulthood, individuals continue to engage with others in classrooms, while also potentially forming relationships on college campuses, in shared housing, at work, during recreational activities, and online.For adults, the workplace, local communities, and the school gates may be where interactions typically occur, and in retirement, within local communities and care residences.Across all life stages, interactions can be superficial and fleeting but with age individuals begin to develop deeper friendships.During adolescence, intimate contact can take the form of romantic relationships, and such relationships may become even more meaningful during adulthood.Although most of the individuals we develop relationships with will be similar to ourselves in some regard, many will also be different from us.That is, our friends, family members, and partners may belong to different religious, ethnic, cultural, or political groups, they may identify with a different gender or sexuality, and they may experience physical or mental illnesses that we do not.
Intimate intergroup interactions have long been considered the ideal form of contact for prejudice reduction (Paolini et al., in press).A construct in its own right, intimacy typically embodies and encapsulates the elements associated with high-quality contact.Intimate interactions are voluntarily entered into, pleasant, equal, and cooperative (Islam & Hewstone, 1993).Such intimate relationships, more so than casual acquaintances, are likely to exemplify the conditions identified by Allport (1954) as being integral to prejudice reduction, specifically cooperation, common goals, equal status, and authority support, as well as other characteristics thought to augment the contact-prejudice relationship, such as repeated interactions over time and situation and reciprocal self-disclosure.Although individuals prone to prejudice are less likely to engage in intimate intergroup contact (Maunder et al., 2019), there is evidence that they nevertheless benefit from close intergroup associations (see Turner et al., 2020, for a review).
Further supporting the importance of deeper levels of intergroup contact, Pettigrew and Tropp (2006) found that studies specifically assessing the effect of cross-group friendship yielded a significantly larger average effect size on various measures of prejudice compared to mere contact.Moreover, there is evidence that when directly comparing intimate and casual contact, intimate contact is more consequential for reducing prejudice (e.g., Fuochi et al., 2020).Less is known however, about how different quantities of superficial and intimate contact compare in their effects on prejudice.For example, it remains to be seen whether a larger number of less intimate relationships would be as effective in reducing prejudice as a smaller number of highly intimate relationships.
Of the array of intimate relations individuals might experience across their lifespan, cross-group friendship has received the most research attention.Evidence shows that time spent with friends and self-disclosure to a lesser extent are crucial elements of cross-group friendship effects on prejudice reduction (Davies et al., 2011).Further, it has been found that perceptions of the strength of association between individuals and their friends (and their friend's group) as well as the salience of the friendship are associated with intergroup attitudes (Page-Gould et al., 2010).Despite these promising findings, however, there is evidence that as children move toward middle and later adolescence, cross-group friendships are less stable (see Turner & Cameron, 2016, for a review) and occur less frequently (Wölfer & Hewstone, 2018).Even in diverse settings, children from different backgrounds often self-segregate (e.g., McKeown et al., 2016).It is therefore important that researchers identify ways to bolster cross-group friendships.
Beyond cross-group friendship, there is evidence that varying forms of intimate intergroup interactions can influence relations amongst romantic partners, roommates, and families (e.g., Harwood et al., 2005;Van Laar et al., 2008).Such intimate relations have been found to improve explicit and implicit attitudes, attitude strength and accessibility, perceived outgroup variability, empathy, trust, perspective-taking, comfort interacting with the outgroup, intended behavior, and the perceived value of intergroup contact, and to reduce blatant and subtle prejudice, perceived outgroup threat, intergroup anxiety, and endorsement of outgroup discrimination in behavior and government policy (e.g., Davies et al., 2011;Grütter & Tropp, 2019;Harwood et al., 2005;Heinze & Horn, 2009;Husnu et al., 2018;Paolini et al., 2004;Pettigrew, 1997;Turner & Feddes, 2011;Turner et al., 2007Turner et al., , 2013;;Vonofakou et al., 2007).Intimate intergroup relationships have also been found to improve perceptions of various social groups including immigrants, religious groups, ethnic groups, the elderly, and lesbian and gay peers (e.g., Harwood et al., 2005;Heinze & Horn, 2009;Paolini et al., 2004;Wagner et al., 2003), andamong children, adolescents, andadults (e.g., Harwood et al., 2005;Heinze & Horn, 2009;Paolini et al., 2004).In a unique comparison of different forms of intimate contact, Van Laar et al. (2008) found that White students' friendships, romantic partnerships, and roommate relationships with African American and Latinx American during college were independently associated with reduced ingroup bias, intergroup anxiety, intergroup unease and symbolic racism, and increased interethnic competence and outgroup dating at the end of college.
The effects of intergroup interactions are not always positive: superficial negative contact has been found to be detrimental to intergroup bias, and more strongly detrimental than superficial positive intergroup contact (see Schäfer et al., in press for a review).Further, there is evidence that contact with majority group members seems to sedate minority group members' willingness to campaign on their own behalf (Saguy et al., 2009;see Hässler et al., in press for a review).Similar effects have been observed for more intimate forms of contact.For example, Hassler et al. (2020) found that cross-group friendship, compared to superficial positive contact, was more strongly associated with reduced support for social change among sexual and ethnic minorities.As for its positive effects, it can be that intimacy could also amplify the unintended consequences of intergroup contact.At the same time, negative experiences within intimate relations appear to be less consequential for prejudice than positive intimate relations and negative superficial relations (e.g., Fuochi et al., 2020;Graf et al., 2018), curtailing concerns about the impact of potential relationship breakdowns on prejudice.For example, Graf et al. (2018) coded participants' descriptions of contact encounters with individuals in neighboring European countries for the valence and intimacy of the relationship, and found negative and ambivalent encounters to be associated with worse outgroup attitudes when the relationship was described as casual or formal compared to negative experiences in intimate relationships.Further, there is some evidence that negative encounters with unknown outgroup members have a smaller effect on individuals who have close relations with other members of the outgroup (Page-Gould, 2012).One reason may be that group salience-awareness of group membership, which facilitates the projection of positive impressions of one individual group member to positive impressions of the entire group-is heightened when individuals have limited previous experience with the outgroup that creates positive expectations for interactions (Paolini et al., 2014).As well, cross-group friendship in which differences and inequality between groups is discussed and condemned may strengthen both minority and majority members' commitment to striving for social change (MacInnis & Hodson, 2017).
There is also evidence that the positive effects of direct contact can ripple outward to influence perceptions of noncontacted outgroups and the attitudes of individuals not involved in the intergroup relationship (see White et al., in press for a review of indirect contact effects).For example, researchers have shown that participants in contact encounters not only project their attitudes from the encountered outgroup member to their outgroup, but also to other outgroups not involved in the encounter (known as the secondary transfer effect, see Boin et al., in press) and that this can occur for both fleeting and intimate interactions (e.g., Davies et al., 2011;Pettigrew, 1997;Pettigrew & Tropp, 2006;Schmid et al., 2012;Tausch et al., 2010).Further, individuals who are merely aware of or observe an intimate interaction between a member of their ingroup and a member of an outgroup may also report improved perceptions of the outgroup represented in the relationship as well as outgroups not represented (see White et al., in press).For example, extended contact with South Asians through an ingroup member's cross-group romantic relationship has been shown to be associated with higher personal approval of ingroup members dating both South Asians and Black people and with more positive attitudes toward South Asians in general (Paterson et al., 2015).Further highlighting the importance of intimacy, Tausch et al. (2010) found the effect of extended contact with a rival religious group to be stronger when the relationship between the ingroup members was more intimate compared to less intimate.That is, a friend known to have a close relationship with an outgroup member improves perceptions of the outgroup more than a neighbor known to have a close intergroup relationship.The same positive effects have been found for intimate intergroup interactions witnessed vicariously.For example, Husnu et al. (2018) had Turkish Cypriot children read stories highlighting cooperation and friendship between Turkish and Greek Cypriot children for three consecutive weeks, after which they reported improved attitudes and behavioral intentions toward Greek Cypriots.Although vicarious contact experiences seem to be effective regardless of the intimacy of the relationship observed (for review, see Vezzali et al., 2014), Wright et al. (1997) found that observing an interaction between confederates pretending to be close friends reduced favoritism of the ingroup over the outgroup more so than when the confederates acted like new acquaintances or were hostile toward one another.
It is important to note that a potential limitation of intimate contact is that, arguably, the focus on interpersonal rather than intergroup aspects of the relationship that one might expect to occur among friends may result in lower levels of category salience, and in turn less generalized prejudice-reduction (Brown & Hewstone, 2005).There is some evidence, however, from Northern Ireland, that Catholic and Protestant teachers with closer cross-community friends actually reported greater confidence in discussing issues of identity and difference (Hughes et al., 2020).It has also been argued that while interpersonal processes such as developing intimacy, affection and trust are important predictors of more positive outgroup attitudes in the early stages of friendships, intergroup processes, such as the belief that the outgroup friend respect's one's group, become important predictors of positive outgroup attitudes as the friendship continues to develop (Davies & Aron, 2016).In the section that follows, we move beyond intergroup outcomes and consider the ways in which intimate intra-and intergroup contact can be associated with health and psychological well-being.

INTIMATE CONTACT AS A SOCIAL CURE OR SOCIAL CURSE
Research examining the benefits and pitfalls of social ties for physical and psychological health have tended to draw on the social identity approach to health (Haslam et al., 2018), also known as the social cure.The central assumption of the social cure approach is that sharing an identity with fellow group members unlocks a range of psychological and social resources, which address their needs and enable members to deal more effectively with the challenges they face, thereby improving their well-being (Haslam et al., 2018;Wakefield et al., 2019).In terms of psychological resources, social identities provide meaning and purpose to individuals' lives as well as a sense of belonging, connectedness, and intimacy with others (Haslam, Reicher, & Levine, 2012;Neville & Reicher, 2011).In terms of social resources, shared identity increases trust, social influence, and helping behavior and leads to cooperative responses to common threats (Jetten et al., 2009).These effects are thought to be cumulative and multiple group memberships have been shown to provide resilience in times of turmoil and transition across the lifespan: multiple group belonging has been found to be related to increased self-esteem in both children attending primary school and retired older adults (Jetten et al., 2015; see also Steffens et al., 2016) and better mental health in adolescents (Miller et al., 2015).However, not all groups benefit.Some group memberships have unhealthful norms; others are stigmatized, leading to marginalization and rejection of group members and a withdrawal of shared resources, a phenomenon known as the social curse (Kellezi & Reicher, 2012;Wakefield et al., 2019).
The relevance to intimate contact is fourfold.First, only group identification and belonging, which are achieved via intimate and repeated contact (Baumeister & Leary, 1995) and that are themselves sources of intimate contact, can trigger cure or curse processes, whereas mere contact cannot (Bratt, 2015;Sani et al., 2012).Second, intragroup support can provide group members with security, confidence and efficacy, thereby enabling them to overcome the anxiety posed by intergroup contact and engage in unfolding intergroup relations (see also Stevenson et al., 2019;Kauff et al., in press), while shared ingroup identity facilitates social influence, such that the contact experiences of fellow group members are likely to transmit more effectively within a more cohesive group.Third, insofar as ingroup and outgroup members can be seen to share a common identity, this may unlock social cure processes of trust, helping and cooperation, which help support unfolding intimacy between ingroup and outgroup members (White & Abu-Rayya, 2012;White et al., 2014White et al., , 2020)).In effect, it is possible for the boundaries of the ingroup to encompass former members of the outgroup (Reicher et al., 2006) thereby extending social support and protection to new members.Fourth, where contact is experienced as negative, this may trigger marginalization and exclusion.Individuals with strong social connections and identification with their stigmatized low-status ingroup may benefit from ingroup support but become more sensitive to and highly affected by the daily episodes of discrimination and prejudice they are likely exposed to (Begeny & Huo, 2017;Marinucci & Riva, 2020;McCoy & Magor, 2003).Likewise, where engaging in intergroup contact may violate ingroup norms, this may also trigger social curse dynamics of exclusion from one's own group, leading to a withdrawal of ingroup support and undermining health and well-being.In effect, the extent to which individuals might be exposed to or susceptible to a social curse might depend on the nature of the interactions they have, not only with ingroup members but also with outgroup members.
Research examining these processes at an intergroup level have mainly investigated cross-group friendship and demonstrated some positive effects.For example, cross-group friendships have been found to be associated with improved psychological health and well-being of adults with physical disabilities (Bagci et al., 2018), protected psychological well-being of adolescents and children from the negative consequences of perceived ethnic discrimination (Bagci et al., 2014;Benner & Wang, 2017) and negative contact (Paolini et al., 2014), improved college students' institutional belongingness and satisfaction (Mendoza-Denton & Page-Gould, 2008), improved psychological well-being in both minority and majority students, via self-disclosure and self-affirmation (Bagci et al., 2017).There is also some evidence, however, that cross-group friendship can be negatively associated with health and well-being.For example, in their work with secondary school students, Brenick et al. (2018) found that for minority group adolescents, cross-group friendship with majority peers strengthened the association between perceived discrimination and depression, especially in classrooms where intergroup friendship was less supported.This finding seems to suggest that the more outgroup friends one has, the more likely individuals are to be ostracized by the group or feel a shared discrimination on behalf of outgroup friends, at least amongst minority group members.This is an assertion that requires further testing to determine when this might occur and also whether this effect is observed across the lifespan.Relatedly, intimate intergroup contact has been found to indirectly predict outcomes that may have knock-on effects for health and well-being.For example, in the context of cross-group romantic relationships, Paterson et al. (2015) found that while intimate intergroup contact was associated with more positive outgroup attitudes, it also exposed individuals to the disapproval of family and friends.While not examined in the study, such disapproval, at its worst, may result in blame and resentment from valued ingroup members, carrying potentially negative health repercussions (see Kellezi et al., 2019).
There is also some evidence that the social curse of intergroup contact on health and wellbeing may be conditional upon the relative status of the groups interacting.In two studies in Latin America, Eller et al. (2016) found that direct contact-considering also friendship among other indicators-had beneficial effects for physical and psychological health only for indigenous minority groups, whereas health was found to worsen for majority group members.Arguably, these opposing effects can be explained by some of the social curse processes.For example, there may be a perceived norm that majority ingroup members disapprove of outgroup contact, which could, in turn, have negative consequences for the health of majority members who do engage in interactions.Further, it is still unclear if intimate intragroup and intergroup contact yielding to group identification are prompted by specific motives or if they are driven to a same extent by individual differences (e.g., personality traits, attachment style, need to belong, self-esteem) predicting engagement in intimate relations.Future research should focus on proximal processes determining whether cure or curse effects are more likely to unfold and on their distant outcomes, integrating the social identity implications of intergroup intimate contact for majority and minority groups.

MAJORITY AND MINORITY PERSPECTIVES IN INTIMATE INTERGROUP INTERACTIONS
Historically, much of the work on intergroup contact has focused on majority group perspectives (Paolini et al., in press).We argue, however, that considerations of the effects of intimate intergroup contact must take into account the perspectives and experiences of both minority (or marginalized) and majority (or dominant) groups.This is because intimate contact is not a onesize-fits-all process as group status positions may pose distinct barriers to establishing intimate contact and different downstream outcomes of such contact.Minority groups may, for example, be more likely to experience benefits and losses in intimate contact.Understanding these processes may help researchers to identify how to minimize risk for minority groups during contact while maximizing benefits for both groups.
Arguably, minority and majority group members face some similar barriers to establishing intimate contact, and this is evidenced across the lifespan.For instance, sustained residential and social segregation may lead both groups to assume that outgroup members endorse separation between groups.Indeed, Shelton and Richeson (2005) found that both White and Black American college students underestimated outgroup members' desire for more outgroup friends and, in turn, avoided opportunities to form these friendships.Both minority and majority group members may also face pressure from ingroup peers and family members to stay within group boundaries when forming intimate relationships, which can have negative implications for the establishment, development, and longevity of these relationships (Edmonds & Killen, 2009;Wang et al., 2006).These normative effects may be particularly pertinent for minority group members, who have been shown to experience heightened own-group conformity pressure (Contrada et al., 2001;Thai et al., 2014).
Once established, such contact also serves some similar functions, for instance in improving intergroup attitudes and reducing intergroup anxiety.One set of correlational studies of elementary, high school students, and undergraduate students in the United Kingdom found that both disclosing to and receiving disclosure from outgroup members predicted more positive intergroup attitudes among both White and South Asian participants via increased empathy and trust (Turner et al., 2007).Moreover, in studies with the elderly and youth in Singapore, relationship building through playing videogames together for two months decreased intergenerational anxiety and improved intergenerational attitudes for both groups (Chua et al., 2013).There is also evidence that cross-group friendships in adolescence are associated with outcomes beyond prejudice, including stronger academic performance, self-efficacy, and satisfaction in school (Baysu et al., 2014).
In addition to commonalities, minority and majority groups experience distinct threats and needs when engaging in intergroup contact.Unaddressed, such threats could jeopardize the formation or maintenance of intimate relationships or interactions.A large body of work on intergroup threat in interactions among cross-race strangers shows that in adulthood, Black Americans tend to have concerns about being the target of prejudice and desire to be respected, whereas White Americans tend to have concerns about appearing prejudiced and desire to be liked (e.g., Bergsieker et al., 2010;Richeson & Shelton, 2007).For White Americans, this concern may begin to emerge as early as middle childhood (e.g., Apfelbaum et al., 2008).In addition to threats to social image (i.e., appearing prejudiced, appearing moral), majority groups' social position may lend them little insight into minority groups' lives and perspectives.Just as a lack of knowledge about minority groups' experiences can lead them to overestimate commonalities and ignore importance differences (e.g., Pew Research Center, 2019; discussed below), majority group members who have little contact with outgroup members may underestimate their commonalities.They may fail to see important ways in which their lives, experiences, and views intersect with those of minority group members (Mallett et al., 2008;Shelton & Richeson, 2005).To effectively establish intimate contact, each individuals' group-based needs must be addressed, at least to some extent, and majority group members must realize the possibility of existing and new commonalities with minority group members.Approaches to achieve this with children have involved promoting a multicultural perspective through storytelling and this has been found to change interaction behavior (observed through seating choice; McKeown et al., 2017).
Even as minority and majority group members experience similar improvements in attitudes and intergroup anxiety, their status positions in society suggest different downstream consequences of intimate contact.For example, intimate contact can take steps to narrow majority groups' gap in knowledge about minority groups and boost allyship.A set of laboratory studies examined a circumstance in which college students were led to feel socially connected to a peer from another racial-ethnic group and had an opportunity to engage with this person's culture.Here, White and Asian American participants were connected to a Mexican American peer and then freely worked with her to design a music video for a Mexican band (Brannon & Walton, 2013).Compared to multiple control conditions (no social connection, a non-Mexican band, choice was not free), this experience reduced participants' implicit prejudice against Latino/as and, as long as six months later, increased their interest in interacting with other Latino/as via increased interest in the peer's culture.Moreover, discussing racial differences in power with cross-race friends is linked to majority group adults' increased commitment to social justice (Ulug & Tropp, in press).For minority groups, intimate intergroup contact can have positive impacts on their adjustment in contexts where they could be negatively stereotyped.In school contexts, for example, facilitating intergroup friendships can help to reduce ethnic achievement gaps (Phalet & Baysu, 2020).There is also evidence that racial minorities who have more friendships with majority group members (i.e., White and Asian Americans in U.S. contexts) experience an increased sense of belonging to their university (Levin et al., 2003;Mendoza-Denton & Page-Gould, 2008;Shook & Fazio, 2008a, b).Although this work has been done in college settings and has not been tested in K-12 schools or workplaces, for example, it gives insight into the unique value of intergroup relationships for minority groups on their sense of belonging in settings where they might be vulnerable to stereotyping and discrimination.Given these differences in outcomes, researchers who aim to intervene to increase intimate intergroup contact must carefully consider what problem they aim to address, and how addressing such problems could influence both groups.In addition to considering the unique benefits, it is particularly important to consider the risks of intimate intergroup contact for minority groups.
Although minority groups can experience important benefits from intimate contact with majority groups, such contact should not come at a cost to ingroup connections and should be able to foster authentic and understanding relationships.As reviewed in the previous section, positive ingroup connections are shown to boost positive life outcomes and belonging for minority groups across the lifespan (e.g., Hughes et al., 2006;Rivas-Drake et al., 2014a,b;Sellers et al., 1998;Tatum, 2017).Thus, intergroup and intragroup connections should be supported for minority group members.Zero-sum thinking about such relationships could harm the benefits of both kinds of connections.Moreover, minority group members may be at risk for feeling relatively inauthentic in intergroup relationships.Work examining cross-race roommates and cross-race friendships over time in college shows that minority groups tend to feel inauthentic, experience fewer positive emotions, and feel misunderstood by cross-race versus same-race roommates and friends (e.g., Holoien et al., 2015;Shelton et al., ;Trail et al., 2009).Intimate contact should therefore integrate mechanisms that help minority group members feel understood and authentic in such relationships.For example, boosting majority group members' interest in minority groups' perspectives and their ability to engage appropriately in conversations about differences, rather than focusing only on similarities might help both parties feel closer to each other and could help minority group members feel like they can express more of their full selves in such relationships and interactions (see Brannon et al., 2017;Sanchez et al., under review).
Taken together, the findings reviewed demonstrate that intimate intergroup contact can have similar yet distinct impacts for minority and majority group members, and both should be considered when theorizing about the benefits and risks to such contact.Differential barriers to establishing contact, downstream outcomes of contact, and risks within such interactions and relationships carry important implications for researchers who aim to intervene to increase or improve intimate intergroup contact.To better ascertain these barriers, it is vital to adopt a range of methodological approaches.We consider this in more detail in the following section.

MEASURING AND MANIPULATING "INTIMATE" CONTACT
A range of methods has been used to examine intimate intergroup contact throughout the lifespan, each with its own advantages and disadvantages (see O'Donnell et al., in press for a review of technological and analytical advances).The dominant approach has been to administer selfreport scales within cross-sectional designs that ask participants to report on interaction quality and quantity (Davies et al., 2011;Paolini et al., in press).Although these measures are relatively easy to implement and have been important in establishing a reliable link between intimate contact and other variables of interest, there are a number of limitations associated with them.For example, there is little consistency across the literature regarding how intimate contact should be operationalized.Notwithstanding the multiple varieties of intimate contact that occur (e.g., friendships, romantic relationships), there is significant heterogeneity within the way each type of intimate contact is measured.Using the research on cross-group friendships to illustrate, some studies have asked participants to report their cross-group friendships in a broader sense (e.g., Barlow et al., 2009;Levin et al., 2003;Pettigrew, 1997;Vonofakou et al., 2007), while others have requested detailed responses about specific cross-group friendships (e.g., Graf et al., 2018;Herek & Capitanio, 1996;Turner & Feddes, 2011).Cross-group friendships have been measured in terms of number of outgroup friends (e.g., Paolini et al., 2007), percentage of friendship circle who are outgroup members (e.g., Tropp, 2003), time spent with outgroup friends (e.g., Turner et al., 2007), and perceived inclusion of outgroup friend in self (e.g., Eller & Abrams, 2003), among others, with some studies collapsing multiple dimensions into single scales.
On one hand, heterogeneity in the measures used to assess intimate contact and the convergence and consistency of results, demonstrates the robustness of the relationship between intimate contact and other variables.On the other hand, however, the use of different measures of intimate contact may reflect divergence in the way researchers conceptualize intimate contact, precluding a more nuanced understanding of the differential effects different forms of intimate contact may have on outcomes (Islam & Hewstone, 1993), and on each other.
Examining the relationship between different dimensions of contact may be particularly important intimate contact literature, as doing so may shed light on mechanisms through which intimate contact becomes more intimate over time.For example, having contact with more outgroup friends may promote more time spent with these friends, leading to greater perceived closeness, which may bolster the time spent with those friends even further, triggering an intimate contact cycle.As well, examining the interaction between different dimensions of contact may reveal whether a larger number of less intimate relationships produce equivalent outcomes to fewer, more intimate relationships.With this in mind, future research would benefit from a broad consensus on which measures or operationalizations of intimate contact to use under which circumstances and how these processes work across the lifespan.Greater consistency and standardization in the measures of intimate contact, themselves, will prove fruitful for ensuring field-wide cohesiveness in the way intimate contact is studied, enabling researchers to fully examine the effects of different aspects of contact (from casual to intimate) and their effects on a wide range of outcomes.
A deep understanding of intimate contact, however, also requires researchers to go beyond the use of self-report measures, which are arguably susceptible to social desirability, demand characteristics, and other biases or errors that may impede accurate responding (see O'Donnell et al., in press).While previous research has demonstrated that self-reports of intergroup contact are generally valid and accurate (Hewstone et al., 2011), it is important to identify other ways to measure intimate contact and its consequences that do not rely so heavily on subjective self-report measures.Adapting procedures from past studies, the quality of contact between people in real intimate relationships can be assessed using established protocols to examine behavior (e.g., dialogue, nonverbals) in the lab (West & Turner, 2014;West et al., 2015) or during everyday interactions, such as where children and young people sit in the lunchroom or classroom (McKeown et al., 2016(McKeown et al., , 2017)), where young people sit in the classroom, which friendships people choose to feature on their social media (Thai et al., 2016), and how people use public spaces such as beaches (Dixon & Durrheim, 2003).These observations may help us to understand the ways in which intimate contact manifests in behaviors, which can, in turn, be used to predict a variety of outcome variables.
An alternative means of exploring cross-group friendships, particularly in diverse classroom settings but also in the workplace, is through social networks analysis (e.g., Wölfer et al., 2017), whereby individuals in a given context can indicate their relationship (e.g., acquaintanceship through to close friendships) with every other person in that context.It is then possible to examine characteristics of participants as a function of their intra and intergroup friendships as well as the social network in which they are embedded more generally.Unlike traditional self-report measures, because social network analysis allows researchers to examine whether friendships are reciprocated, the measures are arguably more accurate.It also enables researchers to explore a range of cross-group friendships at one time (e.g., on the basis of different ethnicities, religions, nationalities, etc.).Irrespective of how it is measured, however, it is difficult for any measure of intimate contact to address the question of causality.This is particularly important when considering intervention.
To allow for conclusions to be drawn regarding causal relationships between intimate contact and outcomes, researchers have attempted to facilitate the development of cross-group friendships using a number of experimental manipulations in the lab.For example, Page-Gould et al. (2008) adapted the Fast Friends procedure (Aron et al., 1997) to manipulate intimate contact.Here, participants were required to meet their assigned "friend" for one hour, once a week, across three consecutive weeks.These "friendship meetings" included activities that were intended to mimic the escalation of reciprocal self-disclosure and intimacy typically observed in actual intimate relationships.Although engaging in such structured activities may heighten the bond between participants who have had no previous interaction with one another, the prompted nature and brief duration of contact limit the extent to which these relationships approximate intimate contact in the real world.There is also the question of whether participants themselves construe inducted intimate contact as equivalent in intimacy to real world intimate contact.Page-Gould et al. (2008) found that, although perceived closeness with assigned friendship partners (measured using the Inclusion of Other in Self scale; Aron et al., 1992) increased as a result of the manipulation, mean closeness ratings did not exceed even the midpoint of the scale at the end of the third meeting.There is also potential for these experimenter-facilitated interactions to produce negative outcomes, such as anxiety and negative affect (see MacInnis & Page-Gould, 2015).
Further illustrating the problems inherent in trying to make participants be friends in the lab, Davies et al. (2011) found studies experimentally manipulating cross-group friendship to have smaller effects on prejudice than self-reported friendship.One approach to accurately assess the enduring effects of these intimate contact induction manipulations may be to track participants after their participation in these experiments to examine whether they have maintained contact with their contact partners.Such instances would represent successful friendship induction and may yield equivalent effects on outcomes to self-reported friendships.Alternatively, contact researchers may consider experimental manipulations of intimate contact that transpire over a longer timeframe to bolster the chances that meaningful intimate relationships between participants are naturally formed.
Field experiments have been somewhat successful on this front.For example, Shook and Fazio (2008a, b) randomly assigned White university freshmen either a White or Black American roommate.They found that those assigned a Black roommate demonstrated lower intergroup anxiety and implicit racial prejudice toward Black people at the end of their first quarter on campus.It should be noted, however, that such unstructured manipulations of intimate contact cannot guarantee that intimate contact between participants will actually form (Towles-Schwen & Fazio, 2006).Also, apart from the specific setting of college dormitories (where random assignment would otherwise occur even without experimental intervention), there are few other contexts in which long-term intimate contact could be successfully manipulated in this way.Future research could explore these processes across the lifespan; perhaps in school settings where new students can be assigned outgroup friends as part of a "buddy system," or in workplace settings, where employees can be assigned to work in teams comprising outgroup colleagues.Logistically, however, such field experiments may be difficult to accomplish.
In order to gain an accurate understanding of the processes and outcomes of different types of intimate contact in a way that is logistically viable, it may be inevitable that researchers use the aforementioned nonexperimental methods that have come to define the contact literature.We suggest that longitudinal and diary methods like experience sampling (Page-Gould, 2012) may be an effective means of studying intimate contact and its consequences (Christ et al., 2010;Davies et al., 2011;Gaias et al., 2018;Levin et al., 2003;Turner & Feddes, 2011;Van Laar et al., 2005).Although such methods preclude true causal inference, they do allow us to examine intimate contact as it naturally occurs, while providing clues on the directionality and temporal nature of its relationships with focal outcome variables.They also enable us to observe how intimate contact unfolds over time.As a result, researchers may be able to investigate increases in intimacy and closeness over time, as well as potential transition across different types of intimate contact, from least (e.g., personally knowing an outgroup member; Fuochi et al., 2020) to most intimate forms (e.g., romantic partners; Paterson et al., 2019).We now turn to discuss how we might intervene to promote more intimate contact for children, young people and adults.

IMPLICATIONS FOR POLICY AND PRACTICE
Knowing more about the nature and consequences of intimate interactions may have important implications for policy and practice, such as informing intervention strategies.The pronounced effect of intimate relations on prejudice as well as health-related variables suggests that in addition to facilitating mere contact between members of diverse groups, interventions should be structured to allow for the development of closer cross-group relationships.Although not all individuals will naturally become close, repeated opportunities for interaction and the encouragement of reciprocal self-disclosure may increase the likelihood of this eventuality compared to one-off encounters and superficial exchanges (White & Abu-Rayya, 2012;White et al., 2014).Moreover, the development and advertisement of friendship between a few group members can have flow-on effects on other individuals and outgroups (see Boin et al., in press).
For children and young people, it is vital to ensure that opportunities for intergroup contact exist.Greater diversity in preschool, school, and university settings has been shown to predict more positive intergroup attitudes and more inclusive friendship groups (Bagci et al., 2014;Davies et al., 2011;Denson & Bowman, 2013;Gaias et al., 2018;Turner et al., 2013), presumably because they afford children and young people greater opportunity to connect with outgroup members, bolstering the likelihood that they form more intimate relationships with these outgroup members.Along with opportunity, it may be particularly important to promote confidence in contact (Kauff et al., in press;Turner & Cameron, 2016).As youth gain such confidence, they may be more likely to subsequently engage in more intimate, rather than superficial, forms of intergroup contact (Bagci et al., 2019).This is particularly valuable because, as noted earlier, cross-group friendships become less frequent and durable with age (e.g., Turner & Cameron, 2016;Wölfer & Hewstone, 2018).
For adults, evidence suggests that at least three strategies may be beneficial for fostering intimate intergroup relationships, and these strategies have the potential for adaption for use amongst children and youth as well.First, normalizing and reframing intergroup anxiety may be effective for majority group members.This is because the feeling of intergroup anxiety among majority group members could serve as a cue that one is prejudiced or could be seen as prejudiced by others, which may lead to contact avoidance.Schultz et al. (2015) found that White participants who were told that many people feel anxious when interacting with outgroup members but that such interactions would help ease future anxiety subsequently chose to engage with a cross-race interaction partner more often and showed fewer anxious behaviors during the interaction than when participants were given no reframing message.Second, emphasizing unexpected similarities can help dominant group members improve expectations for their relationships with crossgroup friends.White college students who watched videos of cross-race friends discussing how they did not expect to find similarities when they first met, but over time realized they have a lot in common, subsequently had more interracial friendships than those in a control condition (Mallett & Wilson, 2010).This procedure was effective because it normalized worries and low expectations about intergroup friendships and provided a counter-narrative model.Supporting the argument that these strategies may be effective among younger age groups, two surveys in multicultural high schools in the United Kingdom showed that lower levels of intergroup anxiety and knowing about the positive interethnic experiences of peers were both predictors of greater confidence at engaging in cross-ethnic friendships among children from a range of ethnic backgrounds.In turn, this greater confidence was associated with higher-quality cross-ethnic friendships (Bagci et al., 2019).A third strategy, reciprocal self-disclosure, may be effective for establishing intimate social connections between group members (Page-Gould et al., 2008).Self-disclosure can help people feel understood and cared for (Reis & Shaver, 1988), and can allow cross-group partners to find underlying similarities and connections.Moreover, in line with Brannon and colleague's selves in contact framework (2017), we argue that successful efforts to promote intimate intergroup contact should integrate a sociocultural perspective, such that they help people acknowledge the value and importance of individuals' histories, perspectives, and social contexts that give rise to group differences.Thus, strategies that promote self-disclosure should offer a way of meaningfully engaging with group-based differences in experiences and perspectives, in addition to finding and appreciating similarities.
For the elderly, intimate group-based interventions can be implemented for both tackling the negative health consequences of social isolation and promoting younger generations' attitudes toward the elderly.Interventions fostering intimate cross-group interactions among older people in home care settings have been shown to promote a sense of shared identification between the residents that in turn promoted health and well-being (e.g., the "Gentlemen's' club"; Gleibs et al., 2011), as well as cognitive performance (Haslam et al., 2010).On an intergroup basis, intergenerational group reminiscence interventions, providing an intimate transmission of the historical knowledge from the elderly to children (Gibson, 2004), produced the twofold result of boosting the quality of life and diminishing the feelings of loneliness of the older people and of improving younger generations' attitudes toward the elderly (Gaggioli et al., 2014).Crucially, some evidence showed that the improvement of the reciprocal intergenerational attitudes extended outside the intervention settings, with both children and older people reporting an increased intergroup contact in everyday life up to 7 months after the intervention (Wenzel & Rensen, 2000).Each of these approaches can be utilized in the field and across the lifespan, with age and context appropriate materials.

CONCLUSION
In our review, we show that intimate intergroup contact can have a range of benefits for majority and minority group members across the lifespan.We also, however, show that intimate interactions are not only positive and that effects can vary depending on the nature and context of the interactions as well as differ for both majority and minority group members.Recognizing this complexity, we call on researchers to come together to develop a deeper understanding of how casual and more intimate forms of contact might be investigated to determine their effects (good and bad).Doing so will require innovative methodological approaches that embrace complexity, move beyond the two-group paradigm and fully accept the developmental pathways of intergroup contact in applied as well as experimental settings.