Disasters are commonly thought of as events that cause significant loss of life, damage and hardship across communities, although actual definitions of disasters vary (Quarantelli, Reference Quarantelli1998). Australia is characterised by frequent natural disasters of varying type, including flood, cyclone, drought and fire, which have been broadly increasing in frequency since reliable records began (Council of Australian Governments, 2004). In Australia and elsewhere, the frequency and intensity of natural disasters is thought to be affected by climate change. For example, a 2007 Working Group of the IPCC predicted with high confidence that there would be an increase in intensity and frequency of heatwaves and fires as well as floods, landslides, droughts and storm surges in Australia and New Zealand in the 21st century (Hennessy et al., Reference Hennessy, Fitzharris, Bates, Harvey, Howden, Hughes, Salinger, Warrick, Parry, Canziani, Palutikof, van der Linden and Hanson2007), a prediction confirmed more recently by the Commonwealth Scientific Industrial Research Organisation (CSIRO, 2011).

While disasters may cause significant hardship, damage and loss of life, the impact of a disaster depends not only on the type of disaster itself but also on the exposure and vulnerability of the individuals and communities involved (Fothergill & Peek, Reference Fothergill and Peek2004). Research has indicated that children are among the most vulnerable to natural disasters (Wisner, Blaikie, Cannon, & Davis, Reference Wisner, Blaikie, Cannon and Davis2004). For example, upon conducting a review of 160 studies of disaster victims worldwide, Norris, Friedman, Watson, Byrne, and Kaniasty (Reference Norris, Friedman, Watson, Byrne and Kaniasty2002) concluded that children experience the adverse effects of disasters much more than adults and the elderly. The ongoing impact of disasters in Australia is evident in the Kinglake Ranges in Victoria. More than 4 years after their experience of the 2009 Black Saturday Bushfires, ‘The Smouldering Stump Campaign’ was established for the ongoing needs of children. The campaign launch brochure states ‘Children and young people continue to struggle with maintaining “normal routines”, and parents are exhausted and overwhelmed with the issues that face their children’ (Smouldering Stump, 2013). The reasons for this are not clear. However, research has indicated that children tend to rank hazardous events, including natural disasters, as one of their major fears, even prior to a disaster occurring (e.g., Campbell & Gilmore, Reference Campbell and Gilmore2006; Ollendick, King, & Frary, Reference Ollendick, King and Frary1989). Some children have even experienced problems following relatively slight hazardous events, such as those where life is not disrupted in a significant manner and there is no loss of life (Ronan, Reference Ronan1997a, Reference Ronan1997b; Ronan & Johnston, Reference Ronan and Johnston1999). Consequently, researchers have postulated that children's vulnerability to hazardous events occurs in part because it is the realisation of one of their worst fears.

Despite their fears around disasters, children can be empowered to prepare for and respond to disasters via various means, including through school-based disaster education programs (Back, Cameron, & Tanner, Reference Back, Cameron and Tanner2009; Ronan, Crellin, & Johnston, Reference Ronan, Crellin and Johnston2012). As a result, children can become more resilient to disasters. Resilience has been variously defined depending on the level of analysis, which may be the individual, community or ecological system. Most definitions incorporate a stressor and the notion of adaptation, and a speedy return to pre-stressor levels of functioning (Norris, Stevens, Pfefferbaum, Wyche, & Pfefferbaum, Reference Norris, Stevens, Pfefferbaum, Wyche and Pfefferbaum2008). Bonanno (Reference Bonanno2004) defines individual resilience as a person's capacity to maintain overall healthy, stable functioning following stressful life events. From the perspective of communities, Norris et al. (Reference Norris, Stevens, Pfefferbaum, Wyche and Pfefferbaum2008) refer to the ability of communities to withstand hazards and/or recover from disasters.

Attention to the needs of children and youth before, during and after a disaster is imperative, whether considering emergency management or broader sustainability perspectives. A thorough, recent review of Australian emergency management plans (Davie, Reference Davie2013) showed that emergency management planners assume that parents, primary caregivers and the education system will take care of the needs of children in emergencies and disasters in Australia. This is clearly apparent in emergency management plans that refer to schools as being charged with the responsibility of evacuating children and with their welfare in emergency recovery. It is also evident more generally in the lack of specific planning for children in emergency management plans (Davie, Reference Davie2013).

The role of schools in preparing children to face disasters is a critical one. Peek (Reference Peek2008) suggested that children's vulnerability is reduced when they are provided with information and resources, are encouraged to participate in disaster preparedness and response activities, and can access personal and communal support. Children's participation in disaster risk reduction not only builds their resilience to disaster but can also provide benefits for the community as a whole. For example, during the 2004 tsunami disaster in southern Thailand, children played significant roles in assisting their families and communities both during the event and through the aftermath and recovery period (Vanaspongse, Ratanachena, Rattanapan, Chutong, & Intraraksa Reference Vanaspongse, Ratanachena, Rattanapan, Chutong and Intraraksa2007). Children helped adults in temporary shelters, looked after younger children, comforted friends who had lost their families, helped in the clean-up and did housework (Vanaspongse et al., Reference Vanaspongse, Ratanachena, Rattanapan, Chutong and Intraraksa2007). The example of a British schoolgirl in the 2004 tsunami is particularly salient; having recently learned about tsunamis in class, she was able to warn and therefore save a beach full of foreign tourists in Thailand (Back et al., Reference Back, Cameron and Tanner2009). Other studies support the conclusion drawn from this example that children can act to minimise risks; for instance, by working as interpreters and relaying messages to households and communities (Mitchell, Haynes, Hall, Wei, & Oven, Reference Mitchell, Haynes, Hall, Wei and Oven2008). Similarly, a study in Mozambique showed that, through participatory processes, children gained a greater knowledge and understanding of risks and began to minimise those risks within their household and at the community level (Back et al., Reference Back, Cameron and Tanner2009). These studies support an approach to disaster risk reduction that encourages the agency of children and youth, to work toward making their lives safer and their communities more resilient to disasters.

Back et al. (Reference Back, Cameron and Tanner2009) suggest that investing in child-centred disaster risk reduction is important because learning and practising disaster risk reduction while young embeds changed behaviour that can be integrated into adult life. This is of particular significance in light of research that suggests that disaster preparedness rates among adults and children are often low, even in high-hazard areas (e.g., Paton & Johnston, Reference Paton and Johnston2001; Peek & Mileti, Reference Peek, Mileti, Bechtel and Churchman2002; Whitney, Lindell, & Nguyen, Reference Whitney, Lindell and Nguyen2004). For example, in an Australian study, Berry and King (Reference Berry and King1998) examined the tropical cyclone awareness and preparedness of far north Queensland school students. They found that the students had little understanding of cyclone preparedness, including the roles and responsibilities of all community members and the expectations of them in times of disaster.

Internationally, school disaster education is considered to be important to raise knowledge and awareness among students and their families and, most importantly, to encourage preparedness action (Back et al., Reference Back, Cameron and Tanner2009; Friedman, Rose, & Koskan, Reference Friedman, Rose and Koskan2011; Jimerson, Brock, & Pletcher, Reference Jimerson, Brock and Pletcher2005; Ronan et al., Reference Ronan, Crellin and Johnston2012). The empowerment of children to face disasters therefore has far-reaching consequences in the spheres of emergency management and sustainability. However, empirical evaluation of school-based disaster education programs has been limited. Most studies on school-based education programs have relied on cross-sectional, correlational research designs to evaluate their effectiveness (e.g., Ronan & Johnston, Reference Ronan and Johnston2001; Shaw, Shiwaku, Kobayashi, & Kobayashi, Reference Shaw, Shiwaku, Kobayashi and Kobayashi2004). More recent research has sought to employ quasi-experimental methodology. In Reference Ronan and Johnston2003, Ronan and Johnston conducted a study in Auckland, New Zealand using a quasi-experimental methodology, which found that hazards education programs led to changes in knowledge, preparedness, and indicators of emotional resilience. Building on their earlier research, Ronan et al. (Reference Ronan, Crellin and Johnston2012) conducted a quasi-experimental study in Napier, New Zealand, to show that following a brief school education program supplementing a larger community-wide effort, children reported significant gains in preparedness indicators, including increased knowledge as well as increases in physical and psychosocial preparedness.

In Australia, disaster-based education has been repeatedly called to be implemented in schools. For example, the National Inquiry on Bushfire Mitigation and Management asserted that it was the responsibility of all governments to jointly develop and implement national and regionally relevant education programs about bushfire (Ellis, Kanowski, & Whelan, Reference Ellis, Kanowski and Whelan2004). The Australian National Curriculum (Australian Curriculum, Assessment and Reporting Authority [ACARA], 2013) contains three cross-curricular priorities that underpin its rationale. One of these, sustainability, is designed to:

Supporting the cross-curricular priority of the Australian National Curriculum, disaster education is also currently part of the Australian Curriculum in Years 6, 7 and 8, as well as in Senior Secondary Geography; and disaster programs have been developed by emergency management authorities for distribution in schools (Dufty, Reference Dufty2009). In his review of emergency management programs in Australian schools, Dufty (Reference Dufty2009) reported that an animated ‘Flood Investigator’ program was developed by Melbourne Water and was supported with lesson outlines, teacher's notes and worksheets. Similarly, ‘The Floods and You’ program in Tasmania included lessons for primary school students. Other programs noted by Dufty (Reference Dufty2009) included one for high school geography students developed by Wollongong City Council in New South Wales. All the state and territory government websites in Australia name various bodies as being responsible for disaster education of school students (Davie, Reference Davie2013). The Disaster Resilience Education for Schools website — http://schools.aemi.edu.au — is a curriculum-aligned resource for teachers and students that contains publications and interactive learning games designed to prepare children for emergencies. Nonetheless, empirical evaluation of the effectiveness of these programs in building student and family response to and recovery from disasters has remained minimal (Dufty, Reference Dufty2009).

Methods

The literature review focused exclusively upon peer-reviewed literature that was published in English and was designed to be as broad and as inclusive as possible. Four databases were searched: SciVerse Scopus using Scirus (Elsevier, Amsterdam), A + education using the Informit search engine (RMIT, Melbourne), PsycINFO and the Education Resources Information Center (ERIC), using the CSA Illumina search engine (ProQuest, Ann Arbor, MI). These databases were selected based on the relevance of their content to the focus of the literature review.

A list of all search terms expected to exhaustively cover the articles was generated by the authors, based on their salience to the subject matter and the background research that had been previously undertaken (Boon, Reference Boon2011; Pagliano, Reference Pagliano2012). Three search terms were entered into the database search engines at any one time and were employed either as a full text search or in all fields. Where a large number of citations was obtained, a further search was employed using the same search terms but limiting the search to the abstract, title and keywords. Substitutes were employed among the keywords in order to capture the breadth of articles available; for example, ‘disaster’ was substituted for ‘natural hazard’ or ‘hazard’. The keywords used were: Group A: disaster or natural hazard or hazard; Group B: education or school or science or children or teaching; Group C: Australia.

Search Execution and Article Screening

The research was undertaken in May, June and August 2013. The search result of primary interest was the intersection of the three search subsets marked with an asterisk (see Figure 1).

Figure 1: The ‘Exact Match’ search intersection (*).

The citations for all of the articles identified by the intersection of the three search subsets (‘Exact match’) were distributed among two of the authors, who independently reviewed each title to determine the article's potential relevance to the research question. Authors were blinded to each other's appraisal. Those titles for which both reviewers indicated a lack of relevance were excluded. For the remaining citations, the study abstracts were obtained, with the process of independent, blinded review repeated. Again, studies were only excluded if both reviewers indicated a lack of relevance. Finally, the full manuscripts for the retained citations were reviewed using a data collection sheet to further screen and characterise the article, and to extract relevant information about each study.

A similar process was undertaken to screen citations identified by the intersection of any two of the three search terms (‘Near match’). The titles of these articles were again checked by two authors independently reviewing each title to determine the article's relevance to the research question. If either reviewer identified a title for inclusion or further review, the abstract for that study was obtained with the process of independent blinded review by two authors repeated. Again, studies were only excluded if both reviewing study team members indicated a lack of relevance. In summary, citations for all the articles were reviewed to determine each article's potential relevance to the research question. Where articles were deemed to be relevant, the abstract was obtained and reviewed. Finally, the full manuscripts for the retained citations were reviewed to extract relevant information about each study. We used content analysis to interpret each relevant article. This process involves several steps: (1) close reading of the text, (2) relevant coding of subject matter within the text, (3) categorising the codes, and (4) the generation of appropriate frameworks or models from the gathered data. Our purpose was to inductively derive a framework from the ground up for future research pertinent to the study area.

Results

The searches of the four online databases generated 71,075 citations (Table 1). This number of citations was based upon searches of the databases using either all fields or full text. Where a particular search generated a large number of citations, a further search was performed using the same keywords but limiting the search areas to the abstract, title and keywords. This reduced the total number of citations to 2,881.

Table 1: Searches of Online Databases

After reviewing the citations and abstracts as described above, a shortlist of 35 relevant articles was obtained.

The results indicate a paucity of research on school-based disaster education in the Australian context, precluding our purpose to develop a framework for future research and analysis. Only three of the shortlisted thirty-five papers had direct relevance to school-based disaster education in Australia. For example, in 2003, the Australian Geography Teachers Association ‘mapped’ the teaching and learning of hazards or disasters (Kriewaldt et al., Reference Kriewaldt, Butler, Doyle, Freeman, Hutchinson, Parkinson, Terry and Boscato2003). This study indicated that curriculum frameworks in each state and territory had common elements, with education being mainly present in Years 5–6 and more comprehensively in Years 7–10. Hazard or disaster education was evident, particularly in the ‘Studies of Society and Environment (SOSE) — Human Society and Its Environment (HSIE)’ in New South Wales. In the SOSE geography strand, it was found in ‘Place, Space and Environment’ in South Australia, ‘Place and Space’ in Western Australia, and, in the optional Queensland geography syllabus, the strand was ‘Place and Space’. However, this research has not been updated in light of the National Curriculum changes that are taking place across Australia. Moreover, research had not been conducted to determine whether the aforementioned programs were effective in increasing knowledge and awareness among children and/or to encourage disaster preparedness.

In Reference Anderson-Berry2003, Anderson-Berry indicated that Queensland primary school students were utilising a computer-based educational game called Stormwatchers, which provided cyclone awareness education. However, this program had not been formally evaluated for effectiveness.

More recently, Dufty (Reference Dufty2009) prepared a paper that sought to determine how school-based disaster education programs could be made more effective in Australia. In that paper, Dufty discussed the importance of school-based education programs in building disaster resilience. Dufty noted that most emergency management authorities in Australia have developed and implemented education programs through schools, although there was a dearth of research evaluating the effectiveness of these programs. Within this context, Dufty described important theoretical considerations in designing school-based education programs, including understanding how young people learn about hazards at different ages.