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

Re: st: dropped in regression analysis


From   Toyoto Iwata <iwata@med.akita-u.ac.jp>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: dropped in regression analysis
Date   Sat, 14 Aug 2004 13:33:46 +0900

Dear Clive Nicholas,

In the archives, I found a post which seems relevant.

http://www.stata.com/statalist/archive/2003-07/msg00193.html


A portion of your mail (At 03:52 04/08/14 +0100) says, 
> 
> >From my own practical experience, variables tend to be dropped for three
> main reasons:
> 
> (a) they are co-linear with one or more other independent variable(s);
> (b) they are one of a set dummy variables, one of which has to be left out
>     to serve as a comparison group; or
> (c) they simply don't vary enough across observations to be variables!
> 
> Others, I'm sure, could give more precise, mathematical explanations as to
> why variables are dropped from regressions; and, indeed, there may be
> other good commonsense reasons. But these, to me, are the most common.
> 

Anyway, impressive and/or practically thankful mails seem to 
be classified into:
(a) deep information from the company staffs
(b) conversations between expert statisticians (most of these
    I cannot understand, though)
(c) answers of the Stata gurus to questions of beginners.

I really thank to the statalist and contributers for this, 
and hope such mails come out further.   

Toyoto

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index