re: st: panel data

[Buzz Burhans <wsb2@cornell.edu>]

A thought for Stata Corp:

Judging by the frequency a similar question is asked, and having been there 
myself, Devendra's question is a very common one when one first begins 
working with panel data in Stata. It would be a useful, and probably also a 
worthwhile investment of resources  if the introduction section of the 
"Cross Sectional Time Series" volume of the Stata manual was expanded to 
include a discussion of management of time series data, with some 
examples.  In addition, some of the common questions about examining 
various issues related to ts data would benefit from a greater exposition 
there..for instance, examining panel level heteroskedasticity, or within 
panel autocorrelation of errors.  These are common issues, they come up 
frequently on the list, and I would guess that they consume a fair amount 
of tech support time as well.

Additionally, perhaps an exposition of  such points would be worth 
considering for a NASUG presentation, or an SJ article.

Just a thought about a seemingly common area of questions

Buzz Burhans


At 09:24 AM 2/20/2004 -0500, you wrote:

> Hi,
> I have a longitudinal (panel) data set with variables that are constant over
> time (e.g. gender). Some of the entries for an individual are missing
> (actually not entered) and I wish to fill in these missing entries.
>
> ID        Gender        Year
> 1            M                1991
> 1            .                  1992
> 1            .                  1993
> 1            M                1994
> 2            .                   1988
> 2            F                 1989
> 2            .                   1990
> 2            .                   1991
>
> Can you suggest a way to fill in these missing entries?
> Many thanks in advance.
> Devendra Amre MBBS., Ph.D.


Buzz Burhans
wsb2@cornell.edu


*
*   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/