Graphic representation of the amount of correct data per % agree or the quadratic value of Kappa. The pioneer paper, introduced by Kappa as a new technique, was published in 1960 by Jacob Cohen in the journal Educational and Psychological Measurement. [5] Solution: the modeling agreement (z.B. on log-linear or other models) is usually an informative approach. Now calculate p, “Displaystyle,” the average value of the P i-Displaystyle-P_,, “s” and the P e-display style, P_ “bar” that go into the formula for “display style”: Zaiontz, Charles. Cohen`s Kappa. www.real-statistics.com/reliability/interrater-reliability/cohens-kappa/ A good example of concern about the importance of Kappa`s results is contained in a paper comparing visual detection of human biological sample anomalies to automated detection (12). The results showed only a moderate agreement between human and automated advisors for kappa (n-0.555), but the same data showed excellent percentage match of 94.2%. The problem with interpreting the results of these two statistics is: how should researchers decide whether advisors are reliable or not? Do the results indicate that the vast majority of patients receive accurate laboratory results and therefore correct medical diagnoses or not? In the same study, the researchers selected a data collector as the standard and compared the results of five other technicians to the standard. While sufficient data to calculate a percentage agreement are not included in the document, Kappa`s results have been only moderate.

How does the lab head know if the results are of good quality with few discrepancies between trained laboratory technicians or if there is a serious problem and a need for training? Unfortunately, Kappa`s statistics do not provide enough information to make such a decision. In addition, a Kappa can have a confidence interval so wide that it contains everything from good to bad chord. In the edition below, we can see that the “Simple Kappa” indicates the estimated kappa value of 0.389 with its standard asymptomatic error (ASE) of 0.0598. The difference between the observed agreement and the expected independence is about 40% of the maximum possible difference.