topic Within subjects only analysis in SAS Health and Life Sciences

Within subjects only analysis

deleted_user — Sun, 03 Jun 2007 02:20:39 GMT

How can I analyze a design where the sample is randomly divided into three groups? Each group receives one of three treatments and dependent variables are measured. Then, a second trial is conducted where each group is administered a different treatment than they received the first time, and dependent variables are measured. The same procedure is followed a third time.

I have been told that this is a repeated measures design. However, the only examples of repeated measures I can find like this involve the same treatments over time, and between subjects (group) differences are compared.

I need a SAS code to partial out the repeated measures part of the SSEs.

Re: Within subjects only analysis

Rucstat_huadli — Mon, 04 Jun 2007 06:07:06 GMT

Your mean is A Three (Sequences) by Three (Periods) Crossover Design with Repeated Measures?

The Mixed Linear Model,I think, is fitted for this design.

Re: Within subjects only analysis

Rucstat_huadli — Mon, 04 Jun 2007 06:10:48 GMT

PROC MIXED data = data;
CLASS subject trt period seq;
MODEL var = trt seq period / s ddfm=satther;
RANDOM subject(seq);
CONTRAST 'A vs B' trt 1 -1 0/ stdiff cl;
RUN;