Fig. 2: Difference in effect size estimates between self- and independent replications for 16 new discoveries in the social-behavioural sciences, compared with initial confirmation study effect size estimates.
From: RETRACTED ARTICLE: High replicability of newly discovered social-behavioural findings is achievable
Difference in ESs (shapes) between self-replications and independent replications for the 16 discoveries, compared with the self-confirmatory test ES (0.0 on the x axis). The independent replication estimates are fixed-effects meta-analytic estimates of the three independent studies. The meta-analytic estimates with prediction intervals in the bottom panel combine across all 16 discoveries.
