Re: st: VIF after -ologit- command

[Richard Williams <richardwilliams.ndu@gmail.com>]

At 04:52 PM 3/8/2011, Weichle, Thomas wrote:
> Hi Statalisters,
> It appears as if the VIF command does not work after running an ordered
> logistic regression model (-ologit- command).  In the documentation, I
> found the following:  "In our parameterization, no constant appears,
> because the effect is absorbed into the cutpoints."  Does this help
> explain why the VIF may not be valid after running the -ologit- command?

We've had variations of this discussion many times on Statalist. 
Usually (always?) multicollinearity involves the right-hand side of 
the equation (the Xs) and how you test for it isn't much affected by 
whether you are using OLS or logit or ologit or whatever. My guess is 
that -vif- only works after -reg- because other commands don't store 
the necessary information, not because it isn't valid. You could just 
"cheat" and run reg followed by vif even if your dv is ordinal. Or, 
you could download UCLA's -collin- command and use it.  For example,

use "http://www.indiana.edu/~jslsoc/stata/spex_data/ordwarm2.dta";, clear
ologit warm yr89 male white age ed prst
collin  yr89 male white age ed prst if e(sample)

I didn't need to run the -ologit- command first. I could have also just said

use "http://www.indiana.edu/~jslsoc/stata/spex_data/ordwarm2.dta";, clear
collin  yr89 male white age ed prst if !missing(warm)

In other words, when using collin, you just want to make sure that 
the sample selection is the same as what you would get in your logit 
or ologit or ols or whatever it is you are doing analysis.


-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
HOME:   (574)289-5227
EMAIL:  Richard.A.Williams.5@ND.Edu
WWW:    http://www.nd.edu/~rwilliam

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/