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

RE: st: DW stats for pooled data

From   Maarten buis <>
To   stata list <>
Subject   RE: st: DW stats for pooled data
Date   Mon, 14 Dec 2009 14:13:23 -0800 (PST)

-- Anita Sayal wrote:
> Could someone please tell how does one detect autocorrelation problem
> in pooled observations.

Stata will automatically detect (perfect) autocorrelation. You will see
that as it drops one of the variables. As it is perfect autocorrelation,
it does not matter which one; the variables are indistinguishable anyhow.
There are people who worry about high but non-perfect correlation between
covariates. If you are one of those, you can look at it by displaying a 
correlation matrix with -corr-. However, it is good to remember that such 
non-perfect correlations are not a problem, the only possible concern 
could be that such variables will eat a lot of statistical power if the
correlations are extremely high, but that is exactly as it should be. 

Hope this helps,

Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen


*   For searches and help try:

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