1

I am currently working on my dissertation project where my data are essentially all dummies. From my dependent to my independent variables, everything is a dummy variable (0,1), at least for the preliminary model. The data are cross-sectional.

I was wondering what the best model to use would be. I have decided to do a simple regression followed by logit and probit. Does that sounds like it covers everything? I have some interaction terms too. Also, what tests I can use?

gung - Reinstate Monica
  • 132,789
  • 81
  • 357
  • 650
user52466
  • 11
  • 1

1 Answers1

1

Your suggestion doesn't make sense. If your response variable is 0/1, then you should use logistic regression or probit. (If you want help understanding the distinction between them, it may help you to read my answer here: Difference between logit and probit models.) You can include interactions as you like. The model output will come with Wald tests of the parameters without your having to do anything else.

Community
  • 1
gung - Reinstate Monica
  • 132,789
  • 81
  • 357
  • 650