Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: model identification problem

From   Robert A Yaffee <[email protected]>
To   [email protected]
Subject   Re: st: model identification problem
Date   Sun, 31 Oct 2004 20:55:00 -0500

  If you have a simulataneous equation system, you need to test identification with the order or rank condition.  The order condition is necessary for identification, whereas the rank condition is necessary and sufficient for identification. According to Maddala, G.S., if you have g endogenous variables in the system and k = number of (endogenous and exogenous) variables missing from the equation, then for exact identification you need
k=g-1. If k < g-1, then your equation is not identified. 
  The rank condition procedure can be found in G.S. Maddala on pp.350-354.  See references to order condition on p.347. That is, Maddala, G.S.(2001), Introduction to Econometrics, Third Ed., Wiley:NY, pp.347-354.
   Bob Yaffee

2100 Linwood Avenue
Apt 19-W
Fort Lee, NJ
Phone: 201-242-3824
Fax: 201-242-3825
[email protected]

----- Original Message -----
From: David Han <[email protected]>
Date: Sunday, October 31, 2004 7:57 pm
Subject: st: model identification problem

> Hi, all,
> I use -ml- to estimate my model  and the error information is that my
> model is possibly not identified. How can I check the identifiability
> of my model? Thanks a lot!
> Dave
> *
> *   For searches and help try:
> *
> *
> *
n:Yaffee;Robert A
fn:Robert A Yaffee, Ph.D.
org:Research/Statistical Consulting;
adr:;;2100 Linwood Avenue,  Apt 19-W;Fort Lee;N.J.;07024;USA
email;internet:[email protected]

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