[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

From |
"V. Sarafidis" <vs242@cam.ac.uk> |

To |
statalist@hsphsun2.harvard.edu |

Subject |
Re: st: Test for cross-sectional correlation with small T |

Date |
03 Aug 2005 12:08:38 +0100 |

Dear Niklas,

Yes, it is possible but you need to construct the correlation matrix by yourself. One way to do this is to estimate your model, save the residuals as separate variables and then use "matcorr" to construct the correlation matrix (of course, there might be an easier way to do this). Then sum up the elements of this matrix and proceed as shown in the paper.

One important thing is that this test will not reject the null hypothesis of no cross-sectional dependence if the correlation coefficients sum up close to zero. In other words, even if you have large cross-sectional dependence but the (large) off-diagonal elements of the correlation matrix have the opposite sign, the test will not be able to reject the null.

An alternative is to use Frees' test (Journal of Econometrics, 1995). Contrary to the previous test, this one tends to over-reject the null.

Hope that helps.

Vasilis

On Aug 3 2005, Niklas Hoyer wrote:

Dear users,

I tried to use the xttest2 after xtreg y x1 x2 ..., fe or xtgls y x1 x2....

to check for cross-sectional correlation.

Yet, I obtain the error message: "Correlation matrix of residuals is singular. not possible with test"

The problems seems to be, that the number of households in my samples is bigger than the number of periodes. I found an article by M. Hashem Pesaran "General Diagnostic Tests for Cross Section Dependence in Panels", University of Cambridge & USC, June 2004, on this subject. They seem to develop a test that can be used in this case. My question: Is there any possibility to use it in stata? How?

Any help would be greatly appreciated. Thank you.

Niklas Hoyer

-------------------

student at the Ecole Nationale de la Statistique et de l'Administration Economique (ENSAE) in Paris student at the Humboldt-Universität zu Berlin

______________________________________________________________

Verschicken Sie romantische, coole und witzige Bilder per SMS!

Jetzt bei WEB.DE FreeMail: http://f.web.de/?mc=021193

*

* For searches and help try:

* http://www.stata.com/support/faqs/res/findit.html

* http://www.stata.com/support/statalist/faq

* http://www.ats.ucla.edu/stat/stata/

-- If I could get my membership fee back, I'd resign from the human race. Fred Allen (1894 - 1956) * * For searches and help try: * http://www.stata.com/support/faqs/res/findit.html * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

**References**:**st: Test for cross-sectional correlation with small T***From:*Niklas Hoyer <removed>

- Prev by Date:
**st: Test for cross-sectional correlation with small T** - Next by Date:
**st: Generated regressor problem** - Previous by thread:
**st: Test for cross-sectional correlation with small T** - Next by thread:
**st: Generated regressor problem** - Index(es):

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