I'd like to run a regression so that I can determine the standardized Beta-weights of five categorical IVs (some with more than 2 levels) on a categorical DV (binary). I've ruled out the following techniques for various reasons:

  • no to standard linear regression (non-continuous variables)
  • no to loglinear regression (study design was repeated measures)
  • no to multinomial regression (only one outcome variable with two categories)

Is there a way to get around these analysis confounds that are resulting from my study design? Any help at all would be appreciated!


You want a nonlinear multilevel model with a binomial link.

These are a generalization of logistic regression to allow for repeated measures.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.