the strength of compliance (low, moderate and high), represented by Fleiss`K and Krippendorffs Alpha-∈ [0.4,0.93] (see below) The assessment of reliability in epidemiological studies is heterogeneous and uncertainty is often not taken into account, leading to inappropriate methodological use. In addition, there is no evidence of the best measure of reliability in different circumstances (in terms of lack of data, distribution of prevalence, and number of assessors or categories). With the exception of a study by Häußler  comparing compliance measures for the particular case of two assessors and binary measures, there is no systematic comparison of reliability measures. That is why our objective was, despite the final rejection, as an appropriate measure of the IRR (Cohen, 1960; Krippendorff, 1980) many researchers continue to report the percentage that programmers agree in their assessments, as an indication of coder compliance. For categorical data, this can be expressed as the number of agreements in observations divided by the total number of observations. For ordinal, interval or report data for which a close but not perfect match may be acceptable, compliance percentages are sometimes expressed as a percentage of assessments that correspond to a given interval. Perhaps the main criticism of consensus percentages is that they do not correct agreements that would be expected by chance and thus overestimate the level of convergence. For example, if programmers randomly judged 50% of subjects as “depressed” and 50% “non-depressed,” the percentage of expected consent would be 50%, although all overlapping assessments were random. If, by chance, programmers estimated that 10% of the subjects were depressed and 90% were not depressed, the percentage of expected consent would be 82%, although this apparently high degree of concordance is still due exclusively to chance. Fleiss` kappa, κ (Fleiss, 1971; Fleiss et al., 2003), is a measure of interconceptor compliance used to determine the degree of agreement between two or more evaluators (also known as a “judge” or “observer”) when the assessment method, known as a response variable, is measured on a categorical scale.
In addition, Fleiss` Kappa is used if: (a) the evaluated objectives (z.B. Patients in a doctor`s office, learners who engage in a driving test, customers in a mall/center, burgers in a fast food chain, crates provided by a delivery company, chocolate bars in an assembly line) are randomly selected from the interested population instead of being specially selected; and (b) the evaluators who evaluate these objectives are unclear and are randomly selected from a larger population of evaluators. . . .