** Panel probit and logit models

* RE probit : based on the integration method
* RE logit  : based on the integration method
* FE logit  : based on the conditional likelihood function
* FE probit is not available: using dumies leads to biased estimators.

set memory 30m
set more off

log using xtprobit_logit.smcl, replace

use union.dta, clear
*use http://www.stata-press.com/data/r8/union, clear

iis idcode
tis year

** random-effects probit model
xtprobit union age grade not_smsa south southXt, re
mfx compute

quadchk
* # of points to use in the quadrature approximation of the integral (this checkup is important.)

** random-effects model
xtlogit union age grade not_smsa south southXt, re
mfx compute

quadchk
* # of points to use in the quadrature approximation of the integral (this checkup is important.)

** conditional fixed-effects logit model
xtlogit union age grade not_smsa south southXt, fe
mfx compute
* The above mfx gives "no effect".  But, the following will work!  
* *** clogit is the same as the conditional FE logit. ***
mfx compute, predict(pu0) nose


log close