The effect of audience–character similarity on identification with narrative characters: A meta-analysis

Identification, the experience of a psychological merging between the self and a character, is a key mechanism underlying the power of narratives to influence attitudes, beliefs, intentions, and behaviors in story-consistent ways. Similarity between audience members and characters has been intuitively thought to be an antecedent of identification, but experimental studies have yielded inconsistent findings regarding the effectiveness of manipulating similarity on eliciting identification. The current meta-analysis synthesized and quantified the evidence from 39 studies (k = 50, N = 11,077) and investigated several potential causes of heterogeneity at both the narrative and study levels. The data revealed a small but significant and robust overall effect of similarity on identification (g = 0.19, 95% CI [0.10, 0.28], p < .001), with little evidence of publication bias. A notable narrative-level moderator was type of similarity, with manipulations of psychological similarity yielding larger effects than manipulations of objective similarity. In addition, study design emerged as a significant study-level moderator, with the similarity–identification effect being stronger in studies that manipulated similarity using a within-subjects design than those that used a between-subjects design. Insights gained from this meta-analysis can help to address some ill-defined aspects of the similarity–identification hypothesis, contributing to a better understanding of involvement with narrative characters. Practically, the results can inform the design of more effective targeted and tailored narrative messages that are intended to engage and persuade audiences using the tactic of incorporating similar characters.