Date:         Wed, 28 Jan 1998 09:14:00 +1000
Reply-To:     Hockey Richard <rhockey@MATER.ORG.AU>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         Hockey Richard <rhockey@MATER.ORG.AU>
Subject:      Re: proc tabulate and multiple response variables
Comments: To: Ralph Brands <brinton@UNIXG.UBC.CA>
Content-Type: text/plain; charset="us-ascii"

I received some handy SAS macros from SI in Australia a while back,  I
don't  have them here but they are available from this web site
http://www.publichealth.uwa.edu.au/links/multitab.zip
They basically use proc tabulate creating an indicator variable for each
possible response.
Hope this helps
 ----------
From: Ralph Brands
To: SAS-L@UGA.CC.UGA.EDU
Subject: proc tabulate and multiple response variables
Date: Wednesday, 28 January 1998 3:44AM
 
I recently asked a question about analysis of multiple response data in
SAS.
Having digested a lot of very helpful information from several readers,
and tried to understand Proc Freq, I now understand some situations
(referred to in SPSS as multiple-dichotomy groups), but am still stuck
on
this type (which SPSS calls a multiple-response groups):
 
Respondents are asked to list which of 5 states they have visited. The
data are entered as follows: (each state can appear in ANY column)
 
case  state1   state2   state3
1     NY    CAL      .
2     MASS  .     .
3     CAL      MASS  .
4     MI    WA    MASS
5     WA    .     .
 
We want to know what percentage of respondents reported visiting NY,
what
percentage reported visiting CAL etc, ie the desired output gives the
following information:
 
NY    1  20%
CAL   2  40%
MASS  3  60%
MI    1  20%
WA    2  40%
 
I have the feeling that this CAN be done in Proc Tabulate but haven't
got
it yet. Can someone provide a suggestion please?
 
Ralph Brands