In my present data, the Variance Inflation Factors suggest lack of substential multicolliniearity (<1,7). However, the condition number of 28 is almost at the critical value of 30.

How do I interpret these confusing results? What may cause the difference?

It has been suggested elsewhere that the difference can be eliminated by mean-centering. Though Belsley 1984 demonstrated that this technique does not solve but only mask collinearity problems.

Belsley (1984), Demeaning Conditioning Diagnostics through Centering, The American Statistician , 38 (2), pp. 73-77

  • $\begingroup$ If it is less than 30, you need not worry $\endgroup$ – Bach Mar 12 '16 at 16:36
  • $\begingroup$ Well, from what I read the rule of thumb says that collinearity starts to be a concern at 15 and a serious problem at 30. $\endgroup$ – mkdir Mar 12 '16 at 17:09

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.