Hi
I came across this post when trying to figure out the best model in a similar situation. Here it's been suggested to use "gender" either as factor or as covariate with a vector of 1s/-1s for each gender. My question is what would be the best option when there are only few subjects of one gender. In my case, I'm comparing 3 groups, each group including around 15 subjects where male subjects consist 20 – 37% of all subjects. Here's the exact numbers:
Group 1: 11 F / 5M
Group 2: 9 F / 2 M
Group 3: 10 F / 6 M
Would choosing "gender" as a factor in this case affect statistical power?
Many thanks in advance
Mahsa
|