Lagrange Multiplier test after xtprobit, re

[Tim Burnett]

Dear all,

I am currently a bit stumped as to the syntax to use, if any, for a Lagrange Multiplier test of var(u_i)!=0 following a random effects probit command. I understand the xttest0 postestimation command is good for xtreg, does a similar command exist for xtprobit ..., re?

Many thanks in advance,

Tim

[Phil Bromiley]

In the manual, it shows xtprobit automatically giving a test that rho is zero. This isn't exactly var(u_i) but it is close. Also, since the model assumes u_i are normal, can't you just calculate a test using the parameter estimate and standard deviation provided?

[Marcel Issa]

Hi Phil,
could you please tell me exactly how you would test between the random effects xtprobit and the pooled probit model?