2
$\begingroup$

Can I run a probit on survival time data? It's discrete-round, and I want to look at whether lagged variables affect the failure event. I am, however, getting negative coefficients for a probit regression, and I'm not sure that should be happening. Help?

$\endgroup$
  • 1
    $\begingroup$ Can you write the model? $\endgroup$ – ocram Apr 30 '13 at 7:36
2
$\begingroup$

Surely, you could run a logit, probit, or a cloglog model; in fact any appropriate binary model works well. The logit model (and to some extent the cloglog) model tend to be used more often, but this is only because of the familiarity of the users with these models.

See the following link for more details: https://files.nyu.edu/mrg217/public/btscs.pdf

$\endgroup$
  • $\begingroup$ It is not because of somebody's familiarity with these models. It is because these models have nice substantive interpretations. The logit model is the proportional odds model (you are effectively modeling the odds of the failure vs. survival in the next discrete time period), and the cloglog model is the proportional hazards model (a discrete version of the semi-parametric Cox model which requires fewer assumptions about the data than the parametric models like Weibull: you can assume that all subjects in the study only differ from each other in their hazard rates by multiplicative factors). $\endgroup$ – StasK Feb 25 '14 at 2:51
  • $\begingroup$ And here's an additional link: iser.essex.ac.uk/resources/survival-analysis-with-stata, see especially Jenkins' lesson 6. $\endgroup$ – StasK Feb 25 '14 at 2:52
  • $\begingroup$ So what's the benefit (difference) of a survival model vs simply running a probit / logit on whether the event happens in this period? $\endgroup$ – Heisenberg Feb 11 '15 at 18:11

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.