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

Re: st: TSCS estimator selection for unequally spaced data

From   David Jacobs <>
Subject   Re: st: TSCS estimator selection for unequally spaced data
Date   Tue, 29 Jul 2008 12:03:50 -0400

I had the same problem with an election analysis for state governores and partially solved it using population averaged models.

Look up -xtgee- models. You also will get the population averaged equivalent of random effects with a robust correction by simply typing ".xtreg depvar explanatory variables, pa r". Also check out the material on random effects in the xtreg chapter. This option is outlined there.

Dave Jacobs

At 03:13 PM 7/28/2008, you wrote:

I have a small TSCS data set with 7 panels (countries) and 55 observations (elections) in total. Elections occur in different years across countries, so no observations occur in the same year. There are no missing data.

What estimator should I use? Xtpsce seems to require shared time values across the panels. Creating a fake time variable may artificially erase any contemporaneous correlation. Xtreg, i(id) re works but I don't think it allow corrections for heteroskedasticity. Any suggestions?


Kenneth F Greene
Assistant Professor
Department of Government
1 University Station - A1800
University of Texas at Austin
Austin TX 78712-0119
Tel. 512-232-7206
Fax 512-471-1061

* For searches and help try:

*   For searches and help try:

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