What is the nature of the change over time? If it is linear then a mixed models analysis looking at change modeled individually for each repsondent would work. This way, the time between assessments does not have to be the same across assessments and missing (at random) data can be handled. This could be done with SAS, SPSS (11), Mlwin, or HLM5 at the least.
Paul R. Swank, Ph.D. Professor, Developmental Pediatrics Medical School UT Health Science Center at Houston -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Frank Rigous Sent: Monday, December 08, 2003 1:09 PM To: [EMAIL PROTECTED] Subject: Classical design? Hallo! I am new to the group and not very experienced in statistics. I thought I had a classical design, two groups with either treatment A or B, measuring the response at 4 time point where the first one is a baseline measurement. I was very astonished not finding any example for this design. My first idea of analysing was to calculate 3 t-tests with the difference to baseline. But as I found (e.g. Altmann) this is not recommended. He suggested a summary statistic - but this seems not possible because the course in time may vary a lot. My next idea was to calculate a model. But I have different questions there: - Should a random effect be added which has to be nested under treatment? - Better look at the changes from baseline or add baseline as covariate? If taking covariate: *Should the covariate be nested under treatment? *What's about the homogenity assumption? - How can I calculate differences at the three time points between the treatment groups from the model? Or: Is there a typical model for this kind of design? Any help is appreciated. Frank Rigous . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . ================================================================= . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
