Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: RE: Chow Test and SSE when using ``pweight``

From   "John GIBSON" <>
To   <>
Subject   Re: st: RE: Chow Test and SSE when using ``pweight``
Date   Wed, 13 Nov 2002 19:49:26 +1300


If you are using -pweight- I guess you have got sample survey data, so there is a high chance of it being heteroscedastic. In this case you should be using Wald rather than LR-type tests, so trying to run it off the SSE of two separate models (even if available) wouldn't be a good idea.

I am puzzled why it is proving hard to use xi: with the pooling dummy interacted with each variable. I have done this with models with more than 50 variables and Stata has coped just fine.


<<< "Jennifer Ng" <> 11/13  6:48p >>>
For the chow test, may be you can run the regression for those two groups
separately and then use their SSR to calculate the F-value.

-----Original Message-----
[]On Behalf Of Xiaoqu Luo
Sent: Tuesday, November 12, 2002 9:29 PM
Subject: st: Chow Test and SSE when using "pweight"

Is there any easy way in STATA to do a Chow Test (when run the same
regression model for two groups, test the difference between the
coefficients for the same independent variable)? I know that I can use
interaction terms with the pooled model but there are many independent
variables and this is not very easy to do. In addition, when I use "pweight"
option with "regress" commond, I cannot get the small table for the summary
of the errors (such as SSE, MSE, etc.). How can I get this table, or I
cannot get them by using "pweight"? A bit confused.

Thanks a lot for your help!

Xiaoqu Luo
Surfy!  Great web search, free web email, and $9.95
unlimited Internet access

Powered by Outblaze
*   For searches and help try:

*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index