I don't know what other listservers you
follow, but we had one such thread recently
on Statalist, starting a week ago, and the topic is recurrent.
In most cases, each thread is a constant (reference
to the Bill Sribney FAQ and the Frank Harrell book)
plus a variable (e.g. Ron�n Conroy's spirited
reply last Tuesday).
If more people read the archives, as you suggest,
and also the FAQ, then such threads might not occur:
a mixed blessing given the animosity which stepwise
procedures attract and the cathartic effect of fulminating
against them.
Nick
[email protected]
[email protected]
> I forget if it was on this listserver or another, but there was an
> interesting thread recently on the problems with automated regression
> procedures, such as best subsets and stepwise. You may want
> to look in the
> archives to see if it was on this list.
*
* 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/
