Re: Hausman test / fixed or random effects model

Posted by Ryan on Dec 28, 2013; 6:00am
URL: http://spssx-discussion.165.s1.nabble.com/Hausman-test-fixed-or-random-effects-model-tp5723718p5723721.html

From what I have read, the Hausman test appears to be about testing the equality of within and between-cluster slopes. The within and between-cluster slopes can certainly be examined in SPSS, and decisions on how to proceed based on those results can be made. 

Side comment: I would not use the Hausman test to determine whether to estimate a RE or FE model. You might consider reading "Multilevel Analysis: An Introduction to Basic and Advanced Multilevel Modeling" (2nd edition) by Snijders and Bosker for clarity on this issue.

Ryan

On Thu, Dec 26, 2013 at 5:03 AM, SPSS_beginner <[hidden email]> wrote:

Hey there,

I would like to implement the Hausman test in SPSS in order to decide which
model to use for my panel data. I understand what the Hausman test does and
I assume that a random effects model will be more appropriate for my data,
but I was told to check my assumptions with the Hausman test. I was also
told to use SPSS- but now I cannot find any information about how to do a
Hausman test in SPSS.
Also is it complicated to analyse panel data with both models in SPSS?
Everytime I google this topic only answers concerning how to do it in STATA
or R come up- this somehow worries me...
I hope somebody can help me- I will be very very thankful for any hint!!!



--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Hausman-test-fixed-or-random-effects-model-tp5723718.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD