Table 1

Post hoc power calculations for m independent experiments with Bonferroni correction k

FigurePanelData structureType of testPower
2C Normality not assumedMann-Whitney55.9
DNormality not assumedMann-Whitney53.2
FNormality not assumedMann-Whitney61.7
GNormality not assumedMann-Whitney10.7
HNormality not assumedMann-Whitney12.3
INormality not assumedMann-Whitney2.6
JNormality not assumedMann-Whitney12.6
3BNormality not assumedMann-Whitney21.5
CNormality not assumedMann-Whitney63.3
DNormality not assumedMann-Whitney36.4
4BNormality not assumedMann-Whitney100
DNormality not assumedMann-Whitney100100
7ANormality not assumedMann-Whitney37.1
BNormality not assumedMann-Whitney90
JNormality not assumedMann-Whitney100100
KNormality not assumedMann-Whitney99.9100
  • From Fisher's χ2 test for combined probabilities, we have that χ2 (df = 2m) ∼ –2ln(p1p2…pm), where pi is the p value for the ith independent experiment. The post hoc expected value of χ2 is just the df + the noncentrality parameter (λ). Thus, for three independent experiments, say λ is given by –2ln(p1p2p3) – 6. The power can then be obtained directly using G*Power 3 software with α = 0.05/k.