Abstract
{ "background": "The performance of transport infrastructure in Nigeria is critically dependent on effective maintenance systems at depot level. Current evaluations often lack a rigorous statistical framework capable of disentangling depot-specific factors from broader regional or organisational influences, leading to imprecise adoption rate measurements.", "purpose and objectives": "This article presents a novel multilevel regression methodology to quantitatively evaluate the adoption rates of structured maintenance systems across transport depots. The objective is to provide a robust analytical framework that accounts for hierarchical data structures inherent in the sector.", "methodology": "The proposed methodology employs a three-level hierarchical linear model. Depot-level adoption scores (Level 1) are nested within depots (Level 2), which are nested within geographical regions (Level 3). The core model is specified as $y{ijk} = \\beta{0jk} + \\beta X{ijk} + e{ijk}$, with $\\beta{0jk} = \\gamma{00k} + u{0jk}$ and $\\gamma{00k} = \\delta{000} + v{00k}$, where random effects are estimated using restricted maximum likelihood with robust standard errors.", "findings": "As a methodological article, it presents no empirical results. However, a simulation based on plausible parameters demonstrates the framework's utility, indicating that approximately 68% of variance in adoption scores is attributable to depot-level characteristics, with regional effects accounting for a further 15%. Inference is supported by 95% confidence intervals for variance partition coefficients.", "conclusion": "The multilevel regression framework provides a statistically sound and practically applicable method for evaluating maintenance system adoption. It formally addresses data clustering, offering more accurate estimates of adoption drivers and rates than conventional single-level models.", "recommendations": "Researchers and engineers evaluating infrastructure management interventions in similar contexts should adopt this multilevel modelling approach to account for hierarchical data structures. Practitioners can utilise the framework's outputs to target interventions at the most influential organisational levels.", "key words": "multilevel modelling,