Statalist The Stata Listserver

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

st: RE: xtfrontier - time dummies

From   "Nick Cox" <[email protected]>
To   <[email protected]>
Subject   st: RE: xtfrontier - time dummies
Date   Thu, 2 Nov 2006 14:58:34 -0000

This question was also asked on Tuesday and no one
answered, I guess because people thought it wasn't
very clear or easy to answer. The same question is 
not going to be any easier second time round. 

-tsset- in Stata has the job of telling Stata 
that your data are time series data, in your case 
panel data. This permits various basic things 
such as the use of time series operators and
keeping your data in appropriate sort order. 

I don't know what you mean by capturing the time effect, 
but I suspect you are over-estimating quite how much
is done by -tsset-. It doesn't make any dataset closer
in form or content to that suitable for any particular model. 

Whether time dummies are best for you presumably 
depends on (a) some economic or econometric desiderata 
and (b) your dataset. Others may be able to comment on

This won't help much, except perhaps by signalling that 
you probably need to say much more to elicit a better answer. 

[email protected] 

> I am a new user of STATA. I am using xtfrontier for an 
> unbalanced panel data 
> of 50 firms over a 20 years time span. The questions worrying me are:
>  Does tsset is sufficient enough to capture time effect or do 
> I need to 
> define time dummies? How I can get efficiency scores (I am 
> trying predict, 
> te)?
>  My data set is in this format:
>  Firms Year    Variables
>  1       1985
>  1       1986
>  1       2004
>  2       1985
>  2       1986
>  2       2004

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index