Date: Wed, 9 Jan 2008 14:45:06 -0700
Reply-To: ViAnn Beadle <vab88011@gmail.com>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: ViAnn Beadle <vab88011@gmail.com>
Subject: Re: Contingency Table Testing Within Rows
In-Reply-To: <945D579F76FD5741A043511E6FE99B571586718F@TLRUSMNEAGMBX32.ERF.THOMSON.COM>
Content-Type: text/plain; charset="us-ascii"
How was this table produced? Is a case a respondent from a particular survey
specifying a particular brand preference? That is, do the row percentages
add to 100%
If so, perhaps a column proportion test (requires putting the brands on top
and the dates on the side) might give you what you want. This is available
from CTABLES.
For each row cell it displays significantly different column cells. Go to
Help and select Tables Option>Test Statistics>Comparing Column Proportions
for a discussion.
-----Original Message-----
From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] On Behalf Of Jim
Moffitt
Sent: Wednesday, January 09, 2008 8:49 AM
To: SPSSX-L@LISTSERV.UGA.EDU
Subject: Contingency Table Testing Within Rows
I have a 5 X 5 contingency table that depicts the percentage of
respondents (based to column totals) that prefer each of 5 brands. The 5
brands are the row labels. The column labels are different survey dates
so Column 1 is Fall 2007, Column 2 is Fall 2006, Column 3 is Fall 2005,
etc. Thus, cell 1 (matrix element 11) contains the proportion of Fall
2007 respondents that preferred Brand A, the next cell to the right
(matrix element 12) contains the proportion of Fall 2006 respondents
that prefer Brand A, the next cell to the right (matrix element 13)
contains the proportion of Fall 2005 respondents that prefer Brand A,
etc.
Each survey was administered to a different random sample, so the column
samples are independent of each other.
What significance test would I apply to identify significant differences
within the rows while controlling for family-wise probability issues? In
other words, I'm looking for a test that will test within the row
vectors so I can say matrix element 11 is significantly different from
14 and 15, matrix element 14 is significantly different from 13 and 15,
matrix element 22 is significantly different from 24 , matrix element 34
is significantly different from 31, 33, and 35, etc.
How would I run such a test in SPSS 14.0?
Do I need a specific SPSS module like Advances Tests to run the required
test?
Thanks for your help.
=======
To manage your subscription to SPSSX-L, send a message to
LISTSERV@LISTSERV.UGA.EDU (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@LISTSERV.UGA.EDU (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
