I am using the GVIF^(1/(2df)) method in my analyses to check for multicollinearity of my (mainly) categorical variables. However, I am struggling with the cut-off values. For the 'regular' VIF several cut-offs have been described in the literature, so for referencing purposes I would like to convert the GVIF^(1/(2df)) values to VIF values.

In a previous post regarding multicollinearity (Which variance inflation factor should I be using: $\text{GVIF}$ or $\text{GVIF}^{1/(2\cdot\text{df})}$?) I read the following comment:

"If we then simply apply the same standard rules of thumb for GVIFˆ(1/(2*Df)) values as recommended in literature for the VIF, we simply need to square GVIFˆ(1/(2*Df))."

Is it really that easy or are there precautionary measures to be taken? Can anyone provide me with the mathematical reasoning for this (I'm not a mathematician), or with a reference?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.