A Multilevel Regression Analysis of Water Treatment System Adoption in Kenya: A Methodological Case Study (2000–2026)

{ "background": "Understanding the drivers of technology adoption in infrastructure projects is critical for sustainable development. In Kenya, the uptake of advanced water treatment systems has been variable, with existing studies often failing to account for the hierarchical structure of influencing factors, from national policy to community-level characteristics.", "purpose and objectives": "This case study presents a methodological evaluation of applying multilevel regression analysis to measure adoption rates of engineered water treatment facilities. Its objective is to demonstrate the analytical framework's utility in isolating and quantifying the influence of factors at different organisational levels.", "methodology": "The analysis employs a three-level hierarchical linear model. The statistical model is specified as $y{ij} = \\beta{0j} + \\beta{1}x{1ij} + ... + \\epsilon{ij}$, where $\\beta{0j} = \\gamma{00} + \\gamma{01}w{1j} + u{0j}$, with levels for individual facilities, administrative counties, and regional blocs. Estimation uses restricted maximum likelihood with robust standard errors to account for heteroskedasticity.", "findings": "The methodological application reveals that county-level governance indicators explain approximately 40% of the variance in adoption rates between regions, a finding significant at the 95% confidence level. Facility-level factors, such as maintenance capacity, were moderated by these higher-level institutional contexts.", "conclusion": "Multilevel regression provides a robust analytical framework for engineering adoption studies, effectively partitioning variance across contextual layers. This approach moves beyond simplistic single-level analyses that can produce misleading inferences.", "recommendations": "Future engineering policy assessments should adopt multilevel modelling to inform targeted interventions. Data collection protocols must be designed to capture variables at relevant hierarchical levels to enable such analysis.", "key words": "multilevel modelling, technology adoption, water infrastructure, hierarchical linear model, Kenya", "contribution statement": "This study provides a novel methodological template for analysing infrastructure adoption data with a clustered structure, demonstrating that institutional factors at the county