Acta Univ. Agric. Silvic. Mendelianae Brun. 2020, 68(1), 281-289 | DOI: 10.11118/actaun202068010281

Non-Parametric Anova Methods Applied on Students' Performance Development in Course of Statistics

Martina Zámková, Martin Prokop, Radek Stolín

Department of Mathematics, College of Polytechnics Jihlava, Tolstého 16, 586 01 Jihlava, Czech Republic

This study intends to use non-parametric statistical methods for comparison of mean values in assessing the changes of overall grades in basic statistics at the College of Polytechnics in Jihlava. The data covering students' grades were transferred to the ordinal scale and evaluated with the use of the corresponding non-parametric statistical methods, due to failing the data normality test. In order to identify statistically significant grade differences over the years 2006-2018, we have used the test of equality of means as well as the relevant analyses of variance, i.e. the Mann-Whitney U test to compare two samples, the Kruskal-Wallis test to compare multiple samples, and the post hoc analysis. All tests that were carried out confirm that on average, women and Finance and Management students tend to achieve better grades in statistics, as opposed to men, and Travel and Tourism students. Mean grades have rather fluctuated over the years. In the beginning (2006-2009) and in the end (2014-2017) of the observed period the grades have been significantly superior to the grades in the in-between stage (2010-2013). The identified fluctuation in grades shall lead to implementation of measures aimed at achieving an overall increase in statistics' education quality and students' grades.

Keywords: College of Polytechnics Jihlava, statistics, success rate, Mann-Whitney U test, Kruskal-Wallis test, non-parametric tests, post hoc analysis

Received: June 29, 2019; Accepted: November 22, 2019; Published: February 27, 2020  Show citation