[R] pglm: probit random model

[alfonso.carfora at uniparthenope.it]

Hi all,

I'm using pglm package to estimate a static random probit model of the type:

Yi,t = B1X1i,t-k +... + BzXzi,t-k + Ci+ Ui,t

Where Yi={0,1} and all explenatory variables are continuous.

the function I use is:

pglm(formula,, Data, family = binomial(probit), model =  
"random",method= "bfgs")


Can I test the exogeneity and zero correlation between the explanatory  
variables and ci? Or the function carries in an approach (as  
Mundlak-Chamberlain) to overcome the endogeneity between regressors  
and effects?

Thanks
Alfonso