Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: testing across different sub-samples


From   "Francesco Trivieri" <francesco.trivieri@unical.it>
To   <statalist@hsphsun2.harvard.edu>
Subject   Re: st: testing across different sub-samples
Date   Fri, 25 Sep 2009 11:58:58 +0200

Dear Martin and John,

thank you very much indeed for your precious suggestions and help.

All the best
Francesco


----- Original Message ----- From: "John Antonakis" <john.antonakis@unil.ch>
To: <statalist@hsphsun2.harvard.edu>
Sent: Thursday, September 24, 2009 11:39 PM
Subject: Re: st: testing across different sub-samples


Right. The random effect, the Uj term, is unobserved.

In any case, it is safe to go with FE which is sure to be consistent.

Best,
J.

____________________________________________________

Prof. John Antonakis
Associate Dean Faculty of Business and Economics
University of Lausanne
Internef #618
CH-1015 Lausanne-Dorigny
Switzerland

Tel ++41 (0)21 692-3438
Fax ++41 (0)21 692-3305

Faculty page:
http://www.hec.unil.ch/people/jantonakis&cl=en

Personal page:
http://www.hec.unil.ch/jantonakis
____________________________________________________



On 24.09.2009 23:17, Martin Weiss wrote:
<>

Yes, I probably got confused b/c this thread has meandered from "RE vs. FE"
to "A vs. B" and back again. The strategy you are recommending is only
applicable if Francesco wants to stick to FE, right? For RE, there is no
convenient representation with dummies, is there?


HTH
Martin


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of John Antonakis
Sent: Donnerstag, 24. September 2009 22:33
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: testing across different sub-samples

Hi Martin:

I thought that he could estimate this with dummy fixed effects using -reg-

So, he simply does:

reg y x1 x2 x3 x4 x5 i.id if group ==0
est store zero
reg y x1 x2 x3 x4 x5 i.id if group ==1
est store one
suest zero one
test [zero_mean = one_mean]

Suest should be able to handle the "i.id" fixed effects, no?

Best,
J.

____________________________________________________

Prof. John Antonakis
Associate Dean Faculty of Business and Economics
University of Lausanne
Internef #618
CH-1015 Lausanne-Dorigny
Switzerland

Tel ++41 (0)21 692-3438
Fax ++41 (0)21 692-3305

Faculty page:
http://www.hec.unil.ch/people/jantonakis&cl=en

Personal page:
http://www.hec.unil.ch/jantonakis
____________________________________________________



On 24.09.2009 18:06, Martin Weiss wrote:

<>

"There is a quick way to do this using -suest-"

Can you give an example? -suest- is not available after -xt- commands...

HTH
Martin

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of John Antonakis
Sent: Donnerstag, 24. September 2009 17:46
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: testing across different sub-samples

Hi Francesco:

What you need to do is a Chow test--i.e., test whether the set of coefficients differ across samples.

There is a quick way to do this using -suest-

Model the fixed effects with dummies, if you have any.

HTH,
John.

____________________________________________________

Prof. John Antonakis
Associate Dean Faculty of Business and Economics
University of Lausanne
Internef #618
CH-1015 Lausanne-Dorigny
Switzerland

Tel ++41 (0)21 692-3438
Fax ++41 (0)21 692-3305

Faculty page:
http://www.hec.unil.ch/people/jantonakis&cl=en

Personal page:
http://www.hec.unil.ch/jantonakis
____________________________________________________



On 24.09.2009 15:08, Francesco Trivieri wrote:

Dear Martin,

thank you again for your answer. I thought about using a Hausman test, but I couldn't figure out how to actually implement it in stata for my case. Please find below a more detailed description of my problem:


I am estimating the following model

xtreg y b1x1 b2x2 b2x3 b4x4 b5x5 (fe or re)

on two different sub-samples (A and B), which can be individuated by a dummy D. In both cases, I compute an index given by the sum of the first three coefficients (b1+b2+b3). Then I need to assess whether the two indexes from the two sub-samples are statistically different. In other words, I need to test whether (b1+b2+b3) in A = ( b1+b2+b3) in B.

Best regards
Francesco


*
*   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/


*
*   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/


*
*   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/

*
*   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