[prev in list] [next in list] [prev in thread] [next in thread] 

List:       sas-l
Subject:    Re: Polytomous regression
From:       Newton Love <newton () MICHELOB ! WUSTL ! EDU>
Date:       1996-02-18 5:14:43
[Download RAW message or body]

In article <4fgp7f$pda@mule.fhcrc.org> dmclerra@cclink.fhcrc.org (Dale
McLerran) writes:
>From: dmclerra@cclink.fhcrc.org (Dale McLerran)
>Subject: Polytomous regression
>Date: 10 Feb 1996 00:34:55 GMT

>Help!  Does anyone know how to fit a polytomous regression model in a
>nonlinear regression.  I have both nominal and ordinal responses.  It
>seems that it should be possible, but I am stumped in trying to figure
>out how to do it.  Any help would be appreciated.

Couldn't you just treat the nomials as classes, and model them with
"dummy" regerssors?  (Remeber the old earnings model)

Pay = b0 + b1*man + b2*woman + b3*years_on_job

Look it up in an econometrics book.

-newt

St. Louis UNIX Users Group: http://dark.wustl.edu/
World-wide UUGs: http://dark.wustl.edu/~newton/othr_uug.html

[prev in list] [next in list] [prev in thread] [next in thread]