Panel data modeling in SFA

[Anam nabeel]

Dear All,

i am trying to estimate the random effect (pitt and lee 1981) and true random effect (Greene 2005) model in SFA with weakly balanced panel data. but when i add any dummy variable stata gives only coefficients without any standard error or probability value.

further i tried to run GLS model with xtreg y x1 x2 x3, re, stata gave error message (insufficient observation) but there are 15000 observations.

can anyone help me in this regard? any guidance will be highly appreciated.

Regards

[Clyde Schechter]

Please read the Forum FAQ for excellent advice about how to post questions that are likely to get a timely and helpful response. Among the things you will learn there:

1. Don't use jargon or abbreviations unless they would automatically be understood by anybody with a college education. This is a multi-disciplinary, international forum. Maybe everybody in your circle knows what SFA is, but many of us out here don't. So you are unnecessarily limiting the people who might respond.

2. Similarly, author and date citations are not useful here. Maybe Pitt and Lee 1981 is a classic in your field and everybody immediately recognizes it. The rest of us do not. If it is important for understanding your Stata problem, then provide a full reference that people can track down, or better still, provide a URL to an ungated version of it if possible. If it isn't important for understanding your Stata problem, don't provide references at all.

3. Just saying that Stata gave an error message and a description of your data set is not sufficient for somebody to troubleshoot from afar. If you have code that is not doing what you want, you need to include in your post:

A. The exact code that is giving you difficulty. Do that by copying from your do-file, log-file or Results window directly into the Forum's editor, and surround it by code delimiters (see FAQ #12 for an explanation of these) so it is maximally readable. And don't edit it in any way. Show it exactly as it ran.

B. The output (including any error messages), and, again, do not edit it in any way, and enclose it in code delimiters.

C. An example of your data, using the -dataex- command so that others can work with it to replicate, and find a solution to, your problem.