Hi. I would like to seek advice on how to conduct sample size calculations for the following: I have a 4 arm incentive-based RCT, 3 treatment arms (Arms T1, T2 and T3) and 1 control arm (arm C), where medical providers are provided different kinds of incentives to change targeted behaviors. This is a between-subjects design, with a nested within-subjects study. Arms T1 and T2 provide two different kinds of incentives. Arm T2 allows to then choose two options on how to spend the incentives provided in it (the revealed preference within part). T3 is an interaction of T1 and T2. The hypothesis is that T3>T1/T2>C. The study has a stratified randomization procedure to allocate providers to the different arms. Strata are based on the languages spoken by the providers.
Part of the problem is that there aren’t similar previous studies from where I can gather effect size, variance, or error estimates. I do have some sample data that has the number of languages spoken by providers. This is an ordered variable (ranging from 1-6 languages spoken by each provider). I now need to figure out how to calculate the necessary sample size. Any guidance would be greatly appreciated.
Joseph Coveney I've referred to a few responses by you on sample size calculations to refine my own thinking on this. Would you happen to have any specific advice on this?
Part of the problem is that there aren’t similar previous studies from where I can gather effect size, variance, or error estimates. I do have some sample data that has the number of languages spoken by providers. This is an ordered variable (ranging from 1-6 languages spoken by each provider). I now need to figure out how to calculate the necessary sample size. Any guidance would be greatly appreciated.
Joseph Coveney I've referred to a few responses by you on sample size calculations to refine my own thinking on this. Would you happen to have any specific advice on this?
