RE: st: Chow Test Statistic

[Ian Li <ian.li@uwa.edu.au>]

Hi Richard,

thanks, that's very helpful!

Cheers
Ian


________________________________________
From: owner-statalist@hsphsun2.harvard.edu [owner-statalist@hsphsun2.harvard.edu] On Behalf Of Richard Williams [richardwilliams.ndu@gmail.com]
Sent: Friday, June 17, 2011 11:27 PM
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: Chow Test Statistic

At 12:56 AM 6/17/2011, Ian Li wrote:
>Dear Statalisters,
>
>I am analysing a linear model of labour market earnings using OLS,
>which can be written as:
>
>y = a0 + B1X1+B2X2+...+BnXn+e
>
>The same equation was estimated for disaggregated sub-groups of the
>data, say, for males and females.
>
>I would like to comnpute the Chow test statistic to see if the beta
>estimates for females differ from those of males, for all the
>variables used in the analysis.
>
>I have read the FAQ on doing that:
>http://www.stata.com/support/faqs/stat/chow.html
>
>which suggests the use of interaction terms to compute the Chow test
>statistic. As I have over 70 variables in the model, this is quite a
>tedious process. Can anyone please suggest an easier way to do this?

As Maarten notes, using interactions really isn't that hard to do in Stata 11.

That same FAQ also shows you how to do a Chow test by estimating
separate regressions for each group and contrasting it with the
pooled regression. A few calculations are required but the number of
variables does not affect how complicated those calculations are.

You can also use the -suest- command. Here is a slightly modified
version of the example that appears in the help for -suest-.

webuse income
regress inc edu exp if male
estimates store Male
regress inc edu exp if !male
estimates store Female
suest Male Female
* Test whether all coeff are same, including intercepts
test [Male_mean = Female_mean], constant coef
* Test whether all the coefficients are same, except intercepts
test [Male_mean = Female_mean], coef

When doing a Chow test, you may or may not want to impose the
constraint that the intercepts are equal. It is pretty common to
allow them to be different while all other coefficients are
constrained to be the same.



-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
HOME:   (574)289-5227
EMAIL:  Richard.A.Williams.5@ND.Edu
WWW:    http://www.nd.edu/~rwilliam

*
*   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/