I'm analysing a study of 4 raters and 34 subjects. I'm looking at the inter rater reliability. For the ordinal results, I did a weighted kappa of Fleiss and for the other continuous results I calculated an ICC (2,1).

My question is if the data has to follow a normal distribution to be able to use the Kappa of Fleiss and the ICC (2, 1) correctly?

  • $\begingroup$ Kappa is a measure of inter-rater reliability. How could such data be normally distributed? They are discrete, usually with a quite small number of categories. $\endgroup$
    – Peter Flom
    Sep 27, 2012 at 11:12
  • $\begingroup$ What about the ICC and the interpretation of the f test $\endgroup$ Jul 19, 2013 at 6:19

1 Answer 1


The Kappa statistics have to do with level of agreement in contingency tables and there is no normal distribution involved. The same is true for intra class correlation.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.