Hi
I tried to search the archives in vain - apologies if I missed anything..
My query:
I have three pairs of raters observing the presence/absence of a behaviour.
Each pair made about 30 observations. I understand that Kappa statistics are
commonly employed for examining inter-rater agreement. John Uebersax noted a
few concerns however and recommended that one could instead directly
calculate the intraclass correlation (Source:
http://ourworld.compuserve.com/homepages/jsuebersax/kappa.htm )
I did stumble upon about a paper explicating the use of ICC for binary data.
Ridout, M. S., Demetrio, C. G., & Firth, D. (1999). Estimating intraclass
correlation for binary data. Biometrics, 55(1), 137-148.
And another that appears to suggest similarly:
http://www.science.ulster.ac.uk/research/psychology/profiles/g_rae/equi_rater_kappa_statistics_inter_correl.pdfI must confess that I do not have the ability to comprehend the mathematical
logic.. Can I can perform the usual ICC procedures in SPSS for binary
ratings? If so, is there anything I should note? I wish to assume that the
three pairs of raters were randomly selected so that I may generalise the
findings to a larger population of inidividuals who can perform the rating
task as well. I am using SPSS v11.5 and v12.
Any advice would be valuable - thanks so much!
chong
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
