EVALUATION OF DEPENDENCE OF OCCURRENCE OF RISK EVENTS IN LOGISTICS ON RISK FACTORS BY MEANS OF SOMERS' D COEFFICIENT

The objective of this article is an empirical measurement of the dependencies of selected ordinal variables in order to obtain incentives for risk management in logistics. The article is aimed at assessing an asymmetric dependency of two variables, namely the dependence of occurrence of risk events in logistics on risk factors. Somers' d coefficient was used to measure the dependence, as it is the only one allowing unilateral dependence measurement. The article provides detailed explanation of the procedure used to calculate the Somers' d coefficient (calculated using partial calculations of the number of concordant and discordant pairs etc.) and its subsequent evaluation in terms of reliability (using significance level 1%, 5% and 10%). The resulting values are presented in several tables, depending on the used evaluation criteria - risk factors in logistics with the greatest dependence with the significance level of 1%, the most important groups of risk factors and risk factors with repetition for at least four risk events. The testing was performed using data collected through the medium of a questionnaire survey conducted in 2010 within the scope of SGS in the Czech Republic and Slovakia.