Date: Wed, 5 Sep 2007 07:28:30 -0700
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Shiling Zhang <shiling99@YAHOO.COM>
Subject: Re: a bounded logit model
Content-Type: text/plain; charset="gb2312"
On Sep 1, 1:11 am, WangGuang...@gmail.com wrote:
> I meet a difficulty when i want to fit a bounded logit model,p(y=1|
> X)=¦Ø(exp(X¦Â))/(1+exp(X¦Â),where ¦Ø is also a parameter, representing the
> ceiling of p.and when ¦Ø=1,the model is a plain logit model.
> The question is how can i estimate ¦Ø, as well as ¦Â£¬with SAS?
> Thanks all
It seems to me you did not tell the story about why/how ¦Ø is defined.
If you could provide more details, I am sure someone on the list can