Psychological differences between south-eastern Australian householders' who intend to leave if threatened by a wildfire and those who intend to stay and defend (original) (raw)

Psychological differences between south-eastern Australian householders’ who intend to leave if threatened by a wildfire and those who intend to stay and defend

J. McLennan a,∗{ }^{a, *}, D. Paton b { }^{\text {b }}, R. Beatson a{ }^{a}
a{ }^{a} School of Psychological Science, La Trobe University, Bundoora, Victoria 3086, Australia
b { }^{\text {b }} School of Psychology, University of Tasmania, Launceston, Tasmania 7248, Australia

A R T I C L E I N F O

Article history:
Received 17 September 2014
Received in revised form
28 November 2014
Accepted 28 November 2014
Available online 29 November 2014
Keywords:
Bushfire
Community safety
Evacuation
Property defence

A B S T R A C T

Disastrous wildfires have occurred often in south-eastern Australia. Following multifatality wildfires in Victoria on 7 February 2009 the national approach to community wildfire safety, ‘Prepare, stay and defend or leave early’, came under intense critical scrutiny. The approach was revised subsequently so as to emphasise leaving as the safest option in the event of a wildfire warning. This study reports findings from a survey of 584 residents of at-risk locations. The majority ( 47%47 \% ) reported that they intended to leave if warned of a bushfire threat. However, a substantial minority ( 24%24 \% ) reported that they intended to stay and defend their home. A further 29%29 \% reported that they intended to wait and see what developed before making a final decision. Those intending to leave differed from those intending to stay and defend in several ways. Those intending to leave were characterised generally by being more concerned about the danger posed by bushfires, they viewed themselves as more vulnerable to bushfire threat and they were worried about their house being destroyed in their absence. Those intending to stay and defend were motivated, mostly, to protect their valued property and they believed that their efforts would be successful. They did not perceive themselves to be risk takers. Those intending to leave generally reported rather low levels of preparations for leaving safely. An appreciable percentage of those intending to stay and defend reported levels of preparations for safe defence which were probably inadequate for safe and effective defence.
© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

1.1. Background and aim

Since European settlement in the 18th century, Australian communities have proved vulnerable to four kinds of disasterlevel natural hazard events: cyclones, storms, floods and bushfires-or wildfires [1]. Over the period 1901-2011, 260 major bushfire events resulted in 825 known fatalities, most ( 88%88 \% ) of which occurred in the south-east of the country [2] 1{ }^{1}. Most wildfire scientists predict that in the future there will be more frequent severe wildfires threatening life and property in

[1]

1. • Corresponding author.
     E-mail address: j.mclennan@latrobe.edu.au (J. McLennan),
     1{ }^{1} Comprising the Australian Capital Territory (ACT); and the States of New South Wales (NSW); Tasmania; and Victoria. ↩︎

several countries, including Australia, for three reasons: (a) global climate change resulting in reduced rainfall and higher temperatures in many regions; (b) an increase in the number of householders choosing to reside in areas of high wildfire risk-often a wildland-urban interface (or WUI [3]); and © government policies limiting fire and land management agencies’ vegetation management activities on budgetary or environmental grounds [4-10].

From 2005 to 2009, fire and land management agencies’ approaches to community bushfire safety in Australian states and territories were shaped largely by the Australasian Fire Authorities Council (AFAC) position statement about able-bodied householders being encouraged to remain on their property to actively defend their home when threatened by a bushfire: “…By extinguishing small initial ignitions, people of adequate mental, emotional, and physical fitness, equipped with appropriate skills, and basic resources, can save a building that would otherwise be lost in a fire…People should decide well in advance of a bushfire whether they will stay to defend them or leave if a bushfire threatens”, [11] p. 6. This position came to be summarised as ‘prepare, stay and defend or leave early’ [12]. It should be noted that such a policy differed from that of most North American wildfire jurisdictions where evacuation of residents threatened by a wildfire was, and remains, the preferred community wildfire safety strategy [13] 2{ }^{2}. The AFAC position statement was developed following investigations into disastrous multi-fatality bushfires in Tasmania in 1967 and in Victoria and South Australia in 1983. It was concluded that (a) civilians were most likely to die because of either the fatal impact of radiant heat, or as a result of a motor vehicle accident, while fleeing at the last moment and (b) suitably prepared homes could be defended against bushfires while providing a safe refuge for people during the passage of the main fire front [19]. However, following the disastrous 2009 Victorian ‘Black Saturday’ bushfires police reports that 113 people had perished in or near their homes [20] the ‘prepare, stay and defend or leave early’ position came under intense critical scrutiny by the Victorian Bushfires Royal Commission (Teague et al. [21]). The position was reviewed subsequently and revised by the Australasian Fire and Emergency Services Council in 2010 to emphasise that leaving before imminent bushfire threat was the safest option for residents: “People usually have two safe options when threatened by bushfire: leaving early or staying and defending adequately prepared properties. Leaving early is always the safest option”, [22] p. 1. In 2012 the policy was further revised to give more emphasis to leaving as the safest option in the event of bushfire threat:
4.2.8 The safest action to protect life is for people to be away from the bushfire or threat of bushfire as early as possible. Leaving a high risk bushfire location is the safest action, and leaving before a bushfire threatens is always safer than remaining until a bushfire starts. Leaving becomes increasingly appropriate with higher Fire Danger Ratings. When bushfires are burning on days where circumstances such as weather conditions, topography or fuel loads may create intense fire behaviour, typically when ‘Extreme’ or ‘Catastrophic’ fire danger conditions are expected, leaving early may be the only safe action, even for people who are prepared to defend well-prepared buildings., [23] pp. 5-6.

Notwithstanding the emphasis by fire agencies on at-risk householders leaving before a bushfire threatens, recent surveys of residents impacted by severe bushfires found that many residents of at-risk areas plan to, and do, stay and defend their property against bushfire attack [24-27]. In this paper we report a study in which households in high bushfire risk areas of south-eastern Australia were surveyed about their intentions if threatened by a bushfire. The aim of the research was to investigate psychological differences between those intending to leave and those intending to stay and defend. 3{ }^{3} The purpose was to assist Australasian fire agencies in shaping their future community bushfire safety endeavours so as to reduce risk to residents threatened by significant bushfire events.

1.2. Previous related research

A considerable amount of social science research has been reported about factors associated with homeowner property preparation to reduce risk of destruction by wildfire [31]. Psychological factors which have been identified include: (a) perception of level of wildfire risk [32-40]; (b) beliefs and values about self-reliance and individual responsibility for wildfire safety [33,34,36,37]; © beliefs and expectations about mitigation activity efficacy [35,36,41,42]; (d) attachment to home [33,36,43,44]; and (e) perceptions of community expectations and norms, and community member connectedness and sense of community [44-47].

Several post-bushfire interview-based studies have identified factors involved in householders’ decisions to either leave or to remain on their property when they were threatened by a bushfire. Tibbits and Whittaker [12] concluded that residents had a much clearer understanding of what was involved in staying and defending their home than what leaving early entailed. McLennan et al. [48] analysed interviews conducted with residents impacted by the 2009 Victorian ‘Black Saturday’ bushfires and found that most residents who left when threatened did so because (a) they had not planned nor prepared to stay and defend their home from a bushfire, and (b) on becoming aware of the bushfire threat they decided it

[1]

1. 2{ }^{2} The preference by North American fire agencies for evacuation of residents in the face of wildfire threat is not without controversy. Several researchers have proposed sheltering in place or staying and defending as alternatives [13-17]. Others have argued against sheltering in place [18]. Currently no Australian fire agency recommends sheltering in place as a safe response to bushfire threat.
   3{ }^{3} The present study was part of a larger Bushfire Cooperative Research Centre project involving a survey of residents of bushfire-prone locations in south-eastern Australia. Findings reported previously have described: (a) householders’ reasons for planning to wait and see what developed following an initial bushfire warning [28]; (b) householders’ intentions if threatened by a bushfire and their levels of completion of a range of bushfire preparation activities [29]; and © determinants of householders’ strengths of intentions to leave if threatened by a bushfire. [30] ↩︎

was too dangerous to remain and left-many at the last moment and under hazardous circumstances. Most of those who stayed to defend their property did so because they wanted to protect their property and believed that their efforts would be successful. 4{ }^{4} Recent interview-based surveys of Australian residents threatened by severe bushfires have confirmed that prior to the fires in question, most householders who left had intended to leave because of the perceived danger posed if a bushfire threatened, while most householders who stayed and defended had planned to do so because they wanted to protect their property and believed that they would be successful [24-27].

These same post-bushfire surveys found generally lower than desirable levels of pre-bushfire planning and preparation by residents-especially those residents who intended to leave if threatened by a bushfire. This suggests a need for agencies to review current approaches to community bushfire safety. One way agencies may be able to improve the effectiveness of their messages is to tailor these more specifically to address the psychological makeups of those residents who intend to leave and those residents who intend to stay and defend. What seems to be lacking is information about the psychological characteristics of those householders who intend to leave compared with those who intend to stay and defend their property in response to a serious bushfire threat warning. The present study is intended to contribute to remedying this knowledge gap.

A central question is what psychological constructs should be used as the basis for comparing householders who intend to leave with householders who intend to stay and defend if threatened by a bushfire? The research reported previously on psychological factors associated with householder preparations to reduce wildfire risk to property was a possible starting point. However, we decided to use theory as a basis rather than simply selecting from an array of possible constructs suggested by that research.

1.3. Theory and community bushfire safety research

Following the 2009 Victorian bushfires, Beatson and McLennan [49] noted that development of community bushfire safety programs in Australia had been driven, mostly, by ‘street wisdom’ gained in the aftermath of bushfire disasters. They suggested that future research aimed at enhancing community bushfire safety would be more useful if guided by theoryespecially theories from the field of health promotion and injury prevention. Beatson and McLennan noted that using theory as a basis can: (a) lead to a better understanding of the causes underlying various problematic behaviours; (b) identify potential mechanisms for changing behaviours; and © suggest why particular interventions are more or less likely to be effective. Among the theories they discussed, Theory of Planned Behaviour (TPB; [50]), and Protection Motivation Theory (PMT; [51]) were both identified as having a well-established supporting evidence base. Both theories have been employed previously by researchers investigating determinants of mitigation activity to protect homes against wildfire attack (TPB: [35,52]; PMT: [35,53,54]). For the present study, we chose to use TPB and PMT as theoretical bases for choosing constructs (or variables) on which to compare householders who intend to leave if threatened by a bushfire with those who intent to stay and defend. This choice was because of the accumulated evidence of the usefulness of TPB and PMT in a range of applications, including wildfire research, compared with other theoretical frameworks.

1.4. Theory of planned behaviour (TPB)

According to TPB [50], the immediate determinants of behaviour are behavioural intentions, which are in turn predicted by attitudes, subjective norms and perceived behavioural control. Attitudes are determined by behavioural beliefs which are the product of outcome beliefs and outcome evaluations. Outcome beliefs concern the perceived likelihood of a given outcome resulting from a particular behaviour, and outcome evaluations concern the desirability of that outcome. Subjective norms are the product of normative beliefs about what one should do and motivation to comply. Perceived behavioural control is the perceived ease or difficulty of performing the behaviour. The utility of the theory has been largely supported in meta-analyses, particularly in relation to health promoting and injury preventing behaviours [55-57].

Several researchers have proposed additional determinants of behavioural intentions for TPB [49]. These include: (a) descriptive norms, that is, beliefs about what others supposedly do [58]; (b) self-identification, that is, how individuals view themselves as persons [59-61]; © self-determination, that is, a sense of personal choice and autonomy when acting [62]; (d) moral norms, that is, feelings of personal responsibility to act in particular ways [63]; and (e) anticipated affect “… the prospect of feeling positive or negative emotions (e.g., exhilaration, regret) after performing or not performing a behaviour”, [63] p. 2987.

TPB is a social-cognitive theory which assumes that the behaviours of interest are under volitional control. It does not give explicit attention to the possible role of affective processes associated with threat [64] which are likely to play an important role in shaping intentions for responding to potential bushfire danger. For this reason, we also drew upon constructs from PMT, which gives threat a central role as a determinant of intentions.

1.4.1. Protection motivation theory (PMT)

According to PMT, [51,65,66] both adaptive and maladaptive responses to threats result from two appraisal processes:

[1]

1. 4{ }^{4} A small number stayed and defended despite having not planned to do because they judged that it was too dangerous to attempt to leave. ↩︎

Table 1
Constructs and item means and standard deviations, F-ratios, probabilities and effect sizes: participants intending to leave versus those intending to stay and defend following a bushfire threat warning a{ }^{a}.

Note: TPB=theory of planned behaviour construct; TPB-E=extended theory of planned behaviour construct; PMT=protection motivation theory construct; (R)=(R)= reverse-scored.
[a] the figures are Cronbach’s alpha reliability coefficients for the two-item variables.
a{ }^{a} Arranged in descending order of effect size (d) absolute magnitude; if the mean score of the stay and defend group is greater than that of the stay and defend group then dd takes a minus value.
b { }^{\text {b }} Except where shown, all items were answered on five-point scales: 1=1= disagree strongly; 2=2= disagree; 3=3= not sure; 4=4= agree; and 5=5= agree strongly.

[1]threat appraisal, and coping appraisal. When both threat appraisal and coping appraisal are high, adaptive behaviours should increase and maladaptive behaviours decrease. The threat appraisal process involves an evaluation of the threat and the rewards of maladaptive responses. The threat evaluation is based on perceived severity and susceptibility. Perceived severity involves beliefs about the seriousness of a threat, while perceived susceptibility involves beliefs about one’s vulnerability to the threat.

The coping appraisal process involves evaluation of response costs, and two efficacy components: response efficacy and self-efficacy. Response efficacy is defined in terms of one’s expectations that the recommended behaviour will effectively reduce the danger. Self-efficacy concerns one’s belief that one has the ability to successfully perform the recommended actions. According to PMT, self-efficacy and response efficacy improve the likelihood that one will adopt an adaptive (or recommended) behaviour, while costs of performing the behaviour reduce this likelihood. Like TPB, PMT assumes that behavioural intentions precede actual behaviour. In PMT, protection motivation is operationalised in terms of these intentions, which are in turn determined by the coping and threat appraisal processes. Three meta-analyses have yielded general support for the model [67-69].

The constructs which make up TPB (in its original and extended forms) and PMT encompass most of the psychological factors identified previously as being associated with individual householder wildfire mitigation intentions and actions. Two additional constructs suggested by that body of research are attachment to home [33,36,43,44]; and community member connectedness and sense of community [44-47]. Finally, because wildfires pose such a danger to life we decided to include as a construct emotional reactivity to threat. All the constructs used to compare householders intending to leave if threatened by a bushfire with those intending to stay and defend, and the items which operationalised them, are shown in Table 1.

2. Method

2.1. Participants

A total of 584 residents of bushfire-prone areas in south-eastern Australia responded to invitations to take part in a survey of community bushfire safety intentions and preparedness: 274 men ( 47%47 \% ) and 310 women ( 53%53 \% ). The mean age of the men was 56.6 years (SD=13.42)(S D=13.42), the mean age of the women was 53.0 years (SD=13.93)(S D=13.93). The majority of these householders ( 47%47 \% ) described their property as being located on urban-bushland fringes (i.e., WUI) of a regional centre or capital city. A further 33%33 \% described their location as a small rural town. The remaining 20%20 \% described their location as being an isolated dwelling on a farm or other large rural property.

2.2. Measures

A survey instrument was constructed in two formats, on-line and reply-paid postal questionnaire, with identical substantive content. Both versions specified that only one member of a household should take part. The survey presented participants with the following scenario and response choice options:

Now imagine that during the fire season you and all those who normally reside with you are at home. It has been declared a day of ‘Extreme Fire Danger’, and there is a Total Fire Ban for your Region of the State/Territory. At about 3 pm you become aware of a warning (on the radio, or a web site, or by email, or text, or telephone) that there is a large bushfire burning out of control and that it will probably hit your location in 1-2 hours. You look outside and see a large plume of smoke being blown toward your property.

What do you think you would most likely decide to do?
a. Leave as soon as you can
b. Stay to defend the home
c. Wait and see what develops, before finally deciding whether or not to leave, or to stay and defend.
(Note that option c. was included to serve as a ‘filter’ to increase the likelihood that those choosing option a. or b. was committed to their choice).

The above scenario in which householders had a time period of about an hour to two hours warning in which to react was developed after consultation with experienced fire agency staff. The scenario depicts a very common circumstance during the bushfire season in south-eastern Australia where fires often begin in remote locations and then spread to threaten farms and population centres if initial suppression attempts fail.

1. c{ }^{c} Bonferroni critical pp value for multiple comparisons =.05/19=.003=.05 / 19=.003.
   d { }^{\text {d }} Cohen’s dd effect size. Large effect: d≥0.8d \geq 0.8; medium effect: d=0.5−0.79d=0.5-0.79; and small effect: d=0.2−0.49d=0.2-0.49.
   e{ }^{e} Score range =2−10=2-10.
   1{ }^{1} Answered on a five-point scale: 1=1= not at all vulnerable; 2=2= minimally vulnerable; 3=3= moderately vulnerable; 4=4= highly vulnerable; and 5=5= extremely vulnerable.
   g{ }^{g} Score range =2−11=2-11.
   h { }^{\text {h }} Answered on six-point scales: 1=1= unlikely; 2=2= somewhat unlikely; 3=3= quite likely; 4=4= very likely; 5=5= extremely likely; and 6=6= almost certain. ↩︎

Table 2
Percentage of householders in each intention-group reporting preparatory actions completed.

[1]Depending on their response choice, participants were directed to one of three sets of items about reasons for their choice of a, b or c above. Because we were unable to locate measures used previously by TPB and PMT researchers which employed items with a specific focus on predicting wildfire safety-promoting intentions under threat we had to write new items. Item-writing followed as closely as possible the description of each theory component as described in the source material cited. We limited ourselves to writing very generic items (for example, not making reference to children or to pets or livestock since they would be irrelevant for an appreciable number of respondents). One effect of this was that a majority

1. a{ }^{a} Bonferroni critical pp values for multiple comparisons: 1. Bushfire safety planning: .05/2=.025; 2. Preparations for leaving: .05/4=.013; 3. Preparations for active house defence: .05/8=.006; 4. Preparations for reducing danger to house: .05/4=.013; 5. Preparations for reducing house vulnerability: .05/ 5=.015=.01.
   b { }^{\text {b }} Large effect: ≥.50\geq .50; medium effect: .3−.49.3-.49; and small effect: .2−.29.2-.29. ↩︎

of theory components were operationalized as single item variables. 5{ }^{5} Attachment to Home was measured using two items from Jorgensen and Stedman’s Attachment to Place scale [71]. Sense of Community was measured by two items from Long and Perkins’ Brief Sense of Community Index [72]. Emotional Reactivity to Threat was measured by two items from the Emotionality Scale of the EASI III Personality Inventory [73]. All items are shown in Table 1.

The final section of the questionnaire asked householders to indicate which actions described on a bushfire preparation checklist they had undertaken. The checklist of bushfire preparation activities was developed for research purposes by McLennan and Elliott [29], [74]. The items cover: bushfire safety planning; preparations for leaving; preparations for active house defence; preparations for reducing danger to house; and preparations for reducing house vulnerability; and are shown in Table 2.

2.3. Procedure

As noted previously, the present study was part of a larger Bushfire Cooperative Research Centre project investigating aspects of householders’ intentions and behaviour in response to bushfire threat. Senior community bushfire safety staff in the ACT Rural Fire Service, the NSW Rural Fire Service, the Tasmania Fire Service and the Victorian Country Fire Authority were each asked to nominate six locations which they regarded as being notably at-risk of serious bushfire attack but which had not experienced a serious bushfire within the previous 10 years. 6{ }^{6} They were each invited to select two locations on the bushland-urban fringe of regional centres or a capital city, two locations where householders resided in small rural towns, and two locations where most residences were isolated dwellings on farms or other large rural properties. Information published by the Australian Bureau of Statistics [75] indicated that there were approximately 28,500 occupied dwellings in the local government areas proposed. Only a relatively small percentage of these dwellings would be significantly at risk of bushfire because of distance from bushland, house construction and landscaping [76], [77]. Accordingly, it was judged not to be practical (in light of budget and time constraints) to conduct a mail survey of residents, nor to conduct interviews with selected individual householders. Instead, a procedure was adopted in which householders were invited to participate, by completing an on-line survey or requesting from the researchers a reply-paid printed questionnaire, if they considered that their homes were at risk of bushfire attack. The invitations were issued by several means: residents of small towns were mailed an invitation brochure; posters were displayed in rural community gathering points such as libraries and community centres; researchers gave interviews describing the research on local radio stations and wrote news stories for local newspapers; participants were encouraged to bring the research to the attention of neighbours. Data were collected between February and April 2012 (that is, immediately following the 2011/2012 south-eastern Australia bushfire season).

3. Results and discussion

3.1. The sample of householders

A total of 584 householders responded to the invitations to complete the survey. Comparison of participants’ age data with age distributions for the locations [75] indicated that participants who resided in country towns and on rural properties showed similar age distributions to those reported in the 2011 Census data. Participants in the age-range 18-30 were under-represented among participants from bushland-urban fringe locations.

Because the data gathering procedure required participants to ‘opt-in’ to the study in response to general invitations it is not possible to state a meaningful response rate. Those who took part are probably not representative of their wider communities in relation to concern about bushfire risk and preparation for bushfire threat. Because participation required entering an internet link or requesting a reply-paid questionnaire from the researchers, those who responded are likely to be more aware of, and psychologically engaged with, bushfire risk compared with many residents who did not participate.

Of the 584 participants, the majority ( n=273,47%n=273,47 \% ) responded to the bushfire threat scenario by choosing the ‘leave early’ intention option; 139 (24%) chose the ‘stay and defend’ option; 172 (29%) chose the ‘wait and see’ option. As noted previously, the ‘wait and see’ option served primarily as a filter to enable us to exclude householders not committed either to leaving or to staying and defending. Because of space constraints their data are not discussed further here; however, a detailed analysis of their stated reasons for intending to wait and see is reported in McLennan and Elliot [28].

3.2. Psychological differences between the two groups

Table 1 shows the means of the two groups of householders on each of the constructs studied. The results are arranged in descending order of absolute magnitudes of differences in the means, expressed as Cohen’s dd effect size values. We have used the conventional magnitude descriptors for these values: large, 0.8 ; medium, 0.5 ; small, 0.2 [78].

[1]

1. 5{ }^{5} Numerous studies have used single-item measures of TPB variables [55], [70].
   6{ }^{6} The nominated locations were: ACT: Bonython, Duffy, Fisher, Hackett, Holder, Tharwa, Weston. NSW: Captains Flat, Diggers Camp, Hornsby Heights, Kandos, Leura, Nelson Bay, Walla Walla. Tasmania: Bothwell, Deloraine, Dover, Mount Nelson, New Norfolk, Ouse, Port Sorell/Shearwater. Victoria: Beechworth, Delatite, Warrandyte, West Wodonga, Wonga Park, Yackandandah. ↩︎

3.2.1. Large differences (0.8≤d)(0.8 \leq d)

The two groups evidenced large differences on four constructs. Compared with the mean ratings of those intending to stay and defend, the mean ratings of those intending to leave were higher on: perceived safety of the chosen option; perceived preferences of significant others for the chosen option; and perceived survival chances. Those intending to leave reported a higher mean rating of anticipated anxiety about leaving compared with the mean rating of anticipated anxiety about staying and defending. Inspection of written comments by some participants about their choice of options indicated the anxiety reported by those intending to leave was mostly about the expectation that their house would probably be destroyed in their absence.

3.2.2. Medium differences (0.5≤d<0.8)(0.5 \leq d<0.8)

Those intending to leave reported a higher mean rating of the perceived cost of leaving (as “inconvenience”) compared with the mean rating by those intending to stay and defend of the cost (as inconvenience) of staying and defending. The mean level of emotional reactivity to threat of those intending to leave was higher than that of those intending to stay and defend. Those intending to leave reported higher mean ratings of the likelihood that others would take the same action.

3.2.3. Small differences (0.2≤d<0.5)(0.2 \leq d<0.5)

Those intending to leave reported: (a) a higher mean rating of the perceived severity of threat posed by bushfire compared with those intending to stay and defend; (b) a higher mean level of perceived house vulnerability to bushfire; © a higher mean level of belief that others respect those who intend to leave; and (d) a somewhat greater mean strength of intention to leave compared with the mean strength of intention of those who chose the stay and defend option. Those intending to stay and defend reported: (e) a higher mean level of sense of community; (f) a higher level of self-identification as one who avoids dangerous situations; and (g) a higher mean level of attachment to home.

3.2.4. Minor or no differences (d<0.2)(d<0.2)

There were no meaningful differences between those intending to leave and those intending to stay and defend on mean levels of: (a) self-determination; (b) moral norms to choose the selected intention; © self-efficacy-belief in their ability to carry out the intended action; (d) perceived personal vulnerability to injury while carrying out the intended action; (e) perceived behavioural control over carrying out the intended action; or (f) perceived bushfire risk.

3.2.5. Summary: differences in relation to the two theories

In summary, as Table 1 indicates, there were differences between those intending to leave and those intending to stay and defend on five of the eight TPB-E constructs: subjective norms; anticipated affect; attitudes to the intended action; descriptive norms; and self-identification. There were no meaningful differences on self-determination; moral norms; and perceived behavioural control. There were differences between the two groups of householders on four of the six PMT constructs: response efficacy; perceived costs; severity of threat; and rewards. There were no meaningful differences on self-efficacy and threat susceptibility.

3.3. Differences in bushfire preparation activities

Table 2 shows the percentages of both groups who reported that they had carried out each of the bushfire preparation actions listed. The activities are arranged in descending order of Cramer’s VV strength of association effect size within each preparation activity type: planning; preparing for leaving; preparing for house defence; reducing danger to house; and reducing house vulnerability.

Somewhat more residents who intended to stay and defend reported having prepared a household plan for what to do under bushfire threat conditions compared with those who intended to leave. There were no meaningful differences between the two groups on percentages reporting preparations for leaving. Householders intending to stay and defend were much more likely to have undertaken preparations for active house defence and preparations to reduce danger to the house. There was little difference between the two groups in frequency of preparations to reduce house vulnerability.

3.4. Discussion

We acknowledge limitations of the research. It was noted previously that the overall levels of concern about, and engagement with, bushfire risk by participants were probably higher than those of many residents in the communities who did not participate. This limits the extent to which findings can be generalised. It should be noted that the findings come from the Australian policy context in which residents are usually free to choose to evacuate or not and thus may not generalise to situations where mandatory evacuations are usual-as is the case in many North American jurisdictions. The issue of possible differential effects of choice about leaving or staying and defending versus being ordered to evacuate on householder planning, preparation and action seems worthy of further research. Our use of single item measures of many constructs was noted before. Single-item measures have been criticised by some on two grounds: (a) questionable validity because a single item may fail to capture all the aspects of a construct; and (b) greater measurement error. However, in relation to (a), others such as Bergkvist and Rossiter [79], have demonstrated that single item measures do not necessarily

demonstrate lower levels of predictive validity compared with multi-item measures. In relation to (b) measurement error would increase the error variance of differences between variables, thus our findings are likely to be conservative. Finally, we acknowledge that our choice of constructs on which to compare householders who intend to leave and householders who intend to stay and defend may have overlooked other psychological characteristics likely to be important in differentiating between the two groups. Additional studies using alternative constructs and items can shed further light on this possibility.

Having due regard for the limitations noted above, the study has several strengths. Data were provided by residents of areas judged by fire agency staff to be significantly at risk of serious bushfire events. However, the areas had not experienced a serious bushfire within ten years, thus resembling the risk-history of much of south-eastern Australia. The research employed a quasi-experimental design involving quantitative data, and the items used operationalised constructs derived from theoretical frameworks.

The findings suggest that there are important psychological differences between householders who intend to leave and householders who intend to stay and defend if threatened by a bushfire. Those who intend to leave are more likely to report higher levels of concern about the danger posed by bushfires; however, they do not believe that they are more likely to be threatened, compared with those who intend to stay and defend. They are likely to view themselves and their homes as vulnerable in the event of a bushfire and they tend to believe that others understand leaving as the desirable course of action to follow. Notwithstanding, they are likely to be concerned about the likelihood of their home being destroyed in their absence and they view leaving as likely to be inconvenient. Those who intend to stay and defend wish to protect their valued property and they believe that their endeavours will be successful. They do not see themselves as risk takers. They view themselves as being connected to other members of the community.

Despite the higher mean level of reported strength of intention to carry out their chosen option, those intending leave were much less likely to have planned how to implement their intention. Further, while many of those intending to leave were anxious about their house being destroyed in their absence, relatively few had undertaken preparations likely to increase the likelihood that their house would survive undefended. This paradoxical finding warrants further research. In relation to health promotion and illness prevention, Witte [80,81] proposed that a high level of perceived threat coupled with a low level of perceived efficacy is likely to lead to defensive fear-control processes such as denial of risk and maladaptive behaviours such as information-avoidance. There is some support for this explanation in relation to earthquake preparedness [82] and flood preparation [83] and it seems worthy of further investigation in the context of household wildfire safety.

4. Conclusions

1. Householders’ decisions to leave or to stay and defend their property if threatened by a bushfire are, for most, driven by quite different motivations: the former by concerns about personal and family safety, the latter by desire to protect valued property with an acceptance of some degree of risk to safety.
2. Despite the policy position of Australian fire agencies that leaving is always the safest option, an appreciable percentage of residents of communities which are at risk of bushfires are unlikely to heed messages that the safest plan is to leave well before a bushfire threatens: many of these residents will plan to stay and defend their property. This implies that agencies should ensure that their community bushfire safety education endeavours incorporate effective means of (a) motivating householders intending to stay and defend to undertake systematic risk assessments of their specific property and family member circumstances if threatened by bushfire under a range of fire danger weather conditions; and (b) informing them of the requirements for safe and effective defence under these different threat levels (see Penman et al. [84]).
3. There are important overall differences in the psychological characteristics of householders who intend to leave if threatened by a bushfire and those who intend to stay and defend their property which should be taken into account by agency staff when they develop community bushfire safety programs and messages. It is likely that omnibus messages which fail to take these differences into account may be less effective than messages which target specifically those who intend to leave rather than stay and defend their homes, and those who intend to stay and defend their valued properties. Of course, this is an empirical issue and needs to be tested.
4. The findings presented here seem to suggest that householders who intend to leave in the face of a bushfire threat warning do not need reminding of how dangerous bushfires are. What many probably need is: (a) convincing that leaving at the last minute may expose them and their family to unwarranted danger; and (b) encouragement, information and assistance to implement the four basic elements of a sound plan to leave safely-(i) have a safe destination; (ii) have a safe route; (iii) have a kit of valuables and essential documents plus family necessities for 72 h prepacked and ready to go; and (iv) perhaps most important of all, have an agreed-upon trigger to leave.
5. It seems that relatively few householders prepare their house and surrounds so as to reduce the likelihood of ignition by a bushfire in their absence. This may be because: (a) they have not undertaken a realistic assessment of the vulnerabilities of their home to bushfire attack; (b) they do not believe that such preparation would be effective in possibly preventing house loss; © anxiety about bushfire threat may lead to defensive denial and information avoidance; (d) they

do not know what actions to take; (e) the actions are regarded as being too inconvenient-especially those that have to be carried out regularly, such as vegetation pruning and leaf-litter removal, or too costly;-or any combination of (a)-(e). This is an issue that fire agencies could investigate and develop new strategies to target these potential barriers to house threat mitigation.
6. It appears that many of those householders who intend to stay and defend their property are inadequately prepared to do so under the fire danger weather conditions they are likely to encounter [84,85]). Given that an appreciable percentage of residents in at-risk areas almost certainly will stay and defend, agencies may wish to review and broaden their approaches to providing advice and assistance about property risk assessment and adequate preparations for safe and effective defence. At present, there is heavy reliance on making written information available on agency websites and expecting that residents will avail themselves of this information [86]. The recent post-bushfire interview studies cited previously [24-27] suggest that relatively few residents do so. Agencies need to develop new on-line video motivational and information bushfire safety presentations and evaluate their effectiveness, together with on-line household and property bushfire risk assessment tools. Simply making information about the dangers of staying and defending available to those householders who wish to protect their valued properties is unlikely to contribute greatly to community bushfire safety.

Acknowledgements

As Project Manager, Glenn Elliott organised data collection and data entry. The research was supported by a Bushfire Cooperative Research Centre Extension Grant. However, the views expressed are those of the authors and do not necessarily reflect the views of the Board of the funding agency. The research was approved by the La Trobe University Human Ethics Committee. We are grateful to Professor Geoff Cumming for advice on aspects of the statistical analyses. We thank two anonymous reviewers for advice on revising a previous version of the paper.

References

[1] S. Ellis, P. Kanowski, R. Whelan, National Inquiry into Bushfire Mitigation and Management, Council of Australian Governments, Canberra, 2004.
[2] R. Blanchi, J. Leonard, K. Haynes, K. Opie, M. James, F. de Oliveira, Environmental circumstances surrounding bushfire fatalities in Australia 1901-2011, Environ. Sci. Policy 37 (2014) 192-203, http://dx.doi.org/10.1016/j.ensci.2013.09.013.
[3] W. Mell, S. Manzello, A. Maranghides, D. Butry, R. Rehm, The wildland-urban interface fire problem-current approaches and research needs, Int. J. Wildland Fire 19 (2010) 238-251.
[4] A. Aldersley, S. Murray, S. Cornel, Global and regional analysis of climate and human drivers of wildfire, Sci. Total Environ. 409 (2011) 3472-3481, http: //dx.doi.org/10.1016/j.scitotens.2011.05.032.
[5] A. Gill, S. Stephens, Scientific and social challenges for the management of fire-prone wildland-urban interfaces, Environ. Res. Lett. 4 (2009) 034014, http://dx.doi.org/10.1088/1748-9326/4/3/034014.
[6] A. Gill, S. Stephens, G. Cary, The worldwide “wildfire” problem, Ecol. Appl. 23 (2013) 438-454.
[7] K. Hennessy, C. Lucas, N. Nicholls, J. Bathols, R. Suppiah, J. Ricketts, Climate Change Impacts on Fire-weather in South-east Australia, CSIRO, Collingwood, 2006/2013 (http://www.csiro.au/Outcomes/Climate/Adapting/Climate-Change-Fire-Weather.aspx#a3).
[8] Y. Liu, J. Stanturf, S. Goodrick, Trends in global wildfire potential in a changing climate, For. Ecol. Manag. 259 (2010) 685-697, http://dx.doi.org/10.1016/ j.foreco.2009.09.002.
[9] Y. Liu, S. Goodrick, W. Heilman, Wildland fire emissions, carbon, and climate: wildfire-climate interactions, For. Ecol. Manag. 317 (2013) 80-96, http: //dx.doi.org/10.1016/j.foreco.2013.02.020.
[10] F. Moreira, O. Viedma, M. Arianoutsou, T. Curt, N. Koutsias, E. Rigolot, et al., Landscape-wildfire interactions in southern Europe: implications for landscape management, J. Environ. Manag. 92 (2011) 2389-2402, http://dx.doi.org/10.1016/j.jenvman.2011.06.028.
[11] Australian Fire Authorities Council Position Paper on Bushfires and Community Safety, East Melbourne. (http://www.royalcommission.vic.gov.au/ Documents/Document-files/Exhibits/TEN-001-001-0077;), 2005.
[12] A. Tibbits, J. Whittaker, Stay and defend or leave early: policy, problems and experiences during the 2003 Victorian bushfires, Environ. Hazards 7 (2007) 283-290.
[13] S. McCaffrey, A. Rhodes, M. Stidham, Wildfire evacuation and its alternatives: perspectives from four United States communities, Int. J. Wildland Fire ; http://dx.doi.org/10.1071/WF13050\. (published online).
[14] T. Paveglio, M. Carrol, P. Jakes, Alternatives to evacuation-protecting public safety during wildland fire, J. For. 106 (2008) 65-70.
[15] T. Paveglio, M. Carroll, P. Jakes, Adoption and perception of shelter-in-place in California’s Ranch Santa Fe fire district, Int. J. Wildland Fire 19 (2010) 677-688, http://dx.doi.org/10.1071/WF09034.
[16] T. Paveglio, A. Boyd, M. Carroll, Wildfire evacuation and its alternatives in a post-Black Saturday landscape: catchy slogans and cautionary tales, Environ. Hazards 11 (2012) 52-70, http://dx.doi.org/10.1080/17477891.2011.635185.
[17] S. McCaffrey, A. Rhodes, Public response to wildfire: is the Australian “stay and defend or leave early” approach an option for wildfire management in the United States? J. For. 107 (2009) 9-15.
[18] A. Fu, The façade of safety in California’s shelter-in-place homes: history, wildfire and social consequences, Crit. Sociol. 39 (2012) 833-849, http://dx. doi.org/10.1177/0896920512455936.
[19] J. Gledhill, Community self-reliance during bushfires: the case for staying at home. Paper Presented at the Third International Wildland Fire Conference, Sydney, 3-6 October 2003. (http://www.fire.uni-freiburg.de/summit-2003/3-IVFFC/Papers/3-IVFFC-061-Gledhill.pdf/).
[20] J. Whittaker, K. Haynes, J. Handmer, J. McLennan, Community safety during the 2009 Australian ‘Black Saturday’ bushfires: an analysis of household preparedness and response, Int. J. Wildland Fire 22 (2013) 841-849, http://dx.doi.org/10.1071/WF12010.
[21] B. Teague, R. McLeod, S. Pascoe, Victorian Bushfires Royal Commission Final Report, State Government of Victoria, Melbourne, 2009.
[22] Australasian Fire and Emergency Services Authorities Council, Bushfires and Community Safety: Position; version 4.1, East Melbourne. (http:// knowledgeweb.afac.com.au/positions/documents/AFAC_Position_Bushfires_Community_Safety_v4.1.pdf;), 2010.
[23] Australasian Fire and Emergency Services Authorities Council, Bushfires and Community Safety: Position; version 5.0, East Melbourne. (http://www.

afac.com.au/docs/position/bushfires-community-safety.pdf?sfvrsn = 20:j, 2012.
[24] J. Boylan, C. Cheek, T. Skinner, Preliminary Report on the January 2013 Fires in the South-eastern Tasmania Research Project, University of Tasmania Rural Clinical School, Burnie, 2013.
[25] B. Mackie, J. McLennan, L. Wright, Community Understanding and Awareness of Bushfire Safety: January 2013 Bushfires, Bushfire Cooperative Research Centre, East Melbourne, 2013.
[26] J. McLennan, L. Wright, A. Birch, Community Understanding and Awareness of Bushfire Safety: Residents’ Experiences of the October 2013 Bushfires in Three Areas, Bushfire & Natural Hazards Cooperative Research Centre, East Melbourne, 2014.
[27] J. McLennan, Capturing Community Members’ Bushfire Experiences: Interviews With Residents Following the 12 January 2012 Parkerville (WA) Fire, Bushfire & Natural Hazards Cooperative Research Centre, East Melbourne, 2014.
[28] J. McLennan, G. Elliott, “Wait and see”: the elephant in community bushfire safety room?, in: R. Thornton, L. Wright (Eds.), Proceedings of the AFAC/ Bushfire CRC Research Forum, Perth, September 2012, Bushfire Cooperative Research Centre, East Melbourne, 2013, pp. 56-69.
[29] J. McLennan, G. Elliott, L. Wright, Bushfire survival preparations by householders in at-risk areas of south-eastern Australia, Aust. J. Emerg. Manag. 29 (2) (2014) 11-17.
[30] J. McLennan, S. Cowlishaw, D. Paton, R. Beatson, G. Elliott, Predictors of south-eastern Australian householders’ strengths of intentions to self-evacuate if a wildfire threatens: two theoretical models, Int. J. Wildland Fire ; http://dx.doi.org/10.1071/WF13219\. (Published online.).
[31] S. McCaffrey, E. Toman, M. Stidham, B. Shindler, Social science research related to wildfire management: an overview of recent findings and future research needs, Int. J. Wildland Fire 22 (2013) 15-24, http://dx.doi.org/10.1071/WF11115.
[32] H. Brenkert-Smith, P. Champ, N. Flores, Insights into wildfire mitigation decisions among wildland-urban interface residents, Soc. Nat. Resour. 19 (2006) 759−768759-768.
[33] T. Collins, What influences hazard mitigation? Household decision making about wildfire risks in Arizona’s White Mountains, Prof. Geogr. 60 (2008) 508−526508-526.
[34] H. Faulkner, B. McFarlane, T. McGee, Comparison of homeowner response to wildfire risk among towns with and without wildfire management, Environ. Hazards 8 (2009) 38-51.
[35] T. Hall, M. Slothower, Cognitive factors affecting homeowners’ reactions to defensible space in the Oregon coast range, Soc. Nat. Resour. 22 (2009) 95−11095-110.
[36] W. Martin, I. Martin, B. Kent, The role of risk perceptions in the risk mitigation process: the case of wildfire in high risk communities, J. Environ. Manag. 91 (2009) 489-498.
[37] S. McCaffrey, M. Stidman, E. Toman, B. Shindler, Outreach programs, peer pressure and common sense: what motivates homeowners to mitigate wildfire risk? Environ. Manag. 48 (2011) 475-488.
[38] T. McGee, S. Russell, “It’s just a natural way of life…” an investigation of wildfire preparedness in rural Australia, Environ. Hazards 5 (2003) 1-12.
[39] T. Faseglio, M. Carroll, J. Abscher, W. Robinson, Symbolic meanings of wildland fire: a study of residents in the U.S. inland northwest, Soc. Nat. Resour. 24 (2010) 18-33.
[40] S. Schulte, K. Miller, Wildfire risk and climate change: the influence of homeowner mitigation behaviour in the wildland-urban interface, Soc. Nat. Resour. 23 (2010) 417-435.
[41] D. Paton, G. Kelly, P. Burgelt, M. Doherty, Preparing for bushfires: understanding intentions, Disaster Prev. Manag. 15 (2006) 566-575.
[42] G. Winter, J. Fried, Homeowner perspectives on fire hazard, responsibility, and management strategies at the wildland-urban interface, Soc. Nat. Resour. 13 (2000) 33-49.
[43] G. Kyle, G. Theodori, J. Abscher, J. Jun, The influence of home and community attachment on firewise behavior, Soc. Nat.Resour. 23 (2010) 1075-1092.
[44] D. Paton, P. Burgelt, T. Prior, Living with bushfire risk: social and environmental influences on preparedness, Aust. J. Emerg. Manag. 23 (3) (2008) 41−4841-48.
[45] H. Brenkert-Smith, Building bridges to fight fires: the role of informal social interaction in six Colorado wildland-urban interface communities, Int. J. Wildland Fire 19 (2010) 689-697.
[46] M. Frandsen, D. Paton, K. Sakariassen, Fostering community preparedness through engagement and empowerment, Aust. J. Emerg. Manag. 26 (2) (2011) 24−3024-30.
[47] T. Prior, D. Paton, Understanding the context: the value of community engagement in bushfire risk communication and education. Observations following the East Coast Tasmania bushfires of December 2006, Aust. J. Disaster Trauma Stud. (2008). < http://www.massey.ac.nz/ trauma/issues/ 2008-2/prior.htm > .
[48] J. McLennan, G. Elliott, M. Omodei, J. Whittaker, Householders’ safety-related decisions, plans, actions and outcomes during the 7 February 2009 Victorian (Australia) wildfires, Fire Saf. J. 61 (2013) 175-184, http://dx.doi.org/10.1016/j.firesaf.2013.09.003.
[49] R. Beatson, J. McLennan, What applied psychology theories might contribute to community bushfire safety research after Victoria’s “Black Saturday”, Aust. Psychol. 46 (2013) 171-182, http://dx.doi.org/10.1111/1742-9544.2011.00041.x.
[50] I. Ajzen, The theory of planned behavior, Org. Behav. Hum. Decis. Process. 50 (1991) 179-211.
[51] R. Rogers, Cognitive and physiological processes in fear-based attitude change: a revised theory of protection motivation, in: J. Caciopo, R. Petty (Eds.), Social Psychophysiology: A Sourcebook, Guilford, New York, 1983, pp. 153-176.
[52] B. Bates, B. Quick, A. Kloss, Antecedents of intention to help mitigate wildfire: implications for campaigns promoting wildfire mitigation to the general public in the wildland-urban interface, Saf. Sci. 47 (2009) 374-381, http://dx.doi.org/10.1016/j.ssc.2008.09.002.
[53] I.M. Martin, H. Bender, C. Raish, What motivates individuals to protect themselves from risks, Risk Anal. 27 (2007) 887-900.
[54] B.I. McFarlane, T.K. McGee, H. Faulkner, Complexity of homeowner wildfire risk mitigation: an integration of hazard theories, Int. J. Wildland Fire 20 (2011) 921-931, http://dx.doi.org/10.1071/WF10096.
[55] C.J. Armitage, M. Conner, Efficacy of the theory of planned behaviour: a meta-analytic review, Br. J. Soc. Psychol. 40 (2001) 471-499.
[56] R. Cook, D. French, How well do the theory of reasoned action and the theory of planned behaviour predict intentions and attendance at screening programs? A meta-analysis, Psychol. Health 23 (2008) 745-765.
[57] G. Godin, G. Kok, The theory of planned behavior: a review of its applications to health-related behaviors, Am. J. Health Promot. 11 (1996) 87-98.
[58] A. Rivis, P. Sheeran, Descriptive norms as an additional predictor in the theory of planned behaviour: a meta-analysis, Curr. Psychol.: Dev.Learn. Personal. Soc. 3 (2003) 218-233.
[59] J. Christian, C.J. Armitage, D. Abrams, Predicting uptake of housing services: the role of self-categorization in the theory of planned behaviour, Curr. Psychol.: Dev.Learn. Personal. Soc. 22 (2003) 206-217.
[60] J. Rise, P. Sheeran, S. Hukkelberg, The role of self-identity in the theory of planned behaviour: a meta-analysis, J. Appl. Soc. Psychol. 40 (2010) 1085-1105.
[61] D.J. Terry, M.A. Hogg, K. White, The theory of planned behaviour: self-identity, social identity and group norms, Br. J. Soc. Psychol. 38 (1999) 235-244.
[62] A. Rivis, P. Sheeran, C. Armitage, Expanding the affective and normative components of the theory of planned behavior: a meta-analysis of anticipated affect and moral norms, J. Appl. Soc. Psychol. 39 (2009) 2985-3019.
[63] M.S. Hagger, L.D. Chatzisarantis, Integrating the theory of planned behaviour and self-determination theory in health behaviour: a meta-analysis, Br. J. Health Psychol. 14 (2009) 275-302.
[64] N. Weinstein, Testing four competing theories of health-protective behavior, Health Psychol. 12 (1993) 324.
[65] H. Boer, E. Seydel, Protection motivation theory, in: M. Connor, P. Norman (Eds.), Predicting Health Behaviour, Open University Press, Buckingham, 1996, pp. 95-120.
[66] J. Maddux, R. Rogers, Protection motivation theory and self-efficacy: a revised theory of fear appeals and attitude change, J. Exp. Soc. Psychol. 19 (1983) 469−479469-479.

[67] D.L. Floyd, S. Prentice-Dunn, R. Rogers, A meta-analysis of research on protection motivation theory, J. Appl. Soc. Psychol. 30 (2000) 407-429.
[68] S. Milne, P. Sheeran, S. Orbell, Prediction and intervention in health-related behavior: a meta-analytic review of protection motivation theory, J. Appl. Soc. Psychol. 30 (2000) 106-143.
[69] S. Munro, S. Lewin, T. Swart, J. Volinink, A review of health behaviour theories: how useful are these for developing interventions to promote longterm adherence for TB and HIV/AIDS? BMC Public Health 7 (2007) 104.
[70] H.-J. Paek, K. Hilyard, V. Freimuth, J. Barge, L. Mindlin, Theory-based approaches to understanding public emergency preparedness: implications for effective health and risk communication, J. Health Commun.: Int. Perspect. 15 (2010) 428-444, http://dx.doi.org/10.1080/10810731003753083.
[71] B.S. Jorgensen, R. Steadman, Sense of place as an attitude: Lakeshore owners’ attitudes towards their properties, J. Environ. Psychol. 21 (2001) 233-248.
[72] D.A. Long, D. Perkins, Confirmatory factor analysis of the Sense of Community Index and development of a Brief SCI, J. Commun. Psychol. 31 (2003) 279−296279-296.
[73] V. Braithwaite, P. Duncan-Jones, R. Bosly-Craft, M. Goodchild, A psychometric investigation of the EASI-III Temperament Survey in the Australian general population, Aust. Psychol. 36 (1984) 85-95.
[74] J. McLennan, G. Elliott, Checklist Items for Researchers: Householder Preparations for Bushfires, 2011, School of Psychological Science La Trobe University and Bushfire CRC, Melbourne, 2011.
[75] Australian Bureau of Statistics, Community Profiles, Canberra. (http://www.abs.gov.au/websitedbs/censushome.nsf/home/communityprofiles/), 2011.
[76] K. Chen, J. McAneney, Quantifying bushfire penetration into urban areas in Australia, Geophys. Res. Lett. 31 (12) (2004) L12212, http://dx.doi.org/ 10.1029/204G12020244.
[77] A.A.G. Wilson, I. Ferguson, Predicting the probability of house survival during bushfires, J. Environ. Manag. 23 (1986) 259-270.
[78] J. Cohen, A power primer, Psychol. Bull. 12 (1992) 155-159.
[79] L. Bergkvist, J. Rossiter, The predictive validity of multiple-item versus single-item measures of the same constructs, J. Market. Res. 44 (2007) 175-184.
[80] K. Witte, Putting the fear back into fear appeals: the extended parallel process model, Commun. Monogr. 59 (1992) 329-349.
[81] K. Witte, Fear control and danger control: a test of the extended parallel process model (EPPM), Commun. Monogr. 61 (1994) 113-132.
[82] D. Paton, L. Smith, D. Johnston, When good intentions turn bad: promoting natural hazard preparedness, Aust. J. Emerg. Manag. 20 (1) (2005) 25-30.
[83] T. Harries, Feeling secure or being secure? Why it can seem better not to protect yourself against a natural hazard, Health, Risk Soc. 10 (2008) 479-490, http://dx.doi.org/10.1080/13698570802381162.
[84] T.D. Penman, C. Eriksen, R. Blanchi, M. Chladil, A.M. Gill, K. Haynes, et al., Defining adequate means of residents to prepare property for protection from wildfire, Int. J. Disaster Risk Reduct. 6 (2013) 67-77, http://dx.doi.org/10.1016/j.ijdrr.2013.09.001.
[85] R. McRae, J. Sharples, A conceptual framework for assessing the risk posed by extreme bushfires, Aust. J. Emerg. Manag. 26 (2) (2011) 47-53.
[86] C. Eriksen, T. Prior, Defining the importance of mental preparedness for risk communication and residents well-prepared for wildfire, Int. J. Disaster Risk Reduct. 6 (2013) 87-97, http://dx.doi.org/10.1016/j.ijdri.2013.09.006.