sample size calculation for repeated measure ANOVA

[Manije72]

Hi

I want to calculate sample size by SAS for a clinical trial study which use repeated measure ANOVA test. We want to obtain P interaction between time and treatment.

We have intervention and control group and three time point" baseline, 3 month, 6 month".

We want to see the effects of nut on systolic blood pressure (SBP) compare to control group. These are mean and stardand deviation :

nut : baseline :6±3    after three month:5±4   after 6 month: 4±2  

control : baseline :7±3    after three month:6±4   after 6 month: 5±2  

I want to use this code:

proc glmpower data=WellnessMult;
class Treatment;
weight ?;
model WellScore1 WellScore3 = Treatment;
repeated Time 3;
power
effects=(Treatment)
mtest = hlt
alpha = 0.05
power = .9
ntotal = .
stddev = ?
matrix ("WellCorr") = lear(0.85, 1, 3, 1 3 6)
corrmat = "WellCorr";
run;

I have question: what should I put for stddev? what should I put for weight? Is the code correct?

[jiltao]

stddev= is the error standard deviation. Like the RMSE from an ANOVA model.

weight is the cell weight. If you have a balanced data, you should ignore this. Otherwise you would need to have a weight variable in your exemplary data set WellnessMult, representing the sample size "ratio" for each group, then specify that variable in the WEIGHT statement.