paired t-test or repeated measurments ANOVA


New Member

Given a group where you run 2 models to obtain a parameter, and the goal is to compare both parameters. But not all individuals tested in one group are tested in another, so there is some variance.

If using a paired t-test you execute an F-test to see if the variance is acceptable for a paired t-test and if so you run the paired t-test

But if you run you're data on more then 2 models and you choose a repeated measurments anova. Can you check if you can compare groups like the F-test before the paired t-test? Can you still perform this test if not all individuals are tested in both groups?



No cake for spunky
I have never seen a paired t test used when some of the members in the two populations are not the same. I do not believe that is a valid use of that method. I have also never heard of a F test being used to determine whether it is valid to run a paired t test.