1
$\begingroup$

I am trying to calculate the power in my SEM analysis post-hoc. How exactly should I do this? What is the power for the R-squared result of IT-T2 and IT-T3?

Background info: Sample size is 255. IT and SD are two different personality traits that were measured across three points in time (during an intervention). I originally assumed they would influence each other over time, but those paths turned out to be non-significand and I deleted them.

I used calculations with different results. For T3 one calculation came up with 0.9989 and another with 0.1763. Should I test the power here like a student’s t-test or like a multiple regression? I originally assumed SD will also be a predictor for IT but those estimates were not significant, leaving IT-T2 the only predictor of IT-T3. So no multiple regression?

Pictured are standardized estimates in my model:

enter image description here

Improve this question
$\endgroup$

1 Answer 1

Reset to default
2
$\begingroup$

Just don’t. Post hoc power is meaningless. However, if you insist on it, you can compute it correctly by noting that power is the probability of obtaining statistical significance. Therefore:

  1. If H_0 was rejected, post hoc power = 1.0
  2. Otherwise, post hoc power = 0.0
Improve this answer
$\endgroup$
2
  • $\begingroup$ You are right of course. I know you should care about power a priori! However, I am working with data from an old study and can't change that. Not all variables of this old study have been analyzed yet and I'm trying to get as much info out of it as possible. $\endgroup$
    – E_H
    Commented Mar 3, 2019 at 14:56
  • 1
    $\begingroup$ Right. So by your own words, if you’re not doing a new study, power has no relevance. $\endgroup$
    – Russ Lenth
    Commented Mar 3, 2019 at 15:08

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.