Hi,
I'm working on an RCT design that involves one control group and three treatment groups. I'm trying to find the right command to compute the minimum detectable effect size needed to find a significant effect of the treatments administered to each treatment groups, but all the commands I have found are limited to two groups (control and treatment).
The outcomes for each group are proportion of participants who enroll in a specific activity, and I expect an outcome of 0.1 for the control group. The two-sample command I've found looks like this:
power twoproportions .1, n1(1000) n2 (2000) power(.8) alpha(.05)
The control group has 1000 participants and one treatment group has 2000 subjects. I don't know how to include the N sizes for the other two treatment groups.
Any thoughts/suggestions are appreciated!!
Thanks,
I'm working on an RCT design that involves one control group and three treatment groups. I'm trying to find the right command to compute the minimum detectable effect size needed to find a significant effect of the treatments administered to each treatment groups, but all the commands I have found are limited to two groups (control and treatment).
The outcomes for each group are proportion of participants who enroll in a specific activity, and I expect an outcome of 0.1 for the control group. The two-sample command I've found looks like this:
power twoproportions .1, n1(1000) n2 (2000) power(.8) alpha(.05)
The control group has 1000 participants and one treatment group has 2000 subjects. I don't know how to include the N sizes for the other two treatment groups.
Any thoughts/suggestions are appreciated!!
Thanks,
