The Third-Person Effect: A Meta-Analysis of the Perceptual Hypothesis

In this study, we report the results of a meta-analysis concerning the third-person effect's perceptual hypothesis. The hypothesis predicts that people judge the media to exert greater persuasive influence on other people than on themselves. Thirty-two published and unpublished studies with 121 separate effect sizes were examined. The overall effect size between estimated media effects on self and on others was r = .50. Among the 8 moderators investigated (source, method, sampling, respondent, country, desirability, medium, and message), 3 (sampling, respondent, and message) yielded significant effect size variations. Third-person perception in nonrandom and college student samples was significantly larger than in random and noncollege student samples. From a theoretical perspective, these findings may have been due to student participants perceiving themselves to be smarter than other people. A more disturbing explanation would attribute these findings to researchers relying on student samples.