I am going to conduct a pretty simple experiment: one within factor with 8 levels for a single group of participants. I'm trying to determine the apporpriate sample size using GPower based on an effect size of eta square = 0.60 from a previous study.
However, the end sample size is odd. Here are the parameters I entered :
F tests > ANOVA: Repeated measures, within factors > A priori
Effect size = 0.6 => (f = 1.2247449)
alpha = 0.05 / Power = 0.80
Number of groups = 1
Number of measurements = 8
Corr among measures = 0.5
Nonsphericity correction = 1
Result : a sample size of only 2 participants.
What exactly is the problem with my input? Is there another way to compute the appropriate sample size?