<p>Natural hazards impact millions of people globally and lead to billions of dollars of economic loss each year. New Zealand is one of many countries vulnerable to multiple natural hazards including earthquakes, tsunami, volcanic eruption, and high impact weather. Despite these well-known risks, many residents of the capital city of Wellington have taken few steps to be more prepared. This location, therefore, presented an appropriate population for the investigations within this thesis. Decades of social science research has identified a multitude of factors related to preparation behaviour. However, many of these factors, such as risk perception and previous experience of natural hazards, are difficult to manipulate successfully in broad public education campaigns. The first main aim of this thesis was therefore to identify thoughts and beliefs about preparing which predict people’s preparation behaviour, with the aim of identifying specific factors which can be efficiently and effectively targeted in campaigns encouraging preparation. This research was structured using the Theory of Planned Behaviour (TPB) which proposes a specific set of cognitions that affect intention formation as well as beliefs which inform those cognitions. In addition to using this model to structure the investigation, this thesis undertook several refinements and extensions to the model to address inconsistencies within TPB research. The intent of doing so was to provide a set of findings and a questionnaire which are not limited by those inconsistencies and which can be adapted for a range of behavioural contexts. Four empirical studies were carried out, involving a total of 2,298 participants from the general population. Study 1 tested the applicability of this theory to the context of natural disaster preparation and clarified the nature of that context. Using an online survey of 722 residents of the urban Wellington region, Study 1 demonstrated low levels of preparation, identified not “getting around to it” as a main barrier to preparation, and supported a more extensive application of the TPB with the factors explaining approximately 16% of the variance in intentions. Study 2 tested a full TPB questionnaire including both the two-factor distinction (splitting attitudes into instrumental and experiential, norms in descriptive and injunctive, and perceived behaviour control into controllability and self-efficacy) and belief components which are proposed to precede attitudes, norms, and control. This study used an online sample of 603 Wellington residents. All the cognitions within the theory except perceived descriptive norms were significantly associated with either past behaviour or intentions to prepare, explaining approximately 47% of the variance in intentions. This study also included an experimental framing manipulation, demonstrating benefits of referring to “natural hazard” preparation rather than “natural disaster” preparation. Study 3 concluded the development of the TPB questionnaire by assessing intentions, cognitions, and beliefs at one point in time and behaviour one month later to allow for stronger inferences about causality, with a sample of 61 participants across both time points. This study used a different recruitment method than the previous studies: posted survey invites using addresses randomly selected from the electoral roll. Although this method did not produce a more demographically representative sample than the recruitment method used in Studies 1 and 2 as intended, Study 3 reproduced the findings of the “natural hazard” condition in Study 2. Finally, this study identified several key beliefs related to preparing such as the belief that preparing helps people to get through a natural hazard event better, that people can make the effort to prepare, and that preparing can be fun and rewarding. These beliefs offer specific and tangible factors which can be efficiently addressed in public education campaigns. An intervention run previously by the New Zealand National Emergency Agency, the ShakeOut earthquake drill, was retrospectively evaluated in Study 4 by comparing those who did and those who did not participate (N = 480) using the TPB framework. Those who participated in the drill demonstrated better knowledge and use of the protective actions that are the focus of the drill than those who did not participate. Although this intervention was not informed by the findings of the previous studies, drill participants also demonstrated more positive scores for the TPB cognitions and intentions compared to non-participants, although more of the variance in intentions was explained for the latter (approximately 41%) than the former (approximately 31%). Overall, findings from the empirical studies support the recommendation for consistent inclusion of all tested refinements to the Theory of Planned Behaviour (i.e., the two-factor distinction, the inclusion of belief measures, and careful development of behavioural measures). This thesis represents a uniquely thorough test of the Theory of Planned Behaviour to natural hazard preparation with important implications for both the contextual value of the theory and how the theory is applied in research broadly. The research also supports previous findings of the importance of instrumental attitudes and self-efficacy for natural hazard preparation and contributed the novel factor of experiential attitudes as well as identifying new, specific beliefs to target in public education campaigns. These contributions to our understanding of why people prepare for natural hazards can be used to encourage more people in Wellington, in New Zealand, and globally to be more prepared.</p>
