Improving Sociological Acceptance of Monitoring Industrial Production Processes Using Median Absolute Deviation

Abstract

This study investigates the effectiveness and sociological acceptance of Median Absolute Deviation (MAD)-based control charts in monitoring industrial processes. Traditional control charts, such as Shewhart and CUSUM, depend on normality assumptions, making them less effective in detecting variations in non-normal data or environments with outliers. In contrast, MAD-based control charts offer a more robust alternative by enhancing sensitivity to process shifts and minimizing the impact of extreme values. The study explores barriers to adoption, including resistance to change, perceived complexity, and insufficient training. A mixed-methods approach was employed, combining survey responses from 100 stakeholders with a statistical analysis of production data from the Coca-Cola Bottling Company in Ibadan. The analysis utilized descriptive statistics, specifically percentage calculations, and statistical control charts, performed using the R statistical software. Results indicate that MAD-based control charts improve process variability detection and overall quality control compared to traditional methods. However, their adoption is influenced by factors such as training, user familiarity, and managerial support. The study underscores the need for structured training programs, leadership endorsement, and simplified implementation strategies to encourage adoption. It concludes with recommendations for industry-wide acceptance and proposes future research to expand MAD-based monitoring across various manufacturing sectors.

Keywords: Median Absolute Deviation (MAD), Control Charts, Industrial Monitoring, Technology Adoption, Sociological Acceptance, Statistical Methods.