Logit post estimation

[elie kallab]

Hello,

Please find attached my data set and do list.

My logit is : logit out provinceincanada ridingnumbers partyabv university visible_minority unemployment_rate

Everytime I run post estimation command like:
estat vif
rvpplot
rvfplot

It tells me that estat vif not recognized or not valid.

How can I rectify this?

Thank you!

[Clyde Schechter]

These are not valid because they only run after a linear regression, not logistic regression. If you insist on using those tools, switch to a linear probability model and estimate it with -regress-. Bear in mind, however, the assumption of homoscedasticity (which verifying graphically is a principal reason for using -rvfplot- and -rvpplot- is almost always violated in linear probability models.) And the concept is simply not relevant in logistic regression. Bear in mind also that exploring multicolinearity (the raison d'etre of -estat vif-) is basically just a waste of time anyway. See Bryan Caplan's brief explanation at https://www.econlib.org/archives/200...ollineari.html. For the longer version, see the chapter on multicolinearity in Arthur Goldberger's textbook of econometrics.

In short, I would just forget about these and move on.

Added: I strongly agree with Jared Greathouse's comments in #3.

[Jared Greathouse]

Please read the FAQ about how to ask and report on problems/questions. We ask that you don't directly post your data or do files.

EDIT: The simpler issue, upon my reading of the help, is that it only works with OLS.