First, we set the hypothesis that the arithmetic mean ܺ തଵ and ܺ തଶ of basic sets are equal to each other, i.e.ܺ തଵ = ܺ തଶ, so called null hypothesis. It is necessary to examine the hypothesis, and it boils down to examining differences of arithmetic means of the samples. If the difference is random, the hypothesis is accepted, it is true. Conversely, if the difference is significant, the hypothesis is rejected as inaccurate, i.e. the samples don't belong to one and the same basic set