st: Statistical insignificance of sigma2 using xtfrontier

["Dunki Alexander" <Alexander.Dunki@caa.co.uk>]

Hello,

I have a general question about the output from the xtfrontier command
used for Stochastic Frontier Analysis.

The output includes estimates of the variances of composite error
(sigma2), the inefficiency error (sigma u2) and the random error term.
STATA also provides the standard error of this estimate. 

It happens that the variance estimate is not statistically significant.
For example, the estimate may be 2 and the standard error of the
estimate may be 4. 
Would the fact that the estimate is statistically insignificant affect
the accuracy of the efficiency scores obtained using "predict var, te"?

I would have thought that having statiscally insignificant variance
estimates would be detrimental and give efficiency scores which are not
as accurate as those from a model which has a highly statistically
significant variance estimate. What is the correct interpretation?

Thanks,

Alexander

**********************************************************************
Before printing consider the environment.

This e-mail and any attachment(s) are for authorised use by the intended recipient(s) only. It may contain proprietary material, confidential information and/or be subject to legal privilege. If you are not an intended recipient then please promptly delete this e-mail, as well as any associated attachment(s) and inform the sender. It should not be copied, disclosed to, retained or used by, any other party.
Thank you.

Please note that all e-mail messages sent to the Civil Aviation Authority are subject to monitoring / interception for lawful business **********************************************************************

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/