3.1. DRR Education and Community Resilience

DRR education is increasingly recognized as a fundamental component of community resilience in hazard-prone regions ^4,5. Education plays an important role in strengthening risk awareness, preparedness, and adaptive capacity among populations exposed to natural hazards.

Moreover, its importance is reflected in international policy frameworks. The Sendai Framework for Disaster Risk Reduction (2015–2030) identifies education as a key mechanism for enhancing preparedness and fostering a culture of safety across societies ⁶. Within this context, schools are widely regarded as essential institutions for disseminating disaster knowledge and shaping preparedness behaviors among younger generations ⁷.

Empirical studies reveal that school-based disaster education can improve children’s hazard recognition, evacuation knowledge, and preparedness behavior ⁸. Research in countries such as Japan and Indonesia indicates that structured programs combining classroom learning, evacuation drills, and experiential activities enhance students’ preparedness capabilities ⁹. These findings suggest that schools function not only as educational institutions, but also as important actors in strengthening community resilience through knowledge transmission and behavioral training ¹⁰.

However, most studies primarily focus on learning outcomes among children; relatively few have examined whether school-based disaster education produces measurable impact beyond the classroom, particularly on family environments and household preparedness practices.

3.2. Children as Agents of Change and Intergenerational Knowledge Transfer

Recent scholarship increasingly challenges the traditional view of children as passive disaster victims. Instead, children are now recognized as potential “agents of change” capable of influencing disaster preparedness within families and communities ¹¹.

This perspective is grounded in social learning theory and knowledge-sharing frameworks ¹². Social learning theory suggests that behaviors and attitudes are shaped by observation, modeling, and interactions within social environments ¹³. When children acquire preparedness knowledge through school-based education, they may share this knowledge with family members or demonstrate behaviors that parents observe and adopt ¹⁴.

Knowledge-sharing research suggests that information can circulate through social networks and across generations within households. In disaster contexts, such processes may stimulate family discussions about hazards, evacuation strategies, and preparedness measures ¹⁵.

Empirical evidence for these intergenerational mechanisms remains limited but is suggestive. Studies in Japan report that child-focused disaster education activities can stimulate household discussions and encourage preparedness behavior among parents ¹⁶. Similarly, studies involving children with foreign backgrounds reveal that children may function as important transmitters of disaster knowledge to parents who lack direct access to institutional disaster education ¹⁷.

However, systematic quantitative evaluations of child-to-parent knowledge diffusion are scarce, particularly in Southeast Asia. Many studies rely on qualitative observations or perception-based assessments, rather than statistical analyses of pre–post interventions. Consequently, the pathways through which children’s disaster education influences household preparedness remain insufficiently examined.

3.3. Cultural Embeddedness in Disaster Education

In addition to researching intergenerational learning, scholars have emphasized the importance of culturally grounded approaches in disaster education ¹⁸. Preparedness messages embedded within locally meaningful practices, such as oral narratives, songs, rituals, or community performances, may be more readily accepted and remembered ¹⁹.

Studies on indigenous knowledge systems reveal that traditional storytelling and cultural narratives have long functioned as mechanisms for transmitting survival knowledge across generations ^20,21. One widely cited example is the Smong narrative from Simeulue Island, Indonesia, which suggests that strong earthquakes prompt immediate evacuation to higher ground. During the 2004 Indian Ocean tsunami, long-term transmission of this narrative contributed to rapid evacuation behavior and remarkably low casualty rates on the island ²².

The Smong case demonstrates how culturally embedded knowledge can support effective disaster risk reduction through intergenerational transmission ^23,24,25. However, such systems are highly context-specific. As the language, traditions, and social practices of Simeulue differ substantially from those of Nias Island, directly transferring the Smong narrative is unlikely to generate the same cultural resonance.

Recognizing these differences, this study adopted a locally grounded approach by integrating disaster-prevention messages into Maena. Unlike externally introduced teaching materials, Maena functions as an established medium for collective expression and social interaction.

Embedding disaster knowledge in this familiar cultural performance sought to reproduce, in the Nias context, a mechanism similar to that observed in the Smong case: the transmission of preparedness knowledge through culturally meaningful and socially shared practices. Therefore, the Disaster Prevention Maena-based project can be understood as an attempt to create a locally adapted pathway for disaster knowledge transmission.

Recent studies in Indonesia have explored integrating local wisdom and traditional practices into disaster education ²⁶. These initiatives suggest that culturally familiar frameworks can enhance engagement, community participation, and legitimacy of preparedness programs ^27,28.

However, although culturally embedded disaster education has revealed promising results in improving student awareness, few studies have quantitatively examined whether such programs influence preparedness behaviors at the household level. Empirical evidence linking culturally grounded school-based education to intergenerational diffusion remains limited.

3.4. Research Gap and Contribution

Considered together, the literature suggests that school-based disaster education can strengthen community resilience through multiple pathways. Nevertheless, several important gaps remain.

First, most studies focus on children’s educational outcomes rather than household preparedness effects. Second, quantitative evidence of intergenerational knowledge diffusion remains limited. Third, relatively few studies evaluate culturally embedded disaster education using pre–post statistical designs. Fourth, empirical evidence from small island communities in Southeast Asia is scarce.

To address these gaps, this study quantitatively evaluates a culturally embedded disaster education project implemented in elementary schools on Nias Island. By combining matched pre–post survey analyses with child–parent comparisons, this study examines whether improvements in children’s disaster knowledge coincide with measurable changes in parental awareness and household preparedness behaviors ²⁹.

3.5. Integrated Theoretical Framework

This study integrates insights from three theoretical perspectives: social learning theory, intergenerational knowledge transfer, and culturally grounded education.

First, social learning theory suggests that behaviors and attitudes are shaped by observations and interactions within social environments. When children acquire disaster knowledge in school and demonstrate preparedness behaviors, parents may observe these behaviors and incorporate them into household practices ³⁰.

Second, intergenerational knowledge transfer theory emphasizes that knowledge can circulate between generations through everyday communication within families ³¹. In disaster education contexts, children may act as conduits through which institutional knowledge learned at school is transmitted to parents.

Third, culturally grounded education theory highlights that learning embedded within meaningful cultural practices can enhance engagement, legitimacy, and retention of knowledge ³².

Figure 1 presents the conceptual framework of this study. It proposes that culturally embedded disaster education, implemented through the Disaster Prevention Maena-based project, strengthens children’s disaster knowledge and stimulates communication within households. Through these interactions, children’s learning may coincide with increased parental awareness and preparedness behaviors, thereby contributing to household-level resilience.

The framework does not assume direct causality; rather, it illustrates a plausible mechanism linking school-based disaster education with broader household preparedness outcomes. Moreover, it provides a theoretical basis for the empirical analyses presented in Section 7.

5.1. Project Design Matrix

The disaster prevention education initiative was implemented under the framework of JICA’s Grassroots Partnership Support Scheme. The project was designed using a project design matrix to ensure logical coherence between objectives, activities, and expected outcomes.

The overall goal of the project was to strengthen community-level disaster resilience on Nias Island. Recognizing that schools function as central institutions within local communities, the project aimed to foster resilience through a participatory educational model centered on elementary schools.

The project aimed to establish sustainable, participatory, school-based DRR education involving students, teachers, parents, and local education authorities. Rather than limiting intervention to classroom-based knowledge transmission, the project sought to create a collaborative structure in which disaster awareness would circulate between schools and households.

To achieve this purpose, four principal outputs were identified:

1. Developing a locally adapted disaster prevention curriculum;

2. Creating and disseminating “Disaster Prevention Maena” as a culturally embedded educational tool;

3. Implementing systematic evacuation drills in participating schools;

4. Organizing community performances to encourage household-level diffusion.

The underlying logic assumes that culturally embedded disaster education enhances children’s knowledge, stimulates communication within families, and ultimately improves household preparedness.

5.2. Implementation Structure

The program was conducted over a seven-month period from October 2016 to April 2017 and followed a structured, phased implementation process.

Implementation was led by a project team established under JICA’s Grassroots Partnership Support Scheme.

The team included university researchers with expertise in DRR and education. The project leader was one of the authors of this paper and was responsible for the initial institutional design of the program and for supervising the overall implementation progress throughout the project period.

A project researcher was stationed on Nias Island during the implementation period to coordinate daily activities and maintain communication with local stakeholders. Other team members traveled regularly between Nias Island and their home institutions to provide technical support, monitor progress, and participate in key program events.

At the local level, the project closely collaborated with a community-based NGO that had previously engaged in post-earthquake recovery support on Nias Island. This partnership facilitated communication with schools, communities, and local authorities. Further, the program received cooperation from local government agencies, including disaster management and education offices, which supported coordination, outreach, and implementation activities.

The first phase comprised a baseline survey administered to both children and parents. The survey provided diagnostic data regarding disaster knowledge, awareness, and preparedness practices prior to the intervention.

The second phase involved teacher workshops, during which the importance of disaster education was discussed and shared objectives were clarified. These workshops aimed to build institutional and household commitments to the program.

Subsequently, each participating school established a disaster leadership structure, designated responsible personnel, and clarified lines of communication in case of an emergency. Evacuation routes were mapped and formally agreed upon to ensure that preparedness training corresponded to realistic geographic conditions.

The core creative component of the project was developing Disaster Prevention Maena. This process was collaborative and iterative, involving teachers, community members, and students.

Following the creative phase, structured disaster education classes were conducted in classrooms. These classes covered earthquake mechanisms, tsunami generation, evacuation principles, and risk recognition specific to the students’ local environments.

Practical reinforcement was provided through schoolwide evacuation drills conducted in coordination with teachers and local authorities, aiming to transform cognitive understanding into embodied preparedness behavior.

To consolidate learning and encourage broader community engagement, schools participated in a public Disaster Prevention Maena contest and performance event. Local education offices and disaster management authorities were invited, reinforcing institutional recognition.

Finally, a post-intervention survey was administered to evaluate changes in both children’s knowledge and parental awareness.

Through this structured progression, from institutional coordination and diagnosis to instruction, rehearsal, public performance, and evaluation, the project aimed to create an integrated educational ecosystem rather than a one-time instructional event.

The overall implementation of the educational project is summarized in Fig. 2. The figure illustrates the sequential structure of the intervention, including baseline surveys, teacher workshops, disaster education classes, evacuation drills, and the development and performance of Disaster Prevention Maena. By visualizing the program structure, Fig. 2 clarifies how institutional collaboration, educational activities, cultural engagement, and community participation were integrated throughout the intervention period.

5.3. Disaster Prevention Maena: Educational Content and Pedagogical Process

The central pedagogical innovation of the project was the adaptation of Maena, a traditional Nias communal dance, as a medium for disaster education.

The pedagogical development proceeded in six stages. First, students were introduced to the Smong tsunami narrative, a well-known traditional account from neighboring regions emphasizing immediate evacuation after strong ground shaking. This narrative served as a cultural bridge linking indigenous knowledge with contemporary disaster science.

Second, structured instructions were provided on earthquake response behaviors, including self-protection during shaking, tsunami risk, and the need for rapid evacuation to higher ground.

Third, students participated in composing the lyrics, expressing disaster-prevention messages in their own words. This participatory process aimed to enhance understanding and strengthen ownership of the material.

Fourth, teachers and community members collaboratively refined the lyrics into coherent and rhythmically appropriate verses, ensuring both accuracy and performative feasibility.

Fifth, local musicians composed music consistent with traditional Maena rhythms, preserving cultural authenticity.

Sixth, choreography was developed collaboratively. Dance movements were intentionally aligned with key preparedness behaviors, including protecting oneself during shaking, evacuating immediately to high ground, avoiding coastal areas after earthquakes, and recognizing the importance of emergency stockpiling.

By embodying preparedness behaviors through movement, the program sought to transform abstract knowledge into memorable physical expression.

Public performance played an important reinforcing role. When children performed Disaster Prevention Maena at school events and community gatherings (Fig. 3), parents were exposed to disaster-prevention messages in a culturally legitimate and emotionally engaging format. This visibility was expected to encourage household discussion and strengthen intergenerational transmission of knowledge.

Importantly, Disaster Prevention Maena was not a standardized performance imposed uniformly across schools. Each participating school created its own version through collaboration among teachers, students, and community members. Consequently, songs and movements reflected local cultural expressions, and variations emerged across schools and regions.

The lyrics incorporated concrete disaster-prevention messages related to earthquake and tsunami response. Common themes included warnings that strong earthquakes may be followed by tsunamis, the need to evacuate immediately to higher ground, and the dangers of approaching coastal areas after seismic events. Some songs emphasized that people should not observe the sea or rivers after an earthquake. Others warned that even if seawater suddenly recedes and fish appear on the exposed seabed, people should not approach the shore, but should evacuate immediately.

Dance movements corresponded to these messages. Some gestures symbolized protecting oneself during shaking, whereas others represented running toward higher ground. Integrating lyrics, rhythm, and movement, Disaster Prevention Maena translated disaster-prevention knowledge into an embodied cultural performance.

As Maena is traditionally performed collectively during community gatherings, its use in disaster education has allowed preparedness messages to circulate naturally across generations. Thus, it functioned not only as a pedagogical tool for children, but also as a social medium through which disaster preparedness discourse entered family spaces.

6.1. Research Design

This study employed a quasi-experimental pre–post survey design to evaluate the changes in disaster-related knowledge and preparedness associated with implementing a culturally embedded disaster education project. Baseline data were collected before the intervention in September 2016, and follow-up surveys were conducted after project completion in April 2017.

The research design focused on assessing changes at two levels: (1) knowledge gains among children participating in a school-based disaster education project (Fig. 4) and (2) changes in disaster awareness and preparedness behaviors among their parents. By comparing the responses collected before and after the project, this study aimed to identify statistically significant improvements in both groups.

As the educational intervention was implemented as part of a community-based development project rather than as a controlled experimental study, the research design did not include a randomized control group. Instead, the study relied on a combination of matched pre–post comparisons and aggregated cross-sectional analyses to examine changes associated with the intervention period.

In addition, although both children and parents completed the questionnaires, individual household-level linkages between specific children and their parents were not always possible because of the cultural naming practices on Nias Island, where family surnames are not commonly used. Consequently, the analysis focused on aggregated comparisons between student and parent groups rather than on strictly matched parent–child pairs.

Despite these limitations, the design remains appropriate for evaluating project-related changes in community-based educational settings. Pre–post survey comparisons allow the identification of statistically significant improvements in knowledge and preparedness indicators, while cross-tabulation analyses provide insights into the potential associations between child-level learning outcomes and parent-level preparedness behaviors.

Thus, although causal inference cannot be definitively established, the research design enables a systematic examination of whether implementing the Disaster Prevention Maena-based disaster education project coincided with measurable improvements in both children’s disaster knowledge and parental preparedness.

6.2. Participants

A total of 1,273 students participated in the surveys, of whom 1,242 completed both pre- and post-surveys and were included in the matched analyses. Additionally, 415 parents completed corresponding questionnaires.

6.3. Questionnaire Instrument

The questionnaire was translated into the Nias language for administration and comprised 17 structured items designed to assess disaster-related awareness, knowledge, communication, and preparedness behaviors. Comparable versions were administered to children and randomly selected parents from each participating school.

The instrument was structured to capture five conceptual domains: (A) demographic background, (B) geographic and hazard awareness, (C) disaster learning exposure, (D) evacuation knowledge and behavioral intention, and (E) household preparedness.

The specific items were as follows:

(A) Demographics and Household Background

1. 1.

   Gender

2. 2.

   Age

3. 3.

   Household composition (family members living together)

4. 4.

   Residence location (coastal area, riverside, mountains)

5. 5.

   Parents’ occupation

(B) Hazard Perception and Learning Experience

1. 6.

   Perceived types of disasters likely to occur in the community

2. 7.

   Prior learning experience regarding earthquakes and tsunamis

3. 8.

   Source of disaster-related learning

(C) Behavioral Knowledge and Communication

1. 9.

   Intended action immediately after an earthquake

2. 10.

   Whether the child discusses school disaster learning with family members

3. 11.

   Person to contact first during a disaster

4. 12.

   Intended evacuation destination in the case of an earthquake (e.g., high ground, home, school)

(D) Cultural and Indigenous Knowledge

1. 13.

   Situations in which Maena is performed

2. 14.

   Experience of participating in evacuation drills

3. 15.

   Knowledge of disaster-related local legends

4. 16.

   Type of disaster associated with known legends

(E) Household Preparedness

1. 17.

   Presence of household emergency supplies

Questionnaire items were developed with reference to perspectives on hazard perception, protection motivation, and intergenerational knowledge transfer. Items related to local legends were included to evaluate the educational integration of culturally embedded disaster narratives within Maena-based interventions.

All outcome variables used in the statistical analyses (Tables 1 and 2) were derived from dichotomized responses to these items. For inferential testing, matched pre–post responses were analyzed using McNemar tests, and associations between child and parent outcomes were examined using chi-square tests of independence.

6.4. Statistical Analysis

Changes in paired categorical responses between the pre- and post-intervention surveys were analyzed using McNemar tests. To facilitate interpretation of the practical magnitude, standardized effect-size indices based on \(2\times 2\) contingency tables were additionally reported as Cramér’s \(V\). Associations between children’s learning outcomes and parental awareness or preparedness outcomes were examined using chi-square tests of independence. Statistical findings were interpreted as associations rather than definitive causal relationships.

School-level descriptive consistency checks were conducted to examine whether the effects were limited only to a small number of schools. Positive directional changes in similar patterns in both children’s and parents’ outcomes were broadly observed across the participating schools, although each school developed its own locally adapted version of Disaster Prevention Maena.

In addition to questionnaire surveys, qualitative observations of program implementation and community participation were conducted during the project period. However, this study does not use these qualitative observations as formal analytical evidence, and the reported findings are based exclusively on the quantitative results of the pre–post surveys.

This decision was made considering potential observer bias. Two authors were directly involved in project implementation, one of whom served as the overall project leader. This dual role is recognized as a possible source of interpretive bias, particularly regarding qualitative observations.

To maintain analytical objectivity, this study limits its empirical claims to transparently reported survey data analyzed through standardized statistical procedures.

6.5. Author Positionality and Ethics

Two authors were directly involved in the project implementation, with one serving as the project leader. This dual role has been acknowledged as a potential source of bias. Standardized instruments and transparent statistical procedures were used to address this issue. The study complied with the JICA ethical guidelines, and informed consent was obtained.

7.1. Participant Characteristics

Among the 1,273 participating students, 1,242 completed both the pre- and post-surveys and were included in matched analyses (mean age \(=9.0\) years, SD \(=1.4\); 46.6% female). Parent respondents (\(n=415\)) had a mean age of 39.5 years (SD \(=7.3\)), and 30.1% were mothers, with fathers comprising the majority.

Most participating families (81.9%) resided in low-lying areas such as coastal or riverside locations, indicating high exposure to tsunami risk. Farming and fishing were reported by 61.4% of households, reflecting socioeconomic vulnerability typical of small island communities.

No statistically significant demographic differences were observed across the participating schools, supporting the comparability of the responses.

The distribution of respondents across participating schools and gender/parental categories is presented in Tables 3 and 4.

7.2. Children’s Knowledge Gains

Table 1 presents the pre–post changes in children’s disaster knowledge. Evacuation site recognition increased from 84.5% before the intervention to 99.7% after the program (\(\chi^2=180.3\), \(p<0.001\), Cramér’s \(V\) \(=0.38\)). The moderate-to-large effect size indicates a substantial improvement in evacuation site recognition. Importantly, the reduction of incorrect or “unknown” responses suggests not only incremental improvement but near-universal acquisition of correct evacuation knowledge.

Understanding earthquakes and tsunamis increased markedly from 23.3% to 55.6% (\(\chi^2=353.5\), \(p<0.001\), \(V\) \(=0.61\)). This represents a large effect size and reflects a significant enhancement in hazard differentiation skills—knowledge critical for appropriate behavioral responses during seismic events.

Recognition of local legends of tsunamis increased from 8.3% to 34.8% (\(\chi^2=114.2\), \(p<0.001\), \(V =0.37\)). Although the absolute percentage remains lower than in other domains, the fourfold increase indicates successful transmission of culturally embedded disaster narratives.

Collectively, these findings are consistent with Hypothesis 1, and demonstrate that the Disaster Prevention Maena-based project produced statistically robust and practically meaningful improvements in children’s disaster knowledge.

7.3. Parental Awareness and Preparedness

Importantly, significant changes were also observed among parents, supporting hypothesis H2 and directly addressing the central research question regarding the intergenerational impact.

7.4. Intergenerational Associations

To examine the intergenerational effects of disaster prevention education, the associations between children’s knowledge gains, parental awareness, and preparedness outcomes were assessed using chi-square tests of independence (Table 5).

Three key child–parent pairs were assessed:

1. Evacuation site recognition \(\times\) Awareness of proper evacuation site

2. Understanding earthquakes and tsunamis \(\times\) Family preparedness discussion

3. Recognition of local legends of tsunamis \(\times\) Household preparedness measures

The results indicated that two of the three child–parent pairs demonstrated statistically significant associations.

The first pair—evacuation site recognition—indicated a weak and non-significant association (\(\chi^2=2.44\), \(p=0.118\), \(V = 0.03\)). This result is likely attributable to a ceiling effect as parental baseline awareness exceeded 96%, leaving limited variability to detect statistical relationships.

In contrast, the second pair demonstrated a statistically significant association. Children’s improved understanding of earthquakes and tsunamis was associated with increased family preparedness discussion (\(\chi^2=41.34\), \(p<0.001\), \(V = 0.16\)). Although the effect size was small, the result suggests that children’s learning stimulates disaster-related communication within households.

The third pair demonstrated a significant association. Recognition of local legends of tsunamis among children was associated with increased implementation of household preparedness measures (\(\chi^2=61.96\), \(p<0.001\), \(V = 0.24\)). This finding indicates that culturally embedded knowledge may be particularly effective in promoting practical preparedness behaviors at the household level.

Across domains, the results suggest that children’s disaster learning is meaningfully associated with parental awareness and preparedness. Notably, culturally grounded content demonstrated stronger associations with behavioral outcomes than basic knowledge domains, indicating the potential importance of contextual and culturally relevant educational materials.

These analyses reflect statistical associations rather than causal relationships, and should not be interpreted as evidence of direct causality.

7.5. Supplementary Descriptive Improvements

Beyond the primary analytical domains, the descriptive results indicated a broader educational impact. For example, the proportion of children reporting participation in evacuation drills increased to 100%, and “unknown” responses regarding residential hazard locations were eliminated from the post-survey.

7.6. Summary of Key Findings

Considered together, these findings reveal a consistent pattern of educational influence extending beyond the classroom. The project was associated not only with substantial improvements in children’s disaster-related knowledge, but also with measurable changes in parental awareness and household preparedness.

The marked increase in household preparedness measures suggests that school-based learning may have contributed to practical preparedness actions within families. The simultaneous increase in family discussions on disaster preparedness further indicates that knowledge acquired by children may have circulated within the household environment during the program period.

These patterns provide an empirical basis for examining how culturally embedded disaster education facilitates the intergenerational diffusion of knowledge and preparedness behaviors. The following section discusses these findings in relation to the existing theoretical perspectives on disaster education, intergenerational learning, and culturally grounded resilience.