Ordered Probit - Partial Effects

[luigi1010]

Hello,

A question regarding generating the partial effects in an ordered probit model.

The dependent variable in my regression is ordinal and can take values 0, 1, 2, 3.

I have successfully estimated the equation and obtained the coefficients and limit points (threshold values)... However I am a little bit stuck when it comes to calculating the partial effects of each independent variable at each level of the dependent variable.

I understand that in a typical binary model, one can use the Proc > Forecast (Fitted Probability/Index) and use the output from that (call it dependentvariablef) to generate a new series with the command genr scalefactor=@dnorm(-dependentvariablef). The average of this newly generated series multiplied by a particular coefficient will give the partial effect of that coefficient.

My question is how one replicates this process in Eviews with an ordinal probit model.

Any help is greatly appreciated.

Many thanks,

Louis.

[luigi1010]

I forgot to mention - I'm running Eviews6.