Projection and Mirror Effects in Cross-group Interactions: A Social Relations Model Study of Similarity Perceptions

Abstract

Unstructured interactions between members of different ethnic groups are increasingly common, yet have received relatively little research attention. The present research examined whether membership of an ethnic majority/minority group and one's ethnic identification predicted perceptions of similarity with cross-group interaction partners. Each participant spoke individually with three members of another ethnic group in round-robin fashion. Analyses using the social relations model revealed that majority/minority status moderated the relationship between ethnic identification and similarity perceptions. Among minority participants, but not those from the majority, stronger ethnic identification was associated with less projection (perceived and assumed dissimilarity). Additional analyses revealed that, across both groups, stronger other-group orientation was correlated with more mirror effects (eliciting similarity perceptions from others).

Keywords:

• Ethnic identification
• Similarity
• Intergroup contact
• Social relations model

Acknowledgments

This research was supported by an institutional grant from the Minority Research Infrastructure Support Program at the National Institute of Mental Health (Thereasa Cronan, PI).

Kate Duangdao is now at University of Nebraska, Lincoln, Nebraska, USA.

Notes

1. In addition to perceiver and target variance, the SRM suggests that perceptions can also be a function of relationship variance. Because the research questions in this report are tied to the perceiver and target variance, analyses on relationship variance are not presented.

2. The scenarios were based on common concerns of students at the time of data collection and each was regularly a topic of conversation in academic and social contexts. Although each represents a somewhat different level of potential for intergroup conflict, no differences were found among the three scenarios for similarity perceptions or metaperceptions (Majority groups: both Fs < 1.25; Minority groups: both Fs < 0.30).

3. The analyses collapse results across the three racial/ethnic categories for the minority groups (African American, Asian American, and Latino American). When the data were disaggregated for the three categories, the pattern of results was the same for each category. Furthermore, the three-way interaction of ethnic identification, majority/minority status, and type of minority group (Asian American vs. Latino American) was not significant for either projection or assumed projection, both ts < |1.36|, ns. African Americans could not be included in this moderator analysis because there was only one group in the sample.