4

I have run a Poisson glm and I want to test for multicollinearity in my data. I have used the vif in R and obtained the following result. How can I interpret this?

 GVIF Df GVIF^(1/(2*Df))

make    1.083927  1        1.041118
cc      1.093900  1        1.045897
area    1.113460  3        1.018073
insgen  1.029887  1        1.014834
insage  2.484052  2        1.255423
ageveh  1.067675  1        1.033284
marital 2.369840  1        1.539428
COOLSerdash
  • 25,317
  • 8
  • 73
  • 123
user40494
  • 425
  • 5
  • 13
  • Often, a rule-of-thumb is stated that the VIF should not exceed 5 or 10. These rules have to be adopted [with caution](http://www.nkd-group.com/ghdash/mba555/PDF/VIF%20article.pdf), however. There are other measures of multicollinearity, see [here](http://stats.stackexchange.com/questions/48822/is-there-a-reason-to-prefer-a-specific-measure-of-multicollinearity) and [here](http://stats.stackexchange.com/a/56664/21054). – COOLSerdash Feb 26 '14 at 08:51
  • 2
    The same issue is discussed here with some more conclusions: http://stats.stackexchange.com/questions/70679/which-variance-inflation-factor-should-i-be-using-textgvif-or-textgvif/96584#96584 – MsGISRocker May 06 '14 at 12:07
  • I have the same question! Could someone ask? Tks –  Jan 19 '15 at 11:02

0 Answers0