estimation of marginal effects with probit models
Posted: Thu Sep 30, 2010 6:22 am
Hi!
I'm making a Probit model but I've troubles with calculating the marginal effects of my variables.
I've one binary variable and the others are continous variables.
I'm using Eviews 7.1.
I've read the eviews users guide and the posts on this forum, however, I'm still not able to calculate the marginal effects.
What have I done?
Proc/Forecast(fitted probabilty/index) and then in the window that pops up, I clicked on index instead of probabiliy.
Is that correct?
Eviews makes a new series if I click on ok.
In the Eviews users guide it is said that
"the auto-series @dnorm(-xb), @dlogistic(-xb), or @dextreme(-
xb) may be multiplied by the coefficients of interest"
If I try this I will get a lot of values for every variable.
I would like to have one forecasted marginal effect for every variable I have.
Can someone help me? Thanks!
I'm making a Probit model but I've troubles with calculating the marginal effects of my variables.
I've one binary variable and the others are continous variables.
I'm using Eviews 7.1.
I've read the eviews users guide and the posts on this forum, however, I'm still not able to calculate the marginal effects.
What have I done?
Proc/Forecast(fitted probabilty/index) and then in the window that pops up, I clicked on index instead of probabiliy.
Is that correct?
Eviews makes a new series if I click on ok.
In the Eviews users guide it is said that
"the auto-series @dnorm(-xb), @dlogistic(-xb), or @dextreme(-
xb) may be multiplied by the coefficients of interest"
If I try this I will get a lot of values for every variable.
I would like to have one forecasted marginal effect for every variable I have.
Can someone help me? Thanks!