Visualization of the Decision Criteria in Testing Statistical Hypotheses on Programming in R (Rstudio)

Bayuk O.A., Denezhkina I.E., Zadadaev S.A.
International Journal of Economics and Business Administration, Volume VII, Special Issue 2, 289-296, 2019
EOI: 10.11220/ijeba.07.06.028
DOI: 10.35808/ijeba/393


Purpose: The case study addresses a development and justification of approaches to visualization of decision criteria in the problems of testing statistical hypotheses for a given distribution law (specifically, checking the distribution normality). Design/Methodology/Approach: The study describes a construction of graphical model that visualizes an application of criteria when testing statistical hypotheses for compliance with a given distribution law. This problem is solved in the language of statistical analysis R in the RStudio environment. Using the standard approach and focusing only on the P-value in relation to the chosen level of significance, the researcher cannot take into account the error of the second kind. However, analyzing the graphical representation of the behavior of the sample under study, one can conclude whether the value of the obtained P-value corresponds to the real assumption that the sample corresponds to a given distribution law. Findings: The research case proposes a new approach to testing statistical hypotheses on the compliance of a sample with a given distribution law, using visualization tools and allowing a researcher having even a little experience with the R language to solve applied problems. Practical implications: The approach does not require an in-depth knowledge of mathematics and programming which can be used by experts in various fields of knowledge to successfully solve applied problems. The text of the article contains working scripts in the language R and graphical illustrations obtained with their help. Originality/Value: The main contribution of this study is to expand the variety of methods for testing statistical hypotheses. The proposed method extends the set of statistical problems successfully solved by means of R.

Cite Article (APA Style)