The examples in Table 1 show that the correlation tables that fill (9) are not necessarily symmetrical. Since the examples are ordinary correlation tables found in practice, it seems that the class of correspondence tables that fill (9) is not trivial. Therefore, the standard kappa error for the data in Figure 3, P = 0.94, pe = 0.57 and N = 222 is the number of subjects evaluated, w the weight of consent or disagreement, po the share of observed concordance, pe the expected share of concordance, pij the fraction of evaluations i by the first evaluator and j by the second, and so the default error to test is that kappa statistics are zero. Third sentence. If all specific cases of symmetric kappa are identical, then applies (9). .