{ "background": "Systemic risk in manufacturing facilities within developing economies is a critical engineering challenge, often addressed through fragmented diagnostic approaches. There is a recognised need for integrated analytical frameworks that account for hierarchical operational structures.", "purpose and objectives": "This working paper develops and evaluates a multilevel regression methodology for the diagnostic assessment of risk reduction interventions within complex industrial plant systems. The objective is to provide a robust statistical tool for quantifying the efficacy of safety and reliability measures across different organisational levels.", "methodology": "A multilevel modelling approach is constructed, nesting production units within plants. The core model is specified as $y{ij} = \\beta{0j} + \\beta{1}x{1ij} + e{ij}$, with $\\beta{0j} = \\gamma{00} + \\gamma{01}z{1j} + u{0j}$, where $i$ denotes units and $j$ plants. Inference is based on restricted maximum likelihood estimation with robust standard errors to account for heteroscedasticity.", "findings": "The methodological evaluation, applied to a diagnostic dataset, indicates that plant-level management system interventions explain approximately 40% of the variance in unit-level risk reduction metrics. The random intercept for plants was statistically significant (p < 0.01), confirming the necessity of the hierarchical structure.", "conclusion": "The multilevel regression framework provides a statistically sound and operationally relevant diagnostic tool for engineering risk management, effectively capturing the nested reality of manufacturing systems.", "recommendations": "Adoption of this modelling approach is recommended for plant engineers and safety analysts to prioritise interventions. Further research should focus on integrating time-series data to model dynamic risk pathways.", "key words": "multilevel modelling, risk diagnostics, industrial safety, reliability engineering, systems analysis, Uganda", "contribution statement": "This paper introduces a novel application of multilevel regression for plant-system risk diagnostics, providing a validated method that disentangles unit-level and plant-level effects on risk reduction