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

Re: Re: st: Problems using predict in xtfrontier

From   "Dascha Orlova" <>
Subject   Re: Re: st: Problems using predict in xtfrontier
Date   Wed, 29 Aug 2007 15:37:48 +0200

>Are you sure they should vary?  I don't have books with me but isn't
>eta summed over all the time periods (for each individual) for
>- -predict var, u- and predict var, m- making the predictions time

Dear Scott,
thank you! You're right, the estimates for u really shouldn't vary if one examines the formula properly. :-)
But I still don't understand what sense time invariant u-estimates make in a time-variant model. For if I use any other way to calculate cost efficiency than -predict var, te- the efficiency won't vary with time because of the invariant u's. And if one considers the logic behind the model, the u's are the only cause of the time-variant efficiency scores...
SFA seems to be too intricate for me, sorry :-(

GMX FreeMail: 1 GB Postfach, 5 E-Mail-Adressen, 10 Free SMS.
Alle Infos und kostenlose Anmeldung:
*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index