[Rasch] Help with clasification sonsistency

rsmith.arm at att.net rsmith.arm at att.net
Sat Jan 26 04:10:49 EST 2008

There is a Rasch person standard error based decision consistency index developed by Sterns and Smith that was presented at IOMW 2006 and AERA 2007.  It is will be published as a chapter in the Smith and Stone book on Standard Setting due out this year form JAM Press.

It is quite simple.  For each person calculate the probability that a person will fall into each of the classification groups on retest using the standard error of the person ability estimate.  You then can calculate the probability of the same and different classification on retest.  Sum those over all of the persons who took the test and you get an estimate of the misclassification rate.  The beauty of the method is that you have a misclassification probability for each person, that can be used to mane informed decisions about the person, not an overall rate that really applies to no one and is useless for informing individual decisions.
Richard M. Smith, Editor 
Journal of Applied Measurement 

P.O. Box 1283 
Maple Grove, MN 55311 USA 
(JAM web site) 
voice: 763-268-2282 (w) 
fax: 763-268-2782 

-------------- Original message from Andrés Burga León <21781 at upch.edu.pe>: -------------- 

Hello to everybody on the list:
I hope somebody could help me with this.
I have a test which have two cut scores (set by a panel of experts using bookmark). So I have tree performance groups.
The question is how could I calculate a classification consistency index for my tree groups?. I suspect ist related to the persons measure error, but I’m not sure about how to do it.
Thank you very much.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailinglist.acer.edu.au/pipermail/rasch/attachments/20080125/6f942291/attachment.html 

More information about the Rasch mailing list