Abstract:

Background:
In flood-prone Japan, individual evacuation behavior plans for vulnerable people during disasters are drawing attention; however, the possibility that anyone can become temporarily vulnerable—forming a hidden disaster-vulnerable population—and may develop disaster-related overconfidence remains underrecognized. University students, particularly recent relocatees with limited local familiarity, are an example of such a population and may face heightened vulnerability during evacuation.

Objective:
To examine whether an interactive learning opportunity—sharing predecessors’ experiences and emotions prior to evacuation planning using a flood hazard map—changes students’ evacuation-related awareness and reduces overconfidence.

Methods:
We enrolled 335 students at University B (Prefecture A) in a face-to-face workshop. Before and after the workshop, participants completed online questionnaires assessing confidence in receiving information and taking evacuation actions, as well as perceived difficulty in both (0–10 scales), in addition to confidence in physical strength, information understanding, and digital-device use (0–5 scales). The workshop explained how to use a flood hazard map (identifying hazards, selecting sites, considering routes, defining a starting point, and recognizing the value of walking routes), followed by risk communication through sharing realistic concerns expressed by peers and predecessors (e.g., landslide-prone paths, river routes, evacuating with older adults, and unfamiliar settings). Participants then conducted evacuation planning individually. Analyses employed χ² tests, Fisher’s exact test, and t-tests.

Results:
Of the 330 complete cases, mean confidence in receiving information increased from 5.70 to 6.22, and confidence in taking evacuation actions increased from 5.03 to 5.73; perceived difficulty slightly declined. Per participant, 0/1/2/3/4 indicators improved in 45.5%/31.2%/17.6%/3.9%/1.8% of cases, respectively. The fully corrected group (all four indicators improved) initially reported higher confidence in physical strength, information understanding, and digital-device use. Prior experience with evacuation planning (considering routes or shelters) was associated with a greater number of corrected indicators. When prioritizing evacuation criteria, the corrected group emphasized safety (outside flood and landslide zones) over proximity. Free-text responses noted alternative routes, avoidance of rivers, wide and clearly visible roads, multiple-hazard scenarios, and route changes based on conditions. Viewing a flood hazard map alone prior to the workshop was associated with only partial correction, suggesting that passive exposure without dialogue is insufficient.

Implications:
Combining evacuation planning with risk communication links factual maps with empathy, reflection, and realistic decision-making. This approach aligns with education for sustainable development and community-based disaster risk reduction and may help reduce evacuation delays driven by overconfidence.

Conclusion;
Evacuation planning conducted after sharing predecessors’ experiences yielded educational benefits—including a reduction in disaster-related overconfidence—and reinforced awareness that anyone may become part of a hidden disaster-vulnerable population. This model is transferable to universities and communities to strengthen disaster literacy and preparedness.

Keywords:

disaster-related overconfidence; hidden disaster-vulnerable population; experiential learning; disaster risk communication; community-based disaster prevention.