UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Impact from an interface wildfire, place attachment, and depression : a survey of two rural communities Handler, Risa Jordana 2007

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2007-0426.pdf [ 8.41MB ]
Metadata
JSON: 831-1.0053635.json
JSON-LD: 831-1.0053635-ld.json
RDF/XML (Pretty): 831-1.0053635-rdf.xml
RDF/JSON: 831-1.0053635-rdf.json
Turtle: 831-1.0053635-turtle.txt
N-Triples: 831-1.0053635-rdf-ntriples.txt
Original Record: 831-1.0053635-source.json
Full Text
831-1.0053635-fulltext.txt
Citation
831-1.0053635.ris

Full Text

IMPACT F R O M A N INTERFACE WILDFIRE, P L A C E A T T A C H M E N T , A N D DEPRESSION: A S U R V E Y OF TWO R U R A L COMMUNITIES by RISA JORDANA H A N D L E R B.Sc, Dalhousie University, 1995 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE F A C U L T Y OF G R A D U A T E STUDIES (Counselling Psychology) THE UNIVERSITY OF BRITISH C O L U M B I A August 2007 © Risa Jordana Handler, 2007 A B S T R A C T Disaster recovery research has overlooked the important role of place attachment in recovery to a natural disaster, particularly its association with the initial impact on residents and their mental health. Although increased depression immediately following a natural disaster is well documented, there is a paucity of research that has examined the long-term effects of natural disasters on depressive symptoms. Thus, I examined the relationship between the initial impact of a wildfire on residents of two rural communities in British Columbia (3.5 years prior), their current perceptions of attachment to their community, and their current symptoms of depression. The 2003 McLure Wildfire provides the context for the present exploratory study. I employed a cross-sectional survey design. A sample of 104 male and female residents, with an average age of 56 years, was recruited from both Louis Creek and Barriere to complete self-report measures. I developed two instruments for use in the present study; one assessed the initial impact of the wildfire and the other perceptions of current place attachment. Factor analyses supported the structural validity of the instruments; three factors represented place attachment dimensions (social, natural, and built environments), and one represented change in attachment. Depressive symptoms were measured using a 10-item short version of the Center for Epidemiologic Studies Depression Scale (Andresen, Malmgren, Carter, & Patrick, 1994). Hierarchical multiple regression (HMR) analyses showed that the greater impact the wildfire had on respondents, the (a) less attached to the built environment they felt, and (b) the greater their symptoms of depression. Furthermore, feeling less attached to the natural environment was related to greater symptoms of depression. Age and marital status were controlled for in H M R analyses predicting attachment and depressive symptoms. Ill Although Louis Creek respondents experienced higher mean levels of impact and depressive symptoms, compared with those from Barriere, once age and/or marital status were controlled for, community differences were not statistically significant. The findings support a multidimensional framework of place attachment. Limitations and implications for further research, including the development of a valid measure of place attachment, are discussed. T A B L E OF CONTENTS A B S T R A C T . . . ii T A B L E O F C O N T E N T S iv L I S T O F T A B L E S vi L I S T O F F I G U R E S . vii A C K N O W L E D G M E N T S viii C H A P T E R 1: I N T R O D U C T I O N 1 NATURAL DISASTERS 2 , INTERFACE WILDFIRES 3 M C L U R E FIRE 2003 ; .....4 PURPOSE OF THE STUDY 6 C H A P T E R 2: L I T E R A T U R E R E V I E W 9 PLACE ATTACHMENT 10 Summary 15 NATURAL DISASTER AND IMPLICATIONS ON ATTACHMENT TO PLACE 15 Summary ...... 22 OPERATIONALIZING SOCIAL AND PHYSICAL DIMENSIONS OF PLACE ATTACHMENT . . . . . . 23 Social Dimension 25 Physical Dimension - Natural Environment..... 29 Physical Dimension - Built Environment 31 DEPRESSION: A N OUTCOME OF NATURAL DISASTERS 32 Natural Disasters and Depression 32 Attachment and Depression :. 33 Summary 36 SUMMARY 36 C H A P T E R 3: M E T H O D 39 PARTICIPANTS AND PROCEDURE .' 39 MEASURES , , 45 Demographic Information......... 45 Impact of the 2003 McLure Fire 45 Place Attachment : 47 Depressive Symptoms 54 Control Variables 55 RESEARCH QUESTIONS 56 DATA ANALYSIS 57 C H A P T E R 4: R E S U L T S .61 PRELIMINARY ANALYSES 61 RESEARCH QUESTION #1 66 RESEARCH QUESTIONS #2 AND #3 69 V CHAPTER 4: DISCUSSION 72 THE RELATIONSHIP BETWEEN IMPACT AND PLACE ATTACHMENT 72 IMPACT FROM THE M C L U R E FIRE AND SYMPTOMS OF DEPRESSION 76 PLACE ATTACHMENT AND DEPRESSION . 77 STRENGTHS AND LIMITATIONS ! 79 SUGGESTIONS FOR FURTHER RESEARCH 83 CONCLUSION 85 REFERENCES ..86 APPENDIX A: REGIONAL MAPS 95 APPENDIX B: QUESTIONNAIRE 97 APPENDIX C: FLYER DISTRIBUTED IN THE COMMUNITY 109 APPENDIX D: CONSENT FOR QUESTIONNAIRE (COVER LETTER)... 110 APPENDIX E: SURVEY REMINDER CARD 112 APPENDIX F: IMPACT FROM THE WILDFIRES: SURVEY ITEM DEVELOPMENT 113 APPENDIX G: PLACE ATTACHMENT: SURVEY ITEM DEVELOPMENT... 114 APPENDIX H: CONTENT ANALYSIS 117 V I LIST OF TABLES Table 1. Demographic information of community residents by Barriere and Louis Creek (and other surrounding communities)..... 40 Table 2. Principal-components analysis of the impact of the 2003 McLure'fire measure. 46 Table 3. Principal-Components Analysis of Place Attachment Measure 51 Table 4. Bivariate correlations and internal consistency estimates for attachment subscales 53 Table 5. Skewness and Kurtosis Values 58 Table 6. Means and standard deviations of study variables for the full sample, Louis Creek and the surrounding area, and Barriere 61 Table 7. Intercorrelations of Variables 62 Table 8. Summary of Hierarchical Multiple Regression Predicting Social Attachment. 66 Table 9. Summary of Hierarchical Multiple Regression Predicting Built Environment Attachment 67 Table 10. Summary of Hierarchical Multiple Regression Predicting Change in Attachment 68 Table 11. Summary of Hierarchical Multiple Regression for Variables Predicting Depression 70 VI I LIST OF FIGURES FIGURE 1: Rebuilt Louis Creek home and surrounding hills 3.5 years after the McLure Fire 5 FIGURE 2: Conceptual.framework of place attachment 13 FIGURE 3: Revised final model of place attachment 53 viii A C K N O W L E D G M E N T S This project would not have been possible without the support from numerous individuals. First and foremost, I would like to thank Dr. Bonita.Long for her consistent mentorship, guidance, encouragement, and endless support. A l l students should be so lucky to have such a positive experience with a supervisor that sticks by their side. I have very much enjoyed our collaborative process throughout this project and I am honoured to be your final student as you prepare for retirement. Thank you to Habib Chaudhury, Richard Young, and Anita Hubley for taking the time to participate on my committee and support me through this process. Thank you to Maria Trache for not only statistical support, but also stimulating conversations and encouraging me to figure out what exactly is 'my story.' This project would not have been possible without the enormous generosity given by certain members of the communities under study. They volunteered their time, patience, and wisdom in helping me disseminate questionnaires and providing a perspective on the communities to which I would not have known otherwise. Thank you also to several other community members who volunteered to participate in this project, welcoming me into their homes, and sharing their stories. This mountain would have seemed impassable i f it were not for certain special friends that both directly and indirectly helped me along the way. Dave, you have no idea how much I have appreciated the consistent comic relief, and emotional shoulder that you have provided, right from the very beginning. Jen, you have been a constant, calling and checking in frequently, and providing a home for me in which to escape. Gary, this thesis had the potential to make no sense i f it was not for your keen way of putting ideas ix together. Leanne, you symbolically peeled me off the floor and consistently conveyed your father's wisdom of "just get 'er done!" A big thank you to my friends who allowed me to set up camp in their homes and offices in order to keep me company during the latter lonely days. To all my friends - including Annelies, John St.L, John S., Yonan -from Vancouver, Victoria, the rest of BC, and Toronto thank you especially to those who have stuck by me, despite my sheer inability to make a committed plan to see them. And finally, a big thank you to my family, especially my father, for always believing in me and encouraging me to never give up. 1 CHAPTER 1: INTRODUCTION Place attachment represents the emotional bond we have to our intimate places-house, community, public places, city, region, or country (Low & Altman, 1992). Places are symbolic of who we are, where we came from, and how we want to continue relating to the world. Moreover, they represent our past, present, and future, and are filled with personal and communal meanings. Attachment is a psychological process that creates a sense of familiarity, stability, security, and belonging (Fullilove, 1996). Disruption to this process and the intimate bond with place can lead to feelings of grief, loss, and despair (Brown & Perkins, 1992; Fried, 1963; Fullilove, 1996). When natural disasters destroy or damage a community, individuals may experience clinical levels of posttraumatic stress, depression, or anxiety following experiences of displacement (e.g., evacuation, destruction of home) and loss of loved ones. In the present study, I explored the relationship between the initial impact of a natural disaster and attachment to place, 3 1/2 years after the disaster occurred. Two rural communities in British Columbia, Canada (Louis Creek and Barriere) experienced an interface wildfire in August 2003, and were the focus of the present study. The communities lost many homes and businesses, including a pulp and paper mill, the area's major economic resource. Some of the communities' members were forced to relocate in search of employment, yet many remained to rebuild their lives. Because of the disruption to the social fabric of the communities and the transformation of the surrounding natural environment, in the present study, I explored both the social and physical dimensions of place attachment. The physical dimension encompasses both the natural and built environment; the natural reflects the natural features of the surrounding landscape and the built reflects material possessions and man-made structures including houses and buildings 2 within the community. Specifically, I examined the relationships between the initial impact of a wildfire on residents from two rural communities, their current degree of attachment to their community, and their current depressive symptoms, 3.5 years following the event. Natural Disasters Natural disasters (i.e., disasters caused by natural forces) are defined as such not because of the ecological transformation they cause to the biophysical environment, but because of the disruption they cause to human lives. Drawing upon various sources such as the World Health Organization, Lopez-Ibor (2005) highlighted several components and impacts of a disaster. Events from the physical environment.. .breakdown of everyday functioning takes place.. .continuity of structures becomes problematic... severe psychological and psychosocial disruption...widespread human, material, or environmental losses.. .exceeds the ability of the affected community to cope... [feelings] of helplessness and threat.. .disruptive effect on social systems.. .tensions between individuals.. .deterioration of the links that unite the population and that generate the sense of belonging to the community... affects the foundations of the world everyone builds for his/her own and where he/she lives.. .loss of the sense of invulnerability becomes obvious.. .the vision of the world, of oneself, of the future, changes, (pp. 1-8) Thus, a natural disaster can be defined as a complex event that involves an interaction among social, physical, and natural environments, and has significant economic, cultural, social, and psychological consequences for individuals and communities. It represents how individuals continually exist in relationship to their surroundings and the effects on individuals when that relationship is disturbed. Lopez-Ibor's (2005) cogent description conveys residents' experiences during the aftermath and the recovery period of a natural disaster (Fullerton & Ursano, 2005). Events that take place during and immediately following a natural disaster -evacuation, relocation, loss of homes, jobs, possessions, and often family and friends -3 contribute to the disruption of residents' psychological processes of attachment (Brown & Perkins, 1992; Fullilove, 1996). Emotional distress and feelings of security, stability, familiarity, and belonging are evoked (Brown & Perkins, 1992; Fullilove, 1996). Thus, a natural disaster and the impending outcomes are appropriate contexts to study place attachment, given that during a natural disaster complex interactions occur between individuals and their environment. Interface Wildfires The most common events portrayed by the media and represented in disaster research are earthquakes, hurricanes, and floods (Norris et al., 2002b). However, as people continue to live close to natural forested areas, they are increasingly at risk of wildfires. As cities grow towards the forest's edge, many families choose to establish homes close to the natural environment. Furthermore, many individuals and families prefer to remain in or move to a rural setting for lifestyle or economic reasons. Thus, a wildfire is a unique form of natural disaster1 with respect to studying the person-environment relationship such as place attachment. The intermingling of human and natural factors changes how fires function in their natural ecosystem. Development, forest, and land management practices, and the complexity of natural factors contribute to the accumulation of hazardous levels of vegetation and fuels, which can result in a wildland-urban interface wildfire (Natural Resources Canada, 2004). Similar to other natural disasters, interface wildfires can create significant changes to the economic, social, and cultural fabric of communities, many of which rely on forests for their 1 A wildfire is a natural disturbance within forest ecosystems that occurs several times throughout the hot weather season without impacting humans. When a wildfire affects humans, it becomes a hazard and ultimately considered a natural disaster. Thus, within this context, a wildfire is considered a natural disaster. 4 livelihood, recreation, and quality of life. Thus, the nature of the relationships between communities and forests are varied and complex (Beckley, 1998). McLure Fire 2003 During the summer of 2003, interface wildfires were at an all-time record high in British Columbia, Canada (Filmon, 2003). The Thompson-Okanagan Valley was the hardest hit area in the province; 50,000 people were evacuated from their homes and 70,000 hectares of land including rangeland and 20,000 hectares of forest were destroyed. Careless human error, strong winds, extreme heat, and lightning sparked the McLure Fire, one of the most significant interface wildfire events in the valley that summer, as well as in BC history (Filmon, 2003). The McLure Fire began July 30 and caused the evacuation of 3800 people (880 of whom were evacuated a second time) and the destruction of 26,420 hectares of land (Filmon, 2003). The fire caused extensive disruption to two rural communities in the area, Louis Creek and Barriere, which are the focus of the present study. In total, 73 homes, 9 businesses, and an industrial park were lost including the communities' major economic resource, the Tolko Fadear Sawmill. The decision not to rebuild the mill immediately created a change in the community structure as many whose livelihoods relied on the mill had to seek employment elsewhere or find new forms of employment. The total cost of damages in Louis Creek and Barriere was estimated at 31.1 million dollars (Ministry of Forests, 2003). Homes and businesses have since been rebuilt, and the natural environment is slowly regenerating. Yet, the community is not the same and the surrounding hills will not recover for many years (see Figure 1). Figure 1. Rebuilt Louis Creek home and surrounding hills 3.5 years after the McLure Fire. The two communities border each other, and are geographically distinct, yet the social and economic boundaries are fluid (R. Cox, 2006). Census information is unavailable for each community; however, the most recent general census (2001) indicated that the area, the Thompson-Nicola Regional District, has a population of 3,257 (Statistics Canada, 2001). Demographic information since the fires is unknown. Barriere is 81 kms north of the major centre of Kamloops; Louis Creek, the smaller of the two communities, is located 6 kms south, adjacent to Barriere. The communities are located at the southern end of the North Thompson River Valley, bordered by mountains, and situated at the confluence of the Barriere and North Thompson rivers (see Appendix A ) . The primary industries are forestry, 6 agriculture (beef cattle ranching), and tourism. During the first and second years following the 2003 McLure Fire, R. Cox (2006) conducted a critical ethnographic study in Louis Creek and Barriere, exploring the impact of the wildfire on the residents' recovery process. The study revealed the inherent complexities of the recovery process following a disaster. Her findings revealed differences in how community residents responded based on the degree the wildfire affected them (e.g., evacuation, property loss). Individual and communal feelings of grief and loss were present up to 2 years following the fires. One of the many themes that emerged was how the wildfire disturbed the residents' relationship (i.e., attachment) with their community, and its meaning to the recovery process. Deeply embedded ties to the natural and built environments became conscious, as did a shift in the social fabric within and between the communities. R. Cox's (2006) research provided an appropriate framework for the present study because her findings illuminated individual differences in the way in which residents were impacted by the fires and how the fires disturbed and/or deepened their emotional ties to the social, natural, and built environments in the community. Purpose of the Study A natural disaster significantly affects the natural and built physical environments and creates enormous stress on residents and the entire social fabric of a community. An interface wildfire is a particularly unique form of disaster as it exposes the relationship between the natural environment and community residents. Many residents from Louis Creek and Barriere experienced immense disruption to their lives during and following the McLure Fire of 2003 (R. Cox, 2006). This disruption took both material (i.e., tangible), as well as symbolic forms. Tangible losses included possessions, houses, livelihood, land, animals, and 7 social relationships. However, disruptions awakened residents to their taken-for-granted relationship to their sociophysical environment. The loss of the Tolko sawmill or loss of surrounding trees, as examples, represented a disruption to the residents' relationship to the past, present, and future. Moreover, the disruption in daily routine represented a loss of unconscious continuity, stability, and comfort. A world that residents moved through on a temporal and spatial level with security, ease, and often a sense of belonging became a new and unfamiliar experience. Post-disaster research has primarily examined and evaluated psychological distress immediately following a disaster, however, there exists a paucity of research on its long-term impact (Norris et al., 2002b). Evidence from R. Cox's (2006) ethnographic study revealed that residents continued to feel distressed and their relationship with place was disturbed 2 years following the event. Thus, a natural disaster, such as the McLure Fire of 2003, is a fitting context to examine residents' attachment to their intimate environment, 3.5 years following the event. Stressors such as evacuation, displacement, and various losses have been shown to negatively impact an individual's mental health. However, researchers have overlooked the association between the nature of these experiences and an individual's relationship with both their social and physical environment (i.e., attachment to their community). This potentially precludes consideration of the initial impact of a natural disaster and its long-term repercussion on individuals' mental health as well as their attachment to place. Thus, the aim of the present study was to examine the relationship between the initial impact of the wildfire on Louis Creek and Barriere residents, and their current attachment to both the social and physical dimensions of their community, and their current depressive symptoms, 3.5 years 8 after the event. A better understanding of these relationships may add to disaster mental health and recovery research, and the development of the construct of place attachment. This relationship was explored through the dissemination of a paper-and-pencil survey in these communities, 3.5 years following the wildfire. Because of the paucity of research in this area, and limited measures of attachment or impact, the present study is exploratory. The following research questions rather than hypotheses were posed. Is there a linear relationship between: (a) the initial impact of the McLure Fire on residents and their current strength of attachment to their community, (b) the initial impact of the McLure Fire on residents and their present depressive symptoms, (c) the residents' current attachment to their community and their present depressive symptoms? Differences between the two communities were explored and the effects of relevant demographic characteristics were accounted for in data analysis. 9 CHAPTER 2: LITERATURE REVIEW The recovery process following a natural disaster, such as a wildfire, is an appropriate context to study person-place relationships, more specifically place attachment. One of the many themes that arose from an ethnographic study conducted in two rural communities following the 2003 McLure Fire was how the fire disturbed, and in some cases strengthened, residents' relationship to their community (R. Cox, 2006). R. Cox's findings also identified the nature of residents' attachment to their natural and built environments, as well as a shift in the social system within and between the two communities during the first 2 years of the recovery process. After reviewing the literature on person-place bonds within the context of natural disasters, I determined that there were two interrelated dimensions of place attachment that warranted detailed exploration—the social and physical (attachment to the natural and built environment). Furthermore, feelings of loss and distress were apparent in the two communities up to 2 years following the fire (R. Cox, 2006); yet, to date, the majority of disaster research has explored only the immediate impact of disasters on resident's psychological health. Research on the relationship between place attachment and mental health is dearth. In the present study, I explored the long-term impact of a wildfire on resident's current relationship to their social and physical community, and whether the impact of the wildfire related to residents' present mental health (i.e., symptoms of depression). In this chapter, I present a conceptualization of place attachment drawn from various sources, and review research that explores place attachment as an important and symbolic construct within the context of a natural disaster. I then examine the various dimensions of ( ' 10 place attachment (i.e., social, physical), and how they have been operationalized. Finally, I review depression within the context of natural disasters and place attachment. Place Attachment There is a proliferate amount of research and literature written about place attachment and person-place relationships in general. The complexity and multidimensional nature of this bond has created widespread inconsistency in how this relationship has been both conceptualized and operationalized. Many researchers have concurred that choosing a theoretical framework to guide their research is a challenge because a consensus on a definition and appropriate measures of attachment are lacking (Brehm, Eisenhauer, & Krannich, 2004; Cross, 2004; Lalli, 1992; Pretty, Chipuer, & Bramston, 2003). Depending on the discipline of study, dimensions of place attachment vary. However, there exists an essence to this person-place relationship that cuts across all disciplines. The essence of the concept of attachment draws on Heidegger's concept of Dasein, or being-in-the-world, which denotes that people do not exist apart from the world, but is in an intimate and inescapable relationship (Casey, 1993; Seamon, 2000). From a psychological perspective, place attachment is studied within the context of attachment theory in interpersonal relationships (Chawla, 1992; Fried, 2000; Giuliani, 2003). This theory suggests that intimate emotional bonds that develop between a child and another human being lies at the nucleus of a child's world and provides a sense of familiarity, protection, and security; and fosters creative expression, social affiliation, and environmental exploration. Place also provides an individual with these very same features (Fullilove, 1996). The primary attachment figure, person, or place, can then act as a source of stability, which is deemed vital to healthy development. Place can provide a solid foundation that 11 allows for continued maturation, differentiation, and development of self and identity in the world (Chawla, 1992). Similarly, Relph (1976), a human geographer and phenomenologist, stated: To have roots in a place is to have a secure point from which to look out on the world, a firm grasp of one's own position in the order of things, and a significant spiritual and psychological attachment to somewhere in particular, (p. 38) Environmental psychologists would likely agree with Relph, who also suggested that, "to be attached to places and have profound ties with them is an important human need" (p.38). There is evidence that a strong relationship exists between attachment to place and well-being (Rowles, 1990). Though Rowles' research focused on elderly populations (>65 years), he offered insight into aspects of place attachment in rural communities. He described three elements pertaining to a sense of insideness involving intimate involvement within one's immediate community, which is clearly different from their interaction with the outside world. Physical insideness represents the familiarity of one's surroundings that develops with long-term residence. Social insideness represents a sense of belonging, knowing and being known, and social integration within the community. Autobiographical insideness represents layers of meaning developed over time, from past and present affiliation to the community. It is intertwined with a sense of self and personal identity. Brown and Perkins (1992) explored disruptions to resident's attachment to their place, within the context of an event such as a natural disaster. They provided the following comprehensive definition: Place attachment involves positively experienced bonds, sometimes occurring without awareness, that are developed over time from the behavioral, affective, and cognitive ties between individuals and/or groups and their sociophysical environment. These bonds provide a framework for both individual and communal aspects of identity and have both stabilizing and dynamic features (p. 284). 12 This definition reflects the complexity and multidimensional nature of the place attachment • construct. Similar to Rowles' (1990) theory, it implies that place attachment exists on a continuum, has a temporal and dynamic quality, considers how the cognitive and behavioral contribute to the affective ties of an individual's relationship to their social and physical environment, and plays a role in the development of individual and community identity. This definition also reflects the sociocultural origins of place attachment. Low (1992) defined attachment as "a symbolic relationship with one's place that is developed through culturally shared meanings to a particular space" (p. 165). This relationship becomes pertinent and conscious among residents in a rural community that has been affected by a natural disaster (R. Cox, 2006). To summarize the conceptual framework guiding the present study, I drew on Brown and Perkins' (1992) definition of place attachment. I identified both the social and physical (including natural and built) dimensions of place attachment, and took into account residents' affective, cognitive, and behavioral ties to both their social and physical environment, processes categorized as modalities of place experience (Weisman, Chaudhury, & Moore, 2000) (see Figure 2). I also included a general dimension of place that represents attachment . to the community, in general. 13 Construct Place Attachment Dimensions Natural Built Modalities c Behavior Figure 2. Conceptual framework ofplace attachment. Two interrelated constructs that are often confused are place identity and sense of place, yet they aid in defining each other and the concept of place attachment (West, 2003). Place identity represents how place is a symbolic extension of one's self, and sense of place represents familiarity through experience and knowledge of a place. Individuals cannot form an emotional bond to a place without developing both knowledge and familiarity of the place and an identity with the place (West, 2003). Similarly, an individual identifies with their physical environment through a sensevof attachment and belongingness (Jorgensen & Stedman, 2001; Proshansky, 1983). Thus, an individual's attachment to their locale is strongly related to developing aspects of place identity (Twigger-Ross & Uzzell, 1996). 14 Fullilove (1996) placed these psychological processes (attachment, identity, and familiarity) within a psychology of place framework. These processes allow the individual to continually develop a relationship with their environment, to provide one with feelings of security and stability, and contribute to a sense of belonging and well-being (Fullilove, 1996). Place dependence, another related concept, is a functional form of attachment as it reflects the degree to which a place meets the needs and goals of the individual (Shumaker & Taylor, 1980; Williams & Vaske, 2003). Place refers to a space that has been given meaning and significance through personal, group, or cultural processes (Low & Altman, 1992). Attachment to a place can incorporate many levels and contexts. An individual or group can have strong relationships to their intimate possessions (Belk, 1992), house of residence (Giuliani, 1991; Hidalgo & Hernandez, 2001), neighborhood (Twigger-Ross & Uzzell, 1996), public place (Low, 1992), recreational setting (Eisenhauer, 2000; Jorgensen & Stedman 2001), community (Hummon, 1992), city (Hidalgo & Hernandez, 2001), region (Cuba & Hummon, 1993), or nation (Vorkinn & Riese, 2001). A n individual can develop a positive bond to one or all of these places simultaneously, or have a negative bond with their residence and positive feelings towards their community. Places can be known and directly experienced or symbolic and not tangibly experienced, and often do not have easily specified boundaries (Brown & Perkins, 1992). Phenomenologists speak of places as home, which captures the essence of experience and meaning in place (Manzo, 2003). For the purpose of the present study, place refers to the known and experienced geographical community. Many studies have highlighted how such locales continue to provide a significant locus of sentiment and meaning for the self (Cuba & Hummon, 1993; Feldman, 1990; Goudy, 1990; Hummon, 1992), and acknowledged that 15 communities carry psychological significance for its residents (Christakopoulou, Dawson, & Gari, 2001). Summary Place attachment is a complex and multidimensional construct making it difficult to conceptualize and operationalize. At its simplest, place attachment reflects an emotional bond to an intimate place such as the community within which one lives. This community reflects general, social, and physical dimensions of attachment, taking into consideration the modalities of affect, cognition, and behavior. Within the context of attachment theory, a community can provide a source of stability, security, familiarity, and belongingness for an individual. Researchers have posited that an individual's relationship to place (i.e., community) is symbolic, exists on a continuum, is dynamic, develops over time, contributes to the development of identity, and reflects an extension of one's self. Similar processes such as place identity, sense of place, and place dependence may contribute to peoples' attachment to their community. The relationship to place is often a taken-for-granted bond. Thus, a disruption to place, such as a disastrous wildfire, may disturb the psychological process of attachment to one's community and may bring this relationship into an individual's conscious awareness. Place attachment thus provides an appropriate conceptual framework to explore its relationship with the experience of disruption from a natural disaster. Natural Disaster and Implications on Attachment to Place Although place attachments may develop slowly through time and space, it can be instantaneously disrupted (Brown & Perkins, 1992). The disruption to a resident's community through natural disasters can create a process of emotional suffering from severed attachment, destabilization in new environments, and becoming unsure of one's self 16 and what to expect from others (West, 2003). In such circumstances, individuals may reinvent themselves, their surroundings, and the material and symbolic relationships that shape their old and new identities. Ironically, disruptions to place attachment give rise to a conscious awareness of how fundamental it is to the experience and meaning of everyday life (Brown & Perkins, 1992). Although accounts of disaster and recovery have been shared through research and literature for decades (Erikson, 1976), the conceptualization of person-place relationships, more specifically, the implications of a natural disaster on residents' attachment to place, surprisingly, are often overlooked in natural disaster recovery research. To date, I located only one research study (Schwarz, Brent, Phillips, & Danley, 1995) where the focus was on the experience of a natural disaster within the context of a place attachment framework. However, I found four other studies, drawn from larger ethnographic research projects, which examined the impact of natural disasters on communities and later expanded on the theme of place and the impact on an individual's psychological processes of place (Brown & Perkins, 1992; R. Cox, 2006; H. Cox & Holmes, 2000; Skelton, 2004). Through case studies and open-ended exploratory interviews within one year of a disaster (i.e., flooding), Schwarz et al. (1995) examined why individuals and their families from 20 households in four small communities in Missouri, chose to remain in or move away from their small community despite continual threats of floods. They examined preexisting conditions that influenced place attachments and the conditions following the floods that affected the individual's capacity to cope with their losses and reconstruct new attachments to people and places (Schwarz et al., 1995). Their findings highlighted that place attachment and the meaning of home was dominantly related to the natural environment surrounding their home, length of residence in 17 the community, childhood memories, deep family roots, and social interaction with family and friends in the community. There was also a strong sense of safety, security, and familiarity with the community in general. The disruption contributed to the erosion of social ties for some, yet strengthened social ties for others. The findings also illuminated how individual perceptions of attachment differed; some considered their house as home, whereas others were more strongly attached to their community, the land, and the outdoor recreation opportunities as home. This highlights the multidimensional nature of place attachment and provides further evidence for the present study to explore the general, social, and physical (i.e., natural and built environment) dimensions. Schwarz et al. (1995) concluded that residents unconsciously conducted a cost/benefit analysis, similar to the construct of place dependence. For example, some felt that the continuing threat of floods outweighed a strong attachment with the community and the land. However, the findings were based on interviews from two households, which limits confidence in their findings. In addition, the authors defined termination in one's attachment to place as not returning to the community and did not consider that emotional ties to a place may be present on a continuum (Kaltenborn, 1998; Shamai, 1991). In the present study, I consider that place attachment is present on a continuum; that is, it is still possible to have strong attachments to a place, despite forced displacement or voluntary migration. This is shown in the following study of a Caribbean island community forced to relocate due to a volcanic disruption. Skelton (2004) explored the concepts of belonging and sense of place from a cultural geography framework and drew on her larger ethnographic study of Montserrat's political, social, cultural, and environmental crisis. She explored residents' culturally constructed 18 narratives of Harris, their village, as home. Home to several generations of families, Harris was evacuated and partly destroyed by volcanic pyroclastic flows. Similar to the community in Schwarz et al.'s (1995) study, the volcano continued to be a threat and challenged the residents' identity. Prior to the disruption, residents had a very strong sense of attachment to place, both at the scale of the village and the island. This attachment developed through the appreciation of the landscape's beauty, the social friendliness of the community, and the links to the past through the farmland. Most of the participants were forced to relocate following the disruption, however, Schwarz et al. (1995) found that the stories of home and the residents' attachment to their place was a way of coping with the intense loss and devastation, and enabled them to sustain their identity. This highlights the importance of attachment to place and its sustaining qualities. Of note, many place theorists have proposed that place attachment is not a stable construct, but a dynamic one that changes through time and space (Brown & Perkins, 2002). Therefore, following the disaster, the Harris residents may have re-created attachment towards their community in a new way. H. Cox and Holmes (2000) drew on a conceptual framework of sense of place to illuminate how the disruption of residents' dwelling places from an Australian bushfire influenced their social, cultural, and personal identities, and how the person-environment relationship contributed to healing. Similar to Schwarz et al. (1995), H. Cox and Holmes were interested in why individuals chose to stay in a place that held traumatic memories and remained vulnerable to further disasters. Another study, based on H. Cox's 1996 doctoral dissertation, explored residents' experiences of a bushfire in their rural Australian community where more than 700 homes were destroyed and 3 residents lost their lives. The findings 19 highlighted the strength of the relationship the residents had with their natural environment and how this relationship was instrumental in their healing. The findings also implied that residents with varying degrees of loss experienced place similarly. That is, the transformation of the natural environment was intrinsic to everyone's identity. Thus, all residents experienced a sense of uprootedness, despite differences in the nature of loss. This implies that the bushfire created a new sense of community identity, a finding similar to other disaster recovery studies (R. Cox, 2006). Furthermore, this implies a shift in the social fabric of the community. Therefore, H. Cox and Holmes' (2000) research study also illuminates the multidimensional nature of place attachment and provides a basis to allow further exploration of the general, social and physical (i.e., natural environment) dimensions of place attachment. In the present study of the McLure Fire, I explored the relationship between the degree and nature of the impact from the fires and residents' attachment to their general, social, and natural community. H. Cox and Holmes (2000) found that a strong attachment to their place prior to the bushfire, in turn, provided strength to the community and a deeper sense of belonging, which was instrumental to their recovery. The regeneration of the environment helped residents re-establish their sense of place. As H. Cox and Holmes (2000, p.74) suggested, "The most significant factor to which people attributed their recovery was the return of the environment." Of note, this community differed from communities in the previous studies (Schwarz et al, 1995; Skelton, 2004) because the residents had higher socioeconomic status; and length of residence and family roots did not play a major factor in their attachment to their place. They were a community with a high capacity and ease of mobility, who were intentionally drawn away from the urban life, and sought a peaceful rural and natural setting. 20 Similar to the previous studies (H. Cox & Holmes, 2000; Schwarz et al., 1995; Skelton, 2004), R. Cox's (2006) study of the impact of the McLure Fire on the communities of Louis Creek and Barriere revealed that the residents had a deeply embedded relationship with their community, including the natural environment and their social network, and the importance of place in the recovery process. Attachment to the natural environment was revealed immediately following the wildfires through their feelings of distress related to the loss of the surrounding landscape, and later during their recovery when they spent time "re-greening" and nurturing the environment. With respect to the social community, there were differences in the process of how the Louis Creek and Barriere communities were affected based on the diversity and range of losses and experiences. For example, the homogeneity of losses in Louis Creek (e.g., the majority of residents lost their homes; the community lost their mill) ironically created a bond among the residents and contributed to a heightened sense of community identity. The Barriere community, where material and emotional markers of loss were less evident, tended to withdraw and retreat away from these markers (e.g., burned hillsides, destruction of homes, signs of distress). Furthermore, when material losses were significantly different from those of the immediate neighbors, a disruption to the social network within this community was experienced. The present investigation drew on these markers to assess impact of the wildfires and to highlight the diversity and range of impact among the residents of the communities. I also explored potential differences in the relationship between impact and attachment, impact and depression, and attachment and depression for the two communities: Louis Creek and Barriere. R. Cox's (2006) findings revealed that residents differed in how they responded to the disruption based on the degree the wildfire affected them, in addition to 21 community differences. For example, those who experienced evacuation, without any material loss, were more eager to return to a sense of stability and security (3 months to a year after the fire). Others still experienced feelings of displacement and alienation 2 years later, whereas one resident expressed feeling a sense of community for the first time. In the proposed study, impact of the wildfire was measured in terms of material loss, rather than symbolic, because, based on their complexity and emotional nature, symbolic losses (i.e., loss of routine or control) are more difficult to measure retrospectively and might be confounded with current levels of well-being. According to R. Cox (2006), material losses varied among residents and included the following: total loss of and/or damage to possessions, homes, businesses, properties (including ranches), and land (including gardens and greenery); change in the surrounding environment/landscape (including loss of flora and fauna). R. Cox also noted that evacuation and the presence or absence of insurance also significantly contributed to the degree a resident was impacted by the fires. Moreover, it was found that the greater the extent of material losses, the greater the sense of disorientation towards their familiar environment immediately following the fires and in the early stages of recovery (R. Cox, 2006). Two years into the process of recovery, residents, with varying degrees of disruption, described feelings of distress and displacement, and still "lived with an ongoing sense of threat, fear, and uncertainty" (R. Cox, p. 92). A shift occurred within "the material and symbolic markers of their relationship to themselves, others, their immediate context, and the larger world" (p. 128). The recovery process involved re-creating a sense of identity and belonging towards one's self, the community, and the natural environment. R. Cox's (2006) research provided an appropriate framework for the present study as her findings illuminated 22 individual and community differences in the way in which residents were impacted by the fires, and the process of recovery to their sociophysical environment. In addition, the residents' relationship to place suggested that the fires disturbed and/or deepened the emotional ties to the social, natural, and built environments of their community. Summary Similar themes are threaded through these five articles, which highlight the symbolic and material loss of the general, social, and physical dimensions of one's environment following a natural disaster. Moreover, these studies also revealed that residents lived with unfamiliarity, fear, instability, and a discontinuity between their past and present lives, and that individuals and communities attempted to re-create new attachments to their place (i.e., community) following a disruption such as a natural disaster. The authors also described how individuals' emotional connection with their place, particularly the natural environment, provided them with resiliency, strength, and the capacity to cope with the disruption and losses. Although these researchers interviewed residents up to 2 years following the disaster, long-term changes to place attachment following a natural disaster have not been documented. Although these qualitative studies provided an in-depth perspective of the residents' experience of attachment to place or community, they were limited in terms of providing specific associations between demographic information, and the physical and social impacts of the disaster. Moreover, these qualitative studies relied on interviews with only a few of the community residents, making it difficult to gain an understanding or perspective of the range of experiences across a whole community. 23 The previous studies explored the person-place bond immediately after, or up to 2 years following the disaster. Because it takes time to re-navigate old emotional ties and . possibly re-create new connections, examining a community 3 to 4 years post-disruption is unique to both the disaster recovery and person-place relationship literature. Attachment to the social community and the natural and built environments became a more conscious experience following the disaster in each of the studies reviewed. In the present study, I examined (a) the relationship between the initial impact of the wildfires (e.g., loss of home, employment) and their current attachment to the community (general, social, and physical dimensions), (b) the relationship between the initial impact and depressive symptoms, and (c) the relationship between residents' degree of attachment (general, social, and physical dimensions) to their community 3.5 years following the fires, and their current depressive symptoms, among Louis Creek and Barriere residents. Operationalizing Social and Physical Dimensions of Place Attachment Based on findings from the natural disaster and community and place attachment literature, in the present study, I focused on both social and physical dimensions of attachment. Because the wildfires significantly transformed the natural environment surrounding the communities of Louis Creek and Barriere, the emphasis on the physical environment I considered in the present study was the natural or biophysical environment. The ecological transformation of the environment (e.g., burned hillsides) will be visible for many years, thus the natural environment cannot be ignored. The built environment was also considered as a dimension of place attachment because of the material loss of homes, property, and possessions. Because the context of this investigation took place in a rural setting, the social and physical dimensions of the residents' environment were notably 24 interrelated. Thus, the geographic community was defined as incorporating the community in general, the social environment, the surrounding natural environment, and the built environment (see Figure 2). The most difficult task in designing this study was determining how to operationalize place attachment and its social and physical dimensions. Rural and urban sociologists have most commonly examined attachment to the community as a locale and have found that attachment is most strongly associated with social integration in the community. In contrast, most place attachment researchers define the physical as human made structures, such as houses, buildings, or institutions. Few researchers have attempted to extend the sociophysical environment to incorporate the natural environment (Beckley, 2003; Brehm et al., 2004; Cross, 2004; Eisenhauer, 2000; Henwood & Pidgeon, 2001). Some researchers have advocated for the integration of both the community and place attachment literature based on an expressed need to reinforce the interrelatedness of the social and physical dimensions of attachment (Beckley, 2003; Cross, 2004). Other researchers have emphasized the distinctiveness of the two dimensions, yet advocated for incorporating both dimensions into the conceptual framework of community attachment (Brehm, Eisenhauer, & Krannich, 2004) and place attachment (Hidalgo & Hernandez, 2001). Following Hidalgo and Hernandez's (2001) model, I included the general, social, and physical dimensions of place attachment (see Figure 2). In order to determine whether the general, social, and physical aspects of resident's attachment to their community were present in the population of interest, I conducted a factor analysis of the measure of place attachment used in the present study. ' 25 Based on the literature review of natural disasters and place attachment, researchers (Brown & Perkins, 1992; Schwarz et al., 1995; Skelton, 2004) explored pre-disruption attachments to the residents' communities. Qualitative accounts exposed strong attachments to place prior to a disruption, and the possibility of recreating often stronger attachments post-disruption. Scales measuring change in attachment are nonexistent; therefore, I developed several items that reflected a change in a residents' attachment to their community, from pre-disaster to 3.5 years following (e.g., number of visits with neighbors, participation in community projects, time spent outdoors, and increased interest in community compared to before the fires). Because of recall bias, change is notoriously difficult to assess retrospectively.(Aseltine, Carlson, Fowler, & Barry, 1995; Nieuwkerk, Tollenaar, Oort, & Sprangers, 2007); therefore, the construct - change in attachment - was used for descriptive purposes in the present study. I explored possible changes to resident's attachment to their community 3.5 years following the experience of a wildfire, and whether the direction of the change was consistent with current degree of attachment as an indicator of the validity of the attachment measure developed for the present study. Social Dimension The community attachment literature provided insight into affective, cognitive, and behavioral ties to one's social network in their community. A substantial portion of this research has been conducted within rural communities, primarily from a sociological framework (Brehm et al., 2004; Cross, 2003; Goudy, 1990; Rice & Steele, 2001; Theodori, 2001; Theodori & Luloff, 2000). Although the focus of this research has been to determine which variables predict the extent to which residents have emotional ties to their community, the development of a clear conceptualization of community attachment has lagged behind. 26 Empirical evidence has shown that long-term residence and the presence of social ties (i.e., local social involvement with friends, as well as family, organizational membership, and community improvement activities) consistently have been the strongest predictors of emotional ties to a locale (Brehm et al., 2004; Goudy, 1990; Kasarda & Janowitz, 1974; Riger & Lavrakas, 1981; Sampson, 1988; Theodori & Luloff, 2000). Thus, the longer one lives in a community, the more opportunity one has to develop significant memories and social ties. This finding reflects the temporal aspect of Brown and Perkins' (1992) definition of place attachment. Similarly, Relph (1976) asserted that attachment to a place grows through time and is primarily based on the social relationships in the setting. The same researchers revealed that life-cycle stage (i.e., age or presence of children) also played a role, although the results were less consistent. Factors such as community size, density, type, and social class were shown to be weak predictors of community attachment (Goudy, 1990; Kasarda & Janowitz, 1974). Community attachment has been most commonly measured as a dependent variable using three separate items—sense of community, interest in community, sorry to leave (Goudy, 1990; Kasarda & Janowitz, 1974; Theodori & Luloff, 2000). Other studies have incorporated one or two of these items as a community attachment measure (Sampson, 1988; Theodori, 2001) or summed all three for an index (Rice & Steele, 2001). These studies reported important findings with respect to understanding the social network system of a community and social attachment; however, their findings were limited. First, the measurement of community attachment incorporated only one to three items and was found to be inconsistent across the studies. The degree of variability in the responses, which was found to be significant, was only reported in Theodori and Luloff s (2000) study. Second, the 27 social construction of the community as a geographical place was not considered; and, third, the conceptualization of community attachment was not made .explicit. Based on my review of the literature, I have concluded that the attachment construct is more complex than what is presented in the community attachment literature. Although drawing upon a sense of community conceptual framework, Riger and Lavrakas (1981) identified two interrelated dimensions of attachment: physical or rootedness, and social or,bondedness. Variables such as length of residence, home ownership, and expectations to remain were related to the physical or rootedness dimension, whereas the social or bondedness dimension was related to feeling part of one's neighborhood, ability to recognize strangers, and knowing the neighborhood children. Attachment to place contributes to residents' sense of belonging and reflects the sense of who they are in the world (Proshansky, 1983; Twigger-Ross & Uzzell, 1996). Others have developed models of place attachment that incorporate place identity (Cross, 2003; Jorgensen & Stedman, 2001; Williams & Vaske, 2003). Thus, knowing whether residents feel a sense of belonging to their community and whether their community reflects who'they are would provide us with additional clues as to one's degree of attachment to one's community. Although closely related to psychological sense of community, the social cohesion construct incorporates similar aspects of community attachment (i.e., affective, cognitive, and behavioral ties) to one's social network within an individual's community. In fact, Robinson and Wilkinson (1995) used the terms cohesion and attachment interchangeably in their validation study of the Neighborhood Cohesion Index (NCI). The subscale attraction-to-neighborhood reflected commitment to remaining in the neighborhood; the neighboring subscale reflected acts of visiting and making daily connections with neighbors; and the 28 sense of community subscale reflected cognitive, as well as affective, ties to the community. Thus, many of the items from the NCI (e.g., The friendships and associations I have with other people in my neighbourhood mean a lot to me) (Buckner, 1988) reflect the bond residents have to their neighborhoods or community, and were drawn upon for use in the present study. Summary. The community attachment literature has identified important components of both the general and social dimensions of attachment to one's community. With respect to the general attachment dimension, two of the three attachment modalities: affect and cognition [e.g., sorry to leave (affect), sense of community (cognitive), and interest in community (cognitive)] have been frequently used to assess general attachment. Based on the review of the literature, it became apparent that additional elements of attachment, such as an expectation to remain in a locale, rootedness, belongingness, and identity may also provide insight into the strength of attachment to one's community. With respect to social ties, participating in community improvement activities and greater interactions with family, close friends, and neighbors (i.e., visits in the home) reinforce the modality of behavior. Thus, in order to adequately operationalize the general and social attachment dimensions in the present study, I adapted items that reflect affect, cognition, and behavior to measure attachment to the general and social community. Neighborhood cohesion and the NCI (Buckner, 1988) share similar aspects of community attachment (i.e., they reflect the affective, cognitive, and behavioral ties to one's community). Items from these scales were drawn upon to operationalize the social dimension of the attachment construct in the present study (see also Appendix D). 29 Physical Dimension - Natural Environment Within the disciplines of recreation and natural resource management, research on visitors' experiences in wilderness settings have emphasized the importance of the role the natural environment plays in contributing to one's emotional ties to a specific locale (Bricker & Kerstetter, 2000; Eisenhauer, 2000; Hammitt, Backlund, & Bixler, 2006; Jorgensen & Stedman, 2001; Stedman, 2003; Williams, Patterson, Roggenbuck, & Watson, 1992). The focus, however, has been on visiting a specific outdoor setting, such as a lake or a hiking trail. Thus, place symbolizes a particular and special locale visited by an individual. Furthermore, place attachment has been generally conceptualized as commitment to return to the natural setting and/or environmental stewardship, and has been operationalized in terms of attitudes and/or level of environmental concern towards the management of a place (Manzo, 2003). Some place attachment researchers have advocated for incorporating the geological, biophysical, and landscape attributes of regional ecosystems because it may be an important contributor to residents' sense of attachment in rural communities (Beckley 2003; Brehm et al., 2004; Cross, 2004). For the purpose of natural resource management, a few studies have explored residents' attitudes and perceptions towards gradual environmental change and destruction in their rural communities (Kaltenborn, 1998; Marshall, Picou, & Bevc, 2005; Rogan, O'Connor, & Horwitz,-2005; Vorkinn & Riese, 2001). Yet, there is a paucity of studies that have addressed this, unique person-place relationship—between the individual and the intimate regional landscape (identified as part of and surrounding their community)~and residents' feelings of security, stability, familiarity, and belonging to their community. In the present study, I examined the emotional connection residents have to the landscape that is 30 part of and surrounds residents' community by developing specific items reflecting this connection as part of a measure of attachment. There is a plethora of literature on the restorative qualities of the natural environment, representing people's significant emotional relationships with nature (e.g., Kaplan, 1995; Kaplan & Kaplan, 1989; Korpela & Hartig, 1996). The field of ecopsychology offers perspectives on our psychological relationship to the natural world (Roszak, Gomes, & Kanner, 1995). A person's relationship to nature and its constructed meaning symbolizes an individual's sense of self and identity, and shows how experience in and of nature can be integral to one's self-concept (Henwood & Pidgeon, 2001; Manzo, 2003). The place and disaster literature shares similar sentiments. Mayer and Frantz's (2004) conceptualization of feeling connected to nature and the way they operationalized it was found to be the closest fit to the modalities of affect and cognition of the physical attachment dimension in the present study. . Mayer and Frantz (2004) developed the Connectedness to Nature Scale (CNS) to predict environmentally responsible behavior and was constructed to measure one's affective, experiential connection to nature. Their scale was based on previous theories that this connection incorporates dimensions of kinship, belonging, and a sense of embeddedness within the broader natural world. Despite the focus on natural resource management and recreation, Williams and Vaske's (2003) multidimensional conceptualization of place attachment also provided additional insight into the three modalities (affect, cognition, and behavior) of attachment. Their proposed measure was specifically developed to encapsulate the extent individuals feel attached to, and identify with,.an intimate place. I drew upon both 31 measures of attachment to aid in the development of the natural environment dimension in the present study. Physical Dimension - Built Environment Because of the obvious and immediate loss of buildings, homes, and material personal belongings following the wildfire, attachment (i.e., emotional) to the built environment was included as a physical dimension of attachment in the present study. Home can provide a sense of belonging and identity to individuals because it is often the repository of memories and represents personal.relationships with family, friends, and others in the community. Material possessions such as the home, souvenirs, mementos, and other meaningful objects that make up an individuals daily routine are expressions of one's self and provides a source of security and protection (Belk, 1992). Belk also discusses collective or local possessions because they represent the past and present of a group, such as a community in which one resides. Thus, to represent the physical built attachment dimension, I developed items that reflect residents' current emotional ties to their homes, material possessions, and buildings within their community. Summary. The place attachment literature in the recreation and resource management field reinforced the importance of the one's emotional ties to a special wilderness setting. However, there was a dearth of empirical research that addressed an individual's relationship to the surrounding regional landscape within their community, and its impact on their sense of identity, security, stability, and belonging. The field of ecopsychology and the restorativeness of the natural environment further aids in explaining the psychological significance of this relationship. In the present study, I drew upon the conceptual framework of connectedness to nature and a multidimensional model of place attachment to address the 32 modalities—affect, cognition, and behaviour—of this dimension. I also developed items that addressed attachment to the built environment. Depression: An Outcome of Natural Disasters Natural Disasters and Depression There is evidence that natural disasters may lead to psychological problems such as posttraumatic stress disorder, anxiety, or depression (e.g., Norris et al., 2002b). The extent to which individuals experience psychological problems may be a function of their experience of the disaster, for example, the number and types of stressors, such as the severity or degree of exposure and the duration of individual and community disruption (e.g., injury, loss, relocation) (Fullerton & Ursano, 2005; Jones, Ribbe, Cunningham, & Weddle, 2003; Norris et al., 2002b). A comprehensive review of disaster studies conducted by Norris et al. (2002b) revealed that one of the most commonly observed psychological outcomes of a disaster is depression. Although 4% to 5% of the general Canadian population at any given time experiences major depressive symptoms (Public Health Agency of Canada, 2002), natural disasters can exacerbate depression as a consequence of the multiple psychosocial stressors associated with the experience (Norris, Friedman, & Watson, 2002a). Studies that have examined psychological outcomes of natural disasters are limited because researchers typically have collected data immediately following or within one year of the event (e.g., Jones et al., 2003; Tyler & Hoyt. 2000). Only a few studies have assessed psychological symptoms more than one year following the disaster. Although depression tends to decline over time following a disaster, depressive symptoms were elevated for victims of disasters for up to 44 months after the disaster, compared with people who did not 33 experience a disaster (McFarlane, Clayer, & Bookless, 1997; Norris, Perilla, Riad, Kaniasty, & Lavizzo, 1999; Ohta et al., 2003). An important limitation to most disaster studies is the lack of data available on pre-disaster psychological health. As a result, it is difficult to attribute symptoms to the disaster itself due to the. uncertainty of whether pre-existing pathologies persisted prior to the disaster (Seplaki, Goldman, Weinstein, & Lin, 1999). Furthermore, the context has typically been urban and the types of natural disaster have most often been earthquakes, hurricanes, and floods. With respect to demographics, women and parents with children have been found to show greater signs of stress following a disaster, whereas age, socioeconomic status, and marital status have not been consistent predictors of distress following a disaster. Several disaster studies have documented that the process of recovery occurring several years following a disaster have been more stressful than the disaster itself (R. Cox,'2006; Erikson, 1976; Fullerton & Ursano, 2005; Riad & Norris, 1996). In the present exploratory study, I focused on two rural communities that were impacted by a wildfire 3.5 years prior. I examined the relationship between depressive symptoms and the initial impact of the disaster on residents, and the relationship between depression and current levels of attachment towards their community. Furthermore, I examined whether these relationships differed depending on the community because of the differences in how the residents in each of the communities, Louis Creek and Barriere, were impacted. Attachment and Depression Environmental psychologists and place attachment theorists have posited that loss of an intimate meaningful place has an impact on individuals similar to separation from and loss • 34 of a person, which is known to increase depression (Fullilove, 1996). In a natural disaster, the stress of evacuation, relocation, loss of homes, possessions, jobs, family, or friends disrupt an individual's relationship to place, and may contribute to greater depressive symptoms. Although rarely empirically examined, theories suggest that a strong relationship with place increases personal well-being, such as self-esteem and feelings of belonging to one's j community (Fullilove, 1996; Korplea, 1989; Relph, 1976; Twigger-Ross & Uzzell, 1996; West, 2003). Conversely, a disruption to this relationship can be a source of emotional and mental distress. Moreover, there is a plethora of literature that discusses the psychological benefits of nature and how a connection to nature contributes to an individual's well-being (Kaplan & Kaplan, 1989; Kaplan, 1995; Mayer & Frantz, 2004). Fullilove (1996), an inner-city psychiatrist, discussed the implications of \ displacement such as urban renewal, homelessness, and refugees. She asserted that a sense of belonging is necessary for psychological well-being and that a disturbance to the person-place relationship can lead to psychological disorders. From a medical perspective, she stated that: The disorientation and confusion that accompany a massive alteration in a familiar place are experienced as bodily sensations, as well as emotional feelings. Familiar spatial routines are permanently etched on the nervous system and the musculature; the sudden loss of the exterior world that conditioned those emotions is perceived as a loss of the self. (p. 1521) According to attachment theories, when a disturbance in the attachment to a person occurs, there are often feelings of sadness, longing for the past, and difficulties with separation and commitment. This can lead to a long-term phase of re-creating attachments. Similar problems appear to follow the loss of an intimate place (Brown & Perkins, 1992; Fried, 1963; Fullilove, 1996). A n uncontrollable change or loss with the environment to 35 which an individual identifies can represent a loss of continuity, stimulating grief, threaten feelings of stability and security (Fried, 1963), lower self-esteem (Korpela, 1989), and decrease feelings of belonging to one's community (Relph, 1976). West (2003) posited that a person's relationship with place is a critical factor in managing life stresses. In his dissertation, he examined the relationship of place for Native Americans. From interviews with 16 adult Native Americans, he concluded that the loss of land is injurious to their way of life because they lose the cognitive map of their surroundings, become unsure of themselves, and what to expect from others. These feelings, which are often unconscious, can stimulate responses such as grief and depression (West, 2003). - ' The relationship between community attachment and depression in two American rural communities was examined by O'Brien, Hassinger, and Dershem (1994). They used stratified random and cluster sampling to survey 295 residents both from town and the surrounding open-country area. The communities were similar in population but one was considered more viable than the other. A modified version of the CES-D scale (Radloff, 1977) was used to identify symptoms of general distress and was used as a dependent variable in their analysis. Community attachment was measured by the sum of four items: The extent to which they felt the community an ideal place to live, were satisfied with the community as a place to live, have a lot in common with other people living in this community, feel like they fit in this community (Cronbach's alpha = .74); the higher the score the stronger the attachment. Regression analysis revealed that a strong sense of community attachment was associated with lower levels of depressive symptoms. Age was used as a . 3 6 control variable, and accounted for. a significant amount of variance in depressive symptoms. Therefore, I considered age as a control variable in the present study. The authors concluded that their findings highlight the relationship between . depression and rural residents' desire to preserve the local community. Similar to other studies on community attachment, the community attachment construct was measured by summing only four items, and it is unclear from which conceptual framework it was drawn. Moreover, sociocultural factors, such as marital status and social networks, which have been shown to play a role in the development of depression as well as attachment were not accounted for in their analysis. Summary Empirical evidence has shown that depressive symptoms can be present, many years later, in individuals who have experienced a natural disaster (McFarlane, Clayer, & Bookless, 1997; Norris, Perilla, Riad, Kaniasty, & Lavizzo, 1999; Ohta et al., 2003). As part of the conceptual framework of place attachment, researchers have hypothesized that there is a positive association between attachment and well-being, and that a disruption to the psychological process of attachment can create mental and emotional distress. The present study explored the association between depressive symptoms and individuals' current strength of attachment to their social and physical community. The relationship between the initial impact of the wildfire and the residents' current depressive symptoms was also explored, as well as community differences. Summary In 2003, a wildfire caused extensive damage to two rural communities in British Columbia, Louis Creek and Barriere. An ethnographic study conducted in these communities 37 (R. Cox, 2006) illuminated the disruption to residents' attachment to their community, more specifically the social, natural, and built environments, immediately following the disaster. I drew on these findings and previous research on place attachment and natural disasters, as well as various conceptualizations of attachment, to develop a multidimensional framework to guide this study. I delineated place attachment (i.e., attachment to community) as incorporating general, social, and physical (natural and built) dimensions, accounting for affect, cognition, and behavior (modalities of place experience). A n examination of a change in attachment was also extracted from the literature as an important construct to explore. Because very little empirical research has explored the emotional relationship individuals have to their community following the experience of a natural disaster, I examined the relationship between the initial impact of the wildfire (e.g., loss of home, employment) and residents' current attachment to the community (general, social, and physical dimensions). Furthermore, research has shown that symptoms of depression are elevated in a community, immediately following a disaster. Research on long-term consequences of a disaster on individual mental health is scarce. Thus, I examined the relationship between the initial impact of the wildfire and present symptoms of depression, 3.5 years following the disaster. In addition, qualitative accounts have shown that a disruption in residents' attachment to their community can create emotional distress; however, this relationship has rarely been empirically examined. Therefore, I examined the relationship between current attachment to the community and symptoms of depression, 3.5 years following the disaster. Because of the differences in how residents were impacted following the disaster and potential differences in how they processed their recovery, community differences were explored. Evidence shows that long-term residence is a strong predictor of attachment to the social community (e.g., 38 (Brehm et a l , 2004; Kasarda & Janowitz, 1974) and age has shown to contribute to the relationship between community attachment and depression (O'Brien, Has singer, & Dershem, 1994). Moreover, gender, age, and marital status have been shown to contribute to the impact of a disaster on depression and were considered, including length of residence, as control variables in the present study. 39 CHAPTER 3: METHOD Based on theory and previous empirical research, I developed a self-administered paper-and-pencil cross-sectional survey (see Appendix B) to explore the relationship between the initial impact of the McLure Wildfire (3.5 years ago), and residents' current degree of attachment to their community, as well as current symptoms of depression. I also examined the relationship between place attachment dimensions and depressive symptoms. This survey enabled me to sample a broad spectrum of the residents who were, to various degrees, affected by the wildfire. In this chapter, I describe the respondents by summarizing their demographic information and outline the sampling procedures, criteria for inclusion, and recruitment strategies. The development of the self-administered questionnaire and measures used are reviewed. Finally, the research questions are outlined and the strategies for data analysis delineated, including dealing with missing data. Participants and Procedure Participants were 104 residents from the rural communities of Louis Creek and Barriere, in British Columbia, Canada, who were recruited through flyers, posted notices, newspaper announcements, and word of mouth (snow-ball technique). Inclusion criteria were: (a) adults ages 19 and over, (b) able to read and write English, and (c) living in one of the two communities during the wildfires. A l l residents who lived in the two communities were evacuated at the time of the wildfire. Volunteer respondents were either mailed or personally delivered the survey. Of the 191 questionnaires distributed, 149 were hand delivered in person and 42 were delivered by mail. A total of 109 surveys were returned for a 40 response rate of 57.1%. Eighty-five percent of the returned surveys had been hand delivered. Five surveys were excluded from data analysis because their data were incomplete. Fifty-six percent of the respondents (N = 104) were female and 44% male. The age of the respondents ranged from 20 to 87 years, with a mean age of 55.6 years (SD = 12.8). Most of the respondents were married or living with a partner (79%). Of the 76% who reported having children, 38% had children living at home during the fires, whereas 20% still had children living at home 3.5 years after the fires. At least 45% of the respondents had a minimum of a secondary school education. Just over one-third (36%) of the sample was working full-time and 21% were working part-time. Thirty-seven percent of the respondents reported being retired. Forestry was the more frequently reported occupation by those who were working full or part-time (27%, n = 64). For those who reported their total family income, 58%> reported a family income of less than $60,000 and 33%) reported a family income of $60,000 or higher. The majority of respondents (71%) were first generation in their current community (i.e., do not have parents, grandparents, or ancestors from their community), whereas 28% stated that their parents were from their community of residence. The respondents reported a range of 4 to 65 years living in their community, with a mean of 24.5 years (SD = 15.3). The question about length of residence was accidentally omitted in the survey. It was added following initial distribution of the survey; therefore, the sample size was lower for this demographic (n = 86). Overall, 36% (n = 37) of survey respondents reported currently residing in Louis Creek (38% stated they resided there during the fires); 41% (n = 43) stated their current community of residence was Barriere \(44% reported residing there during the fires); and 23% (n = 24) checked Other (18% residence during the fires). Other communities included 41 Kamloops, Darfield, McLure, Squam Bay, Dixon Valley, and Dunn Lake Road. For data analysis purposes, Louis Creek and other residents were combined (n = 58.7%) to further distinguish those who reported residing in Barriere. See Table 1 for more detailed demographic information on each community. Table 1 Demographic information of community residents by Barriere and Louis Creek (and other surrounding communities) (Barriere n = 43; Louis Creek (and other) n = 61, except where indicated) Variable Total % Louis Creek % Barriere % Gender Female 55.8 57.4 53.5 Male 44.2 42.6 46.5 Age M= 55.6 M = 57.2 M=53.2 SD=12.8 SD= 12.7 SD =12.8 Range (20-87) Range (20-87) Range (24-78) Education Primary / Elementary 15.4 14.8 16.3 Secondary / Trades certificate or 74.0 75.5 72.1 diploma / College diploma / Associates degree Bachelor's / Graduate / 10.6 9.8 11.6 Professional Marital status Married / Common - law / 78.8 77.0 81.4 • Living with partner Never married / Separated / 21.2 23.0 18.6 Divorced / Widowed Number of people living in household One 12.5 14.8 9.3 Two 65.4 68.9 60.5 Three or more 22.1 16.4 30.3 42 Variable Total % Louis Creek % Barriere % Have children3 76.0 (n= 103) 73.8(« = 61) 79.1 (« = Children in house now3 (n = 79) (n = 45) (« = 34) None 55.8 55.7 55.8 One 10.6 11.5 9.3 Two or more 9.7 6.6 11.6 Children in house during 2003 fires3 [n = 79) (n = 45) (n = 34) None 38.5 36.1 41.9 One 17.3 14.8 20.9 Two or more 20.2 22.9 1.6.2 Current employment status Full-time 35.6 31.1 41.9 Part-time 21.2 19.7 20.9 Retired 36.5 44.3 25.6 Unemployed 4.8 3.2 9.3 Caregiving responsibilities 1.9 1.6 2.3 If employed full-time or part-time, («=64) (n=36) (n=2%) main occupation3 Forestry 16.3 9.8 25.6 Tourism / Retail / Service sector 12.5 13.2 11.7 Health care / Education / 10.6 9.8 11.7 Government Services Agriculture / Farming / Ranching 8.6 13.1 2.3 Construction / Transportation 6.7 ' 8.2 4.7 Other 13.6 1-3.1 14.0 Household income3 (n=94) (n = 54) (n = 40) Less than $20,000 19.2 23.0 14.0 $20,000 - $59,999 38.5 34.4 44.2 $60,000 - $99,999 22.1 18.1 27.9 $100,000 and more 10.6 13.1 7.0 Family members from the communityb Parents 27.9 23.0 34.9 Grandparents 8.7 8.2 9.3 Ancestors 5.8 6.6 4.7 None of the above 71.2 75.4 65.1 43 Variable Total % Louis Creek % Barriere % Community residence.during 2003 fires 37.5 + (Other = 18.3) 44.2 Current community of residence 35.6 + (Other = 23.1) 41.3 Years living in community (i.e. region)3 (n = 86) (" = 51) (n = 35) M=24.5, SD= 15.3 Range (4 - 65) M=23.6, SD = -16.1 Range (4 - ,65) M=25.8, SD= 14.1 Range (4 - 50) Note. aDoes not sum to 100% due to missing data. A l l responses that applied were checked. Because of the small size of the two communities, non-probability convenience and snowball sampling methods were used in both Louis Creek and Barriere. During a visit to the communities in August 2006,1 was introduced to and maintained connection with leaders in both communities who participated in R. Cox's (2006) ethnographic study. These individuals have resided in their respective community for several years and acted as intermediaries to the broader community. They provided invaluable information and aided in distinguishing newcomers from residents who were present during the McLure Wildfire. Through these contacts, I recruited volunteers to aid in the dissemination of the survey. They were briefed on the importance Of confidentiality and other ethical standards (i.e., coercion, voluntary participation). Participants were recruited through announcements in two consecutive issues of the local newspaper, the North Thompson Star Journal, and flyers were distributed locally within the community (grocery store, restaurant, community bulletin board) (see Appendix C). I received three phone calls from the news announcement and none from the flyers. However, 44 these were important strategies as they created familiarity with the research prior to my visit to the communities. I employed various additional strategies to disseminate the survey. First, I spent 2 weeks and two separate weekends in March 2007 living in the communities in order to develop connections, make my presence known, and further create discussion about the research project. With the help of the intermediaries/volunteers, I visited businesses and community events and I was introduced to several residents, many of whom requested a survey package upon hearing about the study, which was immediately provided. Survey packages included materials in the following order: a cover letter (which stated that choosing to complete the questionnaire provided informed consent and participation was voluntary), demographic questions, several questionnaires (impact,, attachment, depression), and a return prepaid envelope (see Appendix D for cover letter). Second, I recruited three community members who volunteered to aid in the distribution of the survey (approximately 20 to 30 survey packages each). They were encouraged to discuss the research project with friends, neighbors, and/or colleagues and if requested, provided interested persons with a survey package. Surveys were returned either by mail, direct to UBC, or by a drop off procedure to the volunteer. Third, in order to increase my sample from Louis Creek, I spent one weekend explaining my study to the residents. Again, a survey package was provided to those who were interested. In order to increase response rate, a reminder card was mailed 7 to 10 days following the initial distribution, emphasizing the importance of the study (Dillman, 2000) (see Appendix E). A list of names and addresses was created to keep track of returned and unreturned questionnaires. Each personal identifier was linked with a coded ID number, which was 45 vvritten on the back of the questionnaire. The returned coded questionnaires and the list of personal identifiers and the matched coded ID number were kept in a locked filing cabinet in my office at U B C . Measures Demographic Information Several items on the survey were used to assess the demographics of the sample, for example, gender, age, education, marital status, number of children, number of people living in household, employment status, occupation, income, ancestry, community of residence during and following the McLure Fire, and length of residence (see Table 1). Impact of the 2003 McLure Fire Based on findings from R.Cox's (2006) study and previous surveys on the impact of a natural disaster (Maltais, Lachance, Simard, Brassard, & Picard, 2002; Scher & Ellwanger, 2005), I developed the following 20 items to assess how the fire impacted the residents: (a) loss of material possessions to self and neighbors, (b) loss of loved ones due to relocation, (c) loss of pets and livestock, (d) presence of house insurance (e) loss of or interruption to work, (f) experience of health-related problems, (g) loss of or change to the natural environment, and (h) degree of emotional support. See Appendix F for item development of the impact of the wildfire instrument. The items were adapted so that they reflected wildfires and the rural nature of the targeted communities. Twenty items were rated on a five-point Likert-type scale assessing the extent of loss and the extent a particular occurrence was experienced. For details of the complete scale, see Appendix B (Section B). Loss (9 items) was assessed using the following anchors: (1) no loss, (2) minor loss, (3) moderate loss, (4) major loss, and (5) total loss. The respondents 46 were asked to indicate how you would assess the extent of loss to each of the following: pets, job or business, damage to my house, etc. (see Appendix B, section B). The extent a particular occurrence was experienced (11 items) was assessed using the following anchors: (1) not at all, (2) very little, (3) somewhat, (4) to a considerable extent, (5) to a very large extent. Item #10 (I had house insurance at the time of the fires) was reverse scored. The respondents were asked to indicate to what extent you experienced each of the following: I experienced involuntary disruption to my work, up to one year following the fire, I visited a doctor, hospital, or clinic due to fire-related health problems, etc. (see Appendix C, section B). Although the intent was to develop a measure that reflected factual, rather than subjective experiences from the fire, utilizing a five-point Likert-type scale increased the subjectivity of the responses. Exploratory factor analysis was conducted in order to identify an underlying factor structure of the 20-item impact measure (see Table 2). Items were subjected to a principal-components analysis with varimax rotation, and one, two, and three-factor extractions were examined. Based on conceptual clarity, factor loadings, and reliability estimates, I determined that a one-factor extraction was sufficient. Items were dropped, in a stepwise manner, i f the item loaded <.20. This rule maintained a strong internal consistency estimate. Although it is common practice to retain factors that load > .30, it has been noted that this rule of thumb is arbitrary (Garson, D.G., 2007). Three items were dropped based on this criterion: (a) Family members and/or friends moved away from the community, (b) / experienced involuntary disruption to my work, and (c) / received emotional and moral support from my community. The final one-factor extraction accounted for 34% of the total variance. An item-mean score of impact from the wildfire was calculated (possible range of 1 47 to 5) for the 17-item factor; the higher the score, the greater the individual was impacted by the wildfires. The impact measure had a strong Gronbach's alpha, r = .87. Table 2 Principal-components analysis of the impact of the 2003 McLure fire measure (N =104) Scale item Item # Component factor loading Damage to my property 6 .90 Damage to my neighbors' property 8 .85 Damage to my neighbors' house 7 .82 Damage to my house 5 .81 Mental and emotional difficulties 14 .74 View from my house changes 17 .68 Surrounding natural environment changed 18 .59 Adverse physical health effects 12 .58 Visited a doctor, hospital, or clinic 13 .58 Visited a mental health professional 15 .41 Outdoor recreational areas changed 16 .40 Job loss of family members and/or friends 9 .37 ' Loss of livestock 3 .36 Job or business loss 4 .36 House insurance 10 .35 Emotional and moral support from disaster- 20 .23 related services Loss of pets 2 .21 Note. I removed item #14 (I experienced mental and emotional difficulties), created a new scale and correlated it with depression. The correlation was the same with or without this item (with r = .39, without r = .41). Therefore, I retained item #14. Place Attachment Based on a review of the literature and findings from R. Cox's (2006) study, place attachment was defined as attachment to an individual's community. Four dimensions of place attachment were initially included in the development of the attachment measure for the present study: general, social, and two physical dimensions (i.e., the natural and built environments). For the general (6 items) arid social (8 items) dimensions, I adapted items from existing and proposed measures of community attachment (Cross, 2003; Kasarda & 48 Janowitz, 1974), place attachment (Hidalgo & Hernandez, 2001; Williams, 2000), neighborhood cohesion (Buckner, 1988), and sense of community (Davidson & Cotter, 1986). For the natural dimension (8 items), I adapted items from the Connectedness To Nature scale (Mayer & Frantz, 2004), and proposed community and place attachment scales (Brehm et al., 2004; Cross, 2003; Williams, 2000). I chose the items based on their relevance for the targeted rural communities, their reflection of three modalities of attachment (affect, cognition, and behavior), and on a combination of face validity and high item-total correlation coefficients. I also developed six items, specifically for the present study, for the built dimension. The first three items include the following: (a) / feel attached to my personal possessions and belongings; (b) I feel like I belong in my house; (c) When I think about this community, I think about the town's centre. The last three items were developed in conjunction with an open-ended question (item #53) - In your opinion, which buildings in your community would be considered as the most important ones? Name three buildings. Each of the three items asks whether the particular building mentioned is personally meaningful (item #54, Building #1 is personally meaningful to me; item #55, Building #2 is personally meaningful to me; item #56, Building #3 is personally meaningful to me). Items were based on the most prominent themes in R. Cox's (2006) study about the built environment. See Appendix G for item development of this section of the questionnaire. Two of the last three items (#55 and #56) were not included in the final scale because eight respondents named only 1 or 2 buildings leaving items #55 and #56 incomplete. In addition, based on written and oral feedback on the survey, there was some confusion about these items: (a) respondents from Louis Creek had difficulty thinking of three buildings in their community, and (b) respondents had difficulty identifying three buildings that i . • 49 warranted importance to the community. In general, I determined that these items were poorly constructed and were not useful in determining the degree of attachment to the built environment. However, a content analysis was conducted on item #53, the open-ended question, in the survey (see Appendix H). Although not a dimension of place attachment, I included four items to assess change in attachment since the McLure Fire. Because it is widely acknowledged that change is difficult to assess, this scale was included for descriptive purposes. In order to remain as objective as possible, I developed items to reflect a change in behavior - (a) I visit more with my neighbors and/or friends now, compared to before the fire occurred; (b) I participate in more community projects/activities now compared to before the fire occurred, (c) tspend more time doing outdoor recreation activities now, compared to before the fire occurred. The fourth item reflects a change in cognition: (d) I am more interested in knowing what goes on in this community now compared to before the fire occurred. The items reflect a change in an individual's attachment to the community, in general, as well as the social and natural environments. . A total of 28 items, plus the 4 items assessing change, were selected or developed to measure place attachment and represented the general, social, natural, and built attachment dimensions. A l l items were rated on a five-point Likert-type scale from (1) strongly disagree, (2) disagree, (3) neither agree nor disagree, (4) agree, and (5) strongly agree, as suggested by Williams (2000). This is a commonly used response format that measures intensity of the response (Nardi, 2003). Scores ranged from 1 to 5, and the higher the score the stronger the positive attachment, or the greater the change in attachment (i.e., toward stronger attachment). 50 To determine the meaning and relevancy of the four dimensions and change in attachment subscales that made up the place attachment scale, I undertook exploratory factor analysis to determine the underlying factor structure that best fits the construct of place attachment used in this study, and to identify subscale items. Items were subjected to a principal-components analysis with varimax rotation and three, four, five, and six-factor extraction. Based on the underlying conceptual model and internal consistency estimates, a four-factor extraction was used to define the subscales of place attachment. Items were dropped, in a stepwise manner, i f the item did not load greater than .20 on one factor compared with the other factors. Four items were dropped based on the above criteria. Two items (#31 -1 would be sorry if the people I cared about moved away from this community and #37 - When I think about this place, I think about the things I have done here with my friends, family, and people who matter to me) were dropped from the first factor; one item was dropped from the second factor (#45 - The surrounding natural environment is an important contributor to my livelihood) and; one item was dropped from the fourth factor (#48 - When I think about this community, I think about the town's centre). A fifth item (#54) was dropped from the fourth factor because of five missing cases and the internal consistency of this factor significantly increased when this item was deleted. This item was the third item that asked respondents whether the buildings they mentioned in question #53 were personally meaningful. The final four-component extraction converged after five iterations and accounted for 34% (Factor 1), 12% (Factor 2), 10% (Factor 3), and 6% (Factor 4), respectively of the total variance (Table 3). The main difference between the initial conceptual model of place attachment and the new factor structure was that the first factor combines the general and 51 social dimensions of place attachment; this factor was identified as social. The second and fourth factors represent the natural and built environment dimensions, respectively, and are identified as natural and built in a revised final model (see Figure 3). The final model pertains to a population that was affected by a wildfire 3.5 years prior. The third factor, characterizing change in attachment, emerged as a separate factor I identified as change. Change in attachment is not represented in the place attachment model as it reflects all of the dimensions. 52 Table 3 Principal-Components Analysis of Place Attachment Measure (N = 104) Components Scale item Item# 1 2 3 4 28 .84 .19 .02 .09 33 .83 .11 .09 .07 25 . .83 .09 -.06 -.01 32 .79 .15 -.01 .23 29 .74 .24 .07 .23 24 .74 .23 -.05 .14 34 .73 -.01 .19 -.08 27 .73 .27 .02 .16 30 .56 .05 .32 -.29 26 .55 .28 .08 .22 35 .54 .25 .03 .10 36 .52 .15 .14 -.05 42 .12 .85 .04 .11 41 .01 .84 .08 .04 40 .29 .84 .03 -.02 38. .27 .83 -.14 -.11 39 .34 .81 -.08 -.07 43 .17 .65 .29 .06 44 .18 .56 .17 -.02 50 .22 .01 .81 -.13 49 -.01 .07 .77 .03 51 .12 .17 .73 .00 52 -.02 .01 .71 .19 46 .09 -.02 .05 .85 47 .35 -.01 .06 .79 Social Attachment Feel like I belong to this community Feel loyal to the people in my community Feel at home in this community Friendships I have mean a lot to me This community says a lot about who I am I would be sorry to leave this community Feeling of fellowship runs deep with others Plan to remain a resident for the rest of my life I participate in community projects/activities Interested in knowing what goes on in community I visit with my friends/neighbors in their homes Identify with values of the people who live here Natural Environment Attachment Natural environment part of community to which I belong Feel kinship with plants/animals in community Sense of oneness with land that surrounds community Identify with land that surrounds community Attached to land that surrounds community Would like to be involved in activities related to the protection of the environment Environment provides outdoor recreation I enjoy Change in Attachment Participate in more community projects now Visit more with neighbors/friends now More interested in knowing what goes on now Spend more time doing outdoor activities now Built Attachment Feel attached to my personal possessions I feel like I belong in my house Note. Scale items presented were shortened. For the complete item, see Appendix F. 53 Construct Dimensions Modalities Figure 3. Revised final model ofplace attachment. Descriptive statistics were computed for the three attachment factors and one change in attachment factor, and a reliability analysis was conducted (see Table 4). The social and natural subscales had strong internal consistency estimates (Cronbach's alpha of r = .92 and .90, respectively), and the change and built subscales had acceptable estimates (Cronbach's alpha of r = .78 and .75, respectively). The three attachment dimensions and change in attachment were weak to moderately correlated with each of the other factors, and all were positively related (rs ranged from .06 to .46). Regarding the change in attachment factor, the direction of the relationship indicated that an increase in attachment since the wildfire was associated with a greater degree of social and natural attachment currently. Change in attachment was minimally associated with the built environment factor because the items do not reflect a change in attachment to the 54 built environment. The weakest relationships were between the built and natural environment factors (r = .06, ns) indicating that they were measuring fairly independent dimensions of attachment. Table 4 Bivariate correlations and internal consistency estimates for attachment subscales Attachment Factor 1 Factor 2 Factor 3 Factor 4 Variable (Social) (Natural) (Change) (Built) Social - .46** .22* .31** Natural - - .17* .06 Change - - . - .08 Built - • - - -No. of items 12 7 4__ 2 " Cronbach's .92 .90 .78 .75 Alpha • Note. *p<05. **p<.00\ (1-tailed). Depressive Symptoms A 10-item short version of the original 20-item Center for Epidemiologic Studies Depression Scale (CES-D) (Andresen, Malmgren, Carter, & Patrick, 1994) was used to examine the residents' current depressive symptoms (see Appendix C, Section D). It was designed to detect the presence of depressive symptoms, not as a diagnostic tool for clinical depression. The original CES-D is a self-report scale designed to measure depression in community samples and has been shown to discriminate between clinical and community samples; it is considered a valid screening tool for depression (Radloff, 1977). The 20-item CES-D has been used in several disaster studies (U.S. Department of Health and Human Services, 2006; Tyler & Hoyt, 2000; Tremblay, Blanchard, Pelletier, & Vallerand, 2006) and in studies of rural communities (Husaini, Neff, Harrington, Hughes, & Stone, 1980). A 15-55 item short version has also been used in studies of residents affected by disasters in rural communities (Ginexi, Weihs, & Simmens, 2000). The CES-D has good internal consistency (a = .85) and concurrent validity, correlating significantly with a number of other depression and mood scales (Tremblay et al., 2006). Because recruitment of a sufficient sample size was a priority for this study, the 10-item short version of the CES-D was chosen because of its short length. This short version has also been utilized in disaster studies (Seplaki, Goldman, & Weinstein, 1999). It has been demonstrated that the sensitivity and specificity of this version is not significantly reduced from the original scale (Kohout, Berkman, Evans, & Cornoni-Huntley, 1993; Shrout & Yager, 1989). It has shown good predictive accuracy when compared to the 20-item version (Andresen, 1994), and internal consistency is comparable. The responses to each item are scored, on a four-point Likert-type scale, as ranging from 0 (rarely or none of the time), 1 (some or a little of the time), 2 (occasionally or a moderate amount of the time), or 3 (most or all of the time). Scores were calculated by reverse scoring the two positive mood items (7 felt hopeful about the future and I was happy); scale scores were then summed, with a possible range of 0 to 30. Higher scores imply more distress. A score of 10 or greater implies that the individual has symptoms indicative of depression. Andresen et al. (1994) found that a cutoff score of equal to 10 is comparable to the 20-item cutoff score of 16. In the present investigation, Cronbach's alpha for internal consistency was r = .89. However, norms for rural community samples are unavailable for this version. Control Variables In order to determine whether the relationships between impact and depression, impact and attachment, and attachment and depression were not confounded by demographic 56 characteristics, several characteristics were examined for their possible use as control variables. From a review of the literature, length of residence was the most consistent demographic variable shown to contribute to an individual's degree of attachment to place (Brehm etal., 2004; Goudy, 1990; Kasarda & Janowitz, 1974; Riger & Lavrakas, 1981; Sampson, 1988; Theodori & Luloff, 2000). However, to date, length of residence has not been associated with depressive symptoms. Furthermore, length of residence may be related to impact of the disaster as the longer residents lived in the community, the more they may have to lose in the wildfires. Thus, length of residence was considered as a potential control variable for an examination of the relationship between impact and attachment dimensions. Age (trends show a curvilinear relationship with depression), gender (females tend to exhibit greater depressive symptoms), and marital status (mixed results) have been shown to contribute to symptoms of depression following a natural disaster (Norris et al., 2002b). Although there is little evidence that these variables also relate to impact or attachment, it is possible that developmental stage, gender roles, and close relationships also affect impact of the wildfires and current attachment to place. Rowles' (1990) research on place attachment in rural elderly populations suggested strong attachments at this particular life-cycle stage; yet, in general, an age trend has not been documented. Thus, I considered these four variables (i.e., length of residence, age, gender, and marital status) as potential control variables for each of the research questions. Research Questions The following exploratory research questions guided my examination of the relationships between (a) the impact of the 2003 McLure Wildfire on residents of Louis Creek and Barriere, their present attachment to their community, and symptoms of depression, 3.5 years following the disaster; and (b) the relationship between place attachment and symptoms of depression. The extent to which each of these relationships I--varied by community (Louis Creek and Barriere) was also explored. Because of the exploratory nature of this study and to decrease the chances of a Type II errors-values less than or equal to .10, revealing a trend towards significance, were examined and reported. 1. Is there a (statistically significant,/? < .10) linear relationship between the initial impact of the 2003 McLure Wildfire on residents of Louis Creek and Barriere and their present strength of attachment to the community (social, natural, and built dimension) to which they are tied, after controlling for demographic variables? 2. Is there a (statistically significant, p < .10) linear relationship between the immediate impact of the 2003. McLure Wildfire on the residents and their present depressive symptoms, after controlling for demographic variables? 3. Is there a (statistically significant, p < . 10) linear relationship between the residents' present strength of attachment to their community (social, natural, and built dimensions) and their present depressive symptoms, after controlling for demographic variables? Data Analysis Data screening. Data analyses were completed using SPSS software (Statistical Package for the Social Sciences, 2003). The data were initially examined for errors and missing values. Five cases were deleted from the final data set because of a full missed section on the survey or several missed questions throughout the survey. Sporadic missed data were handled in several ways. First, i f a missing value was found in the impact and/or general information sections, it was possible to interpolate based on the partner's survey or my knowledge of the community and where this person was situated. Second, crosstabulation 58 frequencies, medians, and means were examined and used for interpolation, except under one circumstance. Item #47 (7 feel like I belong in my house) was missed across, six surveys possibly due to the nature of how it is situated among the other items. A value of 3 was given in order to alleviate any potential bias because of the short length of the scale. Outliers and normality were detected by visually examining histograms, stem-and-leaf plots, and box plots with error bars. Each case presenting with an outlier was reviewed and scanned for erroneous values. An outlier is defined as a score that is between 1.5 and 3 box lengths from the 75 t h percentile or 25 t h percentile (upper or lower edge of the box). An x extreme score.is defined as a score that is greater than 3 box lengths away from the 75 t h percentile or 25 t h percentile (SPSS, 2003). Furthermore, standardized residuals were examined in the regression analyses to identify outliers that may impact this form of analysis. A residual value of 2.50 to 3.3 was the criterion used in the present study to detect problematic outliers (Tabachnick & Fidell, 1996). It was determined that theoutliers found were legitimate cases within the population and were retained for analysis. In order to test assumptions of normality, skewness and kurtosis values were calculated (as presented in Table 5). These represent the symmetry and peakedness of the distribution. Values close to zero represent a normal distribution. A value close to 2.00 is suggested to cause concern for skewness and kurtosis (Miles & Shevlin, 2001). Skewness and Kurtosis values in the present study were found to be closest to 0 and 1. 59 Table 5 Skewness and Kurtosis Values Variables Skewness Kurtosis Impact -0.14 -1.10 Attachment Social Natural Change Built -0.33 -0.25 0.46 -0.37 0.86 -0.30 0.30 -0.14 Depression 1.14 0.44 Preliminary analysis. Means, standard deviations, and bivariate correlations were computed for the full sample and separately by community (Louis Creek and Barriere) and included demographic variables, as well as impact, the three attachment dimensions, change in attachment, and depressive symptoms. One-way between-groups analyses of variances (ANOVAs) were performed to investigate community (Louis Creek, Barriere) mean differences with respect to depression and impact from the fires. Furthermore, one-way between-groups multivariate analyses of variance (MANOVA) were performed to investigate community (Louis Creek, Barriere) mean differences respect to attachment dimensions (social, natural environment, and built) and change in attachment, and to test demographic data (age and length of residence). Chi-square tests for independence were conducted on gender, education, marital status, and income, to examine differences between the two communities. The purpose of these preliminary analyses was to explore whether mean or categorical differences existed between residents of Louis Creek and Barriere, and to further select control variables for the research questions. 60 In order to determine the appropriateness of the control variables, I examined bivariate correlations between potential control variables and the predictor and criterion variables, and those that were statistically significant were used in the data analysis to control for confounding effects. P-values less than or equal to .10 were examined and reported. Research questions. Hierarchical multiple regression analyses (HMR) were used to test the three research questions; the second and third research questions were tested simultaneously. Ancillary question. Because of the difficulty in measuring change and the lack of evidence regarding the validity of this measure, in addition to the research questions, change in attachment was tested as an ancillary question and used for descriptive purposes. 61 CHAPTER 4: RESULTS In this chapter, I present summaries of the preliminary analyses, the analyses of the research questions, and the ancillary question (i.e., change in attachment). The findings support the importance of the initial impact of the 2003 McLure Fire on residents' current degree of attachment to their built environment, as well as on residents' symptoms of depression, after controlling for demographic variables. Moreover, natural environment attachment accounted for a small but significant amount of variance in depression, after controlling for age and marital status. The relationship between impact and attachment or depression and between attachment and depression did not differ by community (Louis Creek, Barriere) once age and marital status were controlled for. Preliminary Analyses Means, and standard deviations for the study variables are reported for the total sample, and Louis Creek and Barriere communities in Table 6. Pearson product-moment correlation coefficients for each community are reported in Table 7." Across both communities, 31 respondents (25 from Louis Creek, 6 from Barriere) or 30% of the total sample (41% of Louis Creek, 14% of Barriere) had depression scores greater than or equal to 10, meaning they showed symptoms indicative of depression (Andresen et al., 1994). Depressive symptoms among Louis Creek residents were greater than those found by Norris et al. (1999) 30 months following a hurricane (41% vs. 36%, respectively). Respondents from Louis Creek and the surrounding area reported higher scores of depressive symptoms (M= 9.38, SD = 7.23, n = 61) than those in Barriere (M= 4.91, SD = 4.82, n = 43). A one-way A N O V A was performed to investigate differences in the depression score means between the two communities. Results indicated a statistically significant difference, 62 F(\, 102) = 12.48, p < .01. Although norms for community samples are unavailable for the 10-item version of the CES-D used in the present study, Seplaki et al. (1999) reported that their sample of Taiwanese adults (N = 1160, age > 50 years), who experienced an earthquake, had a mean score of 5.60 (SD = 5.73) 15 months following the disaster. The mean scores from Louis Creek respondents were almost four points higher than this Taiwanese sample, and the mean scores from Barriere respondents were slightly lower. Therefore, respondents from Louis Creek and Barriere exhibited depressive symptoms, similar to and greater than a community that experienced a natural disaster. The sample from Seplaki et al.'s (1999) study was drawn from both rural and urban areas of Taiwan. Table 6 Means and standard deviations of study variables far the full sample, Louis Creek and the surrounding area (n =61), and Barriere (n = 43). Total Sample Louis Creek Barriere Variables M SD M SD M SD Impact 2.65 0.77 3.00 0.73 2.15 0.49 Attachment Social 3.84 0.63 3.78 0.64 3.92 0.61 Natural 3.99 0.63 4.03 0.59 3.94 0.70 Built 3.95 0.80 3.85 0.87 4.08 0.67 Change 2.89 0.69 2.95 0.74 2.80 0.61 Depression3 7.53 6.70 9.38 • 7.23 4.91 4.82 Note. 3Based on the mean of the summed scores Table 7 Intercorrelations of Variables - Louis Creek and the surrounding area (n = 61); Barriere (n = 43) Variables 1 2 3 4 5 6 7 8 9° 10 11 12b 1. Depression 29** -.17 . 23** -.07 -.08 -.20* .10 -.01 _ 32*** -.14 .09 2. Impact 3Q** - -.00 .08 _35*** .02 . .06 -.01 -.18* .20* 32*** .06 Attachment 3. Social -.22* -.11 - 34*** 31 *** .20* 23** .03 -.04 -.06 .07 .26** 4. Natural -.12 .26** g 5 * * * - .04 .14 -.05 .17* -.18* -.06 .01 .21* 5. Built .09 -.02 29** .13 - .12 4Q*** -.21* .10 - -.07 .01 .10 6. Change .04 .25* 29** .19 .05 - .14 -.13 -.19* .15 .02 _44* * * Control Variables 7. Age -.18 _ 33** .24* -.02 .14 .01 - - 30*** 32*** .13 -.13 .20* 8. Gender .00 _ 29** -.02 -.14 .08 .11 -.08 - -.06 -.08 .18* -.09 9. Incomed -.05 -.09 .30** .08 .13 29** .08 .03 - 3g*** .11 -.16c 10. Marital Status -.04 -.20* .16 -.02 .19 .34** .21* 27** 59*** - .07 _ 35*** 11. Education -.18 -.21* -.08 _ 33** -.16 .13 27** .28** .31** 32** - -.06 12. Length of .01 _ 33** .14 .06 -.05 -.04 .22* .07 .19f .18 .16 -residence (years)6 Note. Upper right half of the matrix represents Louis Creek (and the surrounding area); Lower left half represents Barriere. * p < .10 level *.* p < .05 level *** p < .01 level (1-tailed). -Gender (0 = male, 1 = female), Income (0 = <$20,000, 1 .= $20,000-60,000, 2 = $60,000-100,000, 3 = >$ 100,000), Marital Status (0 = not partnered, 1 = partnered), Education (0.= primary/elementary, 1 = secondary, trades certificate, college diploma, 2 = Bachelors, graduate degree, professional). Due to missing data, a » = 54, bn = 51, cn = 45 (Louis Creek). dn = 40, en = 35, fn = 33 (Barriere). 64 As anticipated, the sample represented variability in terms of the degree of impact from the wildfires, with an observed range of scores falling between 1.18 (minimal impact) and 4.29 (major impact). Consistent with the events of the 2003 Wildfire, respondents from Louis Creek (and the surrounding area) reported a greater impact from the fires (M= 3.00, SD = .73) than those from Barriere (M= 2.15, SD = .49). This is consistent with the fact that Louis Creek residents lost most Of their homes and businesses, and had more extensive property damage than residents of Barriere. A one-way A N O V A was performed to investigate mean differences on impact between the two communities. Results indicated a statistically significant difference, F ( l , 102) = 43.10,/? < .001, with Louis Creek residents reporting an initial greater impact from the fire than Barriere residents. With respect to place attachment, the means indicated that the total sample felt attached to the social, natural, and built environment of their community. Seventy-five percent of the sample had,a mean social attachment score greater than or equal to 3.5 (item mean scores fall between 1 and 5), while 79% of the sample had a mean natural environment and built attachment scores greater than or equal to 3.5. The mean, median, and mode scores of the social, natural, and built attachment subscales were similar within the two communities. The distribution of the responses on each subscale surrounds the fourth anchor point; thus, on average, the respondents agreed with the statements presented. This suggests that respondents, in both communities, reported being attached to the social, natural, and built environment of their community. With regard to change in attachment, 21%> of the total sample had a mean change in attachment score greater than or equal to 3.5. The mean, median, and mode scores of the change in attachment dimension, in both communities, were distributed closest to the middle 65 of the scale, suggesting that, on average, respondents neither agreed nor disagreed with the statements in this subscale. This indicates that, on average, residents were not more involved or attached to their community now, compared to before the fire occurred. Results of the M A N O V A investigating mean differences of the attachment and change in attachment variables showed that the overall test of mean differences between the two communities was not statistically significant, F (4,199) = 1.57, p = .19, indicating that the communities did not differ on mean levels of attachment or change in attachment. I examined the zero-order correlations (Tables 7) between the variables to determine the strengths and directions of the relationships between depression, impact, attachment dimensions (social, natural, built), change in attachment, and several potential control variables, in each community. In general, zero-order correlations between the criterion (depression) and predictor (impact, social, natural, built attachment, change in attachment) and control variables were low to moderate in magnitude for both communities (rs ranging from -.32 to .29, Louis Creek; -.22 to .30, Barriere), and among predictor variables, low to moderate (rs ranged from -.44 to .40, Louis Creek; -.33 to .65, Barriere). In all instances, the scale ihtercorrelations were below their scale reliabilities. Correlations were in the expected direction, with the exception of the relationship between depression and both built attachment (r = .09) and change in attachment (r = .04) in the Barriere community; however, these correlations are very low. M A N O V A results examining mean differences of age and length of residence showed that the overall test of mean differences between the two communities was not statistically significant, F(2,83) = 2.21,/?.= ns. However, when the results for the dependent 66 variables were considered separately (by way of follow-up univariate analyses), age, F(l,85) = 3.34, p = .07, but not length of residence, F(l,85) = 0.44, ns, was statistically (marginal) significant, suggesting that respondents from Louis Creek, compared with Barriere, were statistically significantly older (M= 57.2, SD = 12.7 years; M= 53.2, SD = 12.8 years, respectively). Chi-square tests for independence did not result in statistically significant differences in gender, education, marital status, and income between the two communities. Research Question #1 . To determine the extent to which initial impact of the wildfire was associated with attachment 3.5 years later, hierarchical multiple regression (HMR) was performed to examine the extent to which the predictor variable (impact) explained the variance in attachment. Three HMRs were conducted with social, natural, and built attachment as the criterion variables. The choice of control variables was determined in two steps. In the first step, I determined which demographic variables were significantly correlated (p < . 10) with both the predictor (impact) and the criterion Variables (social, natural, built). The bivariate correlations revealed that age, gender, income, marital status, education, and length of residence were significantly correlated with impact and at least one of the criterion variables. Thus, in initially testing the relationships between impact and attachment, I controlled for all demographic variables that were significantly correlated with each of the criterion variables. Only those control variables that remained statistically significant (p < . 10) in the initial H M R analyses were retained and included in Stepl for a final H M R analyses. Community (Barriere = 1, Louis Creek = 0) was also included in Step 1 in order to determine whether the relationship between impact and attachment differed depended on which community the residents resided in. Initial impact was included in Step 2, to determine i f 67 impact accounted for additional variance beyond that account for by control or community variables. Social attachment. The results of the final H M R analysis predicting social attachment are presented in Table 8. Impact did not statistically significantly account for additional variance in social attachment to the community, after the effects of age and community were accounted for. Age.made an independent contribution and explained 7% of the variance in social attachment. Age was positively related to social attachment, thus, the older the resident, the more attached he or she felt to the social environment of their community (P = .23, p <.05). Community did not statistically significantly account for additional variance in social attachment. Table 8 Summary of Hierarchical Multiple Regression Predicting Social Attachment (N=104) Variables Entered B SE (3 Step 1 Age .01 .01 .23** Community .17 .15 .13 Step 2 Impact -XT2 TO -.03 Note. The beta values are the unstandardized coefficients from the final simultaneous analysis, each term being corrected for all other terms in the model. SE = standard error, (3 = standardized regression coefficient. Community (l=Barriere, 0=Louis Creek) R2=.07,F(2, 101) = 3.54, p<.05, for Step 1; AR2 = .00, F ( l , 100) = .047, ns, for Step 2. **p < .05 (two-tailed) Natural attachment. The overall F-ratio was not statistically significant for the H M R in predicting attachment to the natural environment, F(2,101) = 0.37, ns; therefore, the beta weights were not examined. 68 Built attachment. The results of the H M R predicting built environment attachment are presented in Table 9. After the variance from age and community were accounted for in attachment to the built environment, impact contributed an additional 6% of variance to this model. Both impact and age significantly contributed to the variance in attachment to the built environment (impact, fj = -.28, p < .05; age, R = .29, p < .01). The final simultaneous model accounted for 17% of the variance in attachment to the built environment, F(3,100) = 6.54, p < .001. The results suggested that the greater the initial impact from the fires, the less residents reported attachment to their homes and personal belongings (i.e., built environment). Moreover, the older the resident, the more they reported feeling attached to their built environment. Community did not significantly contribute to the variance in attachment to the built environment. Table 9 Summary of Hierarchical Multiple Regression Predicting Built Environment Attachment (N 104) .'• ' , Variables Entered B SE • P Step 1 Age .02 .01 .29*** Community .05 .18 .03 Step 2 Impact -.29 .11 -.28** . Note. The beta (B) values are the unstandardized coefficients from the final simultaneous analysis, each term being corrected for all other terms in the model. SE = standard error, p = standardized regression coefficient, Community (l=Barriere, 0=Louis Creek). R2 = .11, F (2, 101) = 6.15,/? < .01, for Step 1; AR2 = .06, F (3, 100) = 6.54,/? < .001, for Step 2. ***p < .01 **p < .05 (two-tailed) Ancillary question. A n additional question was addressed regarding change in attachment. The results of the H M R predicting change in attachment are presented in Table 69 10. Impact did not contribute to this model. However, both marital status and length of residence accounted for 15% of the variance in change in attachment (marital status, B = .23, p < .05; length of residence, R = -.27, p < .01). The results suggested that having a partner and. a shorter amount of time living in the community was associated with a greater attachment now, compared with before the wildfire (i.e., participating more, and being more interested, in the community). Table 10 Summary of Hierarchical Multiple Regression Predicting Change in Attachment (n=86) Variables Entered B SE p Step 1 Marital Status .38 .17 .23** Length of Residence -.01 . .01 __27*** Community -.02 .18 -.02 Step 2 Impact .06 .12 .07 Note. The beta (B) values are the unstandardized coefficients from the final simultaneous analysis, each term being corrected for all other terms in the model. SE = standard error, p = standardized regression coefficient. Marital Status (l=partnered, 0=not partnered), Community (l=Barriere, 0=Louis Creek) R 2 = .15, F (3,82) = 4.94, p < .01, for Step 1; AR2 = .00, F(4, 81) = 3.75,p < .01, for Step 2. **p < .05 , ***p < .01 (two-tailed). Research Questions #2 and #3 Questions 2 and 3 were tested simultaneously through H M R in order to determine the extent to which initial impact of the wildfire and current strength of attachment was associated with the criterion variable, current depressive symptoms. Control variables and community were entered in Step 1. Impact was entered in Step 2 to determine whether impact from the fire accounted for additional variance in depression beyond the control variables and community. The three attachment variables (social, natural, built) were entered 70 in Step 3 to account for additional variance in depression beyond that accounted for by impact, the control variables, and community. Control variables were identified that were significantly correlated, p < .10, with both the criterion variable (depression) and a predictor variable (impact or attachment). Based on the bivariate correlations, marital status (0 = not partnered, 1 = partnered) exhibited statistically significant weak-moderate correlations with depression and impact, while age exhibited a marginal significance with depression and social attachment (see Table 7). The results of the H M R are presented in Table 11. In Step 1, marital status, age, and community, explained a significant proportion of variance in depression scores F(3,100) = 7.38,p < .001). Marital status uniquely contributed to depression ((3 = -.23,p <.01), and age made a marginal significant contribution (P = -.16, p <.10). Community did not contribute any independent variance in depressive symptoms (p = -.16, ns). In step 2, impact accounted for a significant increase in variance in depressive symptoms scores, AR = .08; F(l,99) = 10.69,/7 — .001, and was a significant contributor to depression (p = .39,p <.001), after age, marital status, and community were accounted for. In Step 3, the attachment variables (social, natural, built) contributed a statistically significant increase in variance, AR2 = .06, F(3, 96)= 2.71, p <.05, after impact, marital status, age, and community were accounted for in depression. Only natural environment attachment exhibited a significant (marginal) contribution to current state of depressive symptoms (P = -.18, p <.10). Of note, the final simultaneous model accounted for 32% of the variance in depressive symptoms, R2 = .32, F(7,96) = 6.42, p < .001. < 71 Table 11 Summary of Hierarchical Multiple Regression for Variables Predicting Depression (N=104)-Variables Entered B • SE. P Step 1 • " • " Age -.01 .01 -.16* Marital Status -.38 .14 -.23** Community --21 -14 -.16 Step 2 Impact .34 .09 .39*** Step 3 Social Attachment -.08 .11 -.07 Natural Environment -.19 .10 -.18* Built Environment A2_ '.08 .14 Note. The beta (B) values are the unstandardized coefficients from the final simultaneous analysis, each term being corrected for all other terms in the model. SE = standard error, P = standardized regression coefficient, Marital Status (l=partnered, 0=not partnered), Community (l=Barriere, 0= Louis Creek). R2 = .18, FQ, 100) = 7.38,p < .001, for Step 1; AR2 = .08, F(\, 99) = 10.69, p = .001, for Step 2; AR2 = .06, F (3, 96) = 2.71,p < .05, for Step 3. *p •<. 10. **p<. 01, ***p< .001 (two-tailed) : Therefore, the younger the respondent, respondents who were not partnered, and those who reported a greater extent of impact from the wildfire, reported greater symptoms of depression. Moreover, respondents who felt less attached to the natural environment reported greater symptoms of depression. These findings suggest that multiple variables contribute to symptoms of depression in this sample. The community in which one resided was not a significant predictor of depression after marital status and age were controlled for, even though M A N O V A results revealed significant mean differences in depressive symptoms for Louis Creek and Barriere residents. It may be that community differences in depression were the result of age differences (i.e., on average Louis creek residents were older than Barriere residents). 72 CHAPTER 4: DISCUSSION The purpose of this study was to explore the relationship between (a) the extent residents from two rural communities were impacted by a wildfire 3.5 years prior and their current perceptions of attachment to their community and depressive symptoms, and (b) place attachment dimensions and depressive symptoms. Factor analysis identified a model of place attachment that encompassed three dimensions - attachment to the social, natural, and built environments in communities that have been impacted by a wildfire- as well as a change in attachment scale. Items retained reflected the three modalities of place experience: affect, cognition, and behavior. After controlling for age or marital status, and the effects of community, the results indicated that the greater the initial impact of the wildfire, the less attached to the built environment (i.e., house, personal belongings) respondents reported. The results also indicated that the greater the initial impact of the wildfire, and the less attached respondents were to the natural environment, the greater the symptoms of depression they reported. The communities did not contribute to degree of attachment or depressive symptoms, once the effects of age or marital status were accounted for in the criterion variables. The Relationship Between Impact and Place Attachment Of the three dimensions of attachment, impact from the wildfire was only associated with attachment to the built environment, after controlling for age and the community in which residents reside. This was an interesting finding considering that the built attachment subscale only contained two items and reflected one modality, affect. The results suggest that residents who reported a greater extent of impact from the 2003 McLure Fire were less attached to their house and personal belongings. Because the impact of the wildfire was 73 greater for Louis Creek respondents, this suggests that the relationship between impact and attachment to the built environment can be primarily attributed to the respondents from Louis Creek. These findings are not surprising considering that those who reported the greatest extent of impact and loss (i.e., residents from Louis Creek) were living in new homes and have acquired new belongings and possessions since the fire. Because of place attachment's temporal quality, inhabiting a new home signifies less time to cultivate meaning and memories, thus less time to build a strong attachment to place. Up to one year following the fire, R. Cox (2006) found that residents continued to feel a sense of dislocation with the loss of their homes and disorientation with the loss of personal belongings, suggesting a disruption to their attachment to the built environment. These findings attest to the long-term impact of natural disasters on residents' attachment to place (i.e., the built environment within their community). It is also possible that feeling less attached to material possessions following a significant impact from a disaster may represent a conscious decision to not develop an emotional connection based on a realization of life's impermanent nature. Age also contributed to both built and social attachment. The older the respondent, the more attached he/she felt to their immediate built and social environments. Homes and personal possessions play a particularly significant role in the lives of the older adult and elderly population (Gitlin, 2000; Rubinstein & Parmelee, 1992). Although the average age for Louis Creek respondents was 57 years and Barriere respondents was 53 years, the overall actual range was up to 87 years (78 years in Barriere). To explore this further, I conducted regression analyses, selecting cases with age greater than and equal to 65 years, to determine whether old age was uniquely related to the various dimensions of attachment. The results were not statistically significant. However, age as a predictor of community attachment is 74 consistent with other studies predicting attachment to the social ties with the community (Goudy, 1990; Kasarda & Janowitz; Sampson, 1988; Theodori & Luloff, 2000). Contrary to other studies, length of residence was not a predictor of attachment in the present study (Brehm et al., 2004; Goudy, 1990; Kasarda & Janowitz; Theodori & Luloff, .2000). This may be due to the more recent transient nature of the Louis Creek and Barriere communities. Many young families have relocated to larger urban centres, while an increased number of retirees have moved into the communities seeking an economical, peaceful rural lifestyle. Newcomers may feel attached to the social, natural, and/or built environments of the community in a different way than residents who have lived in the communities for several years. On average, respondents across both communities reported a positive attachment to their social and natural environments. A positive attachment tp the social and natural environment of rural residents' community is consistent with the literature on place attachment within the context of a natural disaster (Brown & Perkins, 1992; H. Cox & Holmes, 2000; Schwarz et al., 1995; Skelton, 2004). Unique to these studies is their illustration of how a strong attachment to place prior to the disaster foreshadowed a strong, or often stronger, attachment to place post-disruption. Due to a lack of comparison data, it is difficult to extrapolate this prediction to the present study, yet it is likely that the respondents in the present study had a positive attachment towards their community prior to the 2003 McLure Fire because of the various meanings they attribute to their home or community (e.g., close to family, lifestyle, employment). It is possible that the impact of the fire disrupted the residents' process of attachment immediately following the disaster, and 3.5 years following was enough time for residents to recreate a similar strength of attachment to 75 that prior to the fire. This may explain why the impact of the fire did not detrimentally or positively relate to respondents' current social and natural attachment 3.5. years following the fire. My visits to the communities and the results of R. Cox's (2006) ethnographic study revealed the integral role the natural environment plays in the everyday lifestyle and identity of the residents. This was also reported in other studies of rural communities (Brehm et al., 2004, H. Cox & Holmes, 2000; Schwarz et al., 1995; Skelton, 2004). H. Cox and Holmes (2000) found that rural residents, who experienced a bushfire, held a similar relationship to their place, despite the individual variability in the degree of impact and loss. They suggested that the natural environment, and its transformation following the bushfire, was intrinsic to the identity of the community as a whole. This may provide an explanation of the very weak relationship between variation in the degree of impact experienced by individuals and their current attachment to the natural environment. Further to this, both length of residence and marital status were strongly associated with a change in attachment to one's community whereas impact did not contribute to this relationship. This suggests that those who have lived in the community a longer period of time and who were not partnered reported less participation in and/or less interest in their community at the time of questionnaire completion, compared to before the fires. Moreover, marital status was found to be a strong predictor of depression, i.e. residents who were not partnered reported greater symptoms of depression. It seems likely that residents who reported greater symptoms of depression were less likely to be interested and/or involved in their community because loss of interest in activities is one of the symptoms of depression. This conclusion is drawn with caution because, on average, respondents reported that they 76 neither agreed nor disagreed when responding to a change in attachment to their community. This may suggest ambivalence when responding to whether they participate more or are more interested in their community at the time of completing the questionnaire, compared with before the wildfire occurred. In addition, retrospective accounts and assessing change in attachment is difficult (Aseltine et al., 1995; Nieuwkerk et al., 2007) and, thus, was used for ' r descriptive purposes only. Impact from the McLure Fire and Symptoms of Depression The results provide support of the relationship between impact from the 2003 wildfire and current depressive symptoms. In the present investigation, the greater extent to which residents were impacted by the wildfire was associated with greater symptoms of depression among respondents, after controlling for age and marital status and community. Moreover, the younger the respondent and those not partnered reported greater symptoms of depression. Several studies have attributed the impact of a disaster to increased symptom levels of depression immediately following a disaster (Ginexi et al., 2000; Jones et al., 2003; Norris et al., 2002b; Tyler & Hoyt, 2000). These studies found that exposure to the disaster and extent of impact was the strongest predictor of postdisaster symptoms of depression, second to predisaster symptoms of depression. However, few researchers have explored longer-term postdisaster consequences and the relationship with the degree of impact from the disaster. Depression has been observed to show a curvilinear relationship with age, yet there is empirical evidence and consistent with the present study, that depression is consistently elevated up to 4 years following a disaster (McFarlane, Clayer, & Bookless, 1997; Norris, Perilla, Riad, Kaniasty, & Lavizzo, 1999; Ohta et al., 2003). The present investigation contributes to the post-disaster literature in that 3.5 years following the disaster, impact from the disaster, above and beyond the effects of marital status and age, contributed to symptoms of depression. The community in which respondents resided was not related to the criterion variables in the H M R analyses that addressed the research questions. However, A N O V A revealed that those who lived in Louis Creek and the surrounding area reported greater symptoms of depression than respondents of Barriere. Moreover, 41% of the Louis Creek respondents scored above the cutoff depression score, whereas only 14% of Barriere respondents scored above the cutoff score. When examining the demographic information, more retirees and individuals live alone in Louis Creek (47.5%> retired, 14.8% live alone), compared with Barriere (34.9% retired, 9.3% live alone), possibly explaining the significance of marital status in the relationship between impact and depression. Interpretation of these results must proceed with caution. First, the CES-D is not meant to be a diagnostic tool for depression; it simply indicates the presence of depressive symptoms. Second, only one measure was used to assess depression and other correlates,' such as health status and social support, were not considered. Other possible outcomes include psychosocial resource loss (Norris et al., 1999), impairment in quality of life (Wang et al., 2000), and increased stress levels (Maida, Gordon, Steinberg, & Gordon, 1989). Furthermore, other variables including chronic and situational stressors, may be contributing to depression. Because screening for pre-existing depressive symptoms prior to the fire was not conducted, we cannot know how much of the depressive symptoms were specifically related to the impact of the fire. Place Attachment and Depression The results support the relationship between place attachment and depression. Attachment to the natural environment was weakly associated with symptoms of depression, 78 once the variance associated with marital status, age, and impact from the fire were controlled for in depressive symptoms. Respondents, who reported a lesser degree of attachment to the natural environment, reported greater symptoms of depression. Attachment to the social and built environments did not statistically significantly contribute to symptoms of depression. Symptoms of depression include decreased energy, decreased interest and pleasure in activities, and cognitive distortions, clouding an individual's judgment and perspective of their intimate surroundings. Thus, respondents may be continuing to reorient (i.e., recreate and rebuild attachment) to their surrounding natural environment, a process discussed by R. Cox (2006). The findings are consistent with the literature on ecopsychology that posits that nature has psychological benefits and contributions to individual well-being (Kaplan & Kaplan, 1989; Kaplan, 1995). Furthermore, the surrounding hillsides in Louis Creek and Barriere (see Figure 1) are still stark reminders of the wildfire, and one could speculate that this reminder may decrease respondents' feelings of attachment to the natural environment, and increase their depressive symptoms. These findings further support the addition of attachment to the natural environment in the conceptual framework of place attachment in rural communities that have been impacted by a natural disaster. The fact that attachment to the social community did not significantly contribute to depressive symptoms is inconsistent with O'Brien et al. (1994), whose findings showed that community attachment (a construct similar to place attachment, yet is specific to attachment to the community in general) was a significant contributor to depression in two American rural communities. However, their measure of community attachment included only four items, reflecting cognition, which differed from the items on the social attachment subscale of the present study. The lack of a relationship between social attachment and depressive 79 symptoms in the present study is also inconsistent with the literature on place attachment, which indicates that place attachment is strongly associated with well-being, and a disruption to this bond can create emotional distress (Fried, 1963; Fullilove, 1996; Korplea, 1989; Relph, 1976; Twigger-Ross & Uzzell, 1996; West, 2003). This literature is based on qualitative accounts of place attachment, whereas the present study attempted to assess degree of attachment 3.5 years post disruption. Considering the respondents, on average, reported a positive attachment to their community (social, natural, and built environments), and just over two-thirds of the sample did not report depression, it is possible that the measure of depression I used did not capture the emotional and/or mental well-being that is present in this sample. Moreover, a measure of subjective well-being or quality of life may have resulted in different findings because attachment to the social and natural environments may be more related to well-being in these particular communities, rather than depression. However, a measure of depression was chosen for the present study to be consistent with the disaster research and literature. Although symptoms of depression still existed, 3.5 years following the disaster, residents have had time to recreate and rebuild their attachments to their community, a process that begun immediately and up to 2 years following the McLure Fire (R. Cox, 2006). Strengths and Limitations The present study extends disaster recovery research by contributing to our understanding of the long-term consequences of a wildfire on depressive symptoms and respondents' attachment to their rural community. It also adds to the literature on place attachment by identifying various dimensions and modalities of attachment and exploring the 80 relationship between place attachment and depression, via survey methodology and the development of attachment and impact instruments. The present study also contributes to our knowledge of the psychological and symbolic losses attributable to wildfires, in contrast to the predominant media focus on economic and material losses (R. Cox, Long, Jones, & Handler, 2006). This skewed portrayal of recovery to a natural disaster contributes to the tendency for disaster relief and emergency management agencies to address the more practical aspects of the recovery process, while failing to acknowledge the symbolic meanings of the losses and residents' attachments to their home and community (R. Cox et a l , 2006). It is important for the disaster relief and recovery community to gain insight and understanding into how "place attachments provide anchors of meaning in residents' lives by symbolizing and sustaining self, family, or home" (Brown & Perkins, 1992, p. 298). A positive attachment, as found in the present study, implies a willingness to invest in the community, thus contributing to the overall community's well-being (Christakopoulou, Dawson, & Gari, 2001). This exploratory study does not provide definitive answers; but provides a perspective on residents' attachment to place more than 3 years after a disaster. The theoretical and practical implications of the role place attachment plays in individual and community well-being are of particular relevance to the field of Counseling Psychology. To date, attachment theory has been the focus of counseling research and practice (e.g., Bartholomew & Thompson, 1995; Lopez, 1995). This theory highlights the existence of affectional ties between an individual and their primary caregiver (or the individual in closest proximity). Environmental psychologists, have argued that individuals develop similar bonds with an intimate place and are also a fundamental need (Chawla, 1992; 81 Low & Altman, 1992). Similar to attachment theory, a disruption to residents' process of attachment to their social network, the natural environment, and the built environment may play a crucial role in their mental health. Although not addressed in the present study, the findings do not preclude the notion that place attachment potentially acts as a mediator between traumatic stress and psychological well-being, another concern of counseling practice (Meyer & Ponton, 2006). The psychological benefits of a deep emotional bond with one's immediate environment have yet to be empirically explored. Thus, this study brings to the forefront a construct that may play a role in various aspects of mental health and well-being, both in general and within the context a natural disaster. Research on natural disasters focuses on crisis counseling during and immediately following the event (e.g., McCarthy & Butler, 2003), as well as directly treating depression and anxiety (Norris et al., 2002b). Yet there is a lack of clinical knowledge in aiding individuals and communities over the long-term. Consideration should be given to aiding residents in re-creating attachments to their home and place, and to think about possible interventions to facilitate this process. In addition to the above contributions, several limitations present themselves throughout this study. First, because the study was conducted 3.5 years following the McLure Fire, responses reflecting the impact of the fires are limited to retrospective accounts and self-report measures, possibly creating a response bias. I attempted to minimize this limitation by asking mostly factual objective questions regarding the initial impact of the wildfire, rather than subjective questions. Second, sampling was purposive, not random, limiting generalizability to the population. By relying on volunteers to respond to the survey, the results are ultimately limited to those who were willing to respond to research conducted on the McLure Fire. Also, residents who were attached to their community may be more 82 likely to respond to such a survey with the idea that it would support the community. Further to this, the results are also limited to those who remained in the communities following the fires. A small percentage of individuals and families, 9% in Louis Creek and 4% in Barriere, (Rob Rutten, personal communication, March 18, 2007) were either forced to relocate for employment or chose to relocate. Third, the scales I developed to determine the impact of the fires and place attachment were created for the purpose of this project, though many of the items were drawn from developed scales. Although the reliability estimates of these questionnaires and their subscales were good, and limited construct validity was examined through factor analytic procedures (i.e., principal components analysis), no causal inferences can be made from the results of the present study because of its exploratory nature and the fact that only relationships among variables were examined (i.e. the results gathered were correlational rather than causal). I was also limited by a lack of information on the pre-existing emotional and mental health in the communities. Therefore, I cannot attribute current symptoms of depression to the disaster itself (Seplaki, Goldman, Weinstein, &.Lin, 1999). Fourth, only one scale was used to measure symptoms of depression. It would have been helpful to use additional constructs such as well-being or social support to determine other outcomes of the fire's impact. However, based on financial and time constraints and ensuring a large enough sample," a decision was made to limit the length of the survey. It is important to note that, upon visiting the communities, speaking with residents, and after viewing the questionnaire responses, the difficulty in distinguishing between residents from Louis Creek and Barriere became apparent. This is congruent with R. Cox's (2006) study. Many individuals determine their community based on the location of their 83 postal box. Furthermore, there are specific areas within both Louis Creek and Barriere that residents identify with. Dixon Valley, Squam Bay, and Dunn Lake Road are three such areas that many residents consider their community, despite them not being geographical communities. Other responses came from residents in nearby communities such as McLure and Darfield and from residents, who were living in the communities during the McLure Fire, yet currently reside in another town close by (i.e., Kamloops). These surveys were included in the study as the individuals still considered Louis Creek and Barriere as their own and express strong connections through social networks and amount of time still spent in the communities. Many of these responses fell under the category of other, and were combined with Louis Creek responses for the final analysis. Suggestions for Further Research With limitations come suggestions for further research. Because of the exploratory nature of this study, a replication of the study is necessary prior to developing implications for practice. Sampling a rural community of similar size and status, that either has experienced a wildfire or not, would provide a means of comparison for the communities in the present study. Furthermore, comparing this sample to that from an urban community, one that also experienced a wildfire disaster would provide us with additional information on the effect of community size, density, and type (Goudy, 1990; Kasarda & Janowitz, 1974). Kelowna, British Columbia, is one such city where preliminary exploratory work was conducted prior to carrying out the present study. The Okanagan Mountain Fire devastated two communities within Kelowna in the same season as the McLure Fire. Preliminary findings suggested that the recovery process experienced by the residents from the rural communities in the present study and Kelowna were vastly different. For example, the areas 84 in Kelowna that were impacted by the fires were of a significantly higher socioeconomic background, had more financial support, and were provided greater attention. In order to gain a deeper understanding of the personal meanings residents attribute to their social, natural, and built community, utilizing a qualitative methodology (e.g., focus groups, grounded theory) specifically exploring the construct of place attachment in these communities, would further strengthen the findings of this survey. Autobiographical insideness, a term coined by Rowles (1990), is central to understanding the layers of meaning and memories that have developed over time and contribute to an individual's sense of identity with their community. Interviewing individuals and families who were forced to relocate from the communities would also provide further insight into the construct of place attachment. In addition, more robust statistical procedures regarding the development of an attachment measure would further advance the potential for inquiry into the construct of place attachment. The present investigation demonstrated that attachment to the social, natural, and built environments were distinct and modestly overlapping dimensions of place attachment, and that the modalities of affect, cognition, and behavior contributed to this attachment. Evidence of construct validity was provided based on the,factor analysis (i.e., principal components analysis) results and the expected relationships with impact, depression, and attachment. However, where a relationship did not exist, it may be explained by the fact that the questionnaire needs stronger construct validity to illustrate a relationship. Further research should focus on identifying and measuring multiple dimensions of attachment. Thus, further instrument development and evidence of construct validity is needed in order to increase our confidence that we are testing the construct of place 85 attachment. This involves the following: (a) establish norms (obtain a community sample as a control group) to aid in establishing a baseline in which to compare scores; (b) consult with experts in the field of place attachment and person-environment relationships to improve upon the content validity of the measure; (c) compare with scores from related and unrelated measures (i.e., sense of community, sense of place, sense of belonging, social cohesion, residential satisfaction) to assess the practical value of this measure. Conclusion In conclusion, the results of the present study provide direction for future research that may contribute to the understanding of long-term consequences of the impact of a wildfire on residents from rural communities. The findings show that impact contributes to symptoms of depression and place attachment, specifically attachment to an individual's immediate built environment, more than 3 years following a natural disaster. This suggests that place attachment should be taken into consideration when assessing impact from a disaster. In addition, a relationship between place attachment, more specifically attachment to the natural environment, and symptoms of depression, suggest a need to further assess the potential role place attachment plays in individual and community well-being. In addition, the impact to the natural environment on residents should be considered in both impact and place attachment measures. These findings also support the multidimensional nature of place attachment (i.e., attachment to the social, natural, and built environments) in a rural community that has experienced a wildfire, and contributes to the development of an instrument to assess this construct, incorporating the three modalities of place experience -affect, cognition, and behaviour. 86 REFERENCES Andresen, E .M. , Malmgren, J.A., Carter, W.B., & Patrick, D.L. (1994). Screening for depression in well older adults: Evaluation of a short form of the CES-D (Center of Epidemiologic Studies Depression Scale). American Journal of Preventative Medicine, 10, 77-84. Aseltine, R.H., Carlson, K.J . , Fowler, F.J., & Barry, M.J. (1995). Comparing prospective and retrospective measures of treatment outcomes. Medical Care, 33(4), AS67-AS76. Bartholomew, K. , & Thompson, J .M. (1995). The application of attachment theory to counseling psychology. The Counseling Psychologist, 23(3), 484-490. Beckley, T .M. (1998). The nestedness of forest dependence: A. conceptual framework and empirical exploration. Society & Natural Resources, 11(2), 101-121. Beckley, T .M. (2003). The relative importance of sociocultural and ecological factors in attachment to place. In L.E. Kruger (Ed.), Understanding community-forest relations (Gen. Tech. Rep. PNW-GTR-566, pp. 105-126.) Portland, OR: U.S. Department of Agriculture. Retrieved July 15, 2006 from http://www.forestry.umt.edu/departments/socn/publications/McCoolCommunityPaper •pdf Belk, R.W. (1992). Attachment to possessions. In I. Altman, & S.M. Low (Eds.), Place attachment (pp. 37-62). New York: Plenum. Brehm, J.M., Eisenhauer, B.W., & Krannich, R.S. (2004). Dimensions of community attachment and their relationship to well-being in the amenity-rich rural west. Rural Sociology, 69(3), 405-429. Bricker, K., & Kerstetter, D. (2000). Level of specialization and place attachment: An exploratory study Of Whitewater recreationists. Leisure Sciences, 22, 233-257. Brown, B.B., & Perkins, D.D. (1992). Disruptions in place attachment. In I. Altman, & S.M Low (Eds.), Place attachment (pp. 279-304). New York: Plenum. Buckner, J.C. (1988). The development of an instrument to measure neighbourhood cohesion. American Journal of Community Psychology, 16(6), 771-791. Casey, E. (1993). Getting back into place: Toward a renewed understanding of the place-world. Indiana: University Press. ' Chawla, L. (1992). Childhood place attachments. In I. Altman, & S.M. Low (Eds.), Place attachment (pp. 63-86). New York: Plenum. 87 Christakopoulou, S., Dawson, J., & Gari, A . (2001). The community well-being questionnaire: Theoretical context and initial assessment of its reliability and validity. Social Indicators Research, 5<5, 321-351. Cox, R.S. (2006). Echoes, transgressions, and transformations: Identity reorientation and the discourse of disaster recovery. Unpublished doctoral dissertation, University of British Columbia, Vancouver, Canada. Cox, H.M. , & Holmes, C A . (2000). Loss, healing, and the power of place. Human Studies, 23,63-78. Cox, R.S., Long, B.C., Jones, M.I., & Handler, R.J. (2006). Sequestering of suffering: Critical discourse analysis of natural disaster media coverage. Manuscript submitted to the Journal of Health Psychology-Cross, J.E. (2003, July). Conceptualizing community attachment. Paper presented at the annual meeting of the Rural Sociological Society, Montreal, Quebec, Canada. Retrieved August 13, 2006 from http://www.ruralsociology.org/annual-meeting/2003/CrossJennifer.pdf Cross, J.E. (2004, August). Improving measures of community attachment. Paper presented at the annual meeting of the Rural Sociological Society, Sacramento, CA. Retrieved August 13, 2006 from http://www.ruralsociology.org/annual-meeting/2004/Cross.pdf Cuba, L., & Hummon, D . M . (1993). A place to call home: Identification with dwelling, community, and region. The Sociological Quarterly, 34, 111-131. Davidson, W.B & Cotter P R (1986). Measuring sense of community within the sphere of the city. Journal of Applied social psychology, 16(1), 608 - 618. Dillman, D.A. (2000). Mail and Internet surveys: The tailored design method. New York: John Wiley. Eisenhauer, B.W. (2000). Attachments to special places on public lands: A n analysis of activities, reason for attachments, and community connections. Society and Natural Resources, 75,421-441. Erikson, K.T. (1976) Everything in its path: Destruction of community in the Buffalo Creek Flood. New York: Simon and Schuster. Feldman, R . M . (1990). Settlement-identity: Psychological bonds with home places in a mobile society. Environment and Behavior, 22(2), 183-229. Filmon, G. (2003). Firestorm 2003: Provincial Review. Prepared for the Province of British Columbia. Retrieved October 31, 2006 from http://www.2003firestorm.gov.bc.ca/ Fowler, F.J., Jr. (2002). Survey research methods (3 r d ed.). Thousand Oaks, C A : Sage. 88 Fried, M . (1963). Grieving for a lost home. In Duhl, L.J. (Ed.). The urban condition: People and policy in the metropolis (pp; 151-171). New York: Basic Books. Fried, M . (2000). Continuities and discontinuities of place. Journal of Environmental Psychology, 20, 193-205. Fullerton, C.S. & Ursano, R.J. (2005). Psychological and psychopathological consequences of disasters. In J.J. Lopez-Ibor, G. Christodoulou, M . Maj, N . Sartorius, & A. Okasha (Eds.), Disasters and mental health, (pp. 13-36). England: John Wiley. Fullilove, M.T. (1996). Psychiatric implications of displacement: Contributions from the psychology of place. American Journal of Psychiatry, 153, 1516-1523. Garson, D.G. (2007). Statsnotes: Topics in Multivariate Analysis. Retrieved August 3, 2007, from North Carolina State University Web site: http://www2.chass.ncsu.edu/garson/pa765/statnote.htm, Ginexi, E .M. , Weihs, K. , & Simmens, J. (2000). Natural disaster and depression: A prospective investigation of reactions to the 1993 Midwest Floods, American Journal of Community Psychology, 28(4), 495-518. Gitlin, L.N.(2000). Adjusting person-environment systems: Helping older people live the "good life" at home. In R.L. Rubinstein, M . Moss, & M . H . Kieban (Eds.),.The many dimensions of aging (pp. 41-53). New York: Springer. Giuliani, M . V . (1991). Towards an analysis of mental representations of attachment to the home. Journal of Architectural and Planning Research,8, 133-146. Giuliani, M.V . (2003). Theory of attachment and place attachment. In M . Bonnes, T. Lee, & M . Bonaiuto (Eds.,), Psychological theories for environmental issues, (pp. 137-170). England: Ashgate. Goudy, W.J. (1990). Community attachment in a rural region. Rural Sociology, 55, 178-198. Hammitt, W.E., Backlund, E.A., & Bixler, R.D. (2006). Place bonding for recreation places: Conceptual and empirical development. Leisure Studies, 25(1), 17-41. Henwood, K. & Pidgeon, N . (2001). Talk about woods and trees: threat of urbanization, stability, and biodiversity. Journal of Environmental Psychology, 21, 125-147. Hidalgo, M.C. , & Hernandez, B (2001). Place attachment: Conceptual and empirical questions. Journal of Environmental Psychology, 21, 273-281. Hummon, D . M . (1992). Community attachment: Local sentiment and sense of place. In I. Altman, & S.M. Low (Eds.), Place attachment (pp. 165-185). New York: Plenum. 89 Husaini, B.A. , Neff, J.A., Harrington, J., Hughes, M . , & Stone, R.H. (1980). Depression in rural communities: Validating the CES-D Scale, Journal of Community Psychology, 5(1), 20-27. Jones, R.T., Ribbe, D.P., Cunningham, P., & Weddle, J.D. (2003). Psychosocial correlates of wildfire disaster: Post disaster adult reactions. Fire Technology, 39, 103-117. Jorgensen, B.S., & Stedman, R.C. (2001). Sense of place as an attitude: Lakeshore owners attitudes toward their properties. Journal of Environmental Psychology, 21, 233-248. Kaltenborn, B.P. (1998). Effects of sense of place on responses to environmental impacts: A study among residents in Svalbard in the Norwegian high Arctic. Applied Geography, 18(2), 169-189. Kaplan, R., & Kaplan, S. (1989). The experience of nature: A psychological perspective. Cambridge: Cambridge University Press. Kaplan, S. (1995). The restorative benefits of nature: Toward an integrative framework. Journal of Environmental Psychology, 15, 169-182. Kasarda, J.D., & Janowitz, M . (1974). Community attachment in mass society. American Sociological Review, 39, 328-339. Kohout, F.J., Berkman, L.F., Evans, D.A., & Cornoni-Huntley, J. (1993). Two shorter forms of the CES-D (Center for Epidemiological Studies Depression) depression symptoms index. Journal of Aging and Health, 5(2), 179-193. Korpela, K . M (1989). Place-identity as a product of environmental self-regulation. Journal of Environmental Psychology, 9, 241-256. Korpela, K . M . , & Hartig, T. (1996). Restorative qualities of favourite places. Journal of Environmental Psychology, 16, 221-233. Lalli, M . (1992). Urban-related identity: Theory, measurement, and empirical findings. Journal of Environmental Psychology, 12, 285-303. Lopez, F.G. (1995). Contemporary attachment theory: An introduction with implications for counseling psychology. The Counseling Psychologist, 23(3), 395-415. Lopez-Ibor, J.J.(2005). What is a disaster. In J.J. Lopez-Ibor, G. Christodoulou, M . Maj, N . Sartorius, & A . Okasha (Eds.), Disasters and mental health (pp. 1-11). England: John Wiley. Low, S.M. (1992). Symbolic ties that bind: Place attachment in the plaza. In I. Altman, & S.M. Low (Eds.), Place attachment (pp. 165-186). New York: Plenum. 90 Low, S.M., & Altman, I. (1992). Place attachment: A conceptual inquiry. In I. Altman, & S.M. Low (Eds.), Place attachment (pp. 1-12). New York: Plenum. Maida, C.A., Gordon, N.S., Steinberg, A. , & Gordon, G. (1989). Psychosocial impact of disasters: Victims of the Baldwin Hills fire. Journal of Traumatic Stress, 2(1), 37-48. Maltais, D., Lachance, L. , Simard, A. , Brassard, A. , & Picard, L . (2002); Difficult.es et effets a long terme d'une catastrophe en milieu rural: Resultats d'une etude combinant les approches qualitative et quantitative, Revue Quebecoise de Psychologie, 23(2), 197— 217. , Manzo, L.C. (2003). Beyond house and haven: Toward a revisioning of emotional, relationships with places. Journal of Environmental Psychology, 23, Al-61. Marshall, B.K. , Picou, J.S., & Bevc, C.A. (2005). Ecological disaster as contextual transformation: Environmental values in a renewable resource community. Environment and Behavior, 37(5), 706-728. Mayer, F.S., & Frantz, C M . (2004): The connectedness to nature scale: A measure of individuals' feeling in community with nature. Journal of Environmental Psychology, 24,503-51.5. McCarthy, M.A. , & Butler, L. (2003). Responding to traumatic events on college campuses: • A case study and assessment of student postdisaster anxiety. Journal of College Counseling, 6, 90-96. McFarlane, A.C. , Clayer, J.R., & Bookless, C.L. (1997). Psychiatric morbidity following a natural disaster: An Australian bushfire. Social Psychiatry and Psychiatric Epidemiology, 32, 261-268. Meyer, D. & Ponton, R. (2006). The healthy tree: A metaphorical perspective of counselor well-being. Journal of Mental Health Counseling, 28(3), 189-201. Miles, J., & Shevlin, M . (2001). Applying regression and correlations: A guide for students and researchers. Thousand Oaks, C A : Sage. Ministry of Forests Protection Branch (2003). Fire review summary for the McLure Fire (K202 72). Retrieved January 27, 2006, from http://www.for.gov.bc.ca/Protect/reports/2003Review/McLure_Fire_Review_K20272 .pdf Nardi, P .M. (2003). Doing survey research: A guide to quantitative methods. Boston, M A : Pearson Education. 91 Natural Resources Canada. (2004, June'09). Wildland-urban interface: General overview. Retrieved December 22, 2006, from http://fire.cfs.nrcan.gc.ca/research/management/wui_e.htm Nieuwkerk, P.T., Tollenaar, M.S., Oort, F.J., & Sprangers, M.G. (2007). Are retrospective measures of change in quality of life more valid than prospective measures? Medical Care, ¥5,199-205. Norris, F.H., Friedman, M.J. , & Watson, P.J. (2002a). 60,000 disaster victims speak: Part II. Summary and implications of the disaster mental health research. Psychiatry, 65(3), 240-260. Norris, F. H., Friedman, M . J., Watson, P. J., Byrne, C. M . , Diaz, E., & Kaniasty, K . (2002b). 60,000 Disaster victims speak: Part I. An empirical review of the empirical literature 1981-2001. Psychiatry, 65, 207-239. Norris, F.H., Perilla, J.L., Riad, J.K., Kaniasty, K., & Lavizzo, E.A. (1999). Stability and change in stress, resources, and psychological'distress following natural disaster: Findings from Hurricane Andrew. Anxiety, Stress, and Coping, 12, 363-396. O'Brien, D.J., Hassinger, E.W., & Dershem, L. (1994). Community attachment and depression among residents in two rural Midwestern communities. Rural Sociology, 59(2), 255-265. Ohta, Y. , Araki, K. , Kawasaki, N . , Nakane, Y. , Honda, S., & Mine, M . (2003). Psychological distress among evacuees of a volcanic eruption in Japan: A follow-up study. Psychiatry and Clinical Neurosciences, 57, 105-111. Pretty, G.H., Chipuer, H . M . , & Bramston, P. (2003). Sense of place amongst adolescents and adults in two rural Australian towns: The discriminating features of place attachment, sense of community and place dependence in relation to place identity. Journal of Environmental Psychology, 23, 273-287. Proshansky, H .M. , Fabian, A . K . , & Kaminoff, R. (1983). Place identity: Physical world socialization of the self. Journal of Environmental Psychology, 3, 57-83. Public Health Agency of Canada. (2002, October 03). A report on mental illnesses in Canada. Retrieved December 23, 2006, from http://www.phac-aspc.gc.ca/publicat/miic-mmac/chap_2_e.html. Radloff, L.S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385-401. Riad, J. & Norris, F. (1996). The influence of relocation on the environmental, social, and psychological stress experienced by disaster victims. Environment and Behavior, 25,, 163-182. 92 Rice, T.W., & Steele, B. (2001). White ethnic diversity and community attachment in small towns. Social Science Quarterly, 82(2), 397-407. Riger, S., & Lavrakas, P.J.(1981). Community ties: Patterns of attachment and social interaction in urban neighbourhoods. American Journal of Community Psychology, 9(1), 55-66. Relph, E. (1976). Place andplacelessness. London: Pion. Robinson, D., & Wilkinson, D. (1995). Sense of community in a remote mining town: Validating a neighbourhood cohesion scale. American Journal of Community Psychology, 23(1), 137-148. Rogan, R., O'Connor, M . , & Horwitz, P. (2005). Nowhere to hide: Awareness and perceptions of environmental change, and their influence on relationships with place. Journal of Environmental Psychology, 25, 147-158. Roszak, T., Gomes, M.E. , & Kanner, A D . (Eds.) (1995). Ecopsychology: Restoring the earth, healing the mind. San Francisco: Sierra Club Books. Rowles, G.D. (1990). Place attachment among small town elderly. Journal of Rural Community Psychology, 11(1), 103-120. Rubinstein, R.L., & Parmelee, P.A. (1992). Attachment to place and the representation of the life course by the elderly. In I. Altman, & S.M. Low (Eds.), Place attachment (pp. 139-164). New York: Plenum. Sampson, R.J. (1988). Local friendship ties and community attachment in mass society: A multilevel systemic model. American Sociological Review, 53(5), 766-779. Scher, C D . (2003). Fire Impact Questionnaire. Unpublished measure. Scher, C D . , & Ellwanger, J. (2005, November). Negative cognitions and disaster impact are related to emotional and physical health following the 2003 southern California wildfires. Poster presented at the annual meeting of the Association for Behavioral and Cognitive Therapies, Washington, DC. Schwarz, B., Brent, R., Phillips, R., & Danley, M . (1995). Disruption in place attachment following the 1993 Big Flood. In Nasar, J., Grannis, P., & Hanyu, K. (Eds.), Environmental design research: Evolutionary and revolutionary changes (pp. 92-98). Oklahoma City, OK: The Environmental Design_Research_Association. 93 Seamon, D. (2000). A way of seeing people and place: Phenomenology in environment-behavior research. In S. Wapner, J. Demick, T. Yamamoto, & H . Minami (Eds.), Theoretical perspectives in environment-behavior research (pp! 157-178). New York: Plenum. Seplaki, C.L., Goldman, N . , & Weinstein, M . (1999). Before and after the 1999 Chi-Chi Earthquake: Traumatic events and depressive symptoms in an older population, Office of Population Research Princeton University, Working Paper No. 2003-02. Retrieved October 31, 2006 from http://opr.princeton.edu/papers/opr0302.pdf Shamai, S. (1991). Sense of place: An empirical measurement. Geoforum, 22(3), 347-358. Shrout, P.E., & Yager, T.J. (1989). Reliability and validity of screening, scales: Effect of reducing scale length. Journal of Clinical Epidemiology, 42, 69-78. Shumaker, S.A., & Taylor, R.B. (1980). Toward a clarification of people-place relationships: A model of attachment to place. In N.R. Feimer, & E.S. Geller (Eds.), Environmental psychology: Directions and perspectives (219-251). New York: Praeger. Skelton, T. (2004). Montserratian narratives of belonging in the midst of disaster and relocation. In J. Besson, & K.F. Olwig (fids.), Caribbean narratives of belonging (pp.77-96). Oxford, U K : MacMillan. Statistical Package for the Social Sciences. (Version 12.0, for Windows) [Computer software]. (2003). Chicago, IL: SPSS. Statistics Canada (2001a). 2001 Census of Canada: Community profiles. Retrieved December 22, 2006 from: http://wwwl2.statcan.ca/english/profil01/CP01/Index.cfm?Lang=E Stedman. R.C. (2003). Is it really just a social construction?: The contribution of the physical environment to sense of place. Society and Natural Resources, 16, 671-685. Tabachnick, B.G., & Fidell, L.S. (1996). Using multivariate statistics (3rd ed.). New York: HarperCollins. Theodori, G.L. (2001). Examining the effects of community satisfaction and attachment on individual well-being. Rural Sociology, 66(4), 618-628. Theodori, G.L., & Luloff, A .E . (2000). Urbanization and community attachment in rural areas. Society and Natural Resources, 13, 399-420. Tremblay, M.A. , Blanchard, C M . , Pelletier, L . C , & Vallerand, R.J. (2006). A dual route in explaining health outcomes in natural disaster, Journal of Applied Social Psychology, 36(6), 1502-1522. 94 Twigger-Ross, C.L. & Uzzell, D.L. (1996). Place and identity processes. Journal of Environmental Psychology, 16, 205-220. Tyler, K . A . & Hoyt, D.R. (2000). The effects of an acute stressor on depressive symptoms among older adults: The moderating effects of social support and age. Research on Aging, 22(2), 143-164. U.S. Department of Health and Human Services. (2006). NIOSH health hazard evaluation report (HETA #2006-0023-3003). New Orleans, L A : Tak, S., & Dowell, C.H. Retrieved October 31, 2006, from http://0-wvvw.cdc.gov.milll.sjlibrary.org/niosh/hhe/reports/pdfs/2006-0023-3003.pdf VanVoorhis, C.W., & Morgan, B.L . (2001). Statistical rules of thumb: What we don't want to forget about sample sizes. Psi Chi Journal of Undergraduate Research, 6(A). Retrieved January 10, 2007, from http://www.psichi.org/pubs/articles/article_182.asp Vorkinn, M . , & Riese, H. (2001). Environmental concern in a local context: The significance of place attachment. Environment and Behavior, 33(2), 249-263. Wang, X . , Gao, L. , Zhang, H. , Zhao, C., Shen, Y. , & Shinfuku, N . (2000). Post-earthquake quality of life and psychological well-being: Longitudinal evaluation in a rural community sample in northern China. Psychiatry and Clinical Neurosciences, 54, 427-433. • Weisman, G.D., Chaudhury, H. , & Moore, K.D. (2000). Theory and practice of place: Toward an integrative model. In R.L. Rubinstein, M . Moss, & M . H . Kleban (Eds.), The many dimensions of aging (pp. 3-21). New York: Springer Publishing. West, R.A. (2003). Place psychological experience: A Native American perspective. Dissertation Abstracts International, 63, 10B. Williams, D.R. (2000). Notes on measuring recreational place attachment. Retrieved June 17, 2006 from http://www.fs.fed.us/rm/value/docs/pattach_notes.pdf Williams, D.R., Patterson, M.E. , Roggenbuck, J.W., & Watson, A .E . (1992). Beyond the commodity metaphor: Examining emotional and symbolic attachment to place. Leisure Sciences, 14, 29-46. Williams, D.R. & Vaske, J.J. (2003). The measurement of place attachment: Validity and generalizability of a psychometric approach. Forest Science, 49(6), 830-840. 95 APPENDIX A : REGIONAL MAPS 96 97 APPENDIX B: QUESTIONNAIRE Project Title! IMPACT FROM A WILDFIRE, COMMUNITY ATTACHMENT, AND WELL-BEING Contents| A. Introduction B. Wildfire 2003 C. Your Community Today D. Your Health E. Background Information This project is supported in part by the Social Sciences and Humanities Research Council of Canada, and has been approved by the University of British Columbia Behavioral Ethics Committee. 98 SECTION A: INTRODUCTION | Thank you for taking the time to answer this questionnaire. Your participation in this project is very much appreciated. Remember; • The responses you provide will be kept strictly confidential. • There are no wrong answers. • Please take the time to respond to each question. If you are unsure of your answer, do your best to choose the response that seems to fit closest to how you are feeling or thinking at the time. • When applicable, please put a check mark ( V ) next to, or below the response choice that best fits your experience. ***Please place the completed questionnaire in the sealed and stamped envelope, and drop it in the mail*** It will take approximately 10-20 minutes to complete this questionnaire. SECTION B: WILDFIRE 2003 The following questions are intended to tell us something about your experience during or immediately after the fire of 2003. Below is a list of potential losses exper ienced by residents as a result of the fire. P lease indicate how you would a s se s s the extent of loss to each of the following (P lease check (V) the appropriate box for each question): No Loss Minor Loss Moderate Major Loss Total Loss Loss 1. FamiK members and'or friends moved a\\a\ from the communitv. 2. Pels .3. Livestock 4. Job or business 5..Damage to my house -6. Damage to my property 7. Damage to my immediate neighbours' house ¥111111! IfBSfi* 8. Damage to my immediate neighbours' properties ; 9. Job or business loss of family members' and/or friends'. ^BjsisiillllliJ Below is a list of exper iences you may have had during or immediately after the fire. P lease indicate to what extent you exper ienced each of the following (P lease check (V) the appropriate box for each question): Not at all Very Little Somewhat To a To a Very Considerable Large Extent Extent 10.1 had house insurance at the time of the fires. 11.1 experienced involuntary disruption to my work. 12. Up to one \ear following the fire, I experienced adverse.physical ;health effects due to the fire, (e.g., asthma flare-ups, difficulty breathing, runin eyes or nose. ,headaches) iiillllifiti^ili 100 Not at all Very Little Somewhat To a To a Very Considerable Large Extent Extent 13. Up to one year following the fire, I visited a doctor, hospital, or clinic due to fire-related health problems. 14. Up to one year following the lire. I experienced mental and or emotional difficulties due to the lire.(e.g.. depressed, feelings of stress or an\iet\. difficult) concentrating) 15. Up to one year following the fire, I visited a mental health professional due to fire-related mental and/or emotional difficulties. 16. In the areas where I enjoyed ... outdoor recreational activities, the natural em ironmeni changed as a consequence of the fire. pSil9^Blli|i 17. As a consequence of the fire, the view from my house changed. 18. At the time of the lire, the surrounding natural emironment changed. ^lll^ilidl 19.1 received emotional and moral support from my community. 20..I received emotional and moral-support from disaster-related services. ii^hiisS 101 SECTION C: YOUR COMMUNITY TODAY This section asks you about your history in the community and the current thoughts and feelings you have about the community in which you presently reside: Please check the geographical community your lived in when the fire occurred. Also indicate which community you presently reside in? 21. During the fires: • Louis Creek • Barriere • Other 22. Presently reside in: • Louis Creek • Barriere • Other 23. Are your parents, grandparents, and/or ancestors from the community you presently reside in? (Check V all that apply.) • Parents • Grandparents • Ancestors • None of the above Below is a list of statements that descr ibe you current exper ience of your community. P lease indicate to what degree you agree or disagree with the following statements (P lease check (V) the appropriate box for each question): Strongly Disagree Neither Agree Agree Strongly Disagree nor disagree Agree 24. I would be sorry to leave this . community. 25.1 feel at home in this community. 26. I am interested in knowing what noes on in this communitv. I l l l l l l l l l l IHU I1 27.1 plan to remain a resident of this community for the rest of my life. 28. I feel like I belong to this communilN. 29. This community says a lot about who I am. 102 Strongly Disagree Neither Agree Agree Strongly Disagree nor disagree Agree 30.1 participate in many different community projects and activities. . 31.1 would be sorry if the people I cared about moved away from this community. 32. The friendships and associations 1 have with other people in this communin mean a lot to me. iiHHI ^ l l l l^ lSBli wiSiwilSSiSill 33.1 feel loyal to the people in my community. 34. A feeling of fellowship runs deep between me and other people in this communin. H i S i i f t 3 i i M l S J l ^ | | | | M 35.1 visit with my friends and neighbours in their homes. 36. I identif} with the lifest\ les and values-of the people who live here. 37. When I think about this place, I think about the things I have done here with my friends, family, and people who matter to me. 3S. I identify with the land and natural features that make up and surround this communitv. fciillillSsSlii! illlSSiSMlItoiSi illlllilS Isillill 39.1 feel attached to the land and natural features that make up and surround this community. 40. 1 often feel a sense of oneness with the land and natural features that make up and surround this communin. 41.1 often feel a kinship with the wild animals and plants that live and grow in this community. ? . 42. 1 think of the surrounding natural em ironmenl (including land, plants, and animals) as part of m\ community to which I belong. IIIIIJ^ p 103 Strongly Disagree Neither Agree Agree Strongly Disagree nor disagree Agree 43.1 would like to be involved in activities related to the protection of the natural environment (including land, plants, animals) that makes up and surrounds this community. 44. The surrounding natural environment provides outdoor recreational opportunities that I en jo. 45. The surrounding natural environment is an important contributor to my livelihood. 46. 1 feel attached to m\ personal possessions and.belongirigs. -> 47. I feel like I belong in my house. 48. When I think about this community. I think about the town's centre. 49.1 visit more with my neighbours -and/or friends now, compared to before the fire occurred. 50.1 participate in more community projects acti\ ilies now compared to before the lire occurred. liftlSlillllifcial^Jill 51.1 am more interested in knowing what goes on in this community now, compared to before the fire occurred. 52. I spend more time doing outdoor recreation acti\ ilies now. compared to before the lire occurred. l l l l l i l l 53. In your opinion, name 3 buildings which represent the essence of this community: ' #1 • ' #2 • #3 54. Building #1 is personally meaningful to me. • Strongly Agree • Agree • Neither Agree nor disagree • Disagree • Strongly Disagree 55. Building #2 is personally meaningful to me. • Strongly Agree • Agree • Neither Agree nor disagree • Disagree • Strongly Disagree 56. Building #3 is personally meaningful to me. • Strongly Agree • Agree • Neither Agree nor disagree • Disagree • Strongly Disagree SECTION D: YOUR HEALTH 105 Below is a list of s o m e of the ways you may have felt or behaved. P lease indicate how often you have felt this way DURING THE PAST WEEK. P l ea se check (V) the appropriate box for e a c h question. Rarely or Some or a Occasionally or a Most or all of none of the little of the moderate amount the time time (less than time of time (3-4 days) (5-7 days) 1 day) (1-2 days) 57. I was bothered b\ things thiil usually don't bother me. 58.1 had trouble keeping my mind on what I was doing. 59. I felt depressed. iilllil• :EjiiBr:dti• 'iisilliSlli25i; 60.1 felt that everything I did was an effort. 61.1 fell hopeful about the future. | f | | | |{IHHI 62. I felt fearful. 63. My sleep was restless. 64.1 was happy. 65. I felt loneh. 66.1 could not "get going". \ 106 SECTION E: BACKGROUND INFORMATION The following questions are intended to tell us something about you and your current living and employment conditions. They also help us understand the social backgrounds of people living in your area. 67. How old are you? years 68. You are • Female • Male 69. What is your marital status? • Married • Single • Common law/living with a partner • Separated/Divorced • Widowed 70. Including yourself, how many people currently live in your household? Indicate exact number (put' 1' i f you live alone) 71. Do you have children? • Yes —> go to next question (#64) • No —» go to question #66 , 72. If yes, how many children do you have currently living at home? Indicate exact number ' (put '0' i f none of your children currently live at home) 73. At the time of the fires in 2003, how many children were living at home? Indicate exact number 107 74. What is your current work situation? • Full-time (employed by others and/or self-employed) • Part-time (employed by others and/or self-employed) • Childcare/caregiving/family responsibilities • Unemployed • Retired • Student 75. If you are currently working full-time or part-time, in what area is your main occupation? • Tourism/hospitality • Forestry • HealthCare • Education • Retail • Band Council/Band Administration • Construction • Agriculture/Farming • Ranching • Service sector • Government Services • Transportation • Law Enforcement/Legal • Other {please specify: ) 76. What is the highest level of education you have completed? • No formal schooling • Some primary/elementary school (Grade 8 and lower) • Completed primary/elementary (Grade 8) • Completed secondary/high school (Grade 12 or G.E.D. or equivalent) • Trades certificate or diploma • College diploma or associates degree • Bachelor's degree • Graduate degree (Master's, doctorate, professional degree) 108 77. What is your estimated total family income (gross from all sources) over the last 12 months? • less than $20,000 • $20,000 - $39,999 • $40,000 - $59,999 • $60,000 - $79,999 • $80,000 - $99,999 • $100,000 - $119,999 • greater than $120,000 THANK YOU FOR TAKING THE TIME TO COMPLETE THIS SURVEY We are interested in comments you may have about this survey. You may record them in the following space and/or on the other side of this page. 110 APPENDIX D: CONSENT FOR QUESTIONNAIRE (COVER LETTER) (Original on U B C Letterhead) March 16,2007 The University of British Columbia Department of Educational and Counselling Psychology, and Special Education Faculty of Education 2125 Main Mall . Vancouver, B.C. Canada V6T 1Z4 Title of project: Impact from a wildfire, community attachment, and well-being. Thank you.for taking the time to read this letter. We are conducting a survey to gain a better understanding of the relationship you have with the community in which you live, and how this may relate to your experience of the 2003 McLure Wildfire and your overall well-being. We are interested in the experiences of individual's (ages 19 and over) from Barriere and Louis Creek who were living in these communities when the McLure Wildfire occurred. . You will be asked to answer several questions pertaining to: (1) your experience during and immediately following the wildfire, (2) the degree to which you currently feel attached to your community, and (3) your current state of well-being. Your participation will help the emergency response community gain a broader ^ perspective as to the potential long-term impact of wildfires on the health and well-being of individuals and communities. By understanding to what extent your community is an important part of who you are, it will enable the emergency response community to make more informed decisions on implementing the appropriate services during the recovery process following a wildfire, particularly in a rural community. The survey may be beneficial to you in that you may gain a better understanding of how you feel towards your community and how this may relate^to the impact of the fires and your current state of well-being. It will take you approximately 10-20 minutes to complete this survey. Answering some of the questions may evoke unwanted memories and feelings pertaining to your experience of the wildfire. We have minimized this risk by not asking you to describe your experience in detail and asking you questions that you are likely tp have been faced with as part of your everyday life over the past 3 V2 years. APPENDIX E: S U R V E Y REMINDER C A R D REMINDER Your responses are important You recently received a QUESTIONNAIRE, exploring the relationship you have with your community and how this may relate to your experience of the McLure fire. If you have already completed and returned it to us, please accept our sincere thanks. If not, please do so today. Your responses are extremely important as they will help to accurately represent the experiences of individuals in your community. Please return the questionnaire no later than FRIDAY, APRIL 13, 2007. Sincerely, Risa Handler 113 APPENDIX F: IMPACT F R O M THE WILDFIRES: S U R V E Y ITEM D E V E L O P M E N T Item Source Assessing the extent of loss of the following: 1. Family members and/or friends moved away from the community Developed specifically for this project 2. Pets Developed specifically for this project 3. Livestock Developed specifically for this project 4. Job or business Maltais et al., 2002* 5. Damage to my house Maltais et al., 2002* 6. Damage to my property Maltais et al., 2002* 7. Damage to my immediate neighbours' house Scher, 2003 8. Damage to my immediate neighbours' property Scher, 2003 9. Job or business loss of family members' and/or friends Developed specifically for this project To what extent the following was experienced: 10.1 had house insurance at the time of the fires. Developed specifically for this project 11.1 experienced involuntary disruption to my work. Maltais et al., 2002* 12. Up to one year following the fire, I experienced adverse physical health effects due to the fires (e.g., asthma flare-ups, difficulty breathing, runny eyes or nose, headaches). Scher, 2003 13. Up to one year following the fire, I visited a doctor, hospital, or clinic due to fire-related health problems. Scher, 2003 14. Up to one year following the fire, I experienced mental and/or emotional difficulties due to the fires (e.g., depressed, overwhelming feelings of stress or anxiety, difficulty concentrating). Scher, 2003 15. Up to one year following the fire, I visited a mental health professional due to fire-related mental and/or emotional difficulties. Scher, 2003 16. In the areas where I enjoyed outdoor recreational activities, the natural environment changed as a consequence of the fire. Developed specifically for this project 17. As a consequence of the fire, the view from my house changed. Developed specifically for this project 18. At the time of the fire, the surrounding natural environment changed. Developed specifically for this project 19.1 received emotional and moral support from my community. Developed specifically for this project 24.1 received emotional and moral support from disaster-related services. Developed specifically for this project *Adapted from the following source: Maltais, Lachance, Simard, Brassard, & Picard, (2002). 114 APPENDIX G: P L A C E A T T A C H M E N T : S U R V E Y ITEM D E V E L O P M E N T Item Construct Aspect Source General Dimension I would be sorry to leave this community. Attachment Affect Kasarda & Janowitz, 1974 I feel at home in this community. Attachment Affect Kasarda & Janowitz, 1974 I am interested in knowing what goes on in this community. Attachment Cognitive Kasarda & Janowitz, 1974 I plan to remain a resident of this community for the rest of my life. Attachment (commitment) Cognitive/ Behaviour . Buckner, 1988 (NCI Scale; .71 item-total) I feel like I belong to this community. Attachment Affect Buckner, 1988 (NCI Scale; .83 item-total) This community says a lot about who I am. Identity Cognitive Cross, 2003 Social Dimension I participate in many different community projects and activities Attachment Behaviour Davidson & Cotter, 1986 I would be sorry if the people I cared about moved away from this community. Attachment Affect Hidalgo & Hernandez, 2001 The friendships and associations I have with other people in this community mean a lot to me. Attachment Affect Buckner, 1988 (NCI Scale; .80 item-total) I feel loyal to the people in my community. Attachment Affect Buckner, 1988 (NCI Scale; .83 item-total) A feeling of fellowship runs deep between me and other people in this community. Attachment Affect/ Behaviour Buckner, 1988 (NCI Scale; .84 item-total) I visit with my friends and neighbours in their homes. Attachment Behaviour Buckner, 1988 (NCI Scale; .70 item-total) I identify with the lifestyles and values of the people who live here. Identity Cognitive Williams, 2000 When I think about this place, I think about the things I have done here with my friends, family, Attachment Cognitive Cross, 2003 115 and people who matter to me. Item Construct Aspect Source Physical Dimension (Natural Environment) I identify with the land Identity Cognitive Williams, 2000 and natural features that make up and surround this community. I feel attached to the Attachment Affect Williams, 2000 land and natural features that make up and surround this community. I often feel a sense of Attachment Affect Mayer & Frantz, 2004 oneness with the land and natural features that make up and surround this community. I often feel a kinship Attachment Affect Mayer & Frantz, 2004 with the wild animals and plants that live and grow in this community. I think of the Attachment Affect/ Mayer & Frantz, 2004 surrounding natural Cognitive environment (including land, plants, and animals) as part of my community to which I belong. I would like to be Attachment Behaviour Brehm, Eisenhauer, & involved in activities Krannich, 2004 related to the protection of the natural environment that makes up and surround this community. The surrounding natural Dependence Cognitive/ Brehm, Eisenhauer, & environment provides Behaviour Krannich, 2004; Cross, 2003 outdoor recreational opportunities that I enjoy. The surrounding natural Dependence Cognitive/ Cross, 2003 environment is an Behaviour important contributor to my livelihood. 116 Item Construct Aspect Source Physical Dimension (Built Environment) I feel attached to my personal possessions and belongings. Attachment Affect Developed specifically for this project I feel like I belong in my house. Attachment Affect Developed specifically for this project When I think about this community, I think about the town's centre. Attachment Cognitive Developed specifically for this project Which buildings in your community would be considered as the most important ones? Identity Cognitive Developed specifically for this project Building # is personally meaningful to me.(x3) Attachment Affect Developed specifically for this project Change Dimension I visit more with my neighbours and/or friends now, compared to before the fire occurred. Attachment Behaviour Developed specifically for this project I participate in more community projects/activities now compared to before the fire occurred. Attachment Behaviour Developed specifically for this project I am more interested in knowing what goes on in this community now, compared to before the fire occurred. Attachment Cognitive Developed specifically for this project I spend more time doing outdoor recreation activities now, compared to before the fire occurred. Dependence Behaviour Developed specifically for this project 117 APPENDIX H: CONTENT A N A L Y S I S Content analysis was conducted on item #53 in the survey - In your opinion, which buildings in your community would be considered as the most important ones? Name three buildings. Three hundred responses were provided. Responses were grouped, tallied, and summarized. The results of the content analysis on item #53 - In your opinion, which buildings in your community would be considered as the most important ones? Name three buildings -revealed that the grocery store, community health centre, post office, and schools (37%, 36%, 27%, and 26% of the 300 responses) were the most common buildings participants stated were the most important in their community. The respondents also deemed the fire hall, bank, and church as important (17%, 16%, and 13% of the 300 responses, respectively). Based on my verbal discussions with a few respondents, there was some confusion regarding the meaning of this question. Some respondents interpreted this question as asking which buildings in the community did the respondent deem important to the community, while other respondents interpreted this question as asking which buildings in the community did the respondent deem personally important. Respondents from Louis Creek were more likely to name 0, 1, or 2 buildings because of the few number of buildings in the community. Page 1 of 1 The University of British Columbia Office of Research Services Behavioural Research Ethics Board Suite 102, 6190 Agronomy Road, Vancouver, B.C. V6T1Z3 CERTIFICATE OF APPROVAL - MINIMAL RISK PRINCIPAL INVESTIGATOR: INSTITUTION / DEPARTMENT: UBC BREB NUMBER: UBC/Education/Educational & Bonita C. Long Counselling Psychology, and Special H07-00254 Education INSTITUTION(S) WHERE RESEARCH WILL BE CARRIED OUT: Institution I Site N/A < N/A Other locations where the research will be conducted: This project will be conducted in two communities. Since the survey is designed as a self-administered instrument, participants can choose to complete the survey at a location most comfortable to them. CO-INVESTIGATOR(S): Risa J. Handler SPONSORING AGENCIES: N/A PROJECT TITLE: Impact from an interface wildfire, place attachment, and depression: A survey of two rural communities CERTIFICATE EXPIRY DATE: March 2, 2008 DOCUMENTS INCLUDED IN THIS APPROVAL: DATE APPROVED: March 2, 2007 Document Name | Version | Date Protocol: Proposal N/A February 2, 2007 Advertisements: Advertisement Version 1 February 2, 2007 Questionnaire. Questionnaire Cover Letter. Tests: Questionnaire Version 2 February 2, 2007 Questionnaire cover letter Version 2 February 2, 2007 The application for ethical review and the document(s) listed above have been reviewed and the procedures were found to be acceptable on ethical grounds for research involving human subjects. Approval is issued on behalf of the Behavioural Research Ethics Board and signed electronically by one of the following: Dr. Peter Suedfeld, Chair • —**"""- — — — ~ " Dr. Jim Rupert, Associate Chair Dr. Arminee Kazanjian, Associate Chair Dr. M. Judith Lynam, Associate Chair U B C hrtps://rise.ubc.ca/rise/Doc/0/3HlM5LTCBLHK3824BNO79P7QF4/fromString.html 8/13/2007 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0053635/manifest

Comment

Related Items