Methodological Evaluation and Multilevel Regression Analysis of Process-Control System Adoption in Tanzanian Industry

The adoption of modern process-control systems in industrial sectors of developing economies is a critical driver of productivity and quality. However, systematic, evidence-based analysis of the factors influencing adoption rates in these contexts remains limited, particularly in Sub-Saharan Africa. This study aims to methodologically evaluate the determinants of process-control system adoption within Tanzanian industry and to quantify their relative influence using a robust statistical framework. A multilevel regression model was specified to analyse firm-level survey data, accounting for nested structures within industrial sectors. The model, yᵢⱼ = β₀ + β₁Xᵢⱼ + uⱼ + eᵢⱼ, where i denotes firms and j sectors, was estimated using maximum likelihood with robust standard errors to correct for heteroscedasticity. Capital investment capacity was the strongest predictor of adoption (β = 0.47, p < 0.001). A significant random intercept variance (σ² = 0.31, 95% CI [0.12, 0.50]) indicated substantial unexplained variation between sectors. Only 34% of surveyed manufacturing firms reported using advanced programmable systems. The adoption of sophisticated control technologies is uneven and strongly linked to financial resources, while sector-specific characteristics exert a significant but poorly understood moderating effect. Industrial policy should integrate targeted financial mechanisms to alleviate capital constraints. Further research must investigate the sector-level random effects to inform tailored sectoral strategies. industrial automation, technology adoption, multilevel modelling, manufacturing, developing economies This paper provides a novel methodological application of multilevel regression to technology adoption in an African industrial context, yielding a new, sectorally-stratified dataset and identifying capital constraints as the primary adoption barrier.