Risk Perception and Disaster Information Verification Among Chinese University Students Exposed to Simulated Social Media Posts

Summary

This randomized factorial experiment evaluated how source cues, information veracity, and emotional tone in simulated disaster-related social media posts affect Chinese university students' risk perception, credibility judgments, verification intentions, sharing intentions, and task-based verification choices.

Abstract

Disaster-related information increasingly circulates through social media during emergencies, but limited experimental evidence has examined how university students evaluate and verify such messages. This randomized 2 × 2 × 2 factorial experiment tested the effects of source cue, information veracity, and emotional tone in simulated disaster-related social media posts among Chinese university students. A total of 426 valid responses were analyzed. Misleading information and high-threat wording increased perceived risk, whereas official-source cues increased perceived credibility and sharing intention. Verification intention was more sensitive to misleading content and urgent wording, but task-based verification behavior showed a clearer association with source cues than with veracity or tone. The binary verification accuracy model provided weaker support than the continuous behavioral score. These findings suggest that risk perception, credibility judgments, intentions, and verification choices may diverge in the context of disaster misinformation. Campus risk communication should therefore emphasize concrete verification routines rather than relying only on credibility cues or general reminders to check information before sharing.

Introduction

Social media has become an important channel for disaster-related communication because it allows rapid dissemination of alerts, eyewitness updates, resource information, and public responses during emergencies. At the same time, the speed and openness of social media create conditions in which inaccurate or unverified disaster information can circulate before formal correction occurs. Recent work on social media data for disaster risk management has noted that social platforms can support preparedness and response, but their usefulness is limited by concerns about data accuracy, misinformation, privacy, and verification¹.

Protocol

Study design

Before any recruitment or data collection, the study protocol was reviewed and approved by the Research Ethics Committee of the University of Malaya (approval ID: UUM20259910). All procedures involving human participants were conducted in accordance with institutional requirements and the Declaration of Helsinki. Participants provided electronic informed consent before entering the questionnaire, and no personally identifiable information was exported for analysis.

This study used a randomized between-subjects online experiment to examine risk perception and disaster information verifi....

Results

Participant characteristics and experimental allocation

After eligibility screening, consent confirmation, completion checks, attention-check screening, and data-quality review, 426 valid responses were retained for analysis. The final sample included 171 male participants, 239 female participants, and 16 participants who preferred not to report gender. The mean age was 21.01 years (SD = 1.96). The sample included 388 undergraduate students and 38 postgraduate students. Partic.......

Discussion

This randomized experiment examined how Chinese university students responded to simulated disaster-related social media posts that varied by source cue, information veracity, and emotional tone. The results showed three main patterns. First, misleading disaster information and high-threat wording increased perceived risk. Second, official-source cues increased perceived credibility and sharing intention. Third, task-based verification behavior was more strongly associated with source cues than with information veracity .......

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
IBM SPSS Statistics	IBM Corp.	Version 27.0	Conduct descriptive statistics, reliability analysis, t tests, chi-square tests, factorial ANOVA, and general linear models
Online survey platform	Wenjuanxing/Questionnaire Star or equivalent online survey system	Online survey system; version not applicable	Administer eligibility screening, electronic informed consent, baseline questionnaire, random assignment, stimulus exposure, post-exposure measures, and verification task
R statistical software	R Foundation for Statistical Computing	Version 4.3	Conduct statistical analysis, robustness checks, logistic regression, and bootstrap-based secondary analyses
Statistical syntax/codebook files	In-house	SPSS/R syntax and codebook	Document variable coding, exclusion rules, composite-score construction, and model specifications
Stimulus screenshot-style interface templates	In-house	Fictional Chinese social media post layout	Present simulated disaster information in a realistic social media format without using real disaster events or real government announcements