[Rasch] inter-rater agreement

Robert Hess Robert.Hess at ASU.edu
Sat Jan 12 06:09:39 EST 2013


Shirin,

The issue is not as simple as the general notion of inter-rater reliability
in a classical sense. There are actually two different levels of questions
that need to be examined. The first we might interpret as between scorer (or
rater) consistency. This refers to the extent to which all raters in general
tend to score high ability students as high, medium ability students as
medium, and low ability students as low (somewhat similar to the notion of
inter-rater reliability but not really the same). The second refers to
within-rater issues. In this area the concerns are tri-fold: 1) rater
consistency: is a rater consistent in their scoring (always scoring high
ability students as high, low ability students as low, etc.); 2) rater
severity: is a rater more severe on students than other raters or is a rater
more lenient than other raters; & 3) if scores are based on more than one
trait is there a halo effect - in other words, does the first trait scored
dictate the level of scores receive by following traits.

There are a whole plethora of articles available that can describe these
influences in much greater detail.

 

Robert Hess

Emeritus Professor of Measurement and Evaluation

Arizona State University

 

From: rasch-bounces at acer.edu.au [mailto:rasch-bounces at acer.edu.au] On Behalf
Of shirin shirazi
Sent: Friday, January 11, 2013 7:11 AM
To: Rasch
Subject: [Rasch] inter-rater agreement

 

Dear List Members,

I have a question about inter-rater agreement obtained through Facets
analysis.

It is said in the literature if observed inter-rater agreement > expected
agreement, then raters behave like rating machines (higher inter-rater
reliability); if we take it right, then what if separation and reliability
of the same sample are high, then can we interpret it as low inter-rater
agreement? should they be in line with each other or do they report two
distinct issues?


Your answers are highly appreciated

Shirin

-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailinglist.acer.edu.au/pipermail/rasch/attachments/20130111/1b742184/attachment.html 


More information about the Rasch mailing list