Analýza přístupů ke konstrukci vybraných regulačních diagramů v prostředí kontaminovaných dat

Abstract

This bachelor’s thesis focuses on analyzing approaches to constructing selected control charts in the context of contaminated data. The aim of the thesis is to evaluate the effectiveness of traditional statistical quality control tools under conditions where the assumption of data normality is violated. The work includes simulation studies comparing the performance of x̄, R and x̄, s control charts in detecting changes in process variability, for both clean and contaminated datasets. The results demonstrate how data contamination affects the ability of control charts to correctly signal process instabilities and suggest which methods are more robust to deviations from the ideal data distribution.