Marginal Effects in Probit
Posted: Thu Apr 23, 2009 5:21 pm
Hello, I have read all the recent posts on generating the marginal effects for probit models and have read the users guide but am still having trouble. My independent variable of interest is a dummy variable if that changes things. Here is my understanding of the steps:
1) Estimate my equation
2) Click Proc/Forecast-Index
--Eviews generates a series named LBWF (lbw is my dep variable) when I click ok.
--Is this the series that I use for the @dnorm business (which the user guide implies)?
--Or do I "make a residual series" (as a former post implied)?
3) Type "show @dnorm(-series)*.381" into the command window (.381 is my coefficient of interest)
4) A series window opens but then I do not know how to interpret this as the marginal effect
Don't I need to have one number at the end that implies the increased probability in LBW from and change in the dummy variable from 0 to 1?
Thanks in advance for your help.
1) Estimate my equation
2) Click Proc/Forecast-Index
--Eviews generates a series named LBWF (lbw is my dep variable) when I click ok.
--Is this the series that I use for the @dnorm business (which the user guide implies)?
--Or do I "make a residual series" (as a former post implied)?
3) Type "show @dnorm(-series)*.381" into the command window (.381 is my coefficient of interest)
4) A series window opens but then I do not know how to interpret this as the marginal effect
Don't I need to have one number at the end that implies the increased probability in LBW from and change in the dummy variable from 0 to 1?
Thanks in advance for your help.