Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Vuong test for nested linear models

From   Greenfish Wu <>
To   "" <>
Subject   Re: st: Vuong test for nested linear models
Date   Sun, 9 Feb 2014 13:24:44 -0800 (PST)

Thank you very much John! That's very helpful!


On Sunday, February 9, 2014 2:35 PM, John Antonakis <> wrote:


You don't need a Vuong test here. Because you are adding one parameter, 
the test of b3 = 0 is the test you want. That is simply the test of the 
significance of the parameter estimate. This is equivalent to the nested 

sysuse auto
nestreg : reg price (mpg) (weight)

The F-change test for adding weight is equivalent to the t-test^2 of the 
parameter, i.e.,

dis (_b[weight]/_se[weight])^2

You can still use the lrtest after OLS.

reg price mpg
est store one
reg price mpg weight
est store two
lrtest one two



John Antonakis
Professor of Organizational Behavior
Director, Ph.D. Program in Management

Faculty of Business and Economics
University of Lausanne
Internef #618
CH-1015 Lausanne-Dorigny
Tel ++41 (0)21 692-3438
Fax ++41 (0)21 692-3305

Associate Editor:
The Leadership Quarterly
Organizational Research Methods

On 09.02.2014 19:33, Cindy wrote:
> Hi, do anyone know how to do Vuong test (Vuong 1989) for model 
selection of
> two nested linear models in Stata? It seems that people use Vuong test
> mostly for non-nested models. But Vuong (1989) covers both nested and
> non-nested models. Following are my two models:
> Model 1: Y=a0+a1X1+a2X2+e
> Model 2: Y=b0+b1X1+b2X2+b3X3+v
> Both models are OLS regressions. I need to test how significant the
> incremental adjusted R2 in model 2 is from adding X3 (my variable of
> interest) to model 1.
> Also, it seems that a likelihood ratio test could be used to test nested
> models, too. But when I looked up the code in stata (lrtest), it says 
> it’s applicable for maximum likelihood models only.
> Thank you very much!!
> --
> View this message in context:
> Sent from the Statalist mailing list archive at
> *
> *   For searches and help try:
> *
> *
> *

*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index