Jean-Pierre Guay wrote:
> A simple question for those who are used to ICC...
>
> What is the difference between the single and the average ICC in ICC
> (3,K)? I mean, why does SPSS calculate both the single and the average
> when i have four judges (so, ICC(3,4))? Is the single used for test-
> retest, meaning that there is only one juge, but it is rating subjects
> two or more times?
The single estimate refers to the reliability of a single judge or
rater. The average estimate refers to the reliability of the average
of several judge's ratings. As far as I know, the reliability of the
average should always be greater than the reliability of a single
judge (Spearman-Brown prophecy).
> And average, is it the average correlation between all pairs of juges?
> Do any of you know the difference between "Absolute agreement"
> and "Consistency" in SPSS? Does this apply to both single and average?
I beleive absolute agreement takes into account differences in rater
means and treats those differences as error variance thereby
decreasing reliability. Consistency does not treat differences in
rater means as error variance and is only sensitive to the degree to
which raters order the objects in the same way. The absolute
agreement estimate shuld always be lower or the same since any
difference in means between raters will reduce reliability. If there
are no differences in means between raters the two estimates should
agree.
Hope this helps,
Chuck
----------------------------------------------
Chuck Cleland
Institute for the Study of Child Development
UMDNJ-Robert Wood Johnson Medical School
97 Paterson Street
New Brunswick, NJ 08903
phone: (732) 235-7699
fax: (732) 235-6189
http://www2.umdnj.edu/iscdweb/
----------------------------------------------
===========================================================================
This list is open to everyone. Occasionally, less thoughtful
people send inappropriate messages. Please DO NOT COMPLAIN TO
THE POSTMASTER about these messages because the postmaster has no
way of controlling them, and excessive complaints will result in
termination of the list.
For information about this list, including information about the
problem of inappropriate messages and information about how to
unsubscribe, please see the web page at
http://jse.stat.ncsu.edu/
===========================================================================
