Prioritizing Renewal Needs of Existing Student Dormitories Through Asymmetric Impact Analysis of Environmental Attributes on Residential Satisfaction

The sustainable renewal of existing university student dormitories requires effective identification of occupant-perceived improvement needs to support targeted interventions and efficient resource allocation. This study develops a renewal need prioritization framework by integrating explainable machine learning and impact-asymmetry analysis (IAA). Based on 1507 valid questionnaires from 10 campuses of seven universities in Guangdong Province, an XGBoost-SHAP model was used to identify key environmental attributes influencing overall satisfaction, followed by IAA to examine their asymmetric effects. The results show that: (1) 17 key attributes were identified from 42 indicators, of which 15 met the criteria for IAA classification and exhibited distinct asymmetric effects; (2) a five-level renewal need prioritization framework was established, with per capita usable area, bathroom type, storage space, study and discussion space, damp and mold prevention, and sound insulation identified as core renewal needs; and (3) improving residential unit comfort through spatial optimization, furniture configuration, and environmental enhancement is an important pathway for enhancing occupant well-being and the long-term adaptability of existing dormitory buildings. This study provides empirical evidence for prioritizing existing dormitory renewal and allocating limited resources more efficiently, contributing to the sustainability of existing university residential environments.