2

I've been trying to learn how to do this analysis but I can't find any information that sheds light on my case and I can't figure out what to do from Hayes' book. I would really appreciate it if someone could help me out.

For my undergraduate thesis, I'm examining how emotion regulation predicts resilience and whether age and gender moderate this relationship. The Emotion Regulation Questionnaire measures two strategies: Cognitive Reappraisal and Expressive suppression. This means I have two predictors and two moderators (one of which is dichotomous, meaning gender).

What model should I use in Process? And considering that only one predictor can be entered, do I run the analysis twice? And would running multiple analyses lead to Type errors I and II, requiring a Bonferroni or Hochberg correction?

Any advice would be greatly appreciated.

Demi
  • 21
  • 1
    Are there any mediators included in your model? If not, then the Hayes PROCESS macro might be overkill. In general in a regression model, a "moderator" is just another name for a variable involved in an interaction term with a predictor of primary interest. In a single regression model you can include as many predictors and interaction terms as you think you need, based on your understanding of the subject matter. The more predictors and interactions, however, the more observations you need to avoid overfitting. – EdM Apr 08 '23 at 07:46
  • Thank you for your response. – Demi Apr 09 '23 at 08:35

1 Answers1

0

You have four regressions (ERQ-CR x Age, ERQ-CR x Gender, ERQ-E x Age, ERQ-E x Gender). Run the moderation macro (model 1) four times, once for each regression. Then correct for multiple comparisons across the four regressions. (This is very easy to do in other software, like R, and also relatively easy to do in SPSS, though you have to make the interaction variables by hand, which can be tricky, depending on how long you've been using SPSS for).

David B
  • 1,532