0

I am running a logistic mixed model with a binary outcome (correct/incorrect).

I have two fixed effect predictors:

  1. Condition (0, ‘low’ vs 1, ‘high)
  2. Sumspq (a continuous questionnaire score)

There is a random intercept for participants. Participants complete trials in each condition and each have a single sumspq score.

The model: Correct ~ sumspq*Condition+ (1|participant)

My question: How do I interpret the regression slopes for the main effects? E.g., is the main effect of sumspq the effect of this variable on the log-odds when Condition is ‘low’ (0)?

I have grand mean centered both Condition and sumspq as recommended in the literature. However, I have read that this only changes the intercept.

SilvaC
  • 512
  • 3
    Check this post: https://stats.stackexchange.com/questions/365907/interpretation-of-fixed-effects-from-mixed-effect-logistic-regression/365918#365918 – Dimitris Rizopoulos Dec 07 '23 at 14:39
  • @DimitrisRizopoulos thank you and apologies as I did not see this post. To double check, the main effect of condition would be the change in log-odds between conditions on average (i.e., across sumspq scores) rather than for the average sumspq score? – SilvaC Dec 07 '23 at 15:53

0 Answers0