“…Meta research projects of the 'Many analysts, one dataset' type, in which many researchers are testing the same hypothesis with the same dataset, demonstrate that researcher degrees of freedom easily lead to entirely different conclusions. In a recent study relying on international survey data, researchers were asked to estimate the association between immigration and public support for government provision of welfare (Breznau et al, 2022). 25% of estimates were significantly negative, 17% were positive, and 58% had a confidence interval including 0.…”