Gravar-mail: On Some Assumptions of the Null Hypothesis Statistical Testing