I was going to use a mixed model (by setting subject as a random effect) to analyze my repeated measures data set which are oridinal, but JMP doesn't allow me to use random effect with oridinal data.
Can anyone let me know how to do it in JMP?
I'm a newbe on mixed models and in my opinion the JMP help and documents do not provide much help. To me the following site is a gift from heaven, http://people.stat.sfu.ca/~cschwarz/CourseNotes/
It provides detailed step by step examples for JMP, R and SAS. Look e.g. at chapters 15 and 16. It deals with these questions.
Depending on how you want JMP to evaluate your data, you either have to change your columns to continuous (makes regression calculations) or nominal. JMP does not seem to accept ordinal factors.
In your case, if "subject" is ordinal in your data set, simply change it to nominal. In the Fit Model window Add " Subject&Random " as Model Effects.