Revisiting the ‘Art of Learning’ bushfire preparedness model

By Christine Eriksen (UOW), Carrie Wilkinson (UOW) and Tim Prior (ETH Zürich)

This blog post presents an assessment and revision of our ‘Art of Learning’ bushfire preparedness model published in 2011. It is based on feedback from a workshop with emergency service personnel who applied the model as a tool to understand successful and unsuccessful attempts at communicating about bushfire to at-risk communities. It was evident from the diversity of scenarios unpacked by the workshop participants that the model provides a flexible framework that practitioners can apply in specific situational contexts. However, changes to the model were deemed necessary to better accommodate the needs of both risk communicators and information receivers.


In 2011 we published the ‘Art of Learning’ model as a response to the considerable effort by emergency services and disaster researchers to understand why people who are aware of the threat of natural hazards still do little or nothing to prepare for such events, while others with the same information to hand actively engage with the risk and prepare. Developing a practical framework to assist the communication of bushfire risk and preparedness to heterogeneous communities at the wildland-urban interface (WUI) is a continuing challenge (Meldrum et al. 2015; McCaffrey 2015; Paveglio et al. 2015; Steelman and McCaffrey 2013).

Research suggests there are cultural, socio-economic, environmental and cognitive reasons behind individuals’ decisions on whether or not to prepare (Eriksen and Gill 2010; McCaffrey et al. 2011; McFarlane et al. 2011; Paveglio et al. 2011). However, there is no single or unambiguous reason. Rather, the crux is the situational context and individual circumstances of each decision-maker. This is particularly the case in peri-urban landscapes characterised by heterogeneous demographics, knowledge and experiences of bushfire, as such diversity often negates the effectiveness of homogeneous and passive engagement techniques that distribute generic risk information to diverse individuals. In focussing on situational context and individual circumstances, the AOL model integrates individual experiential learning processes with cognitive decision cues and risk communication methods. The model thus emphasises the importance of considering how, why and when people learn, to better understand the effectiveness (or not) of traditional risk communication and engagement practices with diverse types of residents.

This blog post presents an assessment and revision of the AOL model. It develops the insights gained from Kolb’s (1984) theory of experiential learning, which informed the basis of the AOL model, and incorporates feedback gathered during a workshop with emergency service staff and volunteers. Workshop participants tested the workability of the AOL model in small groups by reflecting on first-hand experiences of successful and unsuccessful attempts at communicating with at-risk communities. This, in turn, permitted a critical reflection on the strengths and weaknesses of the AOL model.

The ‘Art of Learning’ (AOL) model

The AOL model published in 2011 (Figure 1) integrates three sets of different information relevant to both risk communication and bushfire preparedness:

  1. Kolb’s (1984) learning styles and experiential learning model form the basis of the framework;
  2. Existing means of bushfire risk communication are mapped onto this framework in the places where the techniques best fit with the learning processes described by Kolb;
  3. Key socio-cognitive factors that influence individual decisions about whether or not to prepare properties for bushfire are overlayed where they fit most appropriately with the learning framework and communication techniques.

Figure 1. The original AOL model: a framework for targeting bushfire risk information to learning styles and learning stages (Eriksen and Prior 2011).

The model was presented as a prospective aid for risk communicators by illustrating where typical communication or engagement techniques could fit with stages in an experiential learning process. It connects different communication techniques to cognitive drivers as a means of addressing negative cues or promoting positive decision cues to advance learning and bushfire preparedness. Finally, the model identifies which cognitive preparedness decision cues could influence these learning steps, and when. We suggest that this circulatory model, with no specific beginning or end permits risk communicators to adjust a communication process to suit the diverse knowledge, beliefs, abilities, cognition and level of preparedness of the targeted recipients.


To examine the workability of the AOL model, it was presented to emergency service staff and volunteers in a workshop session, facilitated by the authors, at the Australasian Community Engagement and Fire Awareness Conference hosted by the New South Wales Rural Fire Service (NSW RFS) in Newcastle in May 2010. The conference was attended by natural hazard managers, risk communicators and community educators from a range of Australian fire agencies.

Participants independently formed seven groups – each consisting of six to eight individuals. Groups were asked to identify a successful and/or an unsuccessful first-hand attempt at communicating about bushfire risk and preparedness with at-risk members of the general public. Participants then applied the AOL model to: a) understand why the attempt had or had not been successful, and b) explore how the scenario could be improved or replicated with different community members. In order to determine how user-friendly the model was as a stand-alone tool for trained emergency service staff and volunteers (i.e. could they independently understand the model components and apply them to their real-world scenarios), the authors (as facilitators) provided interpretive support only when asked for by the workshop participants.

Analysis focussed on the extent of the workshop participants’ abilities to understand and apply the model in the context of their first-hand experiences, and to draw on the appropriate aspects of the model to reflect on why their experiences unfolded as they did, and what they could have done differently. The following specifically focuses on the strengths and weaknesses of the AOL model that emerged during the workshop, which point to ways to improve the model. The AOL model has been revised accordingly.

Results and Discussion

Model strengths

It was evident from the diversity of scenarios unpacked by the workshop groups that the AOL model provides a flexible framework that practitioners can apply in specific situational contexts and individual circumstances. Scenarios ranged from engaging residents of a retirement village in hazard identification and risk reduction, capitalising on the recent exposure of a community to bushfire, to encouraging absentee landholders, ‘weekenders’, and youth to be active members of local rural fire brigades.

A common theme of the workshop discussions was the imperative of communicators capitalising on risk immediacy (Experience and Reflection in the model), an issue perceived by practitioners to have the most tangible long-term likelihood of positively influencing individuals’ perceptions of risk and preparedness. Participants identified that a real strength of the AOL model was the clear connection of the Education Mechanisms to elements that influence preparedness, helping practitioners to leverage risk immediacy.

While the cognitive and learning elements of the model were less easy for practitioners to engage with, the list of Education Mechanisms was thorough and provided practical “entry points” for most groups into the AOL model.

Apathetic and complacent attitudes to risk were described by all groups as reducing the success of nearly all attempts at risk communication. A strength of the AOL model was that it clearly demonstrated how these influential factors may be better understood and addressed if practitioners considered how particular Education Mechanisms may be (in)effectual in changing residents’ Conceptualisation of bushfire risk. This will be most effective when tools preceding this learning stage prompt residents to continually Experiment, Experience and Reflect, thus deepening their learning experiences through a feedback loop of adaptation and resilience building.

Model weaknesses

Although the model had a strong theoretical “bottom-up” design focus, the model was presented to risk practitioners in the workshop using an academic “top-down” approach in an attempt to support the participants’ interpretation of the model. This approach meant that certain terms and concepts, part of the everyday vernacular of risk practitioners, were missing in the model. For example, widely used vernacular such as ‘apathy’, ‘incentives’, ‘priorities’, and ‘immediacy of threat’; and tools such as ‘door-knock’, ‘property inspections’, and ‘follow-up’ were not included in the model, although they frequently came up in practitioners’ accounts of (un)successful risk communication scenarios. In contrast, terms included in the model, such as ‘self-efficacy’ and ‘salience’, were not commonly used by the practitioners. Instead, related words, such as ‘confidence’ and ‘immediate’, were used frequently and therefore seem more appropriate words for the model.

The model also covers a lot of information, and both the facilitators and practitioners acknowledged that the AOL model simplifies a complex process. The rigour of the model as a “one-stop-shop” for practitioners developing and delivering bushfire risk information could be strengthened by the inclusion of further decision cues and tools to advance learning and preparedness. However, this would sacrifice visual coherence. As one practitioner commented in the wrap-up discussion:

“I think the internal four sections where you say ‘awareness’ is this and ‘information’ is this, I think that’s valid, but I think having it in a circle and the extra stuff and that, it gets very confusing.”

This is a difficult point to reconcile. Through three tiers the model attempts to illustrate where typical communication techniques could fit with stages in an experiential learning process, and where specific cognitive preparedness decision cues could influence these learning steps. Not all tiers were utilised or understood by the workshop groups and it is difficult to strike a balance between having all “solutions” in the model whilst maintaining its visual coherence. As discussed below, further refinement of the model is needed.

Revisions to the AOL model

Figure 2 provides a revised model and framework for establishing more effective risk communication and engagement. It integrates practitioners’ feedback of the original model’s workability, visual coherence, and vernacular. Table 1 collates extracts from the workshop discussions, which underpin our decision to revise specific elements of the model, as outlined below.

Figure 2. The revised AOL model. Adjustments included in this iteration of the model reflect feedback from risk communicators that increase the real-world applicability of the tool.

The most significant change to the model’s structure is to the outer Preparedness Cycle tier. (Dis)Empowerment was a concept used frequently by all seven workshop groups. Although conceptualised in the AOL model as an Education Mechanism (as part of the Processing Continuum), (dis)empowerment permeated all three tiers of the AOL model. Empowerment was spoken of as being central to the successful communication, acceptance, retention and application of a risk message within a community, and was conceptualised by practitioners as being both a point of entry to, and a point of exit from, the AOL model. The place and conceptualisation of empowerment in the model has therefore been revised in Figure 2 to reflect the prominence of the idea in the workshop discussion with (Dis)Empowerment and related influencing sub-factors Ability to act, Self-confidence, Self-doubt and Agency being prominently situated in the Preparedness Cycle.

Other revisions made to the Preparedness Cycle include: the replacement of Salience with Immediacy of threat and the addition of Previous experience and Incentives as sub-factors to Motivation; the addition of Apathy, Priorities and Values as sub-factors to Action, and the relocation of Ability to Act from a sub-factor of Intention to (Dis)Empowerment.

No revisions were deemed necessary to the Learning Cycle.

The following revisions were made to the Education mechanisms section: Door knocks and Property inspections were added to Raising Awareness; Welcome packages and Follow-up were added to Providing Information; and Role-Play was added to Building Skills.

Finally, a note on the need for reflexive self-assessment by risk communicators was added to the core of the model to emphasise that it is as important for risk communicators to learn to effectively and appropriately communicate, as it is for residents to learn to translate risk information into preparedness actions. In this iteration of the model, the necessity of risk communicators to approach their role with reflexivity is termed Progressive Risk Communication. Practitioners in three groups emphasised that successful risk communication cannot simply be put down to changing the perceptions and attitudes of the community. Rather, practitioners need to look inward, both personally and organisationally, to assess the appropriateness of the tone and manner in which they communicate a message. As one practitioner observed:

“In this diagram we’ve got here, in the ‘information provision’, it’s only got stuff in there like ‘brochures’ etc. Is there a need to put in there something like ‘personal representation’, ‘personal approach’, ‘the human being talking to the human being’?”

This revision to the model also builds on group discussions, which highlighted potential barriers to successful risk communication due to the tone and manner in which some practitioners engage with specific community members. Practitioners of two groups stereotyped residents with a strong environmental conscience or alternative lifestyle as being “tree-huggers”, fundamentally opposed to any hazard reduction measures. Such residents were described as “stupid” or “ignorant” because their lifestyle choices outweigh concerns about bushfire. Interestingly, the groups who perpetuated negative stereotypes in their discussions also engaged the least with the AOL model. These observations highlight how the efficacy of any education mechanisms is influenced as much by the attitude of the communicator, as by the message being communicated. That some workshop participants spoke condescendingly of specific community groups (e.g., “rich bastards” and “hippies”) would suggest the need for agency-led community engagement and communication processes to be conducted objectively in a manner that accommodates the diverse opinions, cultures and lifestyles of the people living in at-risk areas.


The practical application of the AOL model by practitioners illustrates that a better understanding and appreciation of mechanisms that encourage different ways of learning are fundamental to more efficient risk communication practices. The model must both reflect aspects of information receivers, like the ways they learn, and resonate with the risk communicators that use it. Consequently, many of the changes made to the original model reflect the necessity to balance the needs of the information receivers with those of the information providers, including the way information providers can best interact with the public.

The key lessons learnt from this study of the AOL model for further developing the art of engaging members of the general public with official risk rationales include:

  • The everyday relevance and familiarity of terms and concepts influence the model’s applicability as well as its reception by risk practitioners.
  • The AOL model successfully facilitates connection between learning styles and effective education mechanisms/interventions.
  • There is a need for reflexive cues for risk communicators. The efficacy of education mechanisms is influenced as much by the attitude of the information provider, as it is by the attitude of the recipient of the information.
  • Informed decision-making through empowerment is key to successful risk communication.
  • Finally, the generic nature of the AOL model (its applicability to a plethora of risk scenarios and diverse populations) makes it a relevant tool for risk communicators of other natural hazards. A possible next step from this study is to explore the AOL model in various hazard contexts.


Sincere thanks goes to the workshop participants for sharing their experiences, the NSW Rural Fire Service’s Community Engagement team for ongoing research support, and the Australian Centre for Cultural Environmental Research at the UOW for Research Support Funds that enabled the write up of the workshop findings.


  • Eriksen, C. and Gill, N. (2010) Bushfire and Everyday Life: Examining the Awareness-Action ‘Gap’ in Changing Rural Landscapes. Geoforum 41, 814-825.
  • Eriksen, C. and Prior, T. (2011) The Art of Learning: Wildfire, Amenity Migration and Local Environmental Knowledge. The International Journal of Wildland Fire 20, 612-624.
  • Kolb, DA. (1984) ‘Experiential Learning: Experience as The Source of Learning and Development.’ (Prentice-Hall Inc.: Englewoods Cliffs, New Jersey)
  • Meldrum, J. et al. (2015), “Understanding Gaps Between the Risk Perceptions of Wildland-Urban Interface (WUI) Residents and Wildfire Professionals”, Risk Analysis, 35(9), 1746-1761.
  • McCaffrey, S. (2015), Community Wildfire Preparedness: a Global State-of-the-Knowledge Summary of Social Science Research”, Current Forestry Reports, 1(2), 81-90.
  • McCaffrey, S. et al. (2011) Outreach Programs, Peer Pressure, and Common Sense: What Motivates Homeowners to Mitigate Wildfire Risk? Environmental Management 48, 475-488.
  • McFarlane, BL. et al. (2011) Complexity of homeowner wildfire risk mitigation: an integration of hazard theories. International Journal of Wildland Fire 20, 921-931.
  • Paveglio, T. et al. (2011) Symbolic Meanings of Wildland Fire: A Study of Residents in the U.S. Inland Northwest. Society & Natural Resources 24, 18-33.
  • Paveglio, T. et al. (2015), “Categorizing the Social Context of the Wildland Urban Interface: Adaptive Capacity for Wildfire and Community ‘Archetypes’”, Forest Science, 61(2), 298-310.
  • Steelman, T. and McCaffrey, S. (2013), “Best practices in risk and crisis communication: Implications for natural hazards management”, Natural Hazards, 65(1), 683-705.

Leave a Reply

Your email address will not be published. Required fields are marked *