Sentiment polarity and risk amplification in online public opinion: A configurational analysis of university crisis events

Online public sentiment can shift rapidly during university crises, with consequences for institutional governance. Using fuzzy-set qualitative comparative analysis (fsQCA) of 50 university crisis events in China, this study examines how public opinion object, response timeliness, event category, event influence, and school tier combine to shape event-level sentiment valence. Three configurations are associated with comparatively favorable sentiment outcomes: student-centered resonance, media-leveraged responsiveness, and structure-synchronized reputation. Complementary regression analyses identify response timeliness and event category as the most consistent net correlates, while school tier matters in conjunction with other conditions. By integrating configurational and regression evidence, the study extends crisis-communication research on sentiment dynamics and risk amplification and offers a pathway-based framework for institutional response.