[Method of determining the closeness of correlation for curvilinear functions in biomedical research]

A curvilinear association of data is the mostly widespread in nature. A method seldomly used in morphology but, at the same time, the only correct one to estimate the degree of association between the data at their curvilinear dependence by means of calculating correlative ratios is presented. The correlative ratio, unlike the coefficient of correlation, is the universal index demonstrating the degree of interconnection between the data and characterizes the bilateral contingency of the degree of association. In order to illustrate the field of application and to interpret the method in question, the author's morphometric investigation of peripheral myelin fibres is given as an example. Formulae for calculating the corrective association are presented and the properties of the latter are enumerated.