Hello,
I have the following regression: y=ax1+ax2+e
My dependent variable y is a binary choice variable and x1, my
endogenous variable is continuous. x2 represents the rest of my
control variables.
I would like to use the identification through heteroskedasticity
like in Hogan and Rigobon but I don't know if having y as binary and
therefore having to use LPM or Probit make me having a
heteroskedasticity problem anyway.
Does anyone know where does this heteroskedasticity of Probit come
from? could you indicate anything in the literature?
Thank you,
Nirina
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
