0
$\begingroup$

I have two groups of subjects, both undergoing two experimental conditions, and data are collected at three time points. I am interested in the time*cond and the time*cond*group interactions. When I don't include the group*time interaction, the three-way interaction is significant. However, when the non-interesting group*time interaction is included in the model, the three-way interaction is no longer significant.

Why does this happen and which analysis is correct?

Improve this question
$\endgroup$
1
  • $\begingroup$ Do you have equal n? $\endgroup$
    – David Lane
    Commented May 13, 2017 at 13:51

1 Answer 1

Reset to default
0
$\begingroup$

Welcome to the site.

It is almost never correct to include a three way interaction without including all the two way interactions that are part of it. So, you want to include it. You also don't want to get too hung up on significance.

Improve this answer
$\endgroup$
1
  • 1
    $\begingroup$ Thank you. This is what I thought but I was confused about the fact that the F for the three-way interaction dropped sharply when the adding a significant two-way interaction. Why does that happen? $\endgroup$
    – user161246
    Commented May 13, 2017 at 16:01

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.