[Rasch] Pairwise comparisons

Stephen Humphry stephen.humphry at uwa.edu.au
Mon Sep 1 11:28:44 EST 2008


Donald, there are a few ways you could go about this. One is simply to
randomize the order of pairs and then assign a certain proportion to each
respondent. So long as there is overlap, they can be scaled together using
the Bradley-Terry-Luce model.

Another way is to use pairwise comparisons with a subset of the objects and
scale them, then simply have respndents compare new objects with those that
have been scaled. New scale locations can be determined given for any total
score based on the number of favorable comparisons (based on the raw-logit
correspondence for the scaled subset).

Another way is to get crude classifications first, then only compare objects
with others in the same classification and one higher or lower.

I'm not sure what you mean by the 'group-level estimates'. Are objects
grouped?

There is some software you can use for pairwise estimation and analysis of
fit; let me know if it is of interst to you.

Regards,

Steve

Dr Stephen Humphry
Graduate School of Education
University of Western Australia
35 Stirling Highway
CRAWLEY  WA  6009
Mailbox M428
P: (08) 6488 7008
F: (08) 6488 1052
 
CRICOS Provider Code: 00126G

 

-----Original Message-----
From: rasch-bounces at acer.edu.au [mailto:rasch-bounces at acer.edu.au] On Behalf
Of Donald Bacon
Sent: Monday, 1 September 2008 6:25 AM
To: rasch at acer.edu.au
Subject: [Rasch] Pairwise comparisons

Hi Raschers --
   I'm working on a project that involves pairwise comparisons.  There are
so many objects that I'm afraid I'll have too many comparisons for each
respondent to rate (I'm worried about fatigue).  There must be a systematic
way to divide the comparisons, so I can show some comparisons to some
respondents and other comparisons to other respondents.  I don't think I
will need individual estimates of the pairwise comparisons, just group-level
estimates.
   Does anyone know of a citation that might be useful to me?

Thanks --

Don Bacon
Professor of Marketing
University of Denver
_______________________________________________
Rasch mailing list
Rasch at acer.edu.au
https://mailinglist.acer.edu.au/mailman/listinfo/rasch






More information about the Rasch mailing list