Standard analyses fail to show that US studies overestimate effect sizes in softer research


Fanelli and Ioannidis (1) analyzed 82 meta-analyses and found more extreme effect sizes in studies from the United States (US), and when using soft behavioral versus less soft biobehavioral and nonbehavioral methods. Here we present a reanalysis of their data based on standard meta-meta-regressions. Results show overall publication bias across methods, but fail to corroborate Fanelli and Ioannidis’ conclusions that effects from behavioral research and the US are more inflated and more strongly subject to publication bias.

Proc Natl Acad Sci USA, 111(7), 712-713