Date: Thu, 12 Aug 2010 08:22:37 -0700
Reply-To: haiyan chen <haiyan_c@YAHOO.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: haiyan chen <haiyan_c@YAHOO.COM>
Subject: Re: Calculating Incident Rates
In-Reply-To: <201008121415.o7CAki0F026106@willow.cc.uga.edu>
Content-Type: text/plain; charset=iso-8859-1
you have a binomial distribution B(n=4, p=10%); your researcher is interested in p(x=2)={4!/(2!*2!)]*square(10%)*square(90%)=0.486.
--- On Thu, 8/12/10, Billy Thompson <bill.thompson@BROOKS.AF.MIL> wrote:
From: Billy Thompson <bill.thompson@BROOKS.AF.MIL>
Subject: Calculating Incident Rates
To: SAS-L@LISTSERV.UGA.EDU
Date: Thursday, August 12, 2010, 10:15 AM
I researcher approached me with the following question...
We have 4 patients with an event. 2 of the patients have an
abnormal MRI, 2 are normal. The normal population incidence of MRI
abnormalities is 10%.
The question is what is the likelihood of 2 of 4 having the MRI
abnormality, given the population likelihood of 10%, and what are the
confidence intervals.
Another colleague and I feel that CI's are not appropriate and that there
are pieces of the puzzle missing, however I'm always open to suggestions
and the opportunity to learn.
Any help is greatly appreciated.
