Computer-based Simulation of Stochastic Process for Investigation of Efficiency of Statistical Hypothesis Testing in Pedagogical Research

Abstract

Efficiency of Chi-square criterion and Fisher's angular transformation for statistical hypothesis testing are investigated at small samples with computer-based model. The results show that both criteria give satisfactory accuracy even at small samples. Accuracy zones of determination of the false positive rate (type I error) are analysed on the base of stochastic computational experiments. Comparison of sensitivities of investigated tests is shown as a function of samples size for some cases

Keywords

• Chi-square criterion
• Fisher's angular transformation
• small samples
• statistical hypothesis
• Stochastic process