0
$\begingroup$

I am trying to determine what the appropriate post hoc testing for a one way repeated measure between subject ANOVA. Would it be appropriate to use a Tukey test in this scenario?

$\endgroup$
0
$\begingroup$

Well, appropriateness of the test depends upon the comparisons to be made (whether planned or unplanned comparison). As different post-hoc tests have different desirable properties it's better to have a look at them and then decide which best suits you. This article might be of help to start with. If you are looking for post-hoc for between subject factors, this link and for within subject post-hoc this link give extensive resources on the same. And if you are using SPSS, the help file in SPSS GLM-RMANOVA menu has fair bit of explanations on each test.

$\endgroup$
  • $\begingroup$ Thanks for your help! I had taken a look at that page before but it doesn't mention if you can just carry out your post hoc test as you would normally or if it needs to be done somewhat differently to account for the fact that it was a repeated measure scenario. I was wondering if you had any comments on that? $\endgroup$ – sarahmbr Aug 24 '17 at 1:28
  • $\begingroup$ Yes. You can carry out Tukey's test. Essentially, they run pairwise test taking into account the correlation (say, paired t test) and make comparison to adjusted level of significance according to number of comparisons to be made. If you really want to test it out, you can run LSD test (which makes no adjustments) in SPSS and see that, they are nothing but the results of separate paired t tests. For, more details on how Tukey works, you can refer the same link that I mentioned earlier. I hope this helps. $\endgroup$ – Enigma Aug 25 '17 at 2:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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