Dear all,
I would like to ask for your advice.
The situation is like this: I have 1 IV which is measured on a Likert scale and 1 DV which has 4 categories. I also want to test the moderation of several other variables some of which are categorial, dichotomous, or scale.
Can I use Process Macro for such an analysis and if I can should I run several models independently- with one moderator at a time- or should I add all the moderators together?
Thanks for your help!
Regards,
Ilia
Hello Ilia,
According to Hayes' text (p. 587): "PROCESS uses OLS regression for model estimation, it will assume yvar and mvlist have at least interval level measurement properties, except when yvar is dichotomous in the data, in which case logistic regression is used for the model of yvar."
Note: "yvar" is the DV for the model.
So, while other variables may be categorical, the PROCESS macro does not allow for nominal/categorical DVs having more than 2 levels.
Good luck with your work.
Reference: Hayes, A.F. (2022). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach (3rd ed.). New York, NY: Guilford Press.
Given that you are asking a about PROCESS, I'll guess that you are using SPSS. If so, you can estimate your model (including interaction terms) via the NOMREG command (assuming the categories of the DV are not ordered). There is no need for a macro.
- https://www.ibm.com/docs/en/spss-statistics/28.0.0?topic=reference-nomreg
- https://stats.oarc.ucla.edu/spss/dae/multinomial-logistic-regression/
HTH.
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology
- Research Papers