Beyond Single-Cause Explanations: An Integrated Framework for Crisis Learning in Public Organisations: A Survey Experiment in China

Abstract

Drawing on performance feedback, attribution, and accountability theory, this study investigates how the frequency of crises, the type of causal attribution, and the geographic location of an event jointly influence crisis learning among Chinese public sector actors. The study used a between-subjects vignette experiment in which two crisis categories – industrial accidents and public health emergencies – were each varied along a 2 × 2 × 2 factorial design, yielding 16 experimental conditions in total. A total of 557 civil servants within Shandong Province's emergency management system participated in the experiment. Results indicate that high crisis frequency is associated with significantly stronger learning in accident-disaster scenarios, whereas the frequency estimate does not reach statistical significance in public-health scenarios. Internal causal attributions are associated with greater learning, yet this effect does not reach significance on its own. Events occurring within the respondent's province show a modest positive association with crisis learning. In addition, combination comparisons show lower learning scores for low-frequency incidents with external causes and for low-frequency, out-of-province events. These findings point to the value of examining crisis attributes jointly and in context when studying crisis learning. Together, these findings offer an integrated, empirically grounded account of how crisis attributes shape learning and can inform the design of emergency-management learning interventions.