I ran a Probit model to determine which variables/interaction terms are significant. I got these results:

  • Probit: "mino" is not significant. Probit results

  • Whith the margins command (Averadge marginal effect) I obtained this: "mino" seems significant. Margins results

Which p-value should I use ? How should I interpret it ? And for interaction terms how to proceed ?

Thank you,


  • $\begingroup$ Please explain terms/obrat, mino etc $\endgroup$ – Subhash C. Davar May 13 '18 at 14:50
  • $\begingroup$ What for you want to use p-value? $\endgroup$ – Subhash C. Davar May 13 '18 at 15:18

@Louis Ansuez Forget p-values. The results produced by two methods are congruent and show that there is a significant effect of proposed variable (or interaction term) on the dependent variable. See z = - 5.11 in case of log-likelihood and in case of delta-method, it is 5.59. Any clarification !


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.