Date:         Mon, 27 Dec 1999 16:10:29 -0500
Reply-To:     Paige Miller <paige.miller@KODAK.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Paige Miller <paige.miller@KODAK.COM>
Organization: Eastman Kodak Company
Subject:      Re: a question in regression
Content-Type: text/plain; charset=us-ascii

Kattamuri.Sarma@RESPONSEINSURANCE.COM wrote:
>
> Checking the derivatives is only one of the many things you have to do in
> verifying
>    the model, especially if you are fitting a complex model.

Only in certain situations, and certainly not in the wide range of
regression usage. You can only check the derivative for reasonableness
if you know what that derivative should be. In many applications, you
don't know what the derivative should be, there is no fundamental theory
to tell you what range the model coefficients or the derivative should
be.

--
Paige Miller
Eastman Kodak Company
paige.miller@kodak.com
"It's nothing until I call it!" -- Bill Klem, NL Umpire