STUDYING OF INFORMATIVENESS OF THE COMPARISON OF THE AVERAGE SELECTIONS OF EXPERIMENTAL RESULTS
DOI:
https://doi.org/10.34185/1991-7848.itmm.2023.01.041Keywords:
data analysis, non-parametric statistics, statistical homogeneity, computational experiment.Abstract
Criteria of Student, Pearson and Fisher are widely used in the study of the equality of parameters of samples of random variables. In a scientific article by Orlov A.І., the author considers them as "outdated and not corresponding to the realistic formulation of data analysis problems to modern statistical methods." A transition from parametric statistics of Student, Pearson and Fisher to non-parametric statistics of Kolmogorov-Smirnov, from average values and sample variances to shifts and scales (ranges), sums and differences of maximum and minimum values of samples of random variables is proposed.
References
Orlov A.I. Nonparametric Kolmogorov-Smironov omega square agreement statistics and cases of their application. Scientific group Kub GAU, No. 97 (03) 2014, p. 31-45.
Kobzar A.I. Applied mathematical statistics. For engineers and scientists. – M.: FIZMATLIT. 2006. - 816 p.
D. Cox and D. Hinckley, Theoretical Statistics. Ed. "Mir" M. 1978, 557 p.
