Group Distinctiveness, Political Identification, and the Third-Person Effect: Perceptions of a Political Campaign in the 2004 Greek National Election

Abstract

The present study examined the effects of political identification and group distinctiveness on perceptions of media influence during an election campaign. Participants estimated the effect of political communication on self and on voters of two large, nondistinctive political parties and two small, distinctive political parties. Nondistinctive party members showed an ingroup bias (i.e., greater perceived media influence on the outgroup) irrespective of strength of identification compared to the nondistinctive outgroups, whereas they did not show any bias (high identifiers) or even reverse bias (low identifiers) toward the distinctive outgroups. Distinctive party members showed an ingroup bias (irrespective of strength of identification) against the nondistinctive outgroups and an ingroup bias (high identifiers) or no bias (low identifiers) toward the distinctive outgroups. Ingroup assimilation (i.e., lack of difference in perceived influence between self and ingroup) was evident for distinctive party members, but not for nondistinctive party members. Results highlight the importance of group distinctiveness and identification in third-person perceptions.

ACKNOWLEDGMENTS

I would like to thank Mary Beth Oliver and three anonymous reviewers for their helpful comments on an earlier version of this article.

Notes

1. In two experimental studies, Tewksbury (2002) examined the role of group size in third-person perceptions by varying the size of the comparison target group from 1, 10, to 1,000 members. Perceived influence was greater for the 1,000 target group than the 10 or 1 groups. However, group size defined in this way strips from the target group its social relevance—and, consequently, its social identity central to the present study. As Tewksbury notes “…this manipulation of group size is rather artificial. In future research it might be preferable to manipulate size through reference to naturally occurring social groups that are similar to one another but differ in size” (p. 6, footnote 4).

2. At that time, political debates had not taken place among the party leaders and, hence, a relative measure had not been included.

. Measures of perceived gullibility for each party have been omitted due to a technical error.

4. Similar results were also produced after employing: (a) perceived unfairness of a media campaign toward one's party as a covariate, and (b) perceived group distinctiveness as an independent variable (after a median split) instead of group distinctiveness. Interestingly, perceived group distinctiveness, when entered as a covariate, did not influence the effect of group distinctiveness on perception, suggesting that both exert a separate, although related, effect.