Dear Statalisters,
I am testing for the equality of logit regression coefficients across different sub-samples using the -suest- command with -test- and -testnl-.
The help page for the -suest- command mentions: "The estimators should be estimated without vce(robust) or vce(cluster clustvar) options. -suest- returns the robust VCE, allows the vce(cluster clustvar) option". However, I am still having to run the logit command twice- once with cluster(clustervar) so that I can record the log-likelihood statistic to report in my results and once without cluster(clustervar) for the suest command.
Is there any way of getting suest to retrieve the log-likelihood (of the specification using vce (robust) or vce (cluster clustvar)) so that I can avoid running the logit command twice?
Many thanks,
Mihir
I am testing for the equality of logit regression coefficients across different sub-samples using the -suest- command with -test- and -testnl-.
Code:
foreach subsamp in samp1 samp2 { logit y x i.var1 & `subsamp' est store `subsamp' } suest samp1 samp2, cluster(clustervar) test [samp1_y]x = [samp2_y]x
Is there any way of getting suest to retrieve the log-likelihood (of the specification using vce (robust) or vce (cluster clustvar)) so that I can avoid running the logit command twice?
Many thanks,
Mihir