Usually a repeated measures experiment uses equal time intervals. In our case, we used unequal time intervals. Eg. We collected samples at 28, 42, and 75 days. The intervals are not equal (28, 14, and 33 days). Any modern technique to analyze such data?
Thank you, David Eugene Booth
We measured body weight at birth (0), 28, 42, and 75 days of age and calculated the average daily gain (ADG) for over the mentioned time interval.
We analyzed the ADG as repeated measures data using SAS with the syntax:
Proc mixed;
Class ID Trt Day;
Model ADG= Trt Day Trt*Day;
Repeated Day/subject= ID Type=UN;
LSMEANS Trt Day Trt*Day / PDIFF ADJUST = TUKEY;
Run;
However, the time intervals are not equal (b/n birth and 28 is 28; b/n 28 and 42 is 14, and b/n 42 and 75 is 33). Thus, maybe the LS means results with this method are not correct. We are asking for any recommendations to analyze such data
Thank you,
Thanks for your reply I'm not really familiar but it would be Prof Davidian who wrote the book on the subject. I've attached a table of contents and you can download the note version with examples and code.from her website as indicated below. If I can be of further assistance please let me know. Best wishes David Booth
Thank you David Eugene Booth for your recommendation. It is really helpful!
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis