Dyadic interracial interactions: A meta-analysis.

Abstract

This meta-analysis examined over 40 years of research on interracial interactions by exploring 4 types of outcomes: explicit attitudes toward interaction partners, participants' self-reports of their own emotional state, nonverbal or observed behavior, and objective measures of performance. Data were collected from 108 samples (N = 12,463) comparing dyadic interracial and same-race interactions, predominantly featuring Black and White Americans. Effect sizes were small: Participants in same-race dyads tended to express marginally more positive attitudes about their partners (r = .07), reported feeling less negative affect (r = .10), showed more friendly nonverbal behavior (r = .09), and scored higher on performance measures (r = .07) than those in interracial dyads. Effect sizes also showed substantial heterogeneity, and further analyses indicated that intersectional, contextual, and relational factors moderated these outcomes. For example, when members of a dyad were the same sex, differences between interracial and same-race dyads in negative affect were reduced. Structured interactions led to more egalitarian performance outcomes than did free-form interactions, but the effects of interaction structure on nonverbal behavior depended on participant gender. Furthermore, benefits of intergroup contact were apparent: Differences in emotional state across dyadic racial composition disappeared in longer term interactions, and racial minorities, who often have greater experience with intergroup contact, experienced less negative affect in interracial interactions than did majority group members. Finally, there was a significant historical trend toward more egalitarian outcomes across dyadic racial composition for explicit attitudes and for nonverbal behavior; however, participants' emotional responses and performance have remained consistent.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 143 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.