Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • How to test the multicollinearity in generalized ordered logit?

    Hello!
    I want to run the generalized ordered logit model. Does it have a requirement for multicollinearity? If it should test multicollinearity, how to test it?
    In OLS model, we can run "estat vif " command after completing the regression that “reg y x1 x2 ”. Is it same in the generalized ordered logit model?
    Can you help me to understand it? Thank you very much.

  • #2
    Are you talking about the user-written command, gologit2? It should trap collinear predictors and drop them automatically. You wouldn't need to do anything; if it has any collinear predictors, then you'll see an informational message about the dropped variables when you fit your model.

    If for some reason you want to know what the variance inflation factors are for the set of predictors you intend to use in a model, then you can just fit a linear regression model with the predictors and an arbitrary response variable using regress and follow that with estat vif.

    Comment


    • #3
      Originally posted by Joseph Coveney View Post
      Are you talking about the user-written command, gologit2? It should trap collinear predictors and drop them automatically. You wouldn't need to do anything; if it has any collinear predictors, then you'll see an informational message about the dropped variables when you fit your model.

      If for some reason you want to know what the variance inflation factors are for the set of predictors you intend to use in a model, then you can just fit a linear regression model with the predictors and an arbitrary response variable using regress and follow that with estat vif.
      Yes, i will use the gologit2 command.
      Thank you for your help again.

      Comment

      Working...
      X