wu :: forums - Print Page


    
      
        wu :: forums
        (http://www.ocf.berkeley.edu/~wwu/cgi-bin/yabb/YaBB.cgi)
      

        riddles >> hard >> Statistics: Model Selection Question
        
(Message started by: william wu on Oct 23^rd, 2008, 8:53pm)

Title: Statistics: Model Selection Question
Post by william wu on Oct 23^rd, 2008, 8:53pm

The following question is from Christine:

on 10/23/08 at 15:38:44, Christine wrote:

In multivariate regression, one can compare two models where one model has an additional term by using the F-test. This tests the significance of adding the extra term.

Is there a similar test to compare models of the same number of independent variables (IV), but the last IV different in each model?

R2, adjusted R2, R2 pred, and Mallow's Cp all show a difference, but do not test the significance of this difference.

Title: Re: Statistics: Model Selection Question
Post by friendly on Oct 23^rd, 2008, 9:00pm

You can try Vuong's non nested test. There is a R function for that.
http://rss.acs.unt.edu/Rdoc/library/pscl/html/vuong.html

Plus, why do you need to know if one model is significantly better than the other? Usually, if one model has higher adjusted R^2, or any of the statistics, then you should choose that model to do prediction, right?