Date: Sat, 31 Jul 1999 15:32:16 -0500
Reply-To: Annette Towler <towleraj@RUF.RICE.EDU>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: Annette Towler <towleraj@RUF.RICE.EDU>
Content-Type: text/plain; charset="us-ascii"
I am having some problems calculating intraclass correlations for a data
set. I have calculated reliability (ICC(1, K) using Shrout and Fleiss'
equation, but cannot reconcile my figures with some previous calculations
on level of agreement. I am using S & F's ICC (1,1) equation to calculate
agreement i.e. BMS - WMS/BMS + (k-1)WMS.
Here is my ANOVA TABLE:
Sum of Mean
Source DF Squares Square F Value
Model 116 567.68212 4.89381 6.36
Error 6273 4830.21550 0.77000
Corrected Total 6389 5397.89762
here are my calculations:
4.89 - 0.77/4.89 + (116)0.77 = .04
Another student obtained a level of agreement of .09 on this data set, so
one of us is wrong! Can someone point me in the right direction?
Thanks in advance.
466 Sewall Hall, MS 25
Houston, Texas 77005-1892