I need to perform hierarchical regression with dummy variables. I also need to check moderation by introducing in the model interactions of these dummy variables and the moderator. My questions are:
- Do I introduce in the model a number of dummy variables which is equal to the total number of predictors minus 1, making the absent variable the reference category?
- How do I interpret the results, given presence of the reference category?
- For interactions, do I multiply the dummy variables (categorical level) by the moderator (interval level, measured on a 1-5 Likert scale)?
Thanks for your time and any ideas are much appreciated.
Kind regards Muzi