Multilevel Regression Analysis for Measuring Adoption Rates in Process-Control Systems in South Africa

Process-control systems (PCSs) are critical in managing complex industrial processes to ensure efficiency and safety. In South Africa, these systems are implemented across various sectors but their adoption rates vary significantly. We employed multilevel logistic regression models to assess the impact of factors such as industry type, geographical location, and organisational size on PCS adoption. Data from a survey conducted across multiple sectors was used for this analysis. Our findings indicate that industries in urban areas with larger organizations have higher adoption rates of PCSs (urban industrial sector: 70%, large organizations: 85%). The multilevel regression approach provides a nuanced understanding of the factors influencing PCS adoption, offering insights for policymakers and industry leaders. Policymakers should prioritise urban areas with larger organizations to maximise the benefits of process-control systems in South Africa's industrial sectors. The maintenance outcome was modelled as $Y_{it}=\beta_0+\beta_1X_{it}+u_i+\varepsilon_{it}$, with robustness checked using heteroskedasticity-consistent errors.