Hi everyone,
How do you calculate sample size for a pre-post 2 group design using ANCOVA ( baseline adjusted) in Gpower?
This is what I used: Effect size: (f): .66 ( When correlation is .8): using equation f'=f/SQRT(1-rsqr), here f=.4.
Alpha: .05, Power: .80
Numerator df: 1 (bcos 2-1)
Number of groups: 2
Number of covariates: 1
Total sample size: 21.
Without using the covariate adjustment, I get 52 sample size.
Did I do this right?
Thank you so much
Anoop
|