1
$\begingroup$

I am having a difficult time finding an appropriate test for my analysis. Please help -

In my study I have 130 participants who were tested with a binary outcome (pass [1] /fail [0]) before and after an intervention. However, each participant was tested twice in the pre-test and twice in the post-test.

I am trying to determine if the proportion of subjects who pass is higher in the post-test. In this scenario, I think it would be best to use the McNemar's test as we have a binary outcome for a repeated measures test. However, how can I take into consideration the bias between the subjects within each time point?

For clarification: Each Subject wrote the same test twice in the pre-test and the same test twice in the post-test.

Any insight would be helpful. Please let me know if my explanation is clear.

Thanks!

Mon

$\endgroup$
  • 1
    $\begingroup$ You have measurements which are nested within individuals. It is appropriate to analyze this with a multilevel (random effects / mixed effects / hierarchical, they go by several names) model, with a binary outcome. $\endgroup$ – Jeremy Miles Feb 5 at 17:05
1
$\begingroup$

Welcome to CrossValidated.

You have measurements which are nested within individuals. It is appropriate to analyze this with a multilevel (random effects / mixed effects / hierarchical, they go by several names) model, with a binary outcome.

If the results are highly correlated - that is, if almost all of your subjects got the same score in the pre-test, you can safely ignore one of them (you lose a little information, but not much).

$\endgroup$
  • $\begingroup$ That's what i was thinking as well! thanks for your response. I am not sure if you area a Stata user, however, can you let me know if this code looks correct? melogit most_class time || obs_hw_uic where most_class is the test result (0/1) time is the before and after (0/1) obs_he_uic is the subject id number $\endgroup$ – Monica dC Feb 7 at 11:31
  • $\begingroup$ It's a while since I've used Stata - don't you need a colon after obs_hw_uic? $\endgroup$ – Jeremy Miles Feb 8 at 22:17
  • $\begingroup$ looks right apart from that. $\endgroup$ – Jeremy Miles Feb 8 at 22:17
  • $\begingroup$ yes, I added the semi-colon. Thanks for all your help :) $\endgroup$ – Monica dC Feb 12 at 16:36
  • $\begingroup$ Colon, not semi colon. $\endgroup$ – Jeremy Miles Feb 12 at 18:55

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.