Date:         Wed, 10 Feb 2010 13:57:14 -0600
Reply-To:     Robin R High <rhigh@UNMC.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Robin R High <rhigh@UNMC.EDU>
Subject:      Re: Inter- rater agreement for echocardiography measurements in
              SAS
Comments: To: Marieta <marieta.p@GMAIL.COM>
In-Reply-To:  <8a64b019-f91a-4882-9726-9c8b0d81a1e1@f12g2000yqn.googlegroups.com>
Content-Type: text/plain; charset="US-ASCII"

Marieta,

It would help to know more about the data you have, e.g., type of
measurements, what range of values it could take vs what ranges the 3
raters actually have.  ICC is often quite good for the situation you
describe, though there are situations with continuous data where it may
not perform all that well.  And what do you mean by % difference with 3
raters?

Robin High
UNMC





From:
Marieta <marieta.p@GMAIL.COM>
To:
SAS-L@LISTSERV.UGA.EDU
Date:
02/10/2010 12:44 PM
Subject:
Inter- rater agreement for echocardiography measurements in SAS
Sent by:
"SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>



Hi all,

I am looking for a good measurement of inter-rater agreement for
continuous variables implemented in SAS. It may be the ICC or
something else. My client is a physician who doesn't really feel
comfortable with the ICC and last time asked me to do a simple mean %-
difference calculation. I searched the literature for guidelines as to
acceptable degree of variability when looking at the % difference for
these particular measurements but didn't find anything. Any
suggestions are welcome. I have 3 raters and 10 cases per rater.

Thanks!
Marieta