No announcement yet.
  • Filter
  • Time
  • Show
Clear All
new posts

  • Conditional Logit (clogit) post-estimation and diagnostic

    Hello Everyone,
    I am using clogit on a cross-section data with a sample size of about 50k. My model has three alternatives. in the specification, I have state fixed-effects. I have interaction terms in the model. I have a couple of questions-

    1. Are there any problems if I use the predict command to predict the probability of each alternative?

    2. Is there any meaningful way of predicting the Margins?

    3. I am running a two-stage model where the conditional logit is the second stage. Is there any way to correctly estimate the SE and t-test?

    Thanks in advance