[Rasch] Reliability in the Rasch methodology

Mike Linacre (RMT) rmt at rasch.org
Sun Jun 15 05:10:15 EST 2008

Thank you for your question.

You asked:
>In the Rasch world, what is the interpretation for the (a) person 
>reliability, and (b) item reliability?

"Reliability" in both Rasch theory and Classical Test Theory is the 
statistical reproducibility of a set of values (Rasch measures or raw 
scores). It is the hypothesized "true" variance of those values divided by 
their observed variance. CTT computes this for the raw scores of the 
current sample of persons and reports it as the "test reliability". Rasch 
computes it for the Rasch measures (abilities) of the current sample of 
persons (the "person" reliability) and the Rasch measures (difficulties) of 
the current sample of items (the "item reliability").

There is a comparison of Rasch person reliability and CTT test reliability 
at http://www.rasch.org/rmt/rmt113l.htm


Mike L.

Mike Linacre
Editor, Rasch Measurement Transactions
rmt at rasch.org www.rasch.org/rmt/ Latest RMT:  21:4 Spring 2008
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailinglist.acer.edu.au/pipermail/rasch/attachments/20080614/36c92d01/attachment.html 

More information about the Rasch mailing list