Hi,
I am running xtprobit to analyze a panel data set. However, the fitting comparison model (the pooled probit model) does not converge after 1000 iterations but the full model does converge. I was wondering if the estimated coefficients obtained in this case are valid? From my understanding, the comparison model is used only for the likelihood-ratio test of rho=0.
Thanks!
I am running xtprobit to analyze a panel data set. However, the fitting comparison model (the pooled probit model) does not converge after 1000 iterations but the full model does converge. I was wondering if the estimated coefficients obtained in this case are valid? From my understanding, the comparison model is used only for the likelihood-ratio test of rho=0.
Thanks!