Skip to main content
. 2016 Mar 16;5(2):122–124. doi: 10.1007/s40037-016-0256-6

Table 1.

Scenario A (n = 64 per condition) and scenario B (n = 11 per condition) in terms of expected proportions of Type I error prevalence in a pile of statistically significant outcomes.

Scenario A (n = 64) B (n = 11)
Expected rejections of the 20 true null hypotheses 1 1
Expected rejections of the 80 untrue null hypotheses 64 16
Expected proportion of Type I errors in the pile of statistically significant results 1/65 (≈ 0.015) 1/17 (≈ 0.059)