Data set. Type I error of the F-statistic, and Greenhouse-Geisser and Huyhn-Feldt adjusted F-tests in one-way repeated measures designs. 
Tables correspond to the performance of the F-statistic, and Greenhouse-Geisser (F-GG) and Huyhn-Feldt adjusted F-tests (F-HF) in terms of Type I error in one-way repeated measures designs under violation of normality and sphericity. Values of epsilon are fixed according to the Greenhouse-Geisser procedure. The number of repeated measures (3, 4, and 6), sample size (from 10 to 300), non-normal distribution, and epsilon value (from the lower to its upper limit) were manipulated. 
Symbols used:
N = sample size
K = number of repeated measures
F = F-statistic
F-GG = Greenhouse-Geisser adjusted F-test
F-HF = Huyhn-Feldt adjusted F-test
ε = Greenhouse-Geisser epsilon
y1 = asymmetry coefficient 
y2= kurtosis coefficient