Journal Design Engineering Masthead
African Civil Engineering Journal | 12 March 2014

A Bayesian Hierarchical Model for Evaluating Manufacturing Systems Adoption in Senegal

A Methodological Assessment, 2000–2026
M, o, u, s, s, a, S, a, r, r, ,, A, m, i, n, a, t, a, D, i, o, p
Bayesian hierarchical modellingmanufacturing adoptionindustrial policyprobabilistic assessment
Bayesian hierarchical model quantifies regional adoption disparities in manufacturing systems.
Posterior estimates show industrial corridor uptake 2.3 times higher than other regions.
Framework captures unobserved heterogeneity through region-specific parameters.
Probabilistic outputs provide transparent uncertainty for policy targeting.

Abstract

{ "background": "The adoption of advanced manufacturing systems in developing economies is a critical driver of industrialisation, yet robust methodological frameworks for evaluating its progress are lacking. Existing assessments often rely on aggregate indicators that mask regional and sectoral heterogeneity.", "purpose and objectives": "This paper develops and assesses a novel Bayesian hierarchical modelling framework to evaluate the adoption rates of manufacturing plant systems. The objective is to provide a methodologically rigorous tool for capturing spatial and temporal variations in technological uptake.", "methodology": "A Bayesian hierarchical model is constructed, integrating plant-level survey data with regional economic covariates. The core adoption rate for plant $i$ in region $j$ is modelled as $\\text{logit}(p{ij}) = \\alphaj + \\beta X{ij}$, with $\\alphaj \\sim \\text{Normal}(\\mu{\\alpha}, \\sigma^2{\\alpha})$. Inference uses Markov chain Monte Carlo simulation, with posterior distributions summarising parameter uncertainty.", "findings": "The model reveals significant regional disparities, with adoption rates in the industrial corridor approximately 2.3 times higher than in other regions (95% credible interval: 1.8 to 2.9). The hierarchical structure effectively captured unobserved regional heterogeneity, with the hyperparameter $\\sigma_{\\alpha}$ having a posterior median of 0.42.", "conclusion": "The proposed Bayesian hierarchical model offers a superior methodological framework for evaluating manufacturing systems adoption, providing probabilistic estimates that transparently communicate uncertainty and regional variation.", "recommendations": "Policymakers should utilise such probabilistic models to target regional interventions. Future research should integrate this framework with longitudinal data to analyse adoption pathways.", "key words": "Bayesian statistics, hierarchical model, technology adoption, manufacturing systems, industrial policy, probabilistic modelling", "contribution statement": "This paper introduces a novel Bayesian hierarchical modelling framework, specifically tailored to the context of developing economies, which quantifies regional adoption disparities and associated uncertainties more effectively than prior deterministic methods."