In article <ae48fj$njp$[EMAIL PROTECTED]>, Huxley <[EMAIL PROTECTED]> wrote:
>It common knowledge, that for linear function f(x) we can write logit
>trnsformation geting logistics regression: (for binary response)
>logit(p)=log(p/(1-p))=f(x)
>then p=exp[f(x)]/{1+exp[f(x)]}
>For me it's reasonable to use nonlinear function i.e. f(x) is a nonlinear
>function. Am I right.
>If so, could you give me some references to justifice my approach.
>I would appreciate any help
>Huxley

The probability model comes from the user, not from statistics.

If the user's model has a nonlinear function in logistic
regression, it should be there.  It is not the function of
the statistician to make assumptions, although the
statistician can point out that some assumptions make little
difference, and that in other situations, the user must make
assumptions, as they have much effect on the result.


-- 
This address is for information only.  I do not claim that these views
are those of the Statistics Department or of Purdue University.
Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907-1399
[EMAIL PROTECTED]         Phone: (765)494-6054   FAX: (765)494-0558
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to