1

I have 40 "subjects", each of them did 5 types of exams.

The question I wish to test is if there is a difference in variance between the exams.

That is, I would like to run something like a "Bartlett's test", but for "repeated measures".

Does such a test exists?

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
Tal Galili
  • 19,935
  • 32
  • 133
  • 195
  • 1
    What kind of exams are they (do they attempt to measure the same construct, related constructs, or independent constructs)? Do you have multiple exam scores per subject for each exam (ie, sub i took exam j k-times, w/ K>1)? Variance in what sense (internal reliability / measurement error, or between subjects / if the subs are more 'spread out' on 1 dimension than another)? – gung - Reinstate Monica Sep 25 '13 at 21:15

0 Answers0