Mandy,
My preference for running the Hausman test in stata is to include individual means of all the time varying variables in an ols or re model. Then run a Wald test for the joint significance of the individual means. It can be shown that this is algebraically equivalent to the Hausman test. And you don't run into negative covariance problems.
So you should be able to run the identical test in SPSS.
Carl Nelson
---- Original message ----
>Date: Sun, 4 Jan 2009 13:40:13 -0500
>From: "Mandy fu" <[email protected]>
>Subject: st:Hausman test in Stata vs.SPSS
>To: [email protected]
>
>Hi all,
>
>My question is kind of weird: I was wondering if anyone knows if
>there's easy way to do Hausman test after fixed effect and random
>effect models for panel data in SPSS.
>
>I know Stata provides the easiest way to do fixed effect, random
>effect, and then Hausman test. I love this! The thing is, in a project
>using SPSS in all the previous part, I hope to see if there's any way
>to keep using SPSS for the Hausman test after FE and RE models. Any
>comments will be helpful.
>
>I have two reasons to ask statalisters this question :first, most
>people here know the advantages and disadvantages of SPSS and Stata
>well. Second, when searching SPSS forum, I find there's just one post
>about Hausman test and interestingly, someone replies to use Stata to
>do Hausman test instead.
>
>Thanks for your help! I appreciate it.
>
>Mandy
>*
>* 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/
