Having ruled out significant colliearity, I should
also mention that Harrell (2001) provides some useful
guidelines for variable selection:
1. Use the literature to eliminate unimportant
variables.
2. Eliminate variables having narrow distributions.
3. Eliminate variable having a lot of missing data.
4. Consider a type of oblique-rotation principal
components analysis.
SR Millis
Harrell, F. (2001). Regression modeling strategies:
With applications to linear models, logistic
regression, and survival analysis. New York:
Springer-Verlag.
--- SR Millis <[email protected]> wrote:
> Thanks, George. Regarding collinearity, what was the
> highest condition index? And, for any condition
> index
> greater than 30, were there any large
> variance-decomposition proportions (ie, greater than
> .5)?
>
> SR Millis
>
>
> --- georg wernicke <[email protected]>
> wrote:
>
> > hello,
> >
> > my sample size are 874 observation. i used the
> > comand collin and
> > coldiag2 for diagnostics. all the vifs were way
> > below 10, the highest
> > was 6.5.
> >
> > thanks again,
> >
> > georg
>
>
> Scott R Millis, PhD, MEd, ABPP (CN,CL,RP), CStat
> Professor & Director of Research
> Dept of Physical Medicine & Rehabilitation
> Wayne State University School of Medicine
> 261 Mack Blvd
> Detroit, MI 48201
> Email: [email protected]
> Tel: 313-993-8085
> Fax: 313-966-7682
> *
> * For searches and help try:
> *
> http://www.stata.com/support/faqs/res/findit.html
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
Scott R Millis, PhD, MEd, ABPP (CN,CL,RP), CStat
Professor & Director of Research
Dept of Physical Medicine & Rehabilitation
Wayne State University School of Medicine
261 Mack Blvd
Detroit, MI 48201
Email: [email protected]
Tel: 313-993-8085
Fax: 313-966-7682
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
