Discrepancy between theory and practice in extended contact: can (should) we disentangle the two?
Notwithstanding empirical evidence suggesting the strong impact of extended contact, the discrepancy between research on this construct and its translation into naturalistic interventions remains surprisingly large. This also concerns the theory itself and the many forms that extended (or vicarious) contact can take. We argue that this gap between research and how to effectively transform the research findings into real-world interventions severely limits theoretical conclusions about the effectiveness of extended contact and the processes that characterize its effects in naturalistic situations. This is problematic given that the ultimate aim of contact theorists as well as of practitioners dealing with prejudice-reduction is to make positive societal change using contact principles. In our view, this is not therefore simply a matter of translating theory into practice, since in this case theory and practice can hardly be differentiated. We elaborate on this point later.
When considering the implementation of direct contact, there are clear indications stemming from the formulation of the contact hypothesis (Allport, 1954) that face-to-face contact situations should be created following optimal contact conditions (Allport, 1954; Pettigrew, 1998). This already includes practical suggestions for educators and practitioners, such as asking individuals belonging to different groups to interact in a cooperative manner, to achieve common goals, in equal- status groups and in a positive and supportive social climate. But in the case of extended contact, most empirical studies ‘simply’ find that the number of ingroup friends with outgroup friends is associated with lower prejudice. However, this is a piece of descriptive information, rather than a strategy that can be applied; it does not immediately suggest what can be concretely done in order to implement extended contact.
Translating this ‘information’ into actual interventions comes down to how this information is implemented. For example, initial studies examining story reading where ingroup and outgroup characters interact positively (e.g., Cameron & Rutland, 2006; Liebkind & McAlister, 1999) were considered operationalizations of extended contact, whereas at present they are considered as examples of vicarious contact (Brown & Paterson, 2016; Vezzali et al., 2014; see Chapter 3). So in the case of story reading, the implementation followed in practical interventions does not match the extended contact principles: the fact that the positioning of story reading has switched from extended to vicarious contact indicates that its theoretical rationale does not reflect the theoretical definition of extended contact.
Different operationalizations may produce very different effects that are explained by distinct paths. For example, exposing individuals to information provided by a newspaper article on ingroup members’ outgroup friends (another common manipulation of extended contact) is different from obtaining similar information from a social network like Facebook - which implies a different level of personal involvement - or from knowing about contact experiences from direct communication with ingroup members. Although all these possibilities are in line with the basic tenets of extended contact (knowing that ingroup members have positive intergroup relations), it is likely that the different methods by which these are realized activate different processes and have distinct theoretical implications. As such, we argue that it is unrealistic to expect that similar effects between these different forms of extended contact can be acquired and that each new, qualitatively distinct operationalization in the field produces new theory as well. Although we believe this to be true when doing research ‘on the ground’ in general, it is especially the case when considering theory that is difficult to immediately transform into operationalizations. Extended contact is an example of this.
Adding to this argument, a look at existing evidence reveals that applying basic principles of extended contact to naturalistic interventions may not be that easy. In fact, as we noted earlier, most research on extended contact has been correlational (Vezzali et al., 2014), and the few studies that manipulated extended contact provided information that cannot be readily transformed into interventions. Relat- edly, another sore point at the core of extended contact is that common measures of extended contact implicitly assume that individuals are aware of their ingroup members’ cross-group experiences and can promptly quantify them. But is this the case in real life? We will address this point in the following section.
Naturalistic interventions of extended contact
Wright et al. (1997, Study 3) conducted a pioneering pre-post test laboratory study of extended contact. In the first phase of the study, the authors created competition between two minimal groups, each composed of seven people. Next, two participants, one from each of the two competing groups, were asked to take part in a seemingly unrelated study. As part of the study, friendship between the two participants was induced using a task designed to promote personal closeness (Aron et al., 1997). In the last phase, each of the two participants was asked to describe this positive experience with the outgroup member to fellow ingroupers. Results revealed that ingroup members who had not taken part in the intergroup encounter and who were exposed to the positive indirect experience narrated by the ingroup member displayed less bias in measures of outgroup evaluation and resource allocation. Wout, Murphy, and Steele (2010) exposed Black university students to the fictitious social network profile of a White student from their campus. They found that expectations to be liked by the White student, presented as a potential interaction partner, were more positive when the social network profile was more ethnically diverse. This is consistent with our operationalization of extended contact: what is important is not necessarily to know the ingroup member who has intergroup contact but to know that members of the ingroup (even if they are not personally known to individuals) have positive interactions with outgroup members (cf. Vezzali, Hewstone, et al., 2017).