I need to perform an a priori power analysis for a logistic regression. My dependent variable is binary and I have two independent variables (between subjects), one variable with 3 levels (group) and the other with 2 levels (posture). I did not perform any pilot study, but from the literature, I know the ANOVA effect size of posture on the dependent variable (partial eta squared = .22), but I didn't know how to correctly determine the sample size for the logistic regression knowing only the effect size of the ANOVA. I didn't know odd ratio or other parameters to calculate the sample size for logistic regression.Can anyone help me with this? I have alpha, Power, and the ANOVA statistics from the study in literature.
Asked
Active
Viewed 41 times
0
-
Can you clarify what is np2? (by an edit to the post). Also, there are many similar Qs on site, so please do a search. You can start with https://stats.stackexchange.com/questions/162294/multiple-logistic-regression-power-analysis – kjetil b halvorsen Feb 13 '24 at 23:00
-
Thanks, but I only have the following parameters: -alpha = .05 -power= .80, -ANOVA effect size from a previous study (partial eta squared = .22). I'm trying to figure out how to do an a priori sample size estimation. Do you think this is possible? – Tes Feb 13 '24 at 23:37