I have an assignment that gives two ordinal variables: Education (1-poor, 2-medium, 3-high), Wealth (1-poor, 2-medium, 3-rich). The chi-square test rejects the independence of these 2 variables, so my intuition says I should combine them to avoid multicollinearity.

Is there any suggestion to do this? Any input would be greatly appreciated!

  • 1
    $\begingroup$ Multicollinearity is not necessarily a problem, depending on your goal. You should probably give us more information about the question you want to answer ultimately, and about the model where you want to include these variables. Related question: stats.stackexchange.com/questions/255343/… $\endgroup$
    – J-J-J
    Commented Dec 12, 2022 at 4:59


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.