Multicollinearity categorical variables

[Rachel Vos]

Good afternoon!

I'm assessing whether my data is collinear or not. I'm familiar with vif for most of my data, however, I have the category "Industry" as well, which is categorical, consisting of 9 categories. What is the best way to test whether those are collinear?

Thanks in advance!

[Carlo Lazzaro]

Rachel:
before the "best way to test" comes "the best way to share" waht you typed and what Stata gave you back. As per FAQ, it requires CODE delimiters. Thanks.
If not perfect, as reminded many times on ths forum, multicollinearity is rarely an issue (see https://www.hup.harvard.edu/catalog....=9780674175440, Chapter 23).

[daniel klein]

Paul Allison's blog entry on the topic might be useful; especially in the 3rd case.


Edit: Allison still acts as if multicollinearity was a problem; I agree with Carlo in that it is not. The consequences of multicollinearity are accurately reflected in the standard errors so no additional test are needed.