In this chapter are explained the basics and formula of the kappa fleiss, which allows to measure the agreement between several advisors according to category criteria (nominal or ordinal). We also show how to calculate and interpret kappa values with the R. Note that at Fleiss Kappa, you don`t necessarily need the same advice for each participant (Joseph L. Fleiss 2003). The Fleiss Kappa was conducted to determine whether there was a match between the police`s judgment that 23 people in a clothing store were behaving normally, unusual, but not suspicious or suspicious, based on a video clip showing each buyer`s movement through the retail trade. Three non-unique police officers were randomly selected from a group of 100 police officers to assess each person. Each officer evaluated the video clip in a separate room so that he could not influence the decision of the other police officers. To assess a person`s behaviour in the clothing store, each police officer could only choose from one of three categories: “normal,” “unusual, but not suspicious” or “suspicious behaviour.” The 23 people were randomly selected by all buyers who visited the clothing retail trade for a period of one week. Fleiss` Kappa showed that there was a moderate convergence between the officers` judgments, s-557 (95% CI, .389 to .725), p < .0005.
If the number of categories used is small (z.B. 2 or 3), the probability of 2 advisors agreeing by pure coincidence increases considerably. This is because the two advisors must limit themselves to the limited number of options available, which affects the overall agreement rate, not necessarily their propensity to enter into an “intrinsic” agreement (an agreement is considered “intrinsic” if not due to chance). That is why we know that there was a moderate convergence between the officers` judgment with a cappa of 0.557 and a 95% confidence interval between 0.389 and 0.725. We also know that Fleiss` Kappa coefficient was statistically significant. However, we can go further by interpreting each cappa. As with Cohens Kappa, sPSS and R require that the data be structured with separate variables for each coder of interest, as for a variable that presents the empathy assessments in Table 5. If several variables were evaluated for each subject, each variable for each coder would be entered in a new column in Table 5, and CCIs would be calculated in separate analyses for each variable. Subsequent extensions of the approach included versions that could deal with “under-credits” and ordinal scales.  These extensions converge with the intra-class correlation family (ICC), which allows us to estimate reliability for each level of measurement, from the notion (kappa) to the ordinal (or ICC) at the interval (ICC or ordinal kappa) and the ratio (ICC). There are also variations that may consider the agreement by the evaluators on a number of points (for example.B. two people agree on the rates of depression for all points of the same semi-structured interview for a case?) as well as cases of raters x (for example.
B how do two or more evaluators agree on whether 30 cases have a diagnosis of depression, yes/no a nominal variable).