How to calculate conditional probability after probit regression or ordered probit regression?

[Stephen GZ Chen]

Dear all,

I wanna obtain the conditional probability after running probit regression.

The model is: y = constant + z + x1 + x2 + error term
where, y = 1 or 0; z = 1 or 0; x1 and x2 are continuous variables

How do I get the conditional probabilities: prob(y = 1 | z = 1) and prob(y = 1 | z = 0)

I wanna use these two probabilities as new variables.

If y = 0, 1, 2, or 3, I run the ordered probit regression.

How do I get the conditional probabilities: prob ( y >= 1 | z = 1) and prob(y >= 1 | z = 0)

I also wanna use these two probabilities as new variables.

Best,

Chen

[Phil Bromiley]

Look in the documentation - after each routine is a section called postestimation. You want to look for the margins statement section.