New publication by Mark Boukes and colleagues in International Journal of Communication. Read the abstract below.
Focusing on two distinct dimensions of similarity and difference (political identity, political opinions), this study uses a within-subjects experimental design implemented in an online survey to examine preferences for discussion partners and groups that are similar to (same party and same opinion) or different from (different party and different opinion) or that represent a combination of similarity and difference (same party and different opinion, or different party and same opinion) relative to oneself. Participants comprising a diverse national sample (N = 820) completed eight political discussion selection tasks (four discussion partner tasks, four discussion group tasks) by ranking four political discussion options from most to least preferred. Results indicate that complete similarity is not always preferred (in analyses of all ranked discussion groups) and that partisan similarity is preferred over opinion similarity (in analyses of first-ranked discussion partners). More generally, findings reveal that preferences for political discussion depend on the context of the discussion, including whether the discussion involves a single discussion partner or a discussion group and whether preference focuses on individuals’ most preferred option only or relative rankings across a range of options.