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

st: re: prob with xttest2

From   Kit Baum <>
Subject   st: re: prob with xttest2
Date   Sun, 2 Sep 2007 11:11:51 -0400

Abhijit reports

the xttest2 command for cross-sectional correlation is generating the
following reply

Correlation matrix of residuals is singular.
not possible with test

The xttest2 command is looking at the correlation matrix of the residuals, which is N x N, where there are N groups of T observations in the panel. Abhijit says "in my case N=85, T=8." The rank of that correlation matrix thus cannot exceed 8 (by definition, there can be no more than 8 linearly independent vectors in the 85 x 85 matrix). As in the case of -sureg-, which uses this correlation matrix in the "Zellner step", the test cannot be performed for panels with N > T.

Kit Baum, Boston College Economics and DIW Berlin
An Introduction to Modern Econometrics Using Stata:

* For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index