I want to compare modeled species occurrence with observed occurrence data. Using confusion matrix and weighed kappa seems to be a good option. In contrast to the normal kappa the weighted kappa takes into account that the data is interval scaled. However, I am wondering if there is an alternative to weighted kappa measure since I suspect that the weighted kappa measure is scale dependent.

  • 1
    $\begingroup$ Tell us much more about what kind of response you are modelling and what you understand by validation. $\endgroup$ – Nick Cox Jun 7 '15 at 9:10
  • 1
    $\begingroup$ I wouldn't use kappa for interval-scaled data, but perhaps concordance correlation. But the deeper point is that focusing on any single measure is very primitive thinking statistically. It's more interesting and useful to see where the model works and where it doesn't. $\endgroup$ – Nick Cox Jun 7 '15 at 11:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.