0
$\begingroup$

I've run a linear mixed effects model with a null result that would be quite interesting for the field. Is there a test I can run that would bolster the claim that the two groups are indeed equivalent? I know that Bayesian analyses are ideal for this, but I am hesitant because of a lack of previous literature that could provide a prior.

Would simply running a post-hoc power analysis be enough?

Improve this question
$\endgroup$
2
  • 2
    $\begingroup$ A post hoc power analysis is just a transformation of the $p$-value$---$it doesn't tell you anything. However, you might be interested in two one sides tests of equivalence (TOST). $\endgroup$
    – Frans Rodenburg
    Commented Aug 10, 2018 at 0:45
  • $\begingroup$ Thanks for the reply! I've come across TOST, however, I'm not sure if it's possible to implement that given that I'm running a mixed effects model. Have you ever come across someone doing that? $\endgroup$
    – Dave
    Commented Aug 10, 2018 at 17:37

0

Reset to default

Your Answer

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