Religious priming as a research tool for studying religion: evidentiary value, current issues, and future directions

1 The trim and fill correction has been criticized for not sufficiently adjusting for publication bias. Since this meta-analysis was first published a second paper re-analyzed this data with two additional tools that use different approaches to correct for publication bias: Bayesian Biased corrected model and a PET model [21]. The Bayesian model, similar to our analysis, found strong evidence for an overall effect. However, the PET model did not. The authors concluded that there is mixed evidence for the effectiveness of religious priming. Efforts are underway in the field to create and refine new meta-analytic tools that correct for publication bias, and there are ongoing debates — also about PET. Simulations show PET fails to find a true effect in samples with similar effect sizes between 33% and 53% of the time. When the data is heterogeneous, PET’s performance in correctly detecting a real effect declines even further [22]. Religious priming allows researchers to explore the causal impacts of religious concepts on a wide variety of psychological outcomes. We review recent meta-analytic findings, and discuss the impact of methodological variation and convergent effects. We conclude that current evidence supports religious priming as having evidentiary and utilitarian value, but more assessment of moderators and the robustness of these effects across methods and contexts is needed.