I know this is basic, but that's where I'm at. I want to determine a sample size to test a difference between two proportions of 15% for a one tailed test. I am using test family z-tests, statistical test of proportions: difference between two independent proportions. And it is a-priori for g*power. What I don't understand is proportion p1 and proportion p2. If my control group has a treatment result of .57 (57%), I put .57 in proportion 1, correct? And if so, what do I put in proportion 2 .15 (15%) or .42 (42%)?
Thanks so much for your help.
Zeda
Thanks so much for your help.
Zeda