A Meta-Analysis of Socio-Economic Determinants and a Machine Learning Model for Predicting Depression and Anxiety Using the Global Flourishing Study Data

This project has two primary aims. First, to conduct a meta-analysis of 22 independent country samples quantifying the association between key socio-economic determinants (e.g., income, education, employment) and symptoms of depression and anxiety. While the Global Flourishing Study is a single, coordinated data-collection effort, each of the 22 countries represents a distinct population with its own sampling frame, justified cultural adaptations, and unique national context. Treating each country as a separate "study" in a meta-analytic framework allows us to estimate pooled effects while formally quantifying and modeling between-country heterogeneity, a core strength of meta-analysis that would be lost if we simply pooled all respondents into a single regression model. Second, to develop and validate a machine learning (ML) model to predict individual-level risk for depression and anxiety, identifying the most important predictors from a wide range of demographic, social, and economic variables.