Determinants, constraints, and statistical tools in poverty research: A PRISMA-ScR-based scoping review with comparative insights from Nepal, Asia, and Africa

This study evaluates the literature on the determinants of poverty and the statistical approaches used to quantify them. It uses the different databases Scopus, Google Scholar, Springer Nature Link, Wiley Online Library, and Emerald Insight to identify forty-two original and peer-reviewed articles published between 2013-2025, based on the PRISMA-ScR scoping review. Altogether, the review identified 63 variables across 42 studies from 23 countries that focus on the interaction between economic, demographic, socio-cultural, geographical, and environmental factors, and their influence on poverty rates; however, economic factors (33.33%) were found to be most frequently investigated. The most frequently studied variables include household size, education level, age, gender, and employment status. The findings reveal significant regional disparities. African studies are focused on agricultural livelihoods, financial access, household size, employment status, and climate vulnerability, while Asian studies focus more on education level, age, gender, marital status, income, rural/urban status, and remittances, where Nepal shows an integration of structural inequality and environmental effects. Although logistic regression remains the dominating statistical tool, advanced tools, such as Machine learning, multilevel modelling, and panel data models, enhance poverty dynamics research. Multidimensional poverty-diminishing policies are emphasized by this study to address context-specific and structural disparities through education, employment, and social protection. The underexplored factors, such as climate-related shocks and long-term economic trends, can be used for further research, needed to improve poverty alleviation initiatives.