Systematic review and meta-analysis of the evidence for an illusory truth effect and its determinants

Over the last four decades, studies provided evidence that individuals tend to rate statements as being more truthful when they are re-exposed to them, the so-called 'illusory truth effect'. In light of a growing number of studies published since the previous meta-analysis in 2006 and concern of publishing biases, we conduct a meta-analysis on 182 studies and 366 effect sizes (N = 31,184 participants) published from 1977 to 2025. After correcting for small-study effects, we observe a small illusory truth effect (g = 0.37, 95% confidence interval [0.30, 0.44]), with a substantial within and between-study heterogeneity. Here, we show that multiple variables accounted for such heterogeneity, including the type of item, the instructions during the first exposure, the presence of veracity cues, and the duration of presentation on first exposure to the statement. We highlight the importance of the initial exposure and discuss practical implications regarding the current misinformation crisis.