ABSTRACT Traditional control charts assume normality in process data, which is often violated in real‐world applications, particularly for response time data that exhibits significant skewness. This study develops Hybrid Exponentially Weighted Moving Average (HEWMA) control charts specifically designed for skewed response time data using percentile‐based methods implemented in R. A HEWMA chart is proposed that combines EWMA statistics with Sequential Probability Ratio Test (SPRT) principles, utilizing Wald and Maxwell distributions to model skewed response times. Through extensive Monte Carlo simulations, the performance is evaluated using Average Run Length (ARL), Standard Deviation of Run Length (SDRL), and Median Run Length (MRL) metrics. The proposed percentile‐based approach demonstrates superior performance compared to traditional control charts, with standardized ARL 0 values maintained at approximately 370, and ARL improvements of over 20% for various shift magnitudes, ensuring a fair comparison by controlling the in‐control false alarm rate. Industrial application to customer service response time data validates the practical effectiveness of the method, showing significant process improvement and economic benefits. The R implementation provides a comprehensive framework for practitioners to implement these control charts for skewed data monitoring. Future work includes the development of truly adaptive HEWMA charts with dynamic smoothing parameters.
Raza et al. (2026) studied this question.