Positive relationship between education level and risk perception and behavioral response: A machine learning approach

PLOS ONE | https://doi.org/10.1371/journal.pone.0321153 April 3, 2025 9 / 10 PLOS ONE Positive relationship between education level and risk perception and behavioral response 5. Conclusion This study contributes to the growing literature on machine learning in risk research, offering actionable insights to improve public preparedness and resilience. This study examined the impact of education level on risk perception and behavioral response using advanced ML techniques, including Random Forest and XGBoost, with SHAP values for interpretability. The findings highlight education level as a key contextual factor, with higher education linked to more proactive preparatory behaviors and refined risk per- ceptions. Although the models’ predictive accuracy was limited, they effectively identified key factors and emphasized the role of education in shaping risk-related decisions. These insights underscore the importance of context-specific evidence in risk management and policy for- mulation. Future research should expand on these findings by incorporating larger datasets and exploring additional contextual variables to further refine our understanding of risk perception. This study contributes to the growing literature on ML in risk research, offering actionable insights to improve public preparedness and resilience. Supporting information S1 File. Data. (XLSX) Author contributions Conceptualization: Zhichun Zhang. Data curation: Liping Guo. Investigation: Liping Guo, Wenjie Zhou, Kehu Yang. Methodology: Zhipeng Wei. Resources: Zhichun Zhang, Kehu Yang. Validation: Wenjie Zhou.

Writing – original draft: Zhipeng Wei. Writing – review & editing: Zhipeng Wei. References 1. Siegrist M, Árvai J. Risk Perception: Reflections on 40 Years of Research. Risk Analysis. 2020;40(S1):2191–206. https://doi.org/10.1111/risa.13599 2. Mitsushita K, Murakoshi S, Koyama M. How are various natural disasters cognitively represented?: a psychometric study of natural disaster risk perception applying three-mode principal component analy- sis. Nat Hazards. 2022;116(1):977–1000. https://doi.org/10.1007/s11069-022-05708-x 3. Bodas M, Peleg K, Stolero N, Adini B. Risk Perception of Natural and Human-Made Disasters-Cross Sectional Study in Eight Countries in Europe and Beyond. Front Public Health. 2022;10:825985. https:// doi.org/10.3389/fpubh.2022.825985 PMID: 35252099 4. Ge Y, Yang G, Wang X, Dou W, Lu X, Mao J. Understanding risk perception from floods: a case study from China. Nat Hazards (Dordr). 2021;105(3):3119–40. https://doi.org/10.1007/s11069-020-04458-y PMID: 33424123 5. Ng SL. Effects of Risk Perception on Disaster Preparedness Toward Typhoons: An Application of the Extended Theory of Planned Behavior. Int J Disaster Risk Sci. 2022;13(1):100–13. https://doi. org/10.1007/s13753-022-00398-2 6. Fang X, Cao L, Zhang L, Peng B. Risk perception and resistance behavior intention of residents living near chemical industry parks: an empirical analysis in China. Nat Hazards. 2022;115(2):1655–75. https://doi.org/10.1007/s11069-022-05611-5