Bioequivalence and Bioavailability Forum

Hi Angus,

❝ The ANOVA p value was 0.08 so the gender effect was not significant. Under what circumstances would one use the BE data analysis with gender in the model.

This is in my opinion just like a group effect. I wouldn't worry too much about a plain gender effect; in CO studies it will be a between-factor anyway, and genders do differ in terms of size, shape, plasma volume and just about anything else. Gender x Treatment is another matter; it would be a very surprising finding if it comes out significant and could in my opinion be followed up by additional questions; ideally this is only academic - wouldn't want to have a study with [Women, Test] being different from [Women, Ref]= [Men, Test]=[Men, Ref] and so on.
Since we don't deal with group x treatment effects in other types of studies I think one could along the same lines often justify a model without gender effects unless there is a specific reason to think otherwise. Same could be said for hair color, music preference or anamnetic presence of altitude sickness among relatives.