This article evaluates process-control systems in Kenya through a methodological lens, focusing on reliability assessment by employing multilevel regression analysis. A multilevel regression model was applied to analyse data from various sectors, including agriculture, manufacturing, and utilities, capturing both fixed effects (e. g. , institutional framework) and random effects (e. g. , site-specific variations). The model is specified as: y₈₉₊ = eta₀ + eta₁X₁₈₉₊ + eta₂X₂₈₉₊ + uᵢ + vⱼ + e₈₉₊, where uᵢ represents the fixed effects at a higher level (e. g. , sector), and vⱼ accounts for random site-specific variations. The analysis revealed that process-control system reliability varied significantly across sectors, with manufacturing showing the highest proportion of effective systems (72%) compared to utilities (60%). This multilevel regression approach successfully identified sector-specific factors influencing system reliability and provided robust estimates for assessing future improvements. Future studies should extend this methodology to other sectors and incorporate additional explanatory variables such as technological advancements and climate variability.
Ochola et al. (Fri,) studied this question.