Justify your alpha

Abstract

Benjamin et al. proposed changing the conventional “statistical significance” threshold (i.e.,the alpha level) from p ≤ .05 to p ≤ .005 for all novel claims with relatively low prior odds. They provided two arguments for why lowering the significance threshold would “immediately improve the reproducibility of scientific research.” First, a p-value near .05provides weak evidence for the alternative hypothesis. Second, under certain assumptions, an alpha of .05 leads to high false positive report probabilities (FPRP2 ; the probability that a significant finding is a false positive)