Abstract
Abstract
Public avoidance behaviour following radiological events can persist even when authoritative assessments judge the associated exposure to be negligible, posing a persistent challenge for radiological protection communication. Taking Japan's discharge of Advanced Liquid Processing System (ALPS)-treated water from the Fukushima Daiichi site into the Pacific Ocean as a case, this study asks why comparable levels of perceived risk translate into avoidance behaviour for some individuals but not others. Drawing on risk perception theory, the trust literature and dual-process models of information processing, we examine how nuclear risk perception, institutional trust, factual knowledge, acceptance of misinformation and modes of information processing shape whether people avoid Japanese seafood and products, forgo travel to Japan and discourage others from travelling. Using survey data from a quota sample of 1,898 South Korean adults, we estimate hierarchical and moderated regression models of avoidance behaviour. Risk perception was the strongest predictor and established the baseline impulse to avoid, with event-specific accident risk perception exerting a stronger effect than general nuclear risk perception. Institutional trust was associated with lower avoidance and misinformation acceptance with higher avoidance, while the two knowledge measures operated in opposing directions. Systematic and heuristic processing further conditioned the influence of trust, household income and risk perception, indicating that these factors operate as mechanisms linking perception to behaviour rather than as interchangeable predictors. For radiological protection, the results suggest that communication addressing affective risk perceptions, sustained trust-building and active countering of misinformation, tailored to how audiences process information, may be more effective than technical reassurance alone in managing public responses to contested exposure situations. Treating trust, misinformation and information processing as the machinery converting perception into behaviour offers a testable basis for anticipating when public response and technical assessment diverge.