Dear All,
I am looking to test for normality in the Probit equation.
In particular, I need to run an artificial regressions in which the sample
moments have to be regressed against an intercept and the scores from the
Probit model. Under the null the coefficients on the intercepts should
each equal zero.
Scores from the probit equation can be obtained. How does one go about
sample moments? I would like to know you views on how I can output the
3rd and 4th moments as well as the intercept.
I would also like to hear your views on the actual approach that I am
following. My ultimate goal is to perform a parametric two-step sample
selection correction (Heckit). To justify the paramtetric framework, I
have to show that the distributional assumption holds.
In Stata -sktest- will perform a basic test for normality based on the 3rd
and 4th moments (skewness and kurtosis). However, I don't think that this
test can be applied in my case.
Thank you.
Altay Mussurov
PhD candidate
University of Wales Aberystwyth
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
