Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: multicollinearity and heteroskedasticty in panel models


From   Steven Samuels <sjhsamuels@earthlink.net>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: multicollinearity and heteroskedasticty in panel models
Date   Fri, 29 Aug 2008 16:47:44 -0400

Welcome, Lamia! Be sure to read the Statalist FAQ, especially section 3. You will increase your chance of a helpful response if you give details of your specific problem. For example: what kind of design generated your data? A complex survey? What is the purpose of your analysis--estimating fixed effects of particular predictors with others in your model? estimating components of variance? both? ? What are your outcomes? Is there any particular analysis you've seen that you would like Stata to duplicate?

I don't think that anyone can answer your question about heteroskedasticity without such details. Meanwhile, "search heteroskedastic, all" may lead you to some useful information. And, a google search with terms "statalist heteroskedastic" will turn up previous posts.

Collinearity applies to fixed effects, not random ones, so you can discover collinear variables by running -reg- on your fixed covariates and then -estat vif- afterwards. You "account" for it as you would in ordinary regression-opinions and options differ and there is no automatic method. If your program does not drop variables because of multicollinearity, many people would do nothing. Type "search collinear, all" to see some helful resources. A google search on "statalist collinearity" or "statalist multicollinearity" will show you previous posts on the topic.

-Regards,

Steven



On Aug 29, 2008, at 3:35 PM, lamia chourou wrote:


Dear all,
I am new user of STATA. I have a random effect model and I'd like to test the presence of multicollinearity and heteroskedasticty and to account for them. Is there a STATA command that do the job?

Thanks a lot



______________________________________________________________________ _______
Envoyez avec Yahoo! Mail. Une boite mail plus intelligente http:// mail.yahoo.fr

*
* 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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index