Date:         Fri, 11 Jul 1997 14:31:44 EDT
Reply-To:     Anthony_Petrosino@CHSB1.CCMAIL.COMPUSERVE.COM
Sender:       "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From:         Anthony Petrosino <Anthony_Petrosino@CHSB1.CCMAIL.COMPUSERVE.COM>
Subject:      intercoder reliability test

     SPSS members:
 
     I have two files of 8 records each, independently coded by 2 different
     assistants.  I'd like to test the intercoder reliability beyond the
     rate of agreement (% of records agreed on each item).  The item is the
     unit of analysis and the records are coded a1 b1 a2 b2... to designate
     the different coders.
 
     Any ideas?  Should kappa be run and how should the file be constructed
     to permit such a test?
 
     Thanks in advance,
 
     Anthony J. Petrosino, Research Associate
     Criminal History Systems Board