<![CDATA[Background: Flood-prone communities need accessible education that supports preparedness and collective action. Evidence on chatbot-supported flood education in community settings remains limited.Purpose: To assess changes in flood-related knowledge, attitudes, and mitigation practices following PREPARE-BANJIR chatbot use, describe participant engagement, and document the initial formation of community flood-response groups.Methods: A pre-experimental, one-group pretest–posttest study involved 150 residents purposively recruited from flood-prone communities along the Brantas River in Kediri Regency, Indonesia. The intervention comprised six chatbot-based educational modules. Outcomes were assessed using a 48-item questionnaire, while engagement was documented through usage logs. Paired changes were analyzed using Wilcoxon signed-rank tests for knowledge and attitudes and a paired-samples t-test for mitigation scores.Results: Mean knowledge scores increased from 7.79 to 9.93, with a mean change of 2.14 (95% CI: 1.86–2.42). Attitude scores increased from 41.79 to 46.97, with a mean change of 5.18 (95% CI: 4.43–5.93), and mitigation scores increased from 18.55 to 23.73, with a mean change of 5.18 (95% CI: 4.42–5.94). All paired comparisons were statistically significant (p < 0.001). Overall, 79.3% of participants completed all modules. Five flood-response groups, each comprising 30 members, were established.Conclusion: PREPARE-BANJIR use was followed by improved preparedness-related scores and the establishment of community response groups. These preliminary findings warrant controlled evaluations to determine intervention effectiveness and the sustainability of individual and group outcomes.]]>
