How should I treat binary variables
Posted: Fri May 15, 2020 9:45 am
I'm sitting with a dataset with some binary variables that I wish to use as controls in my model where I test two main (continuous) latent outcome variables (in addition to some socio-demographics).
My questions are: (i) Is it ok to use binary variables as predictors in the PLS-model? and (ii) if it is ok, is there any rule of how many you can include to keep reliability of the model.
Any feedback is welcome,
Cheers/Lars
My questions are: (i) Is it ok to use binary variables as predictors in the PLS-model? and (ii) if it is ok, is there any rule of how many you can include to keep reliability of the model.
Any feedback is welcome,
Cheers/Lars