Master’s Thesis at the University of Basrah Examines the Robustness of One-Way ANOVA: A Simulation Study and Applications
The College of Science at the University of Basrah discussed a master’s thesis titled “Robustness of One-Way Analysis of Variance: A Simulation Study and Applications”, which addressed the issue of the classical F-test in one-way ANOVA being sensitive to outliers and deviations from normal distribution.
The thesis, presented by student Asmaa Adnan Sweidan, aimed to explore more robust statistical alternatives, including the nonparametric Kruskal–Wallis test.
The study proposed a robust version of the F-test based on a reweighted M-estimator using Hample’s weighting function, in addition to constructing an approximate distribution for the proposed statistic. This approach enhances the accuracy of results in cases involving outlier data or non-normal distributions.
Department of Media and Governmental Communication
