Abstract
{ "background": "Cost-effectiveness analysis in manufacturing systems within developing economies is often hampered by sparse, heterogeneous data and the need to integrate multi-level operational uncertainties. Traditional deterministic models fail to adequately quantify these uncertainties, limiting robust decision-making for plant investment and policy.", "purpose and objectives": "This study presents a methodological evaluation of a novel Bayesian hierarchical model designed for cost-effectiveness analysis of manufacturing systems. The objective is to provide a robust framework that quantifies uncertainty and borrows strength across related manufacturing units to improve inference where data are limited.", "methodology": "The proposed model is structured as $y{ij} \\sim \\text{Normal}(\\mu + \\alphai + \\beta x{ij}, \\sigma^2)$, with $\\alphai \\sim \\text{Normal}(0, \\tau^2)$, where $y{ij}$ is a cost-effectiveness metric for plant $i$ under condition $j$, $\\alphai$ captures plant-level random effects, and $x_{ij}$ denotes covariates. The methodology was evaluated using simulated data reflecting Rwandan industrial conditions and a case study of agro-processing plants. Posterior distributions were estimated using Markov chain Monte Carlo sampling.", "findings": "The model demonstrated superior performance in uncertainty quantification compared to frequentist alternatives, with 95% credible intervals for incremental cost-effectiveness ratios achieving nominal coverage in simulation. A key finding was that incorporating hierarchical structure reduced the width of credible intervals for plant-specific estimates by approximately 22% on average, indicating significantly improved precision.", "conclusion": "The Bayesian hierarchical model offers a statistically rigorous methodological advance for cost-effectiveness analysis in data-scarce manufacturing contexts. It effectively synthesises information across a system while providing full probabilistic inference on key economic parameters.", "recommendations": "Adoption of this modelling framework is recommended for engineers and policymakers conducting techno-economic evaluations of manufacturing systems in similar developing economies. Future work should focus on integrating time-series data and non-Gaussian likelihoods for broader applicability.", "key words": "Bayesian