-findit collin-
It has VIF, condition number, etc. Often just looking at the correlation matrix or a matrix scatterplot helps, but no magic bullet exists for this problem.
-----Original Message-----
From: "Steven Samuels" <[email protected]>
To: [email protected]
Sent: 12/10/2008 10:27 AM
Subject: Re: st: RE: Collinearity
Regress x on the other covariates and you may detect which ones are
collinear. More than one may be implicated, especially if you have
one too many dummy variables, as Martin suggested.
-Steve
>
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of Ismail
> Ait Saadi
>
> I am running a Logit model with 12 variables, the results says that
> variable
> x was dropped because of collinearity. is there any command that
> can help
> identify the other variable with wich x is correlated, the reason I am
> asking for this is that x is an important variable and I prefer to
> drop
> another variable and keep x. I remember doing this long time ago but I
> forgot the command.
>
*
* 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/
*
* 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/
