Abstract

In some cases when statistical methods of examination of relations and dependences are applied to distributions which are different than normal the results are questioned. One of the reasons is that in the construction of these methods grounded are just a first initial, second central and a compound element. The measures grounded on elements of a higher order as these of asymmetry and excess are usually neglected nevertheless they carry information which in many cases is very valuable. Presented is an idea of simultaneous use of the moments from the first to the fourth order - mean value, mean quadratic digression and the coefficients of asymmetry and excess while comparing several distributions towards a single moment as well as to examine the dynamics of distributions.