A Meta-Analysis Comparing the Effectiveness of Narrative vs. Statistical Evidence: Health vs. Non-Health Contexts

This study is a meta-analysis of primary studies that make a direct comparison between narrative and statistical evidence in both health- and non-health-related communication contexts. The meta-analysis included 50 studies with 65 experimental pairs (k = 65) based on 13,113 (20-1270) participants. We examined the overall persuasiveness of evidence type by computing the correlations (r's) for all pairs, based on the random-effects model, which revealed an effect size of 0.016 (95% CI, -0.014 to 0.045, p = 0.296). Two types of evidence did not significantly differ in effectiveness under either communicative context. The moderation analysis indicated that narrative evidence had a significant advantage over statistical evidence for health messages advocating for prevention behaviors. Compared to non-student samples, the narrative evidence trumped statistical evidence for health-related issues. As communication research continues to investigate the implications for message persuasiveness derived by narrative and statistical appeals, our study suggests that the relative effectiveness is likely a complicated and nuanced matter. Practical implications and limitations have also been outlined.