=========================================================================
Date:         Tue, 18 Jul 2006 14:39:42 -0400
Reply-To:     Paul Mastrangelo <paul@GENSURVEY.COM>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Paul Mastrangelo <paul@GENSURVEY.COM>
Subject:      Analyzing Paired Comparisons

All,

I am investigating approaches to analyzing paired comparisons. For example
in a list of 3 elements, participants would indicate their preference for
A or B, A or C, and B or C. My project has 7 elements, producing 21 paired
comparisons [(7*6)/2]. Obviously, I can report the number of times an
element was preferred across all participants to establish a ranking of
choices, but I wonder if there is a more rigorous analysis. Someone has
suggested Maximum Difference or MaxDiff, and he mentioned software from
Soft Tooth (?). Any information on MaxDiff or comparable SPSS procedures
would be welcomed. Thanks.