Unpacking the Black Box of Training Effectiveness: A Structural Equation Modelling Analysis

Background The global health workforce shortage in low- and middle-income countries (LMICs) remains a critical challenge. The quality of health personnel significantly impacts population health, particularly in LMICs facing severe workforce shortages. While development partners invest in training programmes to improve their quality, rigorous evaluation of these programmes remains rare, with limited evidence on the structural and contextual factors influencing outcomes. This study aims to fill this gap by using structural equation modelling (SEM) to evaluate Japan's Official Development Assistance (ODA) capacity-building training programmes conducted in Japan for LMIC health workforces and to identify key structural factors to guide future workforce development strategies. Methods A cross-sectional questionnaire-based study was conducted from January to June 2023 in Laos and Mongolia. Exploratory factor analysis (EFA) was applied to 97 items, followed by Wilcoxon rank-sum tests and SEM with robust estimation in STATA (BE18). Model refinement excluded unstable/non-significant factors. Results A total of 148 participants were included: 90 from Mongolia and 58 from Lao People's Democratic Republic. The EFA identified six latent factors influencing training effectiveness. Wilcoxon tests showed significant differences (p < 0.0005) in training effectiveness for five factors. During the systematic refinement of SEM, two factors were excluded: Factor 5 (Limited Authority) due to non-significant effects in the Wilcoxon rank-sum test, and Factor 6 (Altruistic Desire) due to construct instability (comprising only two items) and limited contribution to explaining variance. Model fit was evaluated using established cutoff criteria for SEM fit indexes. Conclusions The findings highlighted that training effectiveness is not merely an outcome of the training event itself but a culmination of factors operating before, during, and after the training. The findings offer several key insights into optimising training design and implementation in similar contexts for development partners and national and regional governments.