Date: Tue, 3 May 2011 17:23:24 -0700
Reply-To: Mary Rosenbloom <mary.rosenbloom.sas@GMAIL.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Mary Rosenbloom <mary.rosenbloom.sas@GMAIL.COM>
Subject: Re: Significance Statistics for Tables with Small or 0 cells?
Content-Type: text/plain; charset=ISO-8859-1
I'm also wondering why Fisher's wont work. I thought Fisher's was always
appropriate. It may be underpowered, though.
On Thu, Apr 28, 2011 at 4:05 PM, Dale McLerran <firstname.lastname@example.org>wrote:
> Would you please clarify why Fisher's exact test "won't work at
> all"? Do you run out of memory or is it too time consuming to
> produce Fisher's exact test? Or does it have something to do
> with complex sample design issues that you must account for in
> the analysis (which is why you typically use the SURVEYFREQ
> procedure to produce contingency table test statistics)?
> Dale McLerran
> Fred Hutchinson Cancer Research Center
> mailto: dmclerra@nFoHsCpRaCm.org
> Ph: (206) 667-2926
> Fax: (206) 667-5977
> --- On Thu, 4/28/11, Louise Hadden <louise_hadden@ABTASSOC.COM> wrote:
> > From: Louise Hadden <louise_hadden@ABTASSOC.COM>
> > Subject: Significance Statistics for Tables with Small or 0 cells?
> > To: SAS-L@LISTSERV.UGA.EDU
> > Date: Thursday, April 28, 2011, 1:01 PM
> > I've been asked to produce significance statistics for a number of
> > surveys with very small subgroups. Some of the tables have 0 cells,
> > and some have cells with less than 5 observations in them. Normally I
> > produce CHISQs out of SURVEYFREQ for categorical variables, weighted
> > TTESTS for continuous variables, and output the stats via ODS and apply
> > data-driven stars in my reports where appropriate. In this case however,
> > CHISQs are not appropriate for the small cells and just plain won't work
> > for the 0 cells. (I tried McNemars for the 2 x 2 tables which work, no
> > idea if they are appropriate - but McNemars doesn't work for the 2 x 2
> > tables - these are all comparisons of two groups. Fishers won't work at
> > all. Wald Chis (like the other Chis) also won't work.)
> > Does anyone have a work-around for this? Any suggestions greatly
> > appreciated.