The title is quite self-explanatory - I'd like to know if there's any other parametric technique apart from repeated-measures ANOVA, that can be utilized in order to compare several (more than 2) repeated measures?


1 Answer 1


Multilevel/hierarchical linear models can be used for this. Essentially, each repetition of the measure is clustered within the individual; individuals can then be clustered within other hierarchies. For me, at least, it's more intuitive than repeated-measures ANOVA.

The canonical text is Raudenbush and Bryk; I'm also really fond of Gelman and Hill. Here's a tutorial I read some time ago - you may or may not find the tutorial itself useful (that's so often a matter of personal taste, training and experience), but the bibliography at the end is good.

I should note that Gelman and Hill doesn't have a ton on multilevel models specifically for repeated measures; I can't remember if that's the case or not for Raudenbush and Bryk.

Edit: Found a book I was looking for - Applied Longitudinal Data Analysis by Singer and Willett has (I believe) an explicit focus on multilevel models for repeated measures. I haven't had a chance to read very far into it, but it might be worth looking into.

  • $\begingroup$ I doubt that anyone is going to top this one. Thanks Matt, great answer! $\endgroup$
    – aL3xa
    Commented Aug 5, 2010 at 23:48
  • $\begingroup$ Forgotten bonus: unbalanced data are still useful! $\endgroup$ Commented Aug 18, 2010 at 23:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.