UDC

519.25

Authors

Svetlana A. Parygina*
*Cherepovets State University (Cherepovets, Russian Federation)

Abstract

The article deals with the use of nonparametric methods of mathematical statistics, particularly in nonparametric hypothesis tests, for the empirical data-processing operation of different nature. As an example of practical implementation of these methods we introduce a career-oriented experiment conducted in Cherepovets State University. At different stages of statistical processing of the experiment results the features of static analysis are established, which are related to the specifics of certain data. The rank nonparametric hypothesis tests are used for psychological data processing. The ways of overcoming the series of restrictions of the statistical criteria using, such as a frequent repetition of sample values, the lack of table values for large volume samples, are identified. We attempt to structure the static analysis of ordinal data in accordance with the task of finding significant differences in a large number of characteristics between multiple statistical universes. The alternative static analysis using the Smirnov nonparametric test of homogeneity is carried out. The asymptotically nonparametric test χ² is used for biomedical data processing. This is due to the nominal nature of the most medical and biological characteristics. We demonstrate the approximating of the critical statistics χ² for a large sample and a work with directional hypotheses. The conclusions contain the recommendations for the use of rank nonparametric hypothesis tests for the data processing of different nature: human (in terms of psychological data) and naturalist (in terms of medical and biological data).

Keywords

nonparametric hypothesis test, Kruskal-Wallis test, significant difference, χ2 criterion, contingency table

References

1. Prygina S.A. Sravnitel’naya kharakteristika rangovykh neparametricheskikh kriteriev proverki gipotez i osobennosti ikh primeneniya k obrabotke dannykh razlichnoy prirody [Comparative Characteristics of the Rank-Order Nonparametric Hypothesis Tests and Special Aspects of Their Application to the Data Processing of Different Nature]. Estestvennye i tekhnicheskie nauki [Natural and Technical Sciences], 2015, no. 6, pp. 44–48.
   
2. Parygina S.A. Ob osobennostyakh primeneniya nekotorykh rangovykh neparametricheskikh kriteriev proverki gipotez [On the Features of Application of Some Rank Nonparametric Criteria for Verifying the Hypotheses]. Vestnik Bashkirskogo universiteta [Bulletin of Bashkir University], 2016, vol. 21, no. 2, pp. 252–257.
   
3. Hollander M., Wolfe D.A. Nonparametric Statistical Methods. New York, 1973. 503 p.
   
4. Hajek J. Nonparametric Statistics. San Francisco, 1969. 184 p.
   
5. Borovkov A.A. Matematicheskaya statistika [Mathematical Statistics]. Novosibirsk, 1997. 772 p.
   
6. Niskina N.P., Teyman A.I., Shmerling D.S. Neparametricheskie metody statistiki, osnovannye na rangakh, i ikh primenenie [Nonparametric Statistical Methods Based on Ranks, and Their Application]. Moscow, 1986. 62 p.
   
7. Hettmansperger T.P. Statistical Inference Based on Ranks. New York, 1984. 56 p.
   
8. Sidorenko E.V. Metody matematicheskoy obrabotki v psikhologii [The Methods of Mathematical Processing in Psychology]. Saint Petersburg, 1996. 348 p.
   
9. Orlov A.I. O proverke odnorodnosti dvukh nezavisimykh vyborok [On the Verification of the Homogeneity of Two Independent Samples]. Zavodskaya laboratoriya. Diagnostika materialov [Industrial Laboratory. Materials Diagnostics], 2003, vol. 69, no. 1, pp. 55-60.
   
10. Nikitin Ya.Yu. Asimptoticheskaya effektivnost’ neparametricheskikh kriteriev [Asymptotic Efficiency of Nonparametric Tests]. Moscow, 1995. 235 p.
    
11. Gubler E.V., Genkin A.A. Primenenie neparametricheskikh kriteriev statistiki v mediko-biologicheskikh issledovaniyakh [The Use of Nonparametric Statistics Tests in Biomedical Research]. Moscow, 1973. 141 p.
    
12. Grabar’ M.I., Krasnyanskaya K.A. Primenenie matematicheskoy statistiki v pedagogicheskikh issledovaniyakh. Neparametricheskie metody [Application of Mathematical Statistics in Pedagogical Research. Non-Parametric Methods]. Moscow, 1977. 136 p.
    
13. Fahoome Gail F. Twenty Nonparametric Statistics and Their Large Sample Approximations. Journal of Modern Applied Statistical Methods, 2002, no. 2, pp. 248–268.