Date:         Thu, 9 Dec 2004 15:27:47 -0500
Reply-To:     Jay Weedon <jweedon@EARTHLINK.NET>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Jay Weedon <jweedon@EARTHLINK.NET>
Organization: http://newsguy.com
Subject:      Re: Exact one-way goodness of fit test?
Content-Type: text/plain; charset=us-ascii

On Tue, 23 Nov 2004 11:59:53 -0500, Jay Weedon <jweedon@earthlink.net>
wrote:

>Hi folks,
>
>PROC FREQ offers a one-way chi-square g.o.f. test. With an EXACT
>statement you get an exact p-value as well as the usual chi-square
>distribution-based value. SAS doesn't seem to document the exact test,
>but I presume it's a generalization of the binomial test, based on the
>hypergeometric distribution.
>
>The issue for me is that I can't figure out how to get an exact
>p-value in the presence of a sampling zero. Can anyone either
>
>a. demonstrate how to get SAS to do this, or
>
>b. point me to documentation of such a test that would be sufficient
>for me to write some code to generate the p-value for myself.

Just found out that zero weights are now (Release 9) usable in tests
if you specify /ZEROS in the WEIGHT statement.

JW