0

In my main analysis, I observed that people had better overall performance (0=failed, 1=success) in the manipulation condition than in the control condition. This was a within-subjects design.

This was the model I used:

fit <- glmer(performance~condition + (1|id), data = df, family=binomial())

In the descriptive analysis, I can see that males had better performance than females. I want to examine if that difference is significant, as an exploratory analysis. How can I add gender to the model?

Olivia
  • 121
  • In the same way you would add condition to the model. It would be performance~condition + gender + (1|id). But it depends on if you want to control for condition. – geoscience123 Nov 08 '22 at 01:25
  • If I do that, I'll get the gender effect on the overall performance, right? I'd want to evaluate the gender effect specifically on the manipulation condition. – Olivia Nov 08 '22 at 03:10
  • Are you thinking of interacting condition with gender: performance ~ condition * gender + (1|id)? To learn more about interactions and how to interpret them see eg. 1 and 2. – dipetkov Nov 08 '22 at 09:07
  • @Olivia I think you just want condition ~ gender + (1|id)? – geoscience123 Nov 08 '22 at 21:59
  • Thank you! I wanted an interaction effect like @dipetkov mentioned – Olivia Nov 09 '22 at 16:21

0 Answers0