# Does Multicollinearity between categorical variables affect the predictions in logistic regression?

How do we check for multicollinearity between categorical variables? Does multicollinearity between categorical variables make the logistic regression model unstable?

Per wikipedia:

Multicollinearity does not reduce the predictive power or reliability of the model as a whole, at least within the sample data set

Prediction will not be affected, even for a logistic regression model.

Multicollinearity will make the model coefficients unstable - the standard errors will be greater in the presence of multicollinearity. See here.

You can still use the VIF to assess multicollinearity among categorical variables. the vif function in the car package in r will do this (although, I'm not sure how it works with categorical variables).

Condition Indices can also be used to assess multicollinearity. See here and here.

• Coefficient estimates i logistic regression are not unbiased – kjetil b halvorsen Sep 6 '17 at 17:54