Three researchers wanted to determine the interrater reliability of their field observations regarding aggressiveness of dog encounters at the dog park. Specifically, for each dog-pair, the researchers noted the breeds (or suspected breeds) of each dog, and how aggressive the dogs' first interaction was (1 = not at all aggressive; 5 = moderately aggressive, 10 = they got into a fight). To determine the interrater reliability of their aggressiveness observations, which of the following techniques would be most appropriate and yield the highest (reliable) reliability value?
A) Cohen's kappa
B) Fleiss' kappa
C) Intraclass correlation
D) B or C
E) all of the above
F) none of the above

Question

Quizplus · Accepted Answer

Intraclass correlation is the most appropriate technique for determining interrater reliability of continuous data, such as the ratings of aggressiveness on a scale of 1-10. Cohen's kappa and Fleiss' kappa are better suited for nominal data, such as identifying the breeds of dogs. Therefore, choice C (intraclass correlation) would yield the highest reliability value.

Three Researchers Wanted to Determine the Interrater Reliability of Their