The answer is: E(inverse te_scores)
However, please note that the output in the interval [1; infinity) is common for cost function models (see e.g. the package Frontier, which uses such interval for cost functions, and the interval [0, 1) for production functions). Also, some authors (e.g. in Research Policy 31, 109�124) inverted the program's output in order to obtain efficiencies in the interval [1; infinity)
Nicola
At 02.33 31/07/2007 -0400, "Dascha Orlova" wrote:
>Hi,
>
>I'm estimating a cost function with xtfrontier. After the estimation I predict the efficiency scores using predict te_scores,te.
>
>The predicted scores are in the intervall (1;eternity) and hard to interprete though.
>
>The subject has been discussed here already and I've read:
>http://www.stata.com/statalist/archive/2005-06/msg00578.html
>and
>http://www.stata.com/statalist/archive/2007-06/msg00926.html.
>
>However these threads don't really solve my problem: which way do I compute the MEAN efficiency to make it comparable to other studies? Invertation affects the distribution of the scores, so the results vary between:
>E(inverse te_scores)
>and
>1/(E(originally predicted te_scores))
>
>I hope, I managed to explain my problem :-)
*
* 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/
