"background": "Municipal infrastructure asset management systems are critical for sustainable urban development in sub-Saharan Africa. Prior studies on their adoption have often relied on single-level analytical models, which may not adequately account for hierarchical data structures inherent in municipal governance. ", "purpose and objectives": "This replication study aims to re-evaluate the determinants of system adoption rates by implementing a rigorous multilevel regression analysis. The core objective is to conduct a methodological diagnostic of the analytical approach, testing the robustness of previously reported findings. ", "methodology": "We replicated a multilevel regression analysis using administrative panel data. The model specification was y{ij = \0 + \1Xij + uj + eij, where i denotes assets and j municipalities. Estimation used restricted maximum likelihood with robust standard errors to account for clustering. ", "findings": "The replication confirms a significant positive association between dedicated technical staff and adoption likelihood, with an odds ratio of 2. 45 (95% CI: 1. 78 to 3. 37). However, the methodological diagnostics revealed that the intra-class correlation coefficient was substantially lower than previously estimated, indicating that municipality-level random effects explain less variation than originally reported. ", "conclusion": "While the core directional finding on technical capacity is robust, the methodological reassessment suggests that the hierarchical structure of the data is less influential than the original study implied. This underscores the importance of model diagnostics in multilevel analyses of engineering systems. ", "recommendations": "Future research should prioritise the collection of longitudinal asset condition data to strengthen panel analyses. Practitioners should focus on building technical capacity at the municipal level, as this is a confirmed lever for improving system adoption. ", "key words": "asset management, multilevel modelling, replication study, infrastructure governance, regression diagnostics", "contribution statement": "This study provides a novel methodological diagnostic of hierarchical modelling in infrastructure management research, demonstrating that the significance of municipality-level clustering can be overstated without proper
Mensah et al. (Sun,) studied this question.