Ethnic integration remains a critical challenge for achieving sustainable urban development, particularly in regions where social cohesion and equity have not kept pace with economic progress. This study investigated the determinants of community residential satisfaction in multi-ethnic inter-embedded urban communities in Guangxi, China, a governance model designed to promote spatial, institutional, and cultural integration among diverse ethnic groups. Drawing on embeddedness and social capital theories, an integrative framework links community social networks, social organizations, community development and services, and perceptions of community engagement to community residential satisfaction. Using survey data from 400 residents across eight inter-embedded communities, a dual-stage Structural Equation Modeling (SEM) and Artificial Neural Network (ANN) approach was adopted to capture both causal and nonlinear relationships. SEM results showed that community development and services, social organizations, and perceptions of community engagement significantly enhanced community residential satisfaction, while social networks exerted no significant direct effect. ANN analysis further ranked the relative importance of predictors as community development and services (CDS) > community social organizations (CSO) > perceptions of community engagement (PCE) > community social networks (CSN). These findings indicated that institutional and participatory mechanisms outweighed interpersonal ties in shaping satisfaction. The study contributed by integrating social and institutional embeddedness into sustainable community research, thus demonstrating the utility of the SEM-ANN hybrid approach. Morever, it provided practical insights for strengthening participatory governance, improved community services, and fostered inclusive and cohesive multi-ethnic urban communities.
Xiaona et al. (Tue,) studied this question.