Statalist The Stata Listserver

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

RE: st: frontier models

From   "Gauri Khanna" <>
Subject   RE: st: frontier models
Date   Tue, 28 Nov 2006 08:29:19 +0000

Dear George,

One of the reasons for choosing alternate error distributions is to check how robust your results are. Also, if you flip through Lovell and Kumbhakar's book (2000) "Stochastic Frontier Analysis" you will read that they state that qualititative estimates are not sensitive to the type of distributional assumptions made, however quantitative estimates are but the rankings do not change.

Half normal is widely used in the literature. When I used frontier, I had problems of convergence when using the exponential form. Half normal behaved better.

Hope this helps.


(Ms) Gauri
PhD Candidate
Dept. of Economics
Graduate Institute of International Studies

From: george owuor <>
To: statalist <>
Subject: st: frontier models
Date: Tue, 28 Nov 2006 00:06:26 -0800 (PST)

Dear all,

I am new with frontier. Yet I can see there are three oprions. My model is a production function. What are the key salient reasons for using noemal, exponential or truncated. Please assist.

I am interested to measure inefficiency covariates too.


Do you Yahoo!?
Everyone is raving about the all-new Yahoo! Mail beta.

* For searches and help try:
Express yourself instantly with MSN Messenger! Download today it's FREE!

* For searches and help try:

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