Revisiting the impact of public spaces on the mental health of rural migrants in Wuhan: an integrated multi-source data analysis

Current research on public spaces and mental health often focuses on the independent relationship of one or more social mediators, neglecting the nuanced implications and serial mechanisms inherent in the progressive social process. Using Wuhan city, China, as a study case with multi-source data, this research applies Multilevel Generalized Structural Equation Modeling and deep learning techniques to explore the differential effects of public spaces with varying degrees of publicness (i.e., typical, semi-, and privately owned) on rural migrants’ mental health. Crucially, this study scrutinizes both explicit (social interaction) and implicit (perceived integration) social mechanisms to revisit the relationships. The findings reveal that not all public spaces equally influence mental health, with typical and privately owned public spaces conferring profound benefits. Notably, public spaces impact mental health chiefly through perceived integration instead of through direct social interaction. Social interaction improves mental health primarily by enhancing perceived integration, suggesting that meaningful connections beyond superficial encounters are critical. In particular, we observed significant social effects in typical and privately owned public spaces but limited social functionality in semi-public spaces. This evidence contributes to the knowledge required to create supportive social environments within public spaces, integral to nurturing inclusive urban development.