Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Public perceptions of drinking water risk : a community perspective Yim, Yolanda 2005

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

Item Metadata

Download

Media
831-ubc_2005-0142.pdf [ 8.16MB ]
Metadata
JSON: 831-1.0092103.json
JSON-LD: 831-1.0092103-ld.json
RDF/XML (Pretty): 831-1.0092103-rdf.xml
RDF/JSON: 831-1.0092103-rdf.json
Turtle: 831-1.0092103-turtle.txt
N-Triples: 831-1.0092103-rdf-ntriples.txt
Original Record: 831-1.0092103-source.json
Full Text
831-1.0092103-fulltext.txt
Citation
831-1.0092103.ris

Full Text

P U B L I C P E R C E P T I O N S O F D R I N K I N G W A T E R R I S K : A C O M M U N I T Y P E R S P E C T I V E by YOLANDA YIM B.Sc, McGill University, 2001 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES (Resource Management and Environmental Studies) We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A February 2005 © Yolanda Yim, 2005 ABSTRACT The events of Walkerton made it clear that important changes were necessary to address the safety of our drinking water. Most provinces have since passed new legislation to prevent such tragedies from happening again. There is a need, however, to consider matters from a public perspective if implementation of forthcoming regulations is to be successful. How do people think about drinking water risks? How do they understand its safety? What is the basis of their perceptions? This study addresses these questions from the perspective of a rural community at odds with provincial authorities over the issue of water chlorination. Its specific objectives are to (f) characterize one community's perceptions of drinking water risks and (if) explore the basis of many resident's aversion to chlorine. Three main concepts are explored: mtiritiw toxicology (specifically how people think about the concept of microbial pathogen dose), trust in various levels of authority responsible for managing drinking water, and sense cf place (a summation of people's identification with and attachment to place). A combination of both interview and survey methods were used in this study. Results from the survey were analyzed using SPSS (version 10.0) and form the bulk of the data presented in this study. Results from the survey indicate most respondents' were tolerant toward small amounts of microbial contaminants (indicating sensitivity to microbial pathogen dose). No relationship was found between these views and chlorine aversion, however, suggesting that the reasons behind the chlorine controversy did not revolve primarily around health concerns. Aversion to chlorine was most strongly linked to lack of trust in expert and provincial authorities, in conjunction with increased concerns about resource development. Review of the interview data revealed many residents believed there to be a connection between government interests in logging and the order to disinfect the local water supply. 'Sense of place' played a moderate role in deteirnining both aversion to chlorine and increased tolerance toward microbial contaminants. Disagreements between governments and communities over drinking water can go beyond debates about specific health risks to encompass broader social concerns. These concerns were very real to the citizens of this community, yet perhaps were not adequately recognized as such by provincial authorities. tu TABLE OF CONTENTS Abstract a Table cf Contents iv List cf Tables : ti List cf Figures '. tii A okncnded^rmas : tiii Chapter 1 - The Controversy, the Players, and the Problem 1 1.1 A Controversy Over Water Treatment.... ....1 1.2 The Research Problem 5 Why Erickson? 6 1.3 Specific Research Objectives...... .....8 1.4 Chapter Outline 9 Chapter 2 - Psychological and Socio-cultural Perspectives on Risk 10 2.1 Existing Studies on Drinking Water Risk Perception 10 2.2 The Concept of Risk 10 2.3 Converging Ideas in Risk Perception Research. 11 2.4 Basic Overview of the Field 11 The Psychometric Paradigm. 12 Intuitive Toxicology 13 Limitations of the Psychological Approach 13 Socio-cultural Perspectives 14 2.5 Concepts Addressed in this Study. 15 Intuitive Toxicology: Chemical versus Microbial Contaminants.... 15 Trust J 16 Sense of Place. 17 2.6 Conclusion. 19 Chapter 3 - Measuring Perceptions: Research Approach and Methods 20 3.1 Research Strategy 20 3.2 Qualitative Interviews 22 3.3 Developing the Survey. 24 3.4 Administering the Survey. 29 3.5 Validity, Generalizability and Limitations of this Study 31 Chapter 4 - Results I: Characterizing Community Perceptions 33 4.1 Ratings of Perceived Risk. 33 General Health Risk Ratings 33 Perceived Sources of Risk to Drinking Water. 35 Factor Analysis of Drinking Water Risks 37 4.2 Attitudes and Behaviours Taken in Relation to Chlorine 39 Perceived Chlorine Risk 39 iv Behaviours Taken in Relation to the Controversy 41 Correlation between Drinking Water Variables and Perceived Chlorine Risk 41 4.3 Attitudes Toward Microbial Risks 43 Dose Sensitivity 43 Correlation between Dose Sensitivity and Perceived Chlorine Risk 47 General Attitudes Toward Microbial Contamination 48 Beliefs about Local Water Quality. 49 4.4 Summary of Chapter Findings 50 Descriptive Portrait of Community Perceptions 50 Specific Attitudes Underlying the Chlorine Controversy: 50 Chapter 5 - Results II: Exploring the Socio-cultural Basis of the Chlorine Debate 52 5.1 Trust 52 Trust in Experts, the Government, and Local Levels of Authority. 53 Trust in Information Sources ;. 54 Correlation between Trust Variables and Perceived Chlorine Risk 56 5.2 Sense of Place ; 57 Correlation Between Sense of Place and Perceived Chlorine Risk 60 5.3 Multiple Regression Analysis: Predicting Perceived Chlorine Risk. 61 5.4 Summary of Chapter Findings 63 Chapter 6 - Conclusion 64 6.1 Key Findings of this Research 64 Diverging Perceptions of Microbial Risks 64 Concerns in the Past Led to Conflicts in the Present 65 Protecting Valued Notions of Place 66 6.2 What Happened During the Controversy? 68 6.3 Implications for Implementing Future Drinking Water Policies 70 Conceptual Changes Required at Both Community and Agency Levels 71 Practical Changes 71 Organizational Barriers to Effective Implementation 72 6.4 Directions for Future Research. 73 Appendix I: Interview Protocol 76 Appendix II: Risk Perception and Drinking Water Questionnaire 77 References 92 LIST OF TABLES Table 4.1 Rotated Factor Loadings for Drinking Water Risks 38 Table 4.2 Perceptions of Chlorine Risk 40 Table 4.3 Actions Taken in Relation to the Controversy. 41 Table 4.4 Pearson Correlation Between Drinking Water Variables and Perceived Chlorine Risk - '. 42 Table 4.5 Dose-Response Sensitivity to Microbial Contaminants 45 Table 4.6 Dose-Response Sensitivity to Chemicals and Radiation 46 Table 4.7 General Attitudes Toward Microbial Contamination 48 Table 4.8 Beliefs about Local Drinking Water Quality 48 Table 5.1 Trust in Experts and Various Levels of Authority Responsible for Managing Drinking Water 53 Table 5.2 Pearson Correlation Between Trust Variables and Perceived Chlorine Risk. 57 Table 5.3 Sense of Place 58 Table 5.4 Multiple Regression Analysis: Identifying Predictors of Perceived Chlorine Risk. .62 vi LIST OF FIGURES Figure 3.1 Research Strategy for this Study. 21 Figure 4.1 Perceived Health Risk to Canadian Public 34 Figure 4.2 Perceived Sources of Risks to Drinking Water Quality. 36 Figure 5.1 Sources of Drinking Water Information and Confidence in Those Sources 55 vii ACKNOWLEDGEMENTS This study has taken me on an exciting journey, both academically and personally. It could not have begun without the help of Les Lavkulich, to whom I am ever grateful, for deciding to take a chance on me at a moment's notice. I would like to extend a special thanks to Terre Satterfield, who could not have been a more honest, challenging, and encouraging supervisor. This study also benefited greatly from the help of my committee members, Judy Isaac-Renton and Ralph Matthews, as well as Jennifer Shapka, whose statistical expertise was most appreciated. Sincerest thanks to the community of Erickson, B C , for their cooperation and » participation in this study, as well as the many B C Ministry of Health officials who took the time to help me with my information needs. O n a personal note, I must thank my family, especially my mom, Daniel and Amber, for all their love and humour. Aviva Savelson, my 'partner in crime', I could not have made it through without you. Thank you for your endless support, laughter and friendship. Who knew writing a thesis could be so much fun? Finally, I would like to express my deepest gratitude to Joel, whose spiritual, emotional and intellectual support in many ways made this thesis possible. vui Chapter 1 T H E CONTROVERSY, T H E PLAYERS, AND T H E PROBLEM This study was inspired by events that took place in Erickson, British Columbia, wherein a contentious public1 disagreement arose over a government proposal to begin disinfecting the community's local water supply. This chapter begins with a characterization of the case in Erickson, followed by a discussion of how these events typify a more general problem faced by government agencies responsible for bringing communities into compliance with new drinking water regulations. The purpose and research questions of this study are then outlined, followed by an overview of the chapter contents of this thesis. 1.1 A Controversy Over Water Treatment Erickson is a small rural community located in the southeastern region of British Columbia. The population of the community is about 2,000 people. Erickson's economy is highly dependent on agriculture, particularly fruit farming, as well as some tourism and logging. The majority of people living in Erickson are retirees; most originate from other parts of British Columbia and the Prairies. Almost all of their drinking water comes from Arrow Creek, which flows from a largely undeveloped watershed just northeast of the community. Water from the creek is distributed first to Erickson, and then to the nearby town of Creston. The Columbia Brewery, also located in Creston, is another major user of the water. Over the last 20 years, there have been two documented waterborne disease outbreaks in the Creston area (1985 and 1990). In both cases, the cause of the outbreaks was believed to be beavers, whose feces contained the protozoan parasite Giardia2. As a result, both Creston and Erickson were put on boil-water advisories. Shortly thereafter, Creston began cHorinating its water. The local waterworks utility, hereafter referred to as the Erickson Improvement District (EID), however, did not. The regional health unit at the time, now known as the East Kootenay Service Society (EKHSS), was not in favour of leaving the EID system untreated, 11t is acknowledged that within any society there are many 'publics'. The main public of interest for this study is the community of Erickson, a small, relatively homogeneous population (demographically speaking). 2 Giardia can be transmitted through drinking water, especially surface water sources. Many protozoan cysts like Giardia have been shown to be highly resistant to chorine (Jarroll, 1999). 1 though no basis of enforcement existed at the time, as there were no legally binding provincial standards regarding drinking water. In 1992, the Safe Drinking Water Re§dation was introduced under the authority of the provincial Health A a. According to the regulation, two drinking water standards for all water systems had to be met - one for total coliform3, of which there could be no more than 10 counts per 100ml sample and one for fecal coliform4, for which there had to be zero counts per 100ml sample. The new regulations also required all surface water used for public drinking water supplies be disinfected prior to public consumption. The EID water system was now in violation of both standards, as well as the disinfection regulation, and had been for some time. The EKHSS attempted to engage the EID in a progressive implementation process whereby the highest risk water systems in the region would be brought into compliance with the new regulations first These systems were asked to voluntarily submit an acceptable plan for water treatment to the EKHSS. The EID asked the EKHSS if they would approve a form of water treatment that did not involve the use of chlorine, as many Erickson water users did not want a cUorinated system. The EKHSS agreed to consider a chlorine-free option as long as the system was able to comply with the regulations. During the period of 1992 - 1998, the EID experimented with a number of alternative water treatment technologies, such as UV disinfection. In May 1997, the EID submitted a proposal to install water filtration and UV disinfection units into the homes and business of each EID water user (known as 'point-of-use' technology). However, the EKHSS rejected this proposal for a number of financial, operational, and legal reasons and invited the EID to submit another plan. By February 1998, the EID had still not submitted a plan that met the approval of the EKHSS, and the local medical health officer at the time, hereafter referred to as the 'MHO', issued an order to the EID to begin disinfecting its water supply. The order was issued under the authority of the Health A ct, and required the EID to 3 Conforms are a group of bacteria, most of which are not harmful to humans, the presence of coliform bacteria in drinking water can serve as a general indicator that a system's water treatment system may not be functioning properly (US Environmental Protection Agency, Drinking Water Standards for Regulated Contaminants Webpage, total coliform rule, n.d.) 4 Fecal conforms are a subgroup of coliform bacteria associated with fresh mammalian feces (US Environmental Protection Agency, Drinking Water Standards for Regulated Contaminants Webpage, total coliform rule, n.d.) 2 install a fully operational water disinfection system within the next 20 weeks. The order also outlined four intermediate deadlines to ensure the E I D was on schedule to meet to the 20-week deadline. By July 1998, a new M H O had been appointed to the local health region and decided to take a more assertive approach to the situation in Erickson. The 20-week deadline for the disinfection order had passed and the E I D was now in contravention of the order. This M H O issued another disinfection order. Although both orders did not specify chlorine had to be used, it was and remains to be the cheapest and most widely used form of water disinfection. The M H O has stated5 that he never specified they had to use chlorine although he can see how, due to financial constraints, they may have believed the order to disinfect their water was in reality an order to chlorinate. The E I D eventually complied with the E K H S S order and installed a cWorination unit near the water intake area in April 1999. During this period, a citizen's group was formed, which eventually became the Erickson Water Users Society (EWUS). E W U S was strongly opposed to the idea of chlorine, citing the adverse effects of cMorinated by-products (CBPs) on human health. As such, E W U S did not view chlorine was an acceptable form of water treatment for the community. Local farmers were concerned about the effects of chlorinated water on their crops. Members of E W U S wanted an alternative and felt they were being forced into accepting chlorine because there were no other economically feasible options that would meet the approval of the E K H S S . In the absence of other options, on April 29 t h 1999 a new citizen's group called the Water Action Group (WAG) staged a 55-day blockade of the water intake area to prevent chlorine from entering the water supply. In the meantime, E W U S joined forces with the Sierra Legal Defense Fund to seek legal action against the disinfection order. After a few months in and out of court, the judge requested all parties resolve the matter out of court and allowed Erickson's water situation to remain status quo. 5 Personal communication, Jan 24, 2004. 3 While EWUS and the MHO attempted to resolve their differences, the town of Walkerton, Ontario was heading for disaster. In May 2000, seven people died from drinking water that had been contaminated with a pathogenic strain of fecal coliform. This signal event propelled the MHO to discontinue negotiations with EWUS. In the media, the MHO warned that Erickson was 'shades of Walkerton' (Pynn> 2001) and called on the provincial government to take over Erickson's water delivery service in order to avoid a Walkerton-style outbreak in British Columbia. On January 10th 2001, the government turned control of the EID water service over to the Regional District of Central Kootenay (RDCK). A government appointed receiver assumed financial control of the EID waterworks and was assigned the task of putting together a community approved water treatment plan. Over the next year, a focus group made up of a cross-section of the community assembled to discuss possible treatment options. On February 9th 2002, a community-wide referendum was held asking: A re you infavour cf a water treatment facility which includes filtration, dtravdet dkirfeaim and measures toperniz operation without trace chlorine at a total annual esdrmted cost cf $385 for the average resident} The community voted 80% in favour. Six months later, the receiver negotiated a contract to build a ten million dollar membrane filtration plant for Arrow Creek, the cost of which was to be divided equally between the community, the province, and the federal government. Community response to the agreement was not positive. Many believed the consultations insufficient; the focus group had not come to a consensus on the type of filtration, its total cost, or the parameters of the engineering bid that would be selected to build the plant. Farmers were worried that the capacity of the plant was too small for their agricultural requirements, which average around 30 million litres per day in the summer. Other details that had not been resolved included: how to deal with the open reservoir that supplied water to the Town of Creston (a cost that had not been factored into the engineering contract), how to work out cost-sharing measures with the town, how to incorporate Sullivan Creek (the back-up supply when the Arrow Creek system needed servicing), and the need to install cosdy back-flow prevention valves throughout the system. As of November 27th 2003, chlorine was flowing in Erickson - a temporary measure - while the filtration plant is being built. The plant is scheduled to open February 1st 2005. 4 1.2 The Research Problem The case in Erickson illustrates some of the real-life difficulties involved in bringing communities into compliance with new public health regulations. Specifically, it demonstrates how confucting views of drinking water risks can create serious difficulties for public health officials responsible for implementing such regulations at the conimunity level. Clearly, failure to adequately address and resolve such differences carry significant costs for all parties involved. In the end, the conflict in Erickson took over a decade to resolve, required the construction of a ten million dollar water treatment facility, and left a very tenuous relationship between the government and the community. This last point is perhaps the most significant outcome of the controversy, as the potential now exists for future difficulties to arise as the government tries to implement other regulations within the community. Clearly, these are not the desired outcomes of an effective regulatory process from both a government and community perspective. The case in Erickson indicates a need for a better understanding of the basis of conflicting perceptions of drinking water risk. Following the events of Walkerton, all provinces and territories initiated reviews of their existing legislation and policies in order to prevent such tragedies from happening again. Ontario, Alberta, Quebec, and now British Columbia have all passed new legislation to strengthen and ensure the safety of public drinking water supplies. In 2003, British Columbia introduced its new Drinking WaterF^vtBCtimAa(pWPA), which supersedes the authority of previous the Safe Drinking Water Regulation. Under the new act, appointed public health officials known as drinking water protection officers (DWO) have been granted broad powers to issue terms and orders on water systems in violation of the regulations. At the same time, the responsibilities of water system purveyors have also increased. Water purveyors are now required to provide better public notification, conduct system-wide risk assessments, and engage in more stringent reporting of water quality test results. The implications of these changes are most significant for public health officials working directly with water purveyors and local communities. Along with the increased regulation of all water systems will be an increased need for public health officials to work collaboratively with communities and 'act together on water quality issues' (Office of the Provincial Health Officer, 2003). This is not a new idea in public health and is better known in the field as 'progressive enforcement'6. Yet current approaches to progressive enforcement do not always result in effective collaboration. Public perceptions are also highly consequential in determining the outcome of such enforcement efforts. This raises important questions about implementing new drinking water policies in the future. How are governments to manage and address public perceptions, especially when these views conflict with their own? What are the reasons behind such perceptions? And, perhaps most importantly, how might governments design implementation strategies that anticipate and account for public perceptions in order to avoid public controversy? The need to address such questions is significant, in light of the formcoming changes toward drinking water regulation in British Columbia. This study uses Erickson as a unique case to explore how one community thinks about and understands drinking water risks. In particular, it examines the basis of one community's aversion to chlorine, by characterizing residents' perceptions of drinking water risks, and relating these views to a number of specific cognitive, social and environmental concerns. The overarching goal of this study is not to judge the accuracy of people's perceptions in Erickson, but rather to identify and explore some of the reasons behind the controversy from a community perspective. WhyEnoksoni Erickson represents an ideal case to study for a number of reasons. First, it provides a real-life community setting through which to explore public perceptions of drinking water risk. Having been a community that has already experienced a public controversy, the case in Erickson also presents a valuable learning opportunity through which to learn how similar situations can be avoided in the future. A gradual method of policy implementation that moves from voluntary to more authoritative approaches to enforcement. 6 Erickson also typifies the kind of community that poses the biggest challenge to drinking water regulation due to inherent conplications associated with small water systems. As observed by the Auditor General of British Columbia (1998/1999), small water systems are often limited by a number of financial and technical constraints not experienced by larger systems. These constraints create significant challenges for those in charge of operating small water systems as such systems often lack sufficient technical expertise or financial resources to perform important infrastructure upgrades or install new technologies. These complications do not necessarily influence public perceptions of risk directly, but may serve to amplify the tension between public health officials and communities when conflicting views arise. This is because of the constraining effect such complications have on the range of management options available to small water systems. When conflicting perceptions arise, there are only so many options available to water system managers that can satisfy both the requirements of the regulations and the perceptions of the systems' users. Compounding this problem is the fact that small water systems are also subject to more , contamination threats than larger ones, simply due to the fact that these systems tend to be located in rural areas. These areas often support a variety of resource-based activities (e.g. logging, milling, agriculture) in close proximity to the source of cornmumtydnnking water supplies thus increasing the chance of water contamination. For these reasons, specialized approaches toward drinking water regulation are required in small, rural communities such as Erickson. The effect of the complications associated with regulating small water systems was evident in the case in Erickson. One could reasonably imagine that if the E I D had access to an unlimited amount of money the controversy in Erickson would not have occurred, as the community would have been able to implement whatever water treatment technologies they desired that met the requirements of the regulations. However, in reality, this was not the case. The EID's ability to implement a desirable form of water treatment was essentially limited by the financial resources available to the E I D . Their status as an unincorporated community further exacerbated this situation by preventing the E I D from qualifying for additional infrastructure funding from the government. The end result was a public controversy 7 developed around the specific issue of chlorination because it was the only affordable water treatment option, though it was not the option that they supported. 1.3 Specific Research Objectives Using qualitative and quantitative techniques, this study examines the relationship between perceived chlorine risk (the main dependent variable of this study), and three concepts identified in the risk literature as important to understanding risk perception. They are inttddw toxioolcg)i (specifically how people think about the concept of microbial pathogen dose), trust in various levels of authority responsible for drinking water management in the province, and sense cf place (a summation of people's identification and attachment to place). The research interests of this study are both exploratory and inferential. The exploratory component of this study is necessary in order to provide a descriptive portrait of residents' perceptions before making inferences as to the relationships between specific concepts. Specifically, the exploratory objectives of this study are to characterize residents': • General perceptions toward a variety of health risks (including drinking water) and specific perceptions toward drinking water risks; • Attitudes and behaviours taken in relation to the chlorine controversy; • Attitudes toward microbial risks; • Level of trust in various expert and government authorities; • Affinity toward Erickson as a place (as measured through the concept of 'sense of place'). Subsequent to establishing these objectives, a second interest of this study is to explore a number of specific relationships between residents' perceptions of chlorine risk and the aforementioned factors. Specifically, the inferential objectives of this study are to examine the relationship between perceived chlorine risk and residents': • Perceptions of other drinking water risks; 8 • Attitudes toward microbial risks, particularly sensitivity to dose in the form of microbial pathogens; • Trust in various expert, government and local authorities responsible for managing drinking water quality, • Residents' sense of place toward Erickson. 1.4 Copter Outline The following chapter locates this study within the broader framework of the risk perception literature. Specifically it outlines the conceptual approach of this study and explains the rationale behind the three concepts of interest addressed in this research (how people think about microbial pathogen dose, trust, and sense of place). In so doing it draws from two areas of the risk perception literature, known as the psychological and socio-cultural perspectives on risk. Chapter 3 outlines the research approach and methods used in this study. Results from a community-wide survey conducted in Erickson are presented in the next two chapters. Chapter 4 provides a statistically descriptive picture of residents' perceptions of drinking water risks, and includes data on residents' perceptions of drinking water risks in general, perceptions of chlorine risk, and attitudes toward microbial contaminants. Chapter 5 takes a more in-depth look at the basis of residents' perceptions of chlorine risk through the concepts of trust and sense of place. Chapter 6 summarizes the main findings of this research and outlines the implications of this study for implementation of drinking water policies in the future. i 9 Chapter 2 PSYCHOLOGICAL AND SOCIO-CULTURAL PERSPECTIVES ON RISK 2.1 Existing Studies on Drinking Water Risk Perception Relatively few risk perception studies have addressed the topic of drinking water specifically. O f the existing studies, most tend to be highly descriptive or emphasize only the aesthetic dimensions of water quality perception. Findings from the descriptive studies suggest that Canadians are not highly concerned about the safety of their drinking water relative to a variety other health risks found in Canadian society (Slovic et al., 1993, Health Canada, 2000). However, when questioned directly about their attitudes toward drinking water, public perceptions seem less optimistic; for example, many people believe their water is likely contaminated with industrial chemicals (Auslander and Langlois, 1993) and that the quality of local water supplies is poorer than technical assessments of those same supplies (McDaniels, 1998). Studies of water aesthetics have identified that certain visual, taste, and odour characteristics of water are likely to have a negative impact on public perceptions of water quality (McGuire, 1995; Jardine et al., 1999). However, no single or combined set of studies has explored how and why people think of drinking water risks as they do. This study begins to address this gap in the literature by drawing on concepts developed within the broader risk literature and transferring them into a drinking water context. 2.2 The Concept of Risk In many ways, the very type of conflict that occurred in Erickson drove the development of the field of risk perceptions over three decades ago. Since then, academic understanding of the risk perception problem (generally termed the 'expert-lay debate') has undergone a number of changes over the years, giving rise to a number of different (often opposing) perspectives on the root cause of such disagreements. For the purposes of this study it is sufficient to note that the difference between each perspective revolves less around disagreements as to the conceptual definition of risk itself but rather the operational definition of risk and subsequent assumptions about the nature and legitimacy of expert versus lay understanding of that risk. 10 The conceptual definition of risk used in this study is any phenomenon that has the potential to cklifier substantial harm, whether cr wt theprobability cf Ms harmevzntuatxng is estwnbie (Lupton, 1999:9). This is admittedly a broad definition and one with which most experts and the public would agree. Where the most disagreement occurs is in the range of variables to include in the operational definition of risk, that is, the factors included in the measurement of 'risk'. As will be argued in this chapter, public definitions of risk involve a range of psychological, social, and cultural factors, in addition to their understanding of technical information. If risk encompasses a range of considerations, this opens up our study of Erickson onto a wide variety of possible areas of exploration. 2.3 Converging Ideas in Risk Perception Research The field of risk perception research was at a crossroads when this study began. While the majority of studies were rooted in psychology, there was another body of work that had gained recognition, known as the socio-cultural perspective. Historically, both fields developed largely independently of each another, with socio-cultural researchers being highly critical of their psychological counterparts. There has in recent years, however, been a trend toward convergence between the two perspectives (Bickerstaff, 2004:827) with some researchers suggesting that a combination of both' psychological and socio-cultural lines of inquiry are necessary for a more complete understanding of risk perception (Pidgeon et al., 1992:124; Weyman and Kelly, 1999). With this development in mind, this study focuses on three concepts of relevance to understanding the chlorine controversy in Erickson, drawing from both psychological and socio-cultural perspectives. Since these concepts build on a progression of thought within the risk literature, this review begins with a brief overview of the key developments in risk perception research, followed by an explanation of the specific concepts addressed in this study. 2.4 Basic Overview of the Field Research into risk perception began in the 1970s, in response to growing public opposition toward proliferating chemical and nuclear technologies. Recognizing a policy dilemma, 11 poticymakers and scientists turned to the emerging field of risk perception in the hopes of finding a way out of the increasingly difficult relationship between expert and public understandings of risk. Much of the foundational work came from the field of psychology, which addressed the cognitive processes by which people think and reason about risk Early psychological studies focused on cognitive biases and errors in decision-making as a possible reason as to why the public thought so differently about risk than experts (see for example Tversky and Kahneman, 1974; 1981). This research was guided by a mode of thinking known as the 'deficit' model of public understanding (Wynne, 1982; Irwin, 1995). The defining characteristic of this model was the clear distinction made between objective or expert 'rationality' and subjective or public 'perceptions' (Bickerstaff, 2004:828). Based on this distinction, researchers (and policymakers) inferred that public controversies over risk were likely due to a perceptual problem on the part of the public - as evidenced by widespread public irrationality over expert approved technologies. While the basis of this model has been largely discredited from an academic perspective, its purchase still remains present in many policy circles (see for example Krewski et al., 2002:1657). The Psychorretric Paradigm By the early 1980s, psychological research had shifted to evaluating and mapping perceptions of environmental and technological risks (Weyman and Kelly, 1999:7). Development of the psychometric paradigm by Paul Slovic and his colleagues at Decision Research provided a major contribution to the field. Borrowing ideas from psychometric scaling, Slovic and his colleagues demonstrated that public risk perception could be understood as a function of qualitative attributes of hazards (Slovic et al., 1984; Slovic, 1987). Two of the most important characteristics were 'dreadedness' and 'unfamiliarity'. Thus hazards deemed highly dreaded and unfamiliar, such as nuclear power, tended to be perceived as highly risky. Later studies conducted in other countries confirmed these factors to be near universal indicators of risk aversion (see for example, Teigen, Brun, and Slovic, 1988; Kleinhesselink and Rosa, 1991; Michitov and Rebrik, 1990). 12 c This research served to establish risk perception as widespread and legitimate. More importantly, it demonstrated that there were clear and consistent patterns in {the way lay people were tiiinking about risk. In contrast to the assumption of the deficit model, these views did not reflect the judgements of a haphazard or irrational public but rather a way of thinking that was consistent, predictable and based on a set of factors that attended to a broader set of concerns than accounted for in traditional expert assessments. At the same time, studies conducted outside the United States also revealed important cross-cultural differences, lending support the idea of a socio-cultural basis to risk perception. Intuitive Toxudogy The psychometric paradigm was primarily criticized for being overly broad; thus serving to describe public perceptions yet failing to explain their root causes (Slovic, 2000:8). The 'intuitive toxicology' approach was one of several approaches developed in response to this criticism. The goal of intuitive toxicology was to elicit the reasoning pathways used by experts and the public when thinking about specific risks. Such studies have since identified important areas of difference between expert and public ways of tiiinking about risk in terms of each perspectives' underlying attitudes, beliefs and causal assumptions each perspective hold toward the concept of risk This research has most extensively explored the topic of chemical risks (Kraus et al., 1992; Slovic et al., 1993) and thus is clearly relevant to this study in terms of understanding the cognitive basis of peoples' perceptions of chlorine risk (a point elaborated further below). Linitations of'thePsydxk^adAppmich A second criticism of psychological studies was their overemphasis on the cognitive dimensions of risk perception to the neglect of important social and contextual factors. Bickerstaff (2004:828-29) argues that inconsistent psychological findings, such as varying results between social groups, cannot be understood without looking at the socio-cultural context in which such perceptions were formed. Others argue that while psychological studies have increasingly pointed to a social basis to risk perception they have not been able to expand on their significance because of their undue focus on the individual as the locus of explanation (Nelkin, 1989; Priest, 2003). In response to their critics, some researchers began to go beyond 13 the psychological roots of perceived risk to address more sociological factors such as gender, race, trust, emotions, and vulnerability and justice (Finucane et al., 2000; Kunreuther, Slovic, and MacGregor, 1996; Satterfield et al., 2004). Socio-cultural Perspectives Socio-cultural researchers are highly critical of psychological studies because of their lack of acknowledgement of the real world contexts in which perceptions are inherendy embedded. Thus, socio-cultural studies explore the social and cultural contexts in which risks are understood and interpreted. These contexts included the institutional, social, and cultural settings in which perceptions are formed. The influence of these contexts on risk perception has been used to explain the large degree of variability in risk perception across social groups. While some of these researchers reject the validity of psychological findings altogether (see for example, Otway and Thomas, 1982; Douglas, 1985:29-39), a growing number of researchers believe socio-cultural perspectives provide a necessary complement to psychological findings (Bickerstaff, 2004:830; Wynne, 1987:360). Socio-cultural studies tend to employ qualitative techniques, such as individual interviews or participant observation, to elicit and contextualize the meanings attached to people's perceptions. These techniques are best suited to capturing the complexities and details of how people understand and make sense of risks within the context of their everyday life. As observed by Lupton (1999:108), while arguments about risk acceptability are debated in high-level discussions amongst policymakers and scientists,they are dealt with by most individuals at the level of the local, the private the everyday and the mtimate'. At the same time, evidence suggests that risk perceptions are also the product of people's communication and interaction with others. Risk positions are confirmed, adapted or created based on people's response to and interactions with surrounding social networks (MacGill, 1989:57). A recurrent theme throughout the socio-cultural: literature is the concept of identity, particularly as it relates to risk selection. The interactive relationship between individuals and the people around them develop and support social cohesion or group identity (MacGill, 57, following the work of Mary Douglas, 1966; 1992). Identities create a sense of unity among 14 people who share common beliefs and values. When valued identities are threatened, social action is taken to protect those identities or ways-of-life. Identity is the basis for understanding the relationship between risk perception and the concepts of sense of place and trust. 2.5 Concepts Addressed in this Study Intuitive Toxicology: Chemical zersus Microbial Contarrinants Intuitive toxicology studies are relevant to understanding the situation in Erickson because it is here that public perceptions of chemical risks (and thus by implication chlorine) have been explored most extensively. The public's negative perceptions toward chemicals in general have been well documented (McCallum et al., 1990; Kraus et al., 1992). Psychometric findings have demonstrated that many of the risks people fear most are related to chemical technologies (Slovic, 1987). O f particular significance is the way people think of the concept of dose. Studies have shown that the public is largely dose insensitm when it comes to judging chemical risks (Kraus et al., 1992; Slovic et al., 1993). In short, many chemicals are regarded as highly toxic or risky regardless of dose. This view contrasts sharply to toxicologists' view of chemicals, whose most important disciplinary tenet is indeed 'the dose makes the poison'. Whether or not chemicals are natural or synthetic also seems to play a role (Kraus et al., 1992). Many people believe natural chemicals to be much safer then synthetic ones, with the exception of the use of chemicals for medicinal purposes (Slovic et al., 1995). Dose in the form of microbial contaminants is a topic of central importance to understanding the chlorine debate, since the presence of potentially liarrnful n^croorganisms in water forms the only basis as to why chlorination is needed in the first place. From an expert perspective, it is generally agreed among public health experts that waterborne pathogens, such as Giardia and Gypctcxpondfam, present a muchgreater risk to public health than the risks associated with getting cancer from chlorinated by-products. 15 Although the public is largely insensitive to concepts of dose when judging chemicals risks, however, it is not at all clear that the same logic would apply to microbial risks. The two, by their very nature, are quite different. Microbes are nattiraUy occurring, living organisms. By contrast, most chemicals used in industry are lifeless, synthetic substances. Thus one of the theoretical questions this study is interested in is: Ifthe public believes natural ahemcak to be less harrrfiithansyntheticoherrkak, ^ thfy bdieie naturally (mmi^ (hernkals} This area has not been adequately explored for drinking water, and remains a main area of investigation for this study. Specifically, it was expected that residents' would exhibit tolerant attitudes toward microbial risks and that these views would be negatively correlated with perceived chlorine risk. Residents were also expected to be quite sensitive to the concept of pathogen dose, thus implying tolerance toward small amounts of microbial contaminants. In other words, low levels of concern toward microbial contaminants should also be associated with increased aversion to chlorine (i.e. increased perceived chlorine risk). Trust Findings from psychological studies have established that the reason why expert-lay conflicts occur is in part due to the different ways in which experts and lay people think about risk. Socio-cultural researchers go a few steps further than this supposition and assert that public judgements about competency and commitment of managing institutions form the real basis of public risk conflicts. As observed by Fessenden-Raden et al. (1987:100), public trust in the risk messenger is often more important that the information the messenger carries. This is because trust enables individuals and organizations to act without full information. The relationship between trust and information, however, is not a two-way relationship. Information, does not build trust,but rather, makes it redundant' (Hansen et al., 2003:119). At the societal level, it has been widely observed that there is a general climate of distrust between the public and society's managing institutions. Slovic's work in this area provides a simple yet compelling observation about the asymmetric nature of trust: difficult to build but easy to destroy. Widespread decline in public trust has been attributed to a number of societal trends, namely, our adversarial judicial system, biases in media reporting, and the rise of NGOs, which have increasingly challenged the credibility of the government and industry 16 (Slovic, 1993). Kasperson et al. (1992:172) observes that public trust has decreased sharply over the last diirty years (within the United States) and attributes the trend to a number of key events including the Vietnam War, reports of major environmental degradation throughout the 1960s and 1970s, Watergate, and the energy crisis of the 1970s. Gearly, these events have shaped the context in which risk managers must now work in when considering the dynamics of how trust is created or destroyed. Socio-cultural researchers have also found some very interesting relationships between risk perception and trust, at the local level. For example, in her study of community perceptions of risk following the discovery of extensive groundwater contamination involving the chemical trichloroethylene (TCE), Fitchen (1987) traced apparency low levels of community concern to heightened levels trust in local officials. Baxter and Lee (2004) come to a similar conclusion when applied to community perceptions of risk near the hazardous waste treatment plant in Swan Hills, Alberta. Baxter and Lee suggest that the positive community identity people had come to associate with the plant contributed to lower levels of perceived risk One interest of this study is to examine community trust in local officials and sources of information as a possible explanation of the apparency low levels of concern residents of Erickson held toward potential microbial contaminants in their drinking water. More specifically this study asks: At what levd of authority, locally wprozinaalty, did they believe drinking witer should be managed) Presumably, many citizens felt the need to find out more about drinking water risks and the effects of chlorine on human health. But what sources cf information did they consider trustworthy} What influence (if any) did these view have on people's perceptions of chlorine} Sense cf Place This last concept explores a small but growing literature on the relationship between people and places, in the context of shaping local environmental values. Researchers working in the fields of human geography and environmental psychology maintain that the notion of place has a significant influence over how people perceive, interpret and value the environment (Cheng et al., 2003:88). More recently, the role of place has been employed in understanding conflicts over natural resources (Coughlin et al., 1999; Griffin, 1999; Weber, 2000). Sense of 17 place has yet to be fully explored in the risk literature, although one can see an immediate connection between people's attachment to place and the desire to protect it if perceived to be at risk There is no consensus as to an exact definition of 'sense of place'. However, most researchers agree it is an umbrella concept that describes the emotional and cultural bond between people and places. Our sense of place is developed though place-based experiences and the meanings attached to those experiences (Tuan, 1979; Chawla, 1992, Low and Altman, 1992:2). A fundamental assumption is that sense of place is not inherent in the natural environment but given 'life' through human interpretation (Basso and Feld, 1996; Williams and Patterson, 1996). Thus, sense of place plays an important role in shaping our place-based identities as well as expectations of appropriate behaviour in those places. It has been observed that place-based identity and behaviour have a dualistic relationship (Wakefield et al., 2001; Stedman, 2002; Kaltenborn, 1998). On the one hand, social actions organized around place issues can become expressions of the community itself, thus creating identity. Conversely, a community with a strong sense of place presumably shares attitudes, beliefs and values in line with their place meanings and thereby becomes motivated to maintain or protect those meanings. Regardless of how identities are formed, conflicts of opinion between social groups over proposed uses of land (or water) may not be about the physical impact of those activities but rather competing place-based meanings, denoting what activities are acceptable or unacceptable in those places. There is an obvious overlap between place-based arguments for understanding natural resource conflicts and the situation in Erickson. It is possible that the conflict over chlorine was not solely due to differences of opinion over chemical versus microbial risks, but the idea of chlorine as a threat to their collective identity. Evidence from the brief review of the problem context (outlined in Chapter 1) and preliminary interviews suggests that the very idea of adding chlorine to the community water supply seemed set off an element of moral indignation - as if adding chlorine to their water was somehow violating a kind of community ethos. Given that the Creston Valley is promoted as a retirement community, it is likely that 18 residents of Erickson feel strongly toward their place because it is where they have chosen to live the rest of their lives. There is also evidence to suggest that older people tend to be more strongly attached to places they choose to retire in (Rubenstein and Parmalee, 1992). Sense of place may well play a significant role in understanding the cHorine controversy. For the purposes of this study, itis hypothesized that strong sense cfplace, withstood to be a aoninmtioncf enrticml attachment, identity and nxaning, mil be associated wth an increased awrsian to chlorine. 2.6 Conclusion The concepts reviewed here reflect a diverse set of perspectives on risk perception yet also reflect a growing trend toward convergence of two areas of risk research that have historically been divided along conceptual and methodological lines. As criticized by many researchers, psychological approaches are strong on methodology and empirical results, yet lack the detailed explorations necessary for a more complete picture of risk perception. Socio-cultural approaches fill this gap by providing a grounded account of social processes underlying risk perception in the context of real-life risk controversies. This study aims to bring together concepts from both areas in order to gain a more complete understanding of community perceptions of drinking water risk. Psychological perspectives and findings from intuitive toxicology in particular present an opportunity to examine how people think and reason about microbial risks, a theme not yet explored in the risk literature. The role of trust and sense of place are also examined as important socio-cultural factors that provide a more contextualized approach to understanding community perceptions of chlorine risk. The following chapter outlines how these concepts were operationalized, measured, and analysed. 19 Chapter 3 MEASURING PERCEPTIONS: RESEARCH APPROACH AND METHODS The desire to bring together ideas from both psychological and socio-cultural perspectives supports the use of a mixed method or 'pragmatic' approach. Pragmatism is a relatively new approach to social research and came as a practical response to the paradigm 'wars' fought between qualitative and quantitative researchers in the 1960s (Tashakkori and Teddlie, 1998:4). Pragmatist researchers believe in the authority of the research problem (not the paradigm or the method) as the primary determinant of method selection (Tashakkori and Teddlie, 20). They do not adhere to one philosophical camp or another but believe researchers can and should be free to choose from a variety of measurement techniques to suit their needs (Creswell, 2003:12). This is not a trivially constructed freedom but based on the observation that qualitative and quantitative paradigms have always shared similar logics of inquiry and compatible beliefs about social behaviour (Brannen, 1992:7; Tashakkori and Teddlie, 1998:11). 3.1 Research Strategy This study adheres to the principles of the pragmatic approach by using the research problem as a guide for method selection. The research problem outlined in Chapter 1 indicated that without a better understanding of how communities perceive drinking water risks, implementation of more stringent regulations in the future would be difficult. These difficulties would be especially apparent for health officials working with small, rural communities such as Erickson. Current developments in the risk perception literature suggest that a combination of psychological (traditionally quantitative) and socio-cultural (more qualitative) approaches is necessary for a more complete understanding of risk perception. With these considerations in mind, the following mixed method research strategy was developed: 20 Figure 3.1 Research Strategy for this Study Qualitative Phase -» QUANTITATIVE PHASE Qualitative data collection, (interviews) to identify key concerns in the controversy Qualitative data analysis repeated once QUANTITATIVE data collection, (mail-out survey) incorporating themes from interview data QUANTITATIVE data analysis using statistical software package (SPSS 10.0) 1 Interpretation of results This model was adapted from Creswell's (2003:213) 'sequential exploratory strategy'. It is one of the most common mixed method strategies and is primarily used to develop a survey instrument. Research begins with initial qualitative interviews to explore key concerns and ideas. These interviews are analysed for emergent themes and followed up by more interviews if necessary. Concepts from this data are then incorporated into a quantitative survey to measure the extent to which these ideas are shared within the larger population. The advantage of this approach is that it allows the researcher to identify and clarify concepts that had not been thought of before and also enables them to tailor the survey instrument to the contextual setting. The remainder of this chapter explains how both types of these data were collected and analyzed. As indicated by the capitalized ' Q U A N T I T A T I V E ' term in the diagram, the analytical emphasis of this study is on results from the quantitative survey. This was because the goal of the study was to gain a better understanding of how people thought about risk at the community leveL It does not mean, however, that the qualitative stage of the study was any less important. As mentioned in Chapter 2, accounting for the socio-cultural aspects of the controversy required qualitative inquiry to take place before concepts of interest could be confirmed and the survey could be designed. 21 3.2 Qualitative Interviews The purpose of the qualitative interviews was to identify some of the key ideas and concerns people held toward the chlorine controversy. These interviews were semi-structured, meaning a specific set of questions was used to focus the interview, but participants were free to answer in their own words (Patton, 1990:283). Interviews could and did follow different trajectories based on ideas raised during the conversation. The first round of interviews was conducted in February 2002. One of the main purposes of this trip was to gain familiarity with the area and make contact with some of the key members of the community. Recruitment of participants followed a snowball sampling technique, which is a procedure used to locate 'information-rich key informants or critical cases' (Patton, 1990:176). A total of eight participants7 were identified using this technique including the chairman of E W U S , a long-standing trustee of the Erickson waterworks utility and several people involved in the blockade. Interviews ranged from tJiirty minutes to an hour and a half and took place in the homes of the participants. Al l participants expressed they were opposed to chlorine. Conversations centred around three main questions: What is your opinion of the chlorine controversy? Are there any situations in which you think the use of chlorine to treat water is appropriate? And, what are your thoughts about why some people in the community support chlorination? Since this was a preliminary field visit, the first round of interviews were recorded though extensive note-taking. A detailed field journal of observations and experiences was also kept. These data were organized thereafter into categories of potentially relevant themes. This involved highlighting or 'coding' statements that seemed to speak to larger factors at play including those shaping people's ideas about chlorine or drinking water risks in general. Three criteria were kept in mind during this process. First, responses were considered important if the participants) spent a long time talking about it, especially if he or she returned to the same topic even after being directed onto another question. Second, responses were noted if they 7 The use of relatively small sample sizes is typical in qualitative research. The purpose of qualitative approaches to sampling is not generalization (as is the case in quantitative inquiry) but to select information-rich cases for in-depth study (Patton, 1990:153). 22 were highly unexpected. Last, special attention was paid to how participants understood supportive chlorine positions. This helped to determine how participants d^tinguished themselves from others and identify the boundaries of their opinions. Two main themes emerged from the first round of interviews. The first was a general sense of distrust in the provincial government. This was characterized by varying explanations as to the true motives behind the government order to disinfect, ranging from questionable ties to the logging industry to political retribution for matters unrelated to public health. A common statement across five of the eight interviews was their dislike for the 'heavy hand of government'. The second theme was a surprising reference to the pristine quality of Erickson's environment. This was mentioned in two of the eight interviews, both within the context of explaining their reasons why they were opposed to chlorine. Statements about the naturally high quality of Erickson's water and that water from a pristine environment need not be treated characterized this response. Results from this first round of interviews confirmed the decision to pay express attention to the concept of trust, as many observations were consistent with the trust studies reviewed earlier. What was not understood was the relationship between references to Erickson's pristine environment and their concerns about chlorine. The concept of sense cf place was discovered upon returning to the literature. In August 2003, a second round of interviews was conducted to elaborate further on this concept. Four semi-structured interviews were conducted with a different set of participants, again using the snowball sampling technique. This time, interview questions centred on two areas of interest: participants' view of Erickson as a place and as a community opposed to chlorine (for a copy of the complete interview protocol see Appendix I). Each interview was tape-recorded and fully transcribed. Two community meetings were also attended at this time: an E I D annual general meeting, and a private meeting between representatives of Erickson and the Creston town council regarding cost-sharing measures for construction of the membrane filtration plant. 23 Analysis of these interviews, using the same technique described above, revealed more insight into people's views of Erickson as a place. Al l participants expressed that they liked living in Erickson because of its rural lifestyle and pristine environment. As well, there was a distinct social component to people's understanding of the controversy. One participant explained that the reason why the community was able to keep chlorine out for so long was because there was a core group of good organizers with a strong commitment to their cause. This was seen as a good tiling, as it made him feel proud to live in a place where people stand up for what they believe. Another participant explained that many people living in Erickson were long time residents who did not want to see change in their community. This is consistent with a comment made by a participant in the first round of interviews claiming that people who support chlorine are not long time residents. These findings led to the decision to develop a number of sense of place questions for inclusion in the survey. 3.3 Developing the Survey The quantitative portion of the study aimed to measure the social and psychological basis of residents' attitudes toward o!rinking water at the wider community level. The specific objectives of the survey were two-fold: to characterize residents' perceptions of chlorine risk as well as attitudes toward microbial risks and examine the influence of dose sensitivity, trust, and sense of place on perceived chlorine risk These objectives dictated that five main topics be covered in the survey: 1. Perceived sources of risk to drinking water 2. Dose sensitivity 3. Trust 4. Sense of place 5. Perceived chlorine risk (dependent variable) The survey was organized into four sections. The first was called 'Risk Perception' and contained the perwiml sources of risk to drinking water index. The second section, called 'Thinking about Drinking Water' included questions about trust and dose sensitivity. Section three was tided "Water and Your Community' and contained questions specific to the local community setting. These included questions about peroriied oblorine risk and sense of place. Miscellaneous questions designed to assess respondents' attitudes toward microbial risk, local water quality and actions 24 taken in relation to the controversy were also included in this section. Demographic information was gathered in the last section, entitled 'Background Information' (for a complete copy of the survey see Appendix II). The following section explains how each variable was conceptualized and measured in the survey. Permved chlorine risk was treated as the main dependent variable while the remaining four were considered independent variables. The survey was titled 'Risk Perception and Drinking Water Questionnaire'. Most of the content and format of the survey was self-designed although pre-existing questions from other published studies were also incorporated where appropriate. The response format for most questions was close-ended, using a 4-point Likert scale ranging from 'strongly disagree' to 'strongly agree', except where indicated otherwise. Variable 1 - Pendwd sources (frisk to drinking zmter This variable measured how risky a variety of 28 activities or situations were thought to be to drinking water. Items in the index reflected many of the current drinking water concerns shared by health experts, government agencies, and the public. Respondents were asked to indicate the degree of risk to drinking water quality they associated with each of the 28 items. A general health risk perception index comprised of 17 potential risks to health was also included to located where people placed drinking water relative to a number of other health risks. Variable 2 - Dose sensitmty Ten questions related to dose sensitivity were included in the survey. These questions were adapted from intuitive toxicology studies on chemicals mentioned earlier. Since one of the main interests of this study was to ascertain how sensitive respondents were to the concept of dose in the form of microbial risks some wording changes to the original questions (on chemicals) were necessary. After some deliberation, the term 'disease-causing microorganisms' was decided as the best substitute for the word 'chemicals'. A n example of a modified dose question is: 25 Strongly Disagree Disagree Agree Strongly Agree There is a safe level of exposure to disease-causing microorganisms. 1 2 3 4 Three unmodified questions related to chemicals and radiation were also included for comparative purposes. Variable 3 - Trust Trust in local versus provincial authorities was measured in two ways. First, four questions were asked regarding the level of authority (local versus non-local) they believe drinking water should be managed. A n example of one of these questions is: Strongly Disagree Disagree Agree Strongly Agree I trust the provincial government's ability to manage any risks posed by waterborne pathogens. 1 2 3 4 Second, respondents were asked to indicate how much information they received from a variety of sources about local drinking water quality. This style of question was adapted from the Slovic et al. (1995) study. Responses were rated on a 5-point scale, ranging from 'almost no information' to 'a lot of information'. A n example of what this question looked like is: Using the scale below, please iridhite how much irformation you haw gotten from that source that relates to drink ing voter quality in E rickson: Almost no A lot of information information The Provincial Ministry of Health The Creston Valley Advance Medical doctors 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 26 A mixture of local and regional, formal and informal sources of information was represented by this list. After rating how much information they received, respondents were asked to indicate how much confidence they had in each source. Variable 4 - Sense of place Designing a set of questions to measure sense of place was the most challenging part of the survey to develop. This was due to the lack of consensus within the sense of place literature as to what defines a good measure of sense of place. As noted by Williams and Vaske (2003), problems of item selection and reliability, the dimensional structure of the construct (is sense of place multidimensional or one-dimensional?) and evidence of construct validity (how well do questions actually measure the theoretical idea of 'sense of place'?) characterize this dilemma. A review of the place literature reveals that most researchers believe sense of place to be a multidimensional construct Qorgensen and Stedman, 2001; Stedman, 2002; Kaltenborn8, 1998; McAndrew, 1998). O f these studies, Jorgensen and Stedman explore its specific dimensions most thoroughly and thus was used as a guide for designing this part of the survey. Seven questions were borrowed directly from Jorgensen and Stedman. These were conceptualized along three dkatmiom: place identity (^a.t degree to which a person's identity is defined by a pkce), place attachment (the emotional bond between people and places) and place aependence (the degree to which a certain place meets the needs and desires of a person, given a range of other alternatives). Five additional questions were constructed, incorporating ideas from the qualitative interviews. These were labelled 'symbolic' questions, as they accessed ideas about how meaningful Erickson was as a place. Examples of each of these questions are: 8 In this study Kaltenborn first assumes that sense of place is a unidimensional construct, but by the end suggests that it is probably multidimensional. 27 Strongly Disagree Disagree Agree Strongly Agree Everything about this place is a reflection of me. (place identity) 1 2 3 4 1 really miss this place when I'm away for even a short length of time (place attachment) 1 2 3 4 This place is the best area for doing the things 1 enjoy most, (place dependence) 1 2 3 4 For me, personally, many of the stories that make up my life are intimately connected to this place, (symbolic question) 1 2 3 4 Pevodwd chlorine risk - Dependent variable The main objective of this study was to identify the underlying factors behind residents of Erickson's apparent aversion to chlorine. Five questions were use to measure their attitudes toward chlorine, which subsequendy formed the dependent variable of this study. These questions were direct statements taken from the interviews. An example of one of these statements is: Strongly Disagree Disagree Agree Strongly Agree I try hard to avoid drinking chlorinated water, even in Creston. 1 2 3 4 Misodlaneous questions Miscellaneous questions included measures of attitudes toward microbial contamination and local water quality such as, 'I don't worry too much about getting sick from my water because there are so many other things in my life I have to deal with' and T think it's fine to drink the water in Erickson as is'. Several questions about actions taken in relation to the controversy were also included such as, Did you participate in the blockade at the water intake, in the spring of 1999?' (yes/no). 28 Demographic variables Demographic information including age, income, education and gender was collected at the end of the survey. Respondents were also asked where they were bom and how long they had lived in Erickson. 3.4 Administering the Survey In February 2003, a preliminary survey was piloted tested with a convenience sample of thirty UBC students from the Department of Resource Management and Environmental Studies (RMES)9. After presenting the research interests of this study and a brief synopsis of the controversy, students were asked to complete the survey and provide feedback as to its strengths and weaknesses. Specific feedback was sought on the visual appearance of the survey, length of time taken to complete it, clarity of questions, and logic of the questions given the concepts of interest to this study. Several wording and layout changes were made to me survey based on this session. The eligible sampling population was defined as all people living within the geographic boundaries of Erickson10 who are over the age of 19. Names and addresses of potential participants were obtained using the 2002 Creston area phonebook. Identifying the sample was somewhat of a challenge, as Erickson is an unincorporated community and thus did not have its own section in the phonebook. To determine which households were actually located in Erickson it was necessary to look up each entry in the phonebook and check if that address was actually within the boundaries of Erickson. Eight hundred and thirty-four addresses were identified this way and entered into an Excel spreadsheet. The reliability of this number was verified with the office manager of the Erickson waterworks utility who confirmed that this number was very close to the registered number of water users in Erickson. Using a sampling analysis tool in Microsoft Excel, a random list of 250 households was selected as potential participants in the survey. The decision to survey 250 households was based on cost 9 Ideally, the survey should have been pilot tested with a group of people more representative of the Erickson population, but for reasons of time and cost considerations this was not possible. 1 0 This was defined as the area east of 25 Ave., south of Erickson St. and west of (but not including) Lakeview-Arrow Creek Rd. The Goat River forms the southern boundary of Erickson. As Erickson is an unincorporated community, defining its borders was a challenge. The boundaries selected for this study were determined following consultation with the Creston Chamber of Commerce and the Creston Town Hall. 29 considerations and an anticipated response rate of 30% (considered average for a mail-out survey). Two hundred and fifty surveys were mailed out in July 2003. Each survey package contained an introductory cover letter, one 15-page survey, and a self-addressed stamped envelope. To minimize gender bias, respondents were asked in the introductory letter to have the adult who most recendy celebrated his or her birthday fill out the survey. This served as a simple way of ensuring that the gender characteristics of the population were reflected in the sample. Two weeks after the initial mailing follow up postcards were sent to households that had not replied. This was repeated again two weeks later. The response rate of the survey was 33% (n=82). This response rate produced a confidence interval of plus or minus 10% at a 95% confidence level. The implications of this result on the generalizability of this study are discussed at the end of this chapter. Demographic comparisons between the survey sample and the larger Erickson population were not possible, due to the unincorporated status of the community. Statistics from Creston were used as the next best alternative (British Columbia Ministry of Health Services, 2004; Statistics Canada, 2001). Respondents from Erickson had a similar percentage of males as the Creston population (50% vs. 49%); they were older than people in Creston (median age of 57 versus 45.5 in Creston) and well educated (91.5% had high school education or higher versus 75.3% in Creston). The median income bracket of the Erickson sample was $30 000 - $39 999, which is slighdy higher than the median Creston income of $23 935. Al l the results from the survey were analyzed using the statistical software program SPSS version 10.0. Simple descriptive statistics such as graphing means and calculating response percentages were employed to address the first objective of this study, which was to characterize respondents' perceptions of risk. Particular attention was paid to results from the first section of the survey which included the penmedsources cf risk to drinking mtter variable. More in-depth statistical procedures were required to address the second objective of this study, which sought to learn more about the relationships between the variables. Principal 30 components factor analysis was used to identify and construct composites indices of the perceived sources of risk to drinking water, dose, trust, sense of place variables as well as the perceived chlorine risk scale. Cronbach alpha coefficients were used to establish the reliability of each scale (the specifics of these calculations are explained in the appropriate areas of the chapters below). The relationships between these variables were explored using Pearson correlations and step-wise multiple regression techniques. 3.5 Validity, Generalizability and the Limitations of this Study Validity in the social sciences is a measure of instrument accuracy. It asks: to what extent does an instrument measure that which it is supposed to measure (Sproull, 1995:74)? Two types of validity estimates are applicable to this study. They are face validity and construct validity. Face validity refers to the appearance of a measure 'on the face of it' (Cozby, 1993:32); it asks how well do the questions appear to relate to the concepts of interest? Face validity in this research was primarily established during the pilot study when students were asked to evaluate the logic and clarity of the questions. The use of direct statements lifted from the qualitative interviews also helped to establish face validity. For example, most of the perrniedoblorine risk questions were developed this way. Construct validity is defined as the extent to which a measure 'relates to other variables in a meaningful way* (Cozby, 1993:32). Cozby observes that construct validity is seldom established in a single study, but built up over time as other studies lend insight into the theory behind particular constructs of interest. For this research, construct validity was established more for some variables than others. For example, questions pertaining to dose had a higher construct validity than those used to measure sense of place because of their successful use in other studies. By successful I mean that these questions have consistently been used to distinguish between groups of people (i.e. expert and non-expert ways of tiiinking about dose) and that these distinctions have been replicated across several studies. The same cannot be said of the sense of place questions, since the concept remains largely unexamined in the context of understanding natural resource controversies. Similarly, construct validity could not be established for the trust questions, as they were largely self-designed and thus have not been used in other studies. 31 Generalizability in survey research is largely determined by the sampling procedures used and the'sample size (Tashakkori and Teddlie, 1998:71). The randomized selection of potential respondents in this study combined with efforts to control for gender bias helped to increase sample representativeness. However, the response rate of 33% produced a rather large confidence interval (+/-10% at the 95% confidence level), and thus served to limit the generalizability of these results to the larger community. Nevertheless, similarities mentioned earlier between the sample and the surrounding Creston population suggest at least a satisfactory level of generalizability in terms of demographic qualities (assuming of course that they type of people living in Erickson are similar to those living in Creston). What became apparent while implementing the survey were the constraining effects of research costs and data availability when attempting to work with a rural community. As there was only so much money available to print and send out the surveys only a limited number of surveys and follow up postcards could be sent. Had additional surveys been sent along with the follow up postcards, the return rate would probably have been higher. Erickson's status as an unincorporated community also limited the effectiveness of the sampling procedures as potential respondents had to be identified by hand and survey delivery was dependent on the informal knowledge of the local postal worker to recall each respondent's mailbox number11. Despite these limitations, the findings of this research provide valuable insight into how residents of Erickson viewed and understood risks to their drinking water. The findings of this research are reported in the remaining chapters of this thesis. 1 1 Most rural communities do no have door-to-door mail delivery and instead use rented postal boxes. As these are not published in telephone books I could not address the survey to a specific mailbox. The solution was to ask the local postal workers to look out for my surveys and put them in their respective mailboxes when they arrived. 32 Chapter 4 RESULTS I: CHARACTERIZING COMMUNITY PERCEPTIONS The first main objective of this study was to characterize community perceptions of risks to drinking water. This was addressed by exploring respondents' perceptions of risk to health and drinking water, attitudes and behaviours in relation to chlorine risk, and attitudes toward microbial risk. Taken together, these findings provide a statistically descriptive portrait of the community's perceptions of drinking water risk as well as some explanations of the specific attitttdinal factors underlying the chlorine controversy. This chapter begins with a review of respondents' ratings of perceived risk, both to health and drinking water. This is followed by an examination of the extent to which respondents were averse to cUorine and took actions to oppose it. Lastly, respondents' views toward microbial risk are explored in reference to their knowledge of and sensitivity to dose, general attitudes toward microbial contaminants in the environment, and beliefs about local drinking water quality. 4.1 Ratings of Perceived Risk General Health Risk Rating In the opening sections of the survey, respondents were asked to rate the risk they associated with a variety of health hazards. The purpose of this was to locate respondents' perceptions of drinking water relative to a range of other health concerns and to compare these responses to those of the Canadian public (Slovic et al., 1993). Figure 4.1 presents the response distributions for each of the 17 items. As indicated in this figure most respondents were not highly concerned about dtinking water relative to a number other health concerns prevalent in Canadian society. This probably reflects the fact that by-and-large people believe their tap water is safe to drink unless otherwise alerted to an acute contamination event; one might reasonably expect some increase in perceived drinking water risks over time given the events of Walkerton. Parallel findings have also been reported in other studies, indicating this view is also shared among the larger Canadian population (Slovic et al., 1993, Ffcalth Canada, 2000). The risks respondents were most concerned about included street drugs and asbestos, 33 followed by several lifestyle risks such as heart disease, cigarette smoke and infectious diseases (e.g. AIDS, West Nile virus). Prescription drugs, drinking water and fluoridation • considered least risky to health. were Figure 4.1 Perceived Health Risk to Canadian Public Perceived Health Risk for General Hazard I t e n r e - ^ Response Category (h*=82) Chronic diseases Cigarette smoke Infectious diseases Depletion of the ozone layer Motor vehicle accidents Pesticides in food Moulds in food Nuclear power reactors Mercury in dental fillings Waste incinerators Use of genetically engineered bacteria Indoor air quality Prescription drugs Drinking water Fluoridation of drinking water 20% 40% 60% 80% 100% I Ugh Risk BMacfrateRisk •9jght Risk B No Risk B No Answer These results compared quite well to findings from the Slovic et al. (1993) study12 thereby indicating some consistency of Erickson's beliefs with Canadian society as a whole. Although street drugs and cigarette smoke did not rank in the same order on the two surveys, both appeared among the top five risks to health. Other areas of similarity include the perception that chemicals in the form of prescription drugs are less risky than chemicals in the form of 1 2 It should be noted that the 1993 Slovic et al. study was not replicated exactly, simply due to survey length and time constraints; the onginal list of 33 hazard items was reduced to 18 in this study. 34 street drugs, pesticides, cigarette smoke and waste incinerators. A common explanation of why Erickson was opposed to chlorine was that the community itself was not 'normal' in terms of the way they perceived risks (held by some water quality experts and public health officials). Results from this exercise, however, challenge this idea by demonstrating that, at least in terms of general health risk perception, the respondents' from Erickson were not so different than other Canadians. Permved Sources of Risk ta Drinking Water Following this general portrait of health risks, respondents were asked to rate the risk they associated with 28 potential sources of drinking water contarnination. These items were designed to reflect a variety of water quality issues, identified by water quality experts and governments as matters of concern for public health. This included such activities as clear-cut logging, discharging industrial waste, resident wildlife in watershed, and cattle grazing. Figure 4.2 presents the response distribution for each of the 28 items. As illustrated in the figure, respondents were more concerned about risks from industrial pollution and resource development than microbial hazards. Pollution risks associated with discharging currendy unregulated chemicals and industrial waste drew the greatest percentage of high-risk responses (73.2% and 69.5% respectively). Interestingly, other high risk ratings include those affiliated with a policy decision - 'restricting the public from having a say in environmental protection regulations' - which outranked the risks associated with clear-cut logging, pesticide use in nearby farms, and mining. It is signif icant that a policy decision would be ranked higher than the threat of such activities as clear-cut logging and mining. This suggests that participation in the decision-making process is meariingful in the minds of respondents and that restricting public involvement may actually increase risk perception. 35 Figure 4.2 Perceived Sources of Risks to Drinking Water Quality Perceived Risk to DrinkingWbter Quajily- By Response Category (N=82) nscharg'ng currently unregulated chemicals Discharging industrial waste Restricting public from say in EP regulations Clear-cut logging Pesticide use in nearby farms Mring Use of nearby roads for transport of hazardous waste Discharging treated human sewage (meets regulatory standards) Build-up of bacteria in distribution lines Increasing axTrnerdal access to land Cattle grazing Build-up of algae in distribution lines Spreading manure on fields in WINTER Poor soil drainage Naturally occurring arsenic Spreading manure on fields in SUMVER Recreational boating Steep slopes in watershed Resident wildlife Heavy rainfall Sncrwmelt/Spring runoff Camping Chemically disinfecting water - Ozone Selective logging Water treatment - Merrbrane filtration • Water treatment - UV radiation Water treatment - Sand filtration Using home water filtration systems 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% IHghRiskOlvbderateRskBSIigrtRskBNoreskraNoAnswer Overall, microbial hazards were viewed as only moderately risky to drinking water. These included such activities as discharging treated human sewage, build up of bacteria or algae in the distribution lines, cattle grazing, spreading manure on agricultural fields in winter or summer, resident wildlife in the watershed, heavy rainfall, and spring run-off. Of particular interest was the placement of these last three items, all perceived by water quality experts to be major threats to drinking water quality but here assigned relatively low risk ratings by this sample. The results suggest that an area of potential conflict still exists between the 36 community of Erickson and the government, particularly in terms of differing perceptions of microbial risks. Placement of all the alternative treatment options (other than chlorine) at the bottom of the list is also noteworthy, considering the increased attention given to these technologies in recent years. It is reassuring to see that membrane filtration placed so low on the index since this is the type of water treatment currendy being implemented in Erickson. Similarly, the ranking of UV radiation and sand filtration as least risky to obinking water is not surprising as these were the types of treatment the community had sought from the beginning. Interestingly, none of the items assigned the highest ratings of risk are applicable to Erickson. Namely, there is no current industrial activity in the watershed nor are there plans to be in the future. No substantial resource development is taking place, although the prospect of logging is a possibility in the future. Relevant threats to local drinking water quality include many natural conditions rated relatively low in risk such as resident wildlife in the watershed, steep slopes in the watershed, heavy rainfall events, and spring-run off. FactorAnalysis of'Drinking Water Risks A high degree of intercorrelation between the items suggested that to some extent respondents were thinking of these drinking water items in terms of specific hazard categories. Factor analysis13 was performed to further clarify what these categories might be. The most common method of factor analysis is principle components analysis (PCA). PCA of the 28-item set revealed the presence of three main factors, which explained 30.0% of the variance. Factor 1 contributed 10.2%, factor 2 contributed 10.0% and factor 3 contributed to 9.8 % of the variance. Optimal item loadings for each factor were identified using Varirnax rotation14, to facilitate interpretation of the factors (Table 4.1). 1 3 Factor analysis is a statistical technique used for taking a large set of variables and reducing them to a group of key factors that best represents the data. 1 4 Factor rotation is a way of optimizing the interpretability of results by increasing the likelihood of all the items in factor to have some sort of logical association as well as high degree of intercorrelation. 'Varirnax' means that each factor extracted is unrelated or operating independently of the other factors 37 Table 4.1. Rotated Factor Loadings for Drinking Water Risks Item Factor 1 Factor 2 Factor 3 'Pollution Risks' 'Resource 'Naturally Occurring Development Risks' Risks' Discharging industrial waste Discharging unregulated chemicals Pesticide use in nearby farms Nearby roads transport hazardous waste Clear-cut logging Mining Selective logging Increased commercial access to land Recreational boating Spring run-off Resident wildlife in watershed Steep slopes in watershed Heavy rainfall Bacteria in distribution lines .88 .86 .54 .53 .75 . .74 .66 .61 .49 .73 .69 .68 .59 .48 Cronbach alpha Variance Explained .79 .78 .72 10.2 10.0 9.8 The analysis revealed three main categories of hazards underlying the responses. Factor 1, termed 'pollution risks' consisted of four items pertaining to different types of chemical pollution, including: cliscliarging industrial waste, discharging currently unregulated chemicals, pesticide use in nearby farms, and using nearby roads to transport hazardous waste. Factor 2, termed 'resource development risks', consisted of five types of development activities including: clear-cut logging, mining, selective logging, increased commercial access to land, and recreational boating. Factor 3 was termed 'naturally occurring risks' because all the items referred to natural conditions that pose a threat to drinking water quality, including: spring run-off, resident wildlife, steep slopes, heavy rainfall, and build of bacteria in distribution lines. Inspection of the Cronbach alpha scores15 for each factor indicated a sufficient degree of internal reliability for their use as representative measures of the perceived sources of risk to drinking water index. 1 5 As a rule-of-thumb, the Cronbach alpha coefficient of a scale should be above .7 (Pallant, 2001:85) 38 Interestingly, the classification of concern indicated by these categories appear broader than those considered by public health officials as justification for the disinfection order. Public health officials were prirnarily concerned with naturally occurring risks, notably resident wildlife in the watershed (e.g. beavers). Results from the factor analysis, however, suggest that respondents were also classifying drinking water risks in terms of resource development and pollution in addition to naturally occurring conditions. The relationships between these factors and respondents' perceptions of chlorine risk are explored in the next section of this chapter. 4.2 Attitudes and Behaviours Ta ken in Relation to Chlorine Respondents' attitudes toward chlorine risk were measured through as set of five agree/disagree type questions. Several questions pertaining to behaviours taken in relation to the controversy were also included in the survey. Perceked Chlorine Risk Responses to the five chlorine questions indicated a strong anti-chlorine sentiment shared among respondents (Table 4.2). Specifically, a majority of respondents: • Tried hard to avoid o^inking cHorinated water, even in Creston (65.9% combined agree and strongly agree) • Believed that Erickson needed some form of water treatment just not chlorine (79.3%) • Disagreed that they didn't mind drinking chlorinated water (74.4%) • Questioned the effectiveness of chlorine for the protection of public health (69.5%) 39 Table 4.2 Perceptions of Chlorine Risk Question Strongly Disagree Disagree Agree Strongly Agree No Answer C1.1 try hard to avoid drinking chlorinated water, even in Creston. 4.9 26.8 25.6 40.2 2.4 C2. Erickson should have some form of water treatment, just not chlorine. 8.5 11.0 46.3 32.9 1.2 C3.1 do NOT really mind drinking chlorinated water. 50.0 24.4 19.5 2.4 3.7 CA. 1 question the effectiveness of adding chlorine to our water for the protection of public health. 8.5 20.7 26.8 42.7 1.2 C5. Chlorine adequately disinfects water from a number of contaminants you often find in water. 14.6 34.1 39.0 7.3 4.9 Cell entries are percentages Note for question C5, respondents were evenly split as to whether 'chlorine adequately disinfects water from a number of contaminants often found in water'. This suggests some uncertainty as to benefits and limitations of chlorine. A possible explanation is that while respondents believed chlorine to be an adequate means of protection against some waterborne pathogens, it is not effective against others (such as Giardia and Cryptapvndiwn) and thus was not judged an effective means of protecting public health (consistent with question C4). For analytic purposes, these five questions were combined into one scale termed 'perceived chlorine risk'; the scale had a mean of 3.74 on a 4.0 scale (zero indicating low perceived risk, four indicating high perceived risk or 'aversion to chlorine'). The Cronbach alpha score for these five questions indicated an adequate degree of internal reliability (.74). This scale was then operationalized as a dependent variable, to be used in conjunction with other question sets that comprise the balance of this study. 40 Behaviours Taken in Relation to the Controversy Individual actions taken in relation to the controversy were assessed through four questions. In the preamble, respondents were asked to indicate whether they had participated in one or more of the following actions (Table 4.4). Table 4.3 Behaviours Taken in Relation to the Controversy Question Yes No No Answer B1. Participated in the blockade at the water intake area in spring of 1999? 20.7 76.8 2.4 B2. Displayed a 'No Chlorine' sign on your property? 45.1 51.2 3.7 B3 . Contributed financially to Erickson Water Users Society or the Water Action Group? 45.1 51.2 3.7 B4. Boil your water before drinking it? 32.9 64.6 2.4 Cell entries are percentages Respondents' aversion to chlorine was moderately supported by actions taken in relation to the controversy. About 20% of respondents actually participated in the blockade, while a larger percentage made financial contributions to E W U S or W A G (45.1%). The same percentage also displayed a TSTo Chlorine' sign on their property. Note that within E W U S a rule was made that members involved in the legal proceedings of the dispute should not participate in the blockade. This could explain why only 20% of respondents participated in the blockade. The most surprising result, however, was the high percentage of respondents who admitted they did not boil their water before drinking it, indicating a considerable degree of non-compliance with the long-standing boil-water advisory. Candauon betmen Drinking Water Variables andPeneized Chlorine Risk The relationship between perceived chlorine risk and the three drinking water variables was examined using Pearson product-moment correlation coefficient (Table 4.3). 41 Table 4.4 Pearson Correlation Between Drinking Water Variables and Perceived Chlorine Risk Drinking Water Variable Perceived Chlorine Risk Pollution Risks .072 Resource Development Risks .518" Naturally Occurring Risks .100 ** Correlation is significant at the p < 0.01 level There was a strong, positive correlation between resource development and perceived chlorine risk (r = .518**, p < .01), with higher levels of perceived risk to drinking water quality from resource development (such as logging and mining) associated with higher aversion to cMorine. This was surprising, as the logical relationship between cMorination and resource development is not immediately apparent. Another surprising result was a distinct lack of correlation between naturally occurring risks and perceived chlorine risk, as one would have at least expected a negative relationship between the two (suggesting an 'unusual' tolerance toward naturally occurring risks). The relationship between perceived chlorine risk and resource development begins to make sense, however, when viewed in terms of the local community context. During the interviews, several participants voiced their concerns about logging often in the same breath as their explanations as to why they were opposed to chlorine. "When asked to clarify the relationship between logging and chlorine, reference was made to the questionable relationship between the Ministry of Forests and the Ministry of Health. In the words of one participant, the battle against chlorine was clearly connected to the threat of logging: We believe that the forest industry was behind chlorination. As a group [EWUS] we kind of had to choose our battle. Some people wanted to address logging, some wanted to protest chlorination and some both, but in the end we picked chlorination because we felt that it was our best chance of winning. The logging issue kind of seemed like a done deal and there was nothing we could do to stop it. The 'done deal' referred to the logging license granted to the Creston Valley Forest Corporation (CVFQ in 1997. Formation of the CVFC was controversial. Many residents were opposed to the idea of logging in the watershed. Others were supportive because it left 42 resource management in the hands of a local company and some believed careful logging of the watershed might improve the health of the forest. Residents opposed to logging, however, took the disinfection order as a signal from the government (and the C V F Q that logging would soon begin. These views suggest a significant degree of distrust in the provincial government. Many residents believed that government interests in logging could only be served if the community began treating the water. The rationale was that the Ministry of Forestry would be negligent to approve logging in a watershed that served untreated drinking water to the public and Ministry of Health officials would be hard pressed to explain why they had not done their jobs if a public health crisis had occurred in Erickson. As another participant observed, 'after the logging license was granted the pressure to chlorinate grew. So we started to do a lot of educating about the health effects of chlorine'. Thus for some residents the health risks associated with chlorinated by-products were not seen as a stand-alone reason to oppose chlorine, but as an alternative method to protest logging in the watershed. 4.3 Attitudes Toward Microbial Risks As explained in Chapter 2, the topic of microbial risk is of central importance to the chlorine debate, as the presence of potentially pathogenic microorganisms in water forms the only reason why chlorination is needed in the first place. In this regard, respondents were expected to have generally tolerant attitudes toward microbial risks and that these attitudes would be inversely related to respondents' perceptions of chlorine risk The following section presents the responses to three sets of questions pertaining to microbial risks: sensitivity to pathogen dose, attitudes toward microbial contaminants in the environment, and beliefs about local water quality. Dose Sensitivity Previous intuitive toxicology studies found dose sensitivity (or lack thereof) to be an extremely robust concept in terms of expkining the difference between expert and public understandings of perceived risk It has often been found that the public is largely dose insensitive when judging risks from chemicals as compared toxicologists' view that 'the dose 43 makes the poison' (Kraus et al., 1992; Slovic et al., 1993; MacGregor et al., 1995). This translates into a widespread public belief that any exposure to a toxic substance or carcinogen, no matter how small, is likely to prove harmful. This is a topic of obvious relevance to the chlorine debate. Lack of sensitivity to dose from chemicals provides a logical explanation as to why people might be opposed to small amounts of chlorine. But one can also convert this question and test its apphcability in other dose domains, and thus ask; how tolerant are people to small amounts of microbial contaminants? The question is also theoretically interesting as dose sensitivity has yet to be explored for microbial risks. Table 4.5 presents the responses to these microbial dose sensitivity questions. Overall these findings contradict the parallel evidence for dose sensitivity in regards to chemical risks. That is, unlike chemical risks, dose sensitivity was evident in that most respondents exhibited a significant degree of tolerance toward small amounts of microbial contaminants (Table 4.4). Specifically, a majority of respondents: • Agreed that low level concentrations of microorganisms in water were okay to drink (question D4,64.7% combined agreed or strongly agreed) • Agreed that the more often a person is exposed to disease-causing microorganisms, the more likely they are to get sick (question D5, 58.5%) • Agreed that whether or not a person gets sick from disease-causing microorganisms depends on the dose they are exposed to (question D6,65.8%) • Disagreed that microorganisms are either harmless or dangerous (question D7,65.9%) A slightly lower percentage of respondents: • Disagreed that a trace amount of disease-causing microorganisms would be enough to prevent them from drinking the water (question DI, 50.0%) 44 • Agreed that reducing the concentration of a biological contaminant in a city's dririking water supply also meant that the danger of drinking that water was reduced (question D2,51.2%)1 6 • Agreed that there was a safe level of exposure to disease-causing microorganisms (question D3,54.8%) Table 4.5 Dose-Response Sensitivity to Microbial Contaminants Question Strongly Disagree Disagree Agree Strongly Agree No Answer D1. If a trace amount of disease-causing microorganisms were found in my tap water, I wouldn't drink it. 8.5 41.5 36.6 9.8 3.7 D2. A biological contaminant was found in a city's supply of drinking water in a concentration of 30 parts per million.. .The water was filtered by a process that was able to reduce, but not eliminate the contaminant in the water. Under most circumstances, this means that the danger associated with drinking the water has also been reduced. 1.2 35.4 50.0 1.2 12.2 D3. There is a safe level of exposure to disease-causing microorganisms. 6.1 28.0 52.4 2.4 11.0 D4. Low-level concentrations of microorganisms in water are okay to drink. 7.3 20.7 59.8 4.9 7.3 D5. The more often a person is exposed to a disease-causing microorganism; the more likely he or she is to get sick. 4.9 32.9 56.1 2.4 3.7 D6. Exposure to disease-causing microorganisms would probably make me sick, no matter how low the dose. 7.3 58.5. 26.8 0.0 7.3 D7. Microorganisms are either harmless or dangerous. There really is no "in-between". 11.0 54.9 20.7 3.7 9.8 Cell entries are percentages Importandy, these views contrast sharply to findings from the previously mentioned intuitive toxicology studies on chemicals. For example, in the Kraus et al. study (1992), only 1 8 It should be noted that a high percentage of respondents marked 'no answer' for question D2. This was most likely due to the length of the question. 45 e 34.7% of the public sample agreed there was a safe level of exposure to a chemical that causes cancer (compared to question D3 in which 54.8% of respondents agreed there was a safe level of exposure to disease-causing microorganisms). In the Slovic et al., study (1993), over 73% of the public sample agreed that if a tiny amount of a substance that causes cancer were found in their water they wouldn't drink it (compared to question D l in which only 45.4% of respondents shared a similar view toward trace amounts of disease-causing microorganisms). These comparisons illustrate the stigma surrounding the word 'chemicals' and are particularly surprising considering the already negatively connoted term 'disease-causing microorganisms'. However, respondents' also exhibited an unexpected degree of tolerance toward chemicals (and radiation). Table 4.6 presents the responses to three questions pertaining to chemicals and radiation. As found in other studies, lack of sensitivity to dose was evident in that most respondents disagreed there was a safe level of exposure to a chemical that causes cancer (question D8, 64.7%). However, many respondents also disagreed with questions D9 and D10, suggesting some sensitivity chemical dose. Specifically, a majority of respondents disagreed that exposure to radiation (question D9,70.7%) or a chemical that causes cancer (question D10,65.9%) meant that person would most likely get cancer some day. These results challenge previous findings from other intuitive toxicology studies, in which public respondents were consistendy insensitive to dose from chemicals. Table 4.6 Dose-Response Sensitivity to Chemicals and Radiation Question Strongly Disagree Disagree Agree Strongly Agree No Answer D8. There is a safe level of exposure to a chemical that causes cancer. 23.2 41.5 26.8 1.2 7.3 D9. If a person is explored to radiation, then that person will probably get cancer some day. 8.5 62.2 19.5 2.4 7.3 D10. If a person is exposed to a chemical that can cause cancer, then that person will probably get cancer some day. 3.7 58.5 23.2 3.7 11.0 Cell entries are percentages 46 Qjrrdation between Dose Sensitivity andPerxeiwd Chlorine Risk Factor analysis of the seven microbial dose questions revealed the presence of one factor, which accounted for 29.7% of the variance. This factor, termed 'microbial dose sensitivity' was comprised of three questions: D3 'There is a safe level of exposure to disease-causing microorganisms', D4 'Low-level concentrations of microorganisms in water are okay to;drink', and D6 'Exposure to disease-causing microorganisms would probably make me sick, no matter how low the dose'. Inspection of the Cronbach alpha score revealed an adequate degree of reliability to justify combining the questions into one index (.76). The orientation of question D6 was reversed prior to its inclusion in the index. This was because agreement with question D6 in its original form indicated dose Sensitivity. Interestingly, when correlated with perceived chlorine risk no significant relationship was found (r = .107), suggesting virtually no association between sensitivity to dose from microbial contaminants and perceived chlorine risk This result lends further support to the heretofore-implied idea that residents' opposition to chlorine did not revolve primarily around health concerns. It should be noted that there was one relationship was found when the individual questions included in the 'microbial dose sensitivity' factor were correlated to the individual questions included in the 'perceived chlorine risk' scale. Specifically, there was a moderate, positive relationship between question D5 ('There is a safe level of exposure to disease causing microorganisms') and question CI ('I try hard to avoid drinking chlorinated water, even in Creston), which produced a Pearson correlation coefficient of .254 at a significance level of p <.05. This is a different result than found in previous intuitive toxicology studies on chemicals. In this case, sensitivity to dose (in the form of microbial contaminants) was more closely associated with aversion to chemicals and suggests that the relationship between public perceptions of dose and risk may vary among different risk domains (i.e. the public is largely dose insensitive when it comes to judging chemical risks, but may not be as insensitive when 47 judging microbial risks). There is a need to explore the topic of dose sensitivity more, moroughry, as this finding was only applicable to one set of individual questions. General A tdtudes Towrd Microbial ConUrrmaUon We have established that most respondents' viewed microbial contaminants as only moderately risky to dririking water and were quite tolerant toward small amounts of microbial contaminants, even when presented in terms of 'disease-causing microorganisms'. As such, we would also expect respondents to have generally tolerant views toward microbial contaminants in the environment. The response presented in Table 4.3 confirmed this expectation. Overall, respondents exhibited a low level of concern toward microbial contaminants in the environment. Specifically, a majority of respondents: • Agreed that getting sick from their water was not something they worried about on a daily basis (53.6% agreed or strongly agreed). • Disagreed that water from a pristine source should be treated before dririking it (54.8%) • Disagreed that most microorganisms in the environment were harmful to humans (85.4%). Table 4.7 General Attitudes Toward Microbial Contamination Question Strongly Disagree Disagree Agree Strongly Agree No Answer G3.1 don't worry too much about getting sick from water because there are so many other things in 7.3 26.8 45.1 8.5 12.2 my life that 1 have to deal with. G4. No matter how pristine a watershed seems, the water it 14.6 40.2 31.7 6.1. 7.3 provides should be treated'before drinking it. , G5. Most microorganisms in the environment are harmful to 12.2 73.2 3.7 2.4 8.5 humans. Cell entries are percentages 48 These responses suggest a view of the natural environment as being relatively harmless or benign. Although most respondents acknowledged microorganisms are everywhere in the environment, there are only a few of major concern to public health in everyday life. Living in what is locally perceived to be a pristine environment (such as Erickson) likely minimizes the chances of being affected by these more serious contaminants. It should be noted that although a majority of respondents shared these views, the high percentage of 'no answer' responses suggests some respondents' were uncertain as to their position on these questions. Nevertheless, these views are consistent with respondents' beliefs about local water quality (presented next). Beliefs about Local Water Quality Consistendy, during some of the interviews, participants expressed a certain degree of pride in their water, believing it to be clean and pure ('we have natural high quality water here*) and originating from a pristine environment ('we basically have an untouched watershed, so why spoil it?). By contrast, many of them spoke about chlorine as if it itself was a contaminant ('... for me to go in to drink chlorinated water, it's pretty disgusting actually, just turning on the tap is enough to make you choke and gag"). The essence of these ideas was incorporated into the survey. Responses to these questions are presented in Table 4.4. Table 4.8 Beliefs about Local Drinking Water Quality Question Strongly Disagree Disagree Agree Strongly Agree No Answer E1.1 prefer to drink untreated water. 8.5 24.4 34.1 28.0 4.9 E2. Erickson has naturally high quality water. 3.7 17.1 47.6 . 28.0 3.7 . E3.1 think it's fine to drink the water in Erickson as is. 7.3 22.0 35.4 34; 1 1.2 Cell entries are percentages Most respondents believed that Erickson's water was naturally clean and pure. Over 60% agreed that they prefer to drink untreated water, while 69.5% believed it was fine to drink the 49 water 'as is'. Seventy-five point six percent believed that Erickson had naturally high quality water. These views are consistent with the microbial dose sensitivity data and respondents' belief that water from a pristine environment need not always be treated. 4.4 Summary of Chapter Findings Descriptke Portrait of Community Perceptions This chapter revealed a number of important insights regarding Erickson's perceptions of drinking water risks and attitudes toward chlorine. First, residents' views of general health risks were found to be not so different than other Canadians. Similar ratings of risk for drinking water, chemicals, and other health risks items were consistent with findings from the national study conducted by Slovic et al. (1993). Although the same could not be confirmed for the drinking water risk items (as the index was created for this study), high levels of concern regarding industrial pollutants in drinking water have been found in other water risk perception studies (Ausknder and Langlois, 1999). Second, the extent to which people were averse to the idea of chlorine was widespread among respondents, indicating efforts to oppose chlorine were likely supported by the larger community. Third, the expectation that many residents would have tolerant attitudes toward microbial risks was confirmed. This was reflected in the moderate to low risk ratings for microbial hazards to drinking water, their sensitivity to dose from microbial pathogens, general tolerance toward microbial contaminants in the environment, and positive beliefs about local drinking water quality. Speafk Attitudes Undenting the Chlon^ This chapter also established two important insights into the reasoning behind respondents' aversion to chlorine. Respondents' attitudes toward chlorine did not appear to have any relation (positive or negative) to health concerns related to microbial risks. There was, however, a strong, positive association between the perceived chlorine risk variable and respondents' concerns about resource development. The logical connection behind this relationship became clearer when viewed in terms of the local community context. Statements made during the interviews revealed some residents believed that the order to disinfect was 50 mtimately linked to government interests in logging. Residents who were opposed to the idea of logging believed that by protesting chlorine they might also be able to prevent logging in the watershed. Taken together, the results suggest that the motivating factors behind the controversy were based on more than just disagreements about health. This is problematic because concerns about health form the only substantive arguments as to why chlorine is needed in the first place. However, most respondents did not view microbial risks (from resident wildlife, heavy rainfall events or otherwise) as highly serious threats to drinking water quality and thus had little reason to want chlorine in their water for these reasons. Government efforts to introduce water treatment in the form of chemical disinfection were understandably met with suspicion, given that the community did not view the health concerns associated with microbial risks in the same way many experts do. Even more surprising was the lack of correlation between perceived chlorine risk and dose sensitivity (and naturally occurring risks in general). Thus residents' attitudes toward microbial risks, discrepant as they were to expert views, were not central to explaining their perceptions toward chlorine. Lack of trust, in relation to residents' concerns about logging, seemed to play a larger role in shaping chlorine attitudes. The role of trust and sense of place are explored in further detail in the next chapter. 51 Chapters RESULTS II: EXPLORING T H E SOCIO-CULTURAL BASIS OF T H E CHLORINE DEBATE The second main objective of this study was to explore a number of specific relationships between residents' perceptions of cWorine risk and several potentially relevant psychological and socio-cultural factors. This chapter explores two central socio-cultural concepts (trust and sense of place) in relation to perceived chlorine risk Previous results suggest that lack of trust in the provincial government played a sigriificant role in shaping respondents' aversion to chlorine. This was inferred from the positive correlation between respondents' concern about resource development and perceived chlorine risk, a relationship made clearer through analysis of the interview data. In this chapter, the concept of trust is explored more direcdy by assessing respondents' trust in local versus non-local (e.g. experts and government) authorities as well as their confidence in a variety of drinking water information sources. In Chapter 2, 'sense of place' was also identified as a potentially important concept in understanding public risk perception. Specifically, the idea of 'place-based identity' was proposed as a useful way of mderstanding how people value and hence come to define what are and are not acceptable behaviours in a particular environment. There is an obvious connection between peoples' attitudes toward resource development and specific places; an observation that lends support to the idea that residents' opposition to chlorine was in part due to their attachment to Erickson as a place. Further data for this argument is presented in this chapter. 5.1 Trust One salient feature of residents of Erickson's attitudes thus far is the seemingly low levels of risk they associate with microbial contaminants in water. A review of the trust literature suggests that low levels of concern regarding environmental risks are in part attributable to high levels of trust in local (as opposed to non-local) authorities (Fitchen et al., 1987; Baxter and Lee, 2003). To assess the relevance of these ideas to Erickson, several questions pertaining to trust in local versus non-local authorities as well as their confidence in various 52 sources of drinking water information were included in the survey. The responses to these questions are presented below. Trust in Experts, the Government, and Local Lewis of A uthority Table 5.1 presents the responses to the questions pertaining to trust, included in the Thinking about Drinking Water' section of the survey. Table 5.1 Trust in Experts and Various Levels of Authority Responsible for Managing Drinking Water Question Strongly Disagree Disagree Agree Strongly Agree No Answer T1. Experts are able to make accurate estimates of health risks from harmful microorganisms. 2.4 28.0 61.0 2.4 6.1 T2. I trust what experts say about the risks of UNTREATED water. 25.6 29.3 37.8 3.7 3.7 T3. I trust the provincial government's ability to manage any risks posed by waterborne pathogens. 26.8 51.2 18.3 2.4 1.2 T4. The provincial government can be trusted to properly manage the risks from technologies such as water treatment programs. 25.6 46.3 24.4 2.4 1.2 T5. Local communities can be trusted to properly manage their own water safety programs. 3.7 23.2 52.4 19.5 1.2 Cell entries are percentages These results reflect a general sense of ambivalence toward expert authority. That is, most respondents' viewed expert judgements of risk as credible, although not in its entirety when r^rtaining to matters of local drinking water quality. There was considerable agreement (63.4%) that experts were able to accurately assess the risks associated with harmful microorganisms (question T l ) . However, a majority of respondents (54.9%) disagreed with what experts say about the risks associated with drinking untreated water (question T2). A possible explanation for this is that while respondents' agree with and believe what experts have to say for the most part, when it comes to drinking untreated water they are inclined to 53 disagree because local evidence of pathology is absent, their perceptions of their water as pristine, and the fact that they have been drinking untreated water for many years. Most respondents did not trust the provincial government to adequately manage the risks associated with drinking water and preferred more local forms of dririking water management. Specifically, 78.0% of respondents disagreed that the provincial government could be trusted to handle the risks posed by waterborne pathogens (question T3) and nearly 72% (71.9%) disagreed that the province could properly manage the risks associated with water treatment technologies (question T4). By contrast, the same percentage agreed that local communities could be trusted to manage their own water safety programs (question T5). These results reflect a very negative view of the provincial government. Note for later reference, questions T2, T3 and T4 were combined into one index so as to capture respondents' overall 'trust in expert/government authority. The Cronbach alpha for this index was .73. Trust in Informxtim Sources To assess respondents' view of various sources of dririking water information, respondents' were asked to rate how much information they received from each of 13-sources as well as the level of confidence they had in each source. Figure 5.1 presents the mean ratings of the responses for each source, listed in order of decreasing confidence. The black bars indicate the amount of information obtained from each source about local water quality, while grey bars indicate the level of confidence they associated with each source. Respondents were more confident in local sources of information than other sources, with the exception of the Creston town council. Respondents were most confident in the Erickson Improvement District (EID) waterworks office and this was the source they relied upon the most for information about local water quality. 'Public interest/environmental groups' were ranked second highest in terms of confidence and likely reflects respondents' views toward local citizens groups such as EWUS and WAG. Respondents were more confident in information provided by the local newspaper, 'The Creston Valley Advance', than they were toward other forms of media such as radio and TV. Not surprisingly, medical doctors, friends 54 and relatives, and university scientists ranked quite high in terms of confidence, a finding also reported in other studies of this kind (Solden, 1995; Health Canada, 1993). Figure 5.1 Sources of Drinking Water Information and Confidence in Those Sources Sources of Drinking Water Information and Confidence in Those Sources (N=82) EID waterworks office Public interest/Environmental groups Medical doctors Friends and relatives University scientists Creston Valley Advance Local medical health officer Environment Canada Health Canada TV, radio, other news media Creston town council Ministry of Health Services Private industry Mean I Amount of Information B Confidences in Sources A moderate level of confidence was associated with the local medical health officer (MHO). It is surprising that this source would be rated this high, considering the rancorous debates that took place between the local M H O and members of the community. The results suggest that most respondents do not view the local M H O as an inherently untrustworthy position and recognized that the specific M H O in office at the time was the person with whom they had the most disagreement. This is consistent with statements made during interviews in which two participants expressed their satisfaction and support for the new M H O . Specifically, they referred to his personable demeanour, willingness to work collaboratively with business owners to ensure compliance with the regulations, and his availability should they have any questions or concerns. 55 Respondents were not very confident in the information provided by the provincial Ministry of Health, which was ranked lower than the local MHO, Health Canada, and Environment Canada. This is consistent with the idea that residents' believed there was a 'hidden agenda' behind the disinfection order that was btirnately connected to government interests in logging. Respondent confidence was lowest for private industry, a finding also reported in the Slovic et al. (1993) and Solden (1995) studies. Principal components factor analysis of the 13 items revealed two main factors underlying the data, which explained 50.8% of the variance. The first, tided 'confidence in expert/government sources' consisted of seven expert or government sources: Health Canada, the provincial Ministry of Health, the local MHO, Environment Canada, medical doctors, university scientists, and private industry. These items accounted for 312% of the variance and had a Cronbach alpha score of .87. The second factor, termed 'confidence in local sources', consisted of four sources: The Creston Valley Advance, EID office, 'TV, radio and other news media', and the Creston town council Note that although 'TV, radio and other news media' was a not a local source per se the majority of items in the list did fit this title. The second factor accounted for 19.6% of the variance and had a Cronbach alpha score of .79. Correlation between Trust Variables and Perceived Chlorine Risk In order to determine if there was a relationship between respondents' attitudes toward chlorine and trust, the three trust variables (trust in experts/government authority, confidence in expert/government sources, and confidence in local sources) were correlated with perceived chlorine risk (Table 5.2). There was a strong, negative correlation between 'trust in experts/ government authorities' and perceived chlorine risk (-.552**, p < .01), with lower levels of trust in experts and government associated with higher aversion to chlorine. Low confidence in expert and government sources of information was also associated with higher aversion to chlorine (-291**, p <.01), almough this association was not as strong as the preceding one. There was no association between confidence in local sources of information and perceived chlorine risk. 56 Table 5.2 Pearson Correlation Between Trust Variables and Perceived Chlorine Risk Trust Variable Perceived Chlorine Risk Trust in expert/ government authority Trust in expert/government sources Trust in local sources -.552" -.291" -.006 Correlation is significant at the p < 0.01 level The results suggest that people's aversion to chlorine was strongly linked to their lack of trust in expert or government authorities, and are consistent with results in the previous chapter. Note that this is a variation on findings from the Fitchen et al. (1987) and Baxter and Lee (2003) studies, which concluded that trust in local authorities was one of the main explanatory factors behind low levels of community concern. In this case, lack of trust in non-local authorities appears to be more important than trust in local authorities in shaping residents' perceptions of risk 5.2 Sense of Place In Chapter 2, the concept of 'sense of place' was introduced as a means of gaining deeper insight into the reasons behind aversion to chlorine. During the interviews, it was observed that for some residents the very idea of adding chlorine to their water almost seemed to set off a feeling of moral hdignation. Reactions such as these led to the idea that some residents' viewed the prospect of chlorination as significant threat to the their collective identity. This identity was based on an emotional, psychological, and cultural attachment to the community, and a desire to maintain that sense of community as so imagined. As such, it was hypothesis that having a strong sense of place toward Erickson would also be associated with greater aversion to chlorine. Table 5.3 presents the responses to the 12 'sense of place' questions. 57 Table 5.3 Sense of Place Question Strongly Disagree Disagree Agree S t ! o n & a Agree No Answer •Rlace Attachment ? ,* S1. I feel happiest when I'm in this place. 2.4 14.6 46.3 30.5 6.1 S2. When I'm in this place I feel my most relaxed. 1.2 20.7 46.3 28.0 3.7 S3. I really miss this place when I'm away for even a short length of time. 4.9 24.4 41.5 20.7 8.5 •fplace Identity ~- , • tj- ' *•> i S4. I can really be myself in this place. 0.0 8.5 51.2 35.4 4.9 S5. Everything about this place is a reflection of me. 4.9 35.4 36.6 13.4 9.8 S6. This place says very little about who I am. 8.5 47.6 26.8 8.5 8.5 «!RlaceiDeperidence • S7. This place is the best area for doing the things I enjoy most. 0.0 14.6 50.0 28.0 7.3 S8. I enjoy the routine I have living in Erickson. 0.0 3.7 45.1 47.6 3.7 S9. Every time I pass particular places in Erickson, I am struck by the memory of key events in my life that happened there. 9.8 30.5 41.5 11.0 7.3 S10. As a person who has moved around a lot, if s places in my memory that matter most to me and not Erickson... 19.5 42.7 19.5 4.9 13.4 S11. Most of the important people in my life live (or have lived) in Erickson. 13.4 39.0 26.8 13.4 7.3 S12. For me, personally, many of the stories that make up my life are intimately connected to this place. 7.3 37.8 36.6 11.0 7.3 Cell entries are percentages 58 As expected, most respondents expressed a strong connection between themselves and Erickson as a place. The emotional bond respondents' felt toward Erickson was measured through three questions designed to address place attachment' (questions SI - S3). A substantial majority reported feeling happiest in Erickson (76.8% combined agree and strongly agree), were most relaxed there (74.3%), and really missed Erickson when they were away even for a short length of time (62.2%). The degree to which respondents' personally identified with Erickson was measured through three place identity' questions (questions S4 - S6). Many respondents identified strongly with Erickson. A considerable percentage agreed that they could really be themselves in Erickson (86.6%) and that everything about Erickson was a reflection of who they were (50%). By contrast, many disagreed with the statement, 'this place says very kttle about who I am' (56.1%). Tlace dependence', defined as the degree to which Erickson met the needs and desires of respondents, was addressed through question S7. In this area, a considerable majority of respondents agreed that Erickson was the best area for doing the things they enjoy most (88%). Lastly, five questions capturing the experiential and symbolic qualities associated with place (questions S8 - S12) were included to measure the degree to which respondents' assigned important social meaning to the physical place of Erickson. In this regard, a majority of respondents': • Agreed that particular places in Erickson evoke strong memories of events that occurred there (52.5%) • Disagreed that memories of past places are more meaningful than those associated with Erickson (62.2%) • Disagreed that most of the important people in their lives have lived in Erickson (52.4%) • Agreed that they enjoy the routine they have living in Erickson (92.7%) 59 Although not a majority view, 47.6% respondents' agreed that many of their life stories were mtimately connected to Erickson (question S12). Overall, most respondents possessed a strong sense of place toward Erickson. The Cronbach alpha score for all 12 questions was high (.89), indicating each question seemed to be accessing the same general construct. These questions were combined into one index, termed 'sense of place'. Correlation Between Sense cf Place andPermved Chlorine Risk There was a moderate, positive correlation between 'sense of place' and perceived chlorine risk (.335**, p <.01). That is, having a stronger sense of place toward Erickson was associated with a higher aversion to chlorine. Interestingly, a number of significant relationships were also found between the sense of place index and respondents' beliefs about local drinking water quality (presented in the previous chapter). Having a strong sense of place was also associated with an increased preference to chinking untreated water (.385**, question El), believing that Erickson had naturally high quality water (.459**, question E2), and believing that the water in Erickson is fine to drink 'as is' (.425**, question E3). These results support the hypothesis that the decision to oppose chlorine was partially motivated by a deeper desire to protect their real or imagined collective identity. This identity was embodied in a set of positive values, attitudes, and beliefs held toward the local Erickson environment. It is clear that many residents' believed Erickson to be a place endowed with a naturally clean, pristine environment. This was signalled out as something to be proud of; a befitting complement to the relaxed, rural way-of-life one could experience in Erickson. Being able to drink the water untreated was seen as a 'privilege' and testament to the unspoiled character of the environment. This interpretation of the 'sense of place' data provides a compelling explanation as to why residents were inclined to oppose government efforts to disinfect the local water supply. To the community, the disinfection order carried a powerful underlying message: that the water in Erickson was, in fact, not pure and clean but rather polluted, potentially even dangerous, and 60 by association so to was the place so revered and basic to local identity. That is, the very suggestion of chlorine appeared to contradict all of the positive perceptions of place and environment (as idyllic and pristine) residents associated with Erickson. Submitting to the disinfection order would be an open admission to two things: that the water they had been drinking for years had not been safe to drink after all, and that their 'untouched' environment was no longer or never had been pristine. This was thrown into stark relief by the disinfection order itself which, intentionally or not, depicted Erickson as a setting of questionable environmental purity, knowable only through scientific testing, and potentially rife with pathogens. This was simply not an image of Erickson the community had come to know or were willing to accept from an authority they did not trust. 5.3 Multiple Regression Analysis: Predicting Perceived Chlorine Risk Thus far, our analysis of Erickson has identified a number of significant factors underlying respondents' attitudes and perceptions toward chlorine. Respondents' concerns about resource development, trust in experts and government authorities, and place-based commitments to a positive image of Erickson all appear to have played a role in shaping the reasons behind the chlorine controversy. Interestingly, the most obvious factor - tolerance toward small amounts of microbial contaminants (and naturally occurring risks in general)-had little or no relation to perceived chlorine risk. Each factor carries its own hypothesis or explanation as to the reasons behind the chlorine controversy. The relationship between resource development risks and aversion to chlorine explains the controversy in terms of residents' concerns about logging in the watershed. The negative associations between 'trust in experts and government' and perceived chlorine risk also seems to make sense, especially in relation to community concerns about logging. Why would the community be in favour of chlorination if they had little trust in the authorities telling them to do so? Lasdy, one can understand the relationship between 'sense of place' and perceived chlorine risk when viewed in terms of residents' attachment to the local Erickson environment, as manifest through the images, experiences, and meanings people have come to associate with their home community. It is not clear, however, which factor is most important, since the insights thus far have been limited to analysing relationships between two variables 61 (bivariate correlation). What remains to be established is the relative importance of each factor in relation to perceived chlorine risk. Multiple regression analysis allows for consideration of several different independent variables in relation to one dependent variable. Being a more sophisticated form of bivariate correlation, it expresses the unique contribution of each independent variable to the dependent variable, while controlling for the effect of all the other variables on the relationship. The results of a multiple regression analysis are presented in Table 5.4, with perceived chlorine risk' as the dependent variable and six potentially relevant independent variables. Gender, income, and age were included as demographic controls. Table 5.4. Multiple Regression Analysis: Identifying Predictors of Perceived Chlorine Risk Independent Variable Standardized Regression Coefficient (Beta Value) Significance Gender -.133 .219 Age -.065 .549 Income .073 .456 Resource development risks .355 .001** Trust in experts/government authorities -.398 .000** Sense of place .140 .169 R = .67, R* = .48 Together these variables were strong predictors of perceived chlorine risk (R = .67) and explained a considerable percentage of the variance (48%, as indicated by the R2 value). The two independent variables with the largest regression coefficients (expressed through the standardized beta value) and which produced statistically significant results were 'resource development risks' (j3 = .355, sig.= .001) and trust in experts/government authorities (fi = -.398, sig. = .000). As found earlier there was a positive association between resource development risks and perceived chlorine risk as well as a negative association between trust in experts/government authorities and perceived chlorine risk. Increased concerns about resource development (most likely logging) in combination with decreased trust in expert and government authority were both predictive of greater aversion to chlorine. Also note that 62 upon consideration of all these variables together, the effect of sense of place on perceived chlorine risk became less pronounced. 5.4 Summary of Chapter Findings The results from this chapter confirmed that trust was one of the most important factors underlying the chlorine controversy. Trust was characterized in terms of respondents' attitudes toward local versus non-local authorities as well as the level of confidence associated with various local, government, and expert sources of information. Overall, respondents' held generally negative attitudes toward the provincial government. The Provincial Ministry of Health was viewed as one of the least trustworthy sources of information, second only to private industry in terms of perceived confidence. A n important exception was the moderate confidence rating of the local M H O . This was a surprising result of the chapter, as this was the person with whom the community had the most interaction with and the most disagreement. This result is quite encouraging, as it suggests that with careful communication the relationship between Erickson and the government could be greatly unproved in the future. Results from the multiple regression analysis revealed the factors of trust in expert/government authorities and concerns about resource development to be the strongest predictors of perceived chlorine risk. It is acknowledged that 'lack of trust in expert/ government authorities' could be an artefact of the controversy itself, the end product of years of failed collaborative efforts between the government and the community. However, this hypothesis is not likely in light of Erickson's long-standing concerns about logging in the Arrow Creek watershed. Trust in the provincial government was likely on the decline before the order was even issued and became more pronounced as the controversy unfolded. 63 Chapter 6 CONCLUSION In the years following Walkerton, British Columbia and several other provinces passed new legislation to ensure the safety of public drinking water supplies. Clearly, the success of such commitments will require governments to work more collaboratively with communities, taking into account the needs and concerns of water system purveyors and consumers. Consideration of community perceptions of drinking water risk will be especially important for situations involving small water systems, as their ability to meet the regulations is often limited by a number of financial and circumstantial factors. As illustrated by the case in Erickson, failure to adequately address and resolve government-community disagreements can be costly. The downstream effects of this are not just an inability of governments to realize policy goals, but also decreased chances of public acceptance of other government policies in the future. This study explored how one community thought about and understood risks to drinking water quality. Its specific purpose was to examine the basis of residents' attitudes toward chlorine and, in so doing, shed light on some of the reasons behind the chlorine controversy in Erickson. This was achieved by surveying residents' perceptions of drinking water risks, paying particular attention to the topic of microbial risks, and exploring the relationship between perceived chlorine risk and the role of three orthogonal factors: dose sensitivity, trust, and sense of place. 6.1 Key Findings of this Research Diverging Perceptions ofMicrobial Risks A key finding to emerge from the descriptive data was the considerable tolerance respondents had toward microbial risk. In this regard, we can see why there was disagreement between the community and public health officials, as many of their views regarding microbial risk were opposing. For example, it is generally agreed among public health officials that no matter how pristine a watershed seems, the water it provides should be treated before dririking it. However, most respondents disagreed with this statement and agreed that getting sick from their water was not something they worried about on a day-to-day basis. Similarly, from a 64 public health perspective, it is agreed that naturally occurring hazards such as resident wildlife in the watershed, bacterial growth in the distribution lines, cattle grazing, heavy rainfall events, and the spreading of manure on fields are all high-risk activities to drinking water. Most respondents, however, perceived these hazards as only low to moderate risks to dririking water quality. Concerns m the Past Led to Conflicts in the Present Although conflicting views of microbial risks appears to have formed the most overt basis for disagreement between the community and public health officials, these views exhibited virtually no relationship to perceived chlorine risk. This finding suggested that aversion to chlorine did not revolve primarily around health concerns about microbial contaminants. Instead, aversion to chlorine was most strongly related to lack of trust in expert/provincial authorities in conjunction with increased concerns about resource development. This was one of the most surprising findings of the study and demonstrates how public concerns about past issues can have a significant impact on public perceptions of risk in the present. Researchers studying trust and risk have understood this to be evidence of the relational nature of trust (Wynne, 1987; 1992; Hansen et al., 2003). This view suggests that trust is a product of social interaction, based on experiences between people or between people and institutions. If trust is a 'factor' to be negotiated between parties, it follows that that the viability of future relationships is highly dependent on the quality of the relationship between such parties in the past. According to Wynne, the link between public trust in managing institutions and risk perceptions is intimate and dynamic, risk perceptions being the product of a streamoi public experiences with those institutions. If a relational view of trust is taken, we can begin to understand why some people in Erickson were bringing their concerns about resource development (particularly logging) into their perceptions toward chlorine. For over twenty years prior to the chlorine controversy, the E I D had been making attempts to gain more control over resource management decisions regarding the community's dririking water supply. This involved several applications to the Ministry of Lands to grant them a lease over portions of the Arrow Creek to ensure protection 65 of their drinking water source. These requests, however, were not granted. During the 1980s, members of the community also engaged in a number of government consultation processes regarding integrated watershed planning and land-use agreements. From a community perspective these efforts were successful for a time, as logging did not occur during this period of government consultation. In 1994, however, a moratorium on logging in the Arrow Creek watershed was lifted. By 1997 logging in several local watershed areas, including Arrow Creek, was approved. Many members of the community saw this as a 'loss', and needless to say, community support of the government's system of public consultation was not strong. They had engaged in all the appropriate channels of negotiation yet did not believe their concerns had been adequately addressed. In light of these historical events, we can see how some residents' views of chlorination could not be dissociated from their past experiences with the government over local forest management. Public health efforts to bring about disinfection of the local water supply were not seen as a simple measure to protect public health but rather as another imposition of government standards on an already tenuous community situation. As such, many residents were less than receptive toward facilitating yet another set of government initiatives, in light of their longstanding concerns about resource development and more recent failure to prevent logging in the Arrow Creek watershed. From this perspective, public health officials were not viewed as just employees of the local health department but also as representatives of the provincial government as a whole. Thus aversion to chlorine - its avoidance, its perceived status as a questionable means of protecting public health - could not be separated from the historical and social context in which it was presented. Protecting Valued Notions cf Place Residents' sense of place or place-attachment to Erickson played a moderate role in determining aversion to chlorine as well as their tolerant attitudes toward microbial risks. Although there was a lack of statistical significance between sense of place and perceived chlorine risk within the multiple regression model, the conceptual significance of this emerging idea within risk perception research should not be overlooked. Sense of place is a relatively 66 \ new and difficult concept to define. The discovery of a significant bivariate relationship between perceived chlorine risk and sense of place is interesting in and of itself. The moderate, positive association between sense of place and perceived chlorine risk suggested that, for some residents, the very idea of water chlorination was a threat to positive ideas and images they associated with Erickson as a place. This view made sense, given the lifestyle and position many residents were in; many residents liked the idea of retiring in a (real or imagined) pristine place such as Erickson. As explained in Chapter 5, the messages carried by the government proposal to chlorinate simply did not correspond to the images and perceptions of Erickson residents had come to know and value. The strong, positive association between sense of place and the three questions pertaining to local water quality also fit this model and provide at least a tentative explanation as to the reasons behind the community's tolerance toward microbial risks. Specifically, sense of place was associated with an increased tendency to believe that Erickson had naturally high quality water, preference toward drinking untreated water, and the belief that Erickson's water was fine to drinking 'as is'. The overall conclusion to come from this data is that sense of place had at least some influence over respondents' perceptions of drinking water risk, although the detailed basis of this connection could not be ascertained fully through this study. L These findings suggest that the controversy in Erickson was due to ckVerging views of what disinfection of the local water supply neurit for the community. Presumably, the main concerns of public health officials were to fulfill their duty to protect public health by ensuring community compliance with applicable dtirilring water regulations. From their perspective, what they were asking (and subsequently demanding) the community to do was simply to begin making plans to disinfect their water supply as required by the regulations, for the protection of public health. From the community's perspective, however, this request seemed to cany a number of broader implications. Many saw the disinfection proposal as an imposition of government policy on the community; something the community was less than willing to facilitate in light 67 of their frustration with government policies in the past. A second, albeit less tangible, implication of the disinfection proposal was that something was 'wrong' with the community itself, namely with the pristine nature of the local Erickson environment. This portrayal of Erickson simply did not fit with residents' experiences and values they had come to associate with Erickson as a place. Further, people were not motivated to change these views of Erickson, given the lack of trust many residents had in the government. One of the most important observations to be gathered from this study is the centrality of trust in relation to public acceptance of chlorine. If there had been trust, it is more likely that the community would have been less concerned about logging in the watershed because they would not have been suspicious of the government in the first place. Alternatively, if community concerns about logging had been addressed perhaps an adequate public trust could have been secured. For example, if the government could have ensured that logging would not occur in a specific portion of the watershed the community may have been more willing to accept chlorine. Similarly, if residents' were more trusting of the government they may have been less inclined to believe that a valued aspect of the community (a pristine local environment) was being taken away. This is, of course, all speculation. But the best indicator of distrust is suspicion, and a great deal of this was indeed evident within the community. 6.2 What Happened During the Controversy? The findings from this study also provide important insight into the reactions public health officials encountered during the controversy. When handling the controversy, public health officials emphasized health arguments only, paying express attention to the benefits of chlorine and the provision of relevant scientific information. This served to reduce the debate onto, what was for residents, a secondary technical issue (Le. the safety and prudence of chlorine) leaving the real basis of residents' concerns unattended. These points can be further clarified if we look at the messages carried by the proposal to disinfect the local water supply. The order to disinfect carried with it two messages. The first was that the community was being exposed to dangerous microbial contaminants in water -68 this was the only argument public health officials had to justify the order. The second message was that the community would have to disinfect their water to address this microbial problem. Let us address the first message and the implications of focusing solely on health. As previously established, residents' were quite tolerant toward microbial risks. Since these views diverged from those of public health experts, this explains at least the outward basis of the conflict. The first response of public health officials was to try and explain to the community, using scientific evidence, why the health risks posed by the microbial contaminants in their water was serious. This approach, however, proved ineffectual in changing residents' perceptions of microbial risks, presumably because these views were not rooted solely in scientific arguments but also, from a community point-of-view, value judgements about the nature and quality of the local environment. A more serious implication of focusing solely on health came with the second part of the message: the call to disinfect their water. On this point too, public health officials sought to assuage residents' concerns about chlorine by providing them with further scientific information. They explained how the evidence surrounding the health risks associated with chlorinated by-products was scarce and inconclusive and that the risks associated with waterbome disease were well established and significandy greater than the risks associated with chlorinated by-products. Public health officials hoped that providing and explaining the technical rationale behind the regulations would change local views toward drinking water risks. This was a logical approach based on the disciplinary background of public health officials, but proved ineffectual at resolving the controversy, as the real basis of community concerns were not exclusively rooted in technical considerations. Based on previous interpretations of the relational nature of trust, the order to disinfect was not viewed as a politically neutral regulatory effort to protect public health; instead it was seen as an imposition of government policy on an already sceptical community. As a result, community resistance to chlorine was amplified rather than attenuated as expected. 69 Reliance on scientific information as the sole means of resolving public controversies may serve to actually increase public distrust. In Erickson, this occurred in two-ways. First, failure to address the real basis of residents' concerns about chlorine, namely their concerns about logging in the watershed, served to inadvertendy render these concerns non-legitimate. Subsequent provision of more scientific information as to why the public health officials' view was indeed correct, oruy served to increase people's suspicions of an ulterior motive behind the disinfection order. Second, in the absence of public trust, reliance on scientific information to resolve risk controversies can backfire. As observed by Wynne (1987) this is attributable to the uncertainties in science itself, which is a natural and necessary part of the scientific process. This uncertainty is less problematic for designing regulatory policies than it is for implementing them. In policy design, science plays a descriptive or empirical role by providing a systematic and objective means of assessing physical risk. When implementing such policies, however, science plays a syrnbohc role, a means of justifying risk management decisions to the public. In the absence of public trust this exposure of the internal practices of science can perpetuate government-community conflicts, as both sides dig deeper for more scientific information to support their claims. These back-and-forth technical arguments did indeed occur throughout of the length of the controversy in Erickson. The purpose of this analysis is not to diminish the importance of science within the framework of sound drinking water management. Rather, the critical point is that while scientific information is critical for making informed and safe regulatory choices, it will not be effective if relied on alone as the sole justification of policy decisions. 6.3 Implications for Implementing Future Drinking Water Policies Findings from this study point to several suggestions on how to improve the chances of successful implementation of otinking water polices in the future. Necessary improvements include both conceptual and practical innovations at two levels: the community level, where public health officials actually interact with the community (either through the water purveyor or with members of the community directly) and the agency level, where strategies for 70 / implementation are actually worked out and agreed upon between public heakh inspectors, medical health officers, and regional health authority staff. Conceptual Changes Required at Both Community and A gency L evds Public risk perceptions need to be recognized as legitimate expressions of public concerns. These concerns need not be founded in science to be legitimate. Factors such as trust, community values, a desire to have some input into decisions that will affect their local situation were all considerations of relevance to many residents of Erickson. Overall, this study revealed two main insights about the nature of public perceptions of drinking water risk that could be of major benefit to governments in designing more effective implementation strategies in the future. First, the public is aware that policy interventions do not fall from the sky. They come from somewhere, administered by some agency and, ultimately, by some one. Depending on the quality of the relationship between the administrator and receiver of such policies (in this case, the provincial government and the community), this factor may even overshadow the substantive risk issue at hand. Second, public perceptions of risk are embedded in a complex social environment and this is undeniably the 'ground' on which interventions land. Policy changes often appear to require only physical changes (Le. installation of a water treatment technology), yet invariably also involve social changes as well. In this case, the proposal to disinfect the local water supply was not just asking the community to add a chemical to their water but to alter or even discard important social meanings and beliefs they had come to associate with the community. Practical Changs Practically, these non-technical or social changes at the level of policy could include training public health officials to listen to and, to the best of his or her ability, address public concerns. This implies that public health officials must be prepared to listen to linked problems not direcdy related to his or her area of expertise. In so doingj important social concerns and points of disagreement will be revealed and the basis of such concerns more easily identified. In some communities, there may already be feelings of anger or resentment toward the provincial government, regulatory agents and so on, and in these cases it is 71 important to acknowledge community perspectives as legitimate when (a) identifying the basis of their concerns, and (b) seeking solutions. Practically, at the agency level, risk communication should be included as an important part of the agency's implementation strategy. As all communities are different, it is reasonable to assume that some will be easier to bring into compliance than others. Public health officials may already have a sense of which communities will be more difficult and the health authority should be especially attentive to public concerns in these communities. The risk communication component could include efforts to first identify areas of potential conflict, paying particular attention to the rationale behind such concerns coupled with more extensive efforts to meet with community groups. At all times, the larger community (active or passive) should be kept up to date on the implementation process. Or^rnzational Barriers to E ffeake Implernentatian Thus far it is reasonable to assume that the narrow focus of public health officials' served to perpetuate the controversy in Erickson. The point here is not to cast blame on public health officials but to clarify how this approach masked the actual basis of community concerns; concerns including those not completely unrelated to drinking water quality either (since resource development is an activity that can greatly impact drinking water quality at the source)17. Public health officials made little effort to address these concerns, presumably in part because it was beyond their area of specialization but also because forest management issues are beyond their area of administrative jurisdiction. Thus public health officials could argue, and legitimately so, that Erickson's concerns about logging were simply beyond their area of their mandate. This observation suggests a larger organizational dilemma within government that should be improved if an integrated approach to drinking water management is to be achieved. Water, in all its uses, is an integrated system. Ensuring drinking water quality from the source to the tap requires a number of social, environmental, and economic considerations that 1 7 Source protection is widely recognized as the most cost-effective and crucial component of the 'multi-barrier' approach to drinking water management (see for example, the Walkerton Report or the BC Drinking Water Action Plan). 72 go beyond public health. Al l departments involved in forestry, agriculture, eco-tourism, land-use planning, and industry have the potential to impact on drinking water quality. Yet the assignment of water treatment and end-of-the-pipe water quality regulations to one department, forest management practices into another, watershed protection into yet another do not seem conducive to an integrated approach to dnnking water management. As such, it would seem appropriate if the organizational approach of the government toward ensuring safe dririking water quality more closely reflected the integrated nature of drinking water systems themselves. 6.4 Directions for Future Research Topically, this study offers two unique contributions to risk perception research. This was the first study to explore the concept of microbial dose sensitivity in the context of aVinking water risks. At the end of Kraus et aL's initial intuitive toxicology study on chemicals the question was asked whether lack of dose sensitivity to risks in the form of chemicals could also be extended to microbial risks. That is, would the 'intuitive toxicologist' think of a dangerous microbial pathogen the same way as a toxic chemical? Findings from this study seem to hint that the opposite maybe true in that study participants appeared more tolerant of microbial risks and because the data failed to reproduce the Kraus et al. finding that dose insensitivity is generally predictive of risk aversion. However, this idea was only explored among a small sample population. The topic of microbial dose sensitivity remains an area that requires further research. Is the general public less sensitive to dose in the form of microbial versus chemical risks? What role, if any, does the 'naturalness' factor play in shaping public perceptions of microbial risks? Due to time and budget constraints, this study was only able to investigate the perceptions of microbial risk from a community perspective. A comparative study of both expert and public views toward microbial risk would also be a fruitful endeavour as the results could provide a valuable starting point for discussion between water quality experts and the public in terms of their conceptual, attitudinal and value differences regarding microbial risks to drinking water. The results from this study also demonstrated some very interesting connections between residents' 'sense of place' and their attitudes toward cUorine and microbial risks. However, it 73 was not able to fully articulate what exactly those place meanings were, how these ideas shaped their perception of risk, or contributed to people's decision to oppose chlorine. These shortcomings were likely due to inherent difficulties involved in measuring an abstract, context-dependent concept such as sense of place within a quantitative survey. It is likely that sense of place is a concept that is best measured through qualitative methods. Nevertheless, the results suggest an important connection between 'sense of place' and risk perception. The role of 'place' in shaping peoples' perceptions toward environmental risks presents a promising area of risk perception research. It has been argued that place-based approaches to resource management present a unique framework from which to engage the public in a more contextualized, inclusive approach to resource management (Cheng, 2003). This is because such approaches encourage individuals from diverse backgrounds to work together to solve a problem in relation to a specific local environment. This, in turn, facilitates trust building through the design of joint implementation and monitoring programs. Due to the unique and localized character of drinking water systems, more localized approaches to drinking water management maybe highly beneficial in implementing drinking water regulations at the community level The focus of this study was on one community's perceptions of drinking water risks. In so doing, it demonstrated a number of important observations about risk perceptions that have the potential to aid implementation of new, more stringent drinking water regulations in the future. These findings demonstrated the importance of a number of general principles of risk perception, such as the relational basis of public trust, its influence in shaping publie receptivity toward government policies, and the role of broader social-contextual factors such value judgements about the local environments. However, the details of these insights were confined to the specify community of the Erickson. It is not clear to what extent the content of these findings are characteristic of other communities, either within the Kootenays or regions of the province. It would be worthwhile to conduct a similar study elsewhere in the province, perhaps within a community where there are already signs of potential conflict. Such an endeavour 74 would be highly advantageous at this point in time, in light of the forthcoming changes in drinking water regulation in the province. At present, the new drinking water legislation has only recendy been passed and policy and procedures manuals have yet to be finalized. The government is now at a critical stage of the implementation process, where public concerns can be taken into account and potential conflicts can be addressed before they happen. Ironically, one of the most encouraging data to come out of this study was a somewhat cynical comment by one respondent who said: 'I have to assume this study was funded by the provincial government and is politically motivated. Politically, it is a smart move.' Consideration of public perceptions of risk should not be viewed as an impediment to the regulatory process but a means of easing the process through the development of more effective, collaborative relationships between governments and communities. 75 Appendix I: Interview Protocol Thank you for agreeing to participate in this interview, your help is really appreciated. What I'm interested in, in particular, is hearing your thoughts and opinions on drinking water issues in this area as well as what you think about this community in general. Background Information 1. Why don't you start by telling me a bit about yourself, how did you come to live in Erickson? • A (Mtional questions (if not bmqfpt up in response): • Do you like living here? • Why did you choose Erickson? • What is it about this place that you like? • Have you ever considered moving? Why or why not? Chlorination Issue 2. What was your first reaction when you first heard about the medical health officer's recommendation that Erickson needs to start cHorinating their water? A (Mtional questions (if not brou$Jt up in response): • Are you a member of any local citizen's groups related to drinking water issues? • Are you happy with the decision to build the filtration plant? • What do you diink about the quality of water in Erickson? Sense of Community 3. What do you think about the sense of community here in Erickson? A aauional questions (if not brougpt up in response): • How do you think it compares to the Town of Creston? • Do you consider Erickson your home? Above any other place in the world? 4. Have you been involved in other communities you've lived in? 5. Would you say you're pretty similar or pretty different from other people in the community? 6. What do you see happening in the future to this community? A ckktkmal questions (if not brought up in response): • Do you think it's going to grow? Do you want it to grow? • Is there anything you dislike about Erickson and would like to see changed? Thank you for your time please let me know if you have any other questions about this study. 76 APPENDIX II: MAIL-OUT SURVEY RISK PERCEPTION AND DRINKING WATER QUESTIONNAIRE CONTENTS: i > A. Risk Perception i B . Thinking About Drinking Water Cl Water and Your Community M\V i , D. Background Information 111 Instructions: Thank you for taLking the time to look at this questionnaire. Your cooperation is greatly appreciated. To ensure a random sample, we ask that the person in the household (over the age of 19) whose birthday passed most recently fill out the questionnaire. Please note that by filling out this questionnaire, it will be assumed that your consent has been given to participate in the study. A postage-paid envelope has been provided for the return of your completed questionnaire. Please read each statement or question carefully and then indicate your response by circling the appropriate number. If you wish to comment on any questions or qualify your answers, please feel free to use the space in the margins and/ or to include thoughts on an additional piece of paper. The questionnaire will take about 20 minutes of your time. All responses will be kept strictly confidential. SECTION A : RISK PERCEPTION We would like your opinion on the health risks, if any, that the following items pose to the C a n a d i a n publ ic as a whole. Mark your response by circling the appropriate number. Does not pose a risk to human health Poses P ° s e s , slight risk m o d u e ; a t e to human " s k t o heath hhumf" health Poses high risk to human health 1. Prescriptioh^drugs ,\ \ , * . .1 4 2. Infectious diseases (i.e. AIDS, West Nile vims) 2 3 4 3 Motor vehicle accidents 4. Nuclear power reactors 1 2 3 4 5. Use of genetically engineered bacteria in ,„ t agriculture ' 6. Drinking water 1 2 3 4 7, Asbestos! , ^ V * -8. Fluoridation of drinking water 2 3 4 9. Mercury in dental fillings 4 • a 10. Depletion of the ozone layer 1 2 3 4 M l eigarettlsmoke V,v, ' ' - 1 1* 4 J 12. Moulds in food 2 3 4 13. Indoor air quality 4 14. Street drugs 1 2 3 4 -15. ,Waste incinerators .- *" - ,, 4 16. Pesticides in food 1 2 3 4 78 POSGS Does not Poses m o d e r a t e Poses high pose a risk slight risk m o a e r a r e riskto to human to human . s 0 human health heath JJJJJ health 17. Chronic diseases (i e heart disease, diabetes or cancers caused by lifestyle factors) 79 Listed below are a number of activities taking place in the vicinity of a local water supply. Using the scale below, please indicate whether you think the following activities or items pose a risk to drinking water quality. Circle one number for each question. Does not pose a risk to drinking water quality Poses slight risk to drinking water quality Poses moderate risk to drinking water quality Poses high risk to drinking water quality »18. Discharging treated human sewage that meets regulatory standards ' -19. Build up of bacteria in the water distribution lines 20. Increasing commercial access to land " * 21. Discharging chemicals that are not currently regulated 22. Discharging'industrial waste . i 1 v * v „ y x - ^ i s f v s , ^-**«" >* 23. Recreational boating 24. Cattle grazing mamgagm 1 s i l l t i 25. Steep slopes in watersheds 26. Water-treatment using sand filtration . f 27. Water treatment using U V radiation 28., Resident wilrjlife s^uch as-elk or beaver4 i,r 29. Using home water filtration systems '30. Camping % • 31. Mining .j 1 1 r "*i " 1 1 | | | 1 3' •PIP 8 2 ^ T O ¥ - | 3 ^ V - V 4iX^ 2 2 3 3 4 4 4 3 Z d 80 Does not Poses slight pose a risk risk to to drinking drinking water water quality quality Poses moderate risk to drinking water q u a | i t y Poses high risk to drinking water quality 32 Spreading of manure on agncultural fields in SUMMER 33. Water treatment using membrane filtration 34 Snow melt/Spnng run-off 35. Using nearby roads to transport hazardous waste 36"* Chemically disinfecting water with ozoriel 37. Poor soil drainage <38. Restricting-the^public from hayjng a say ^ in'environrriental'protection regulations \ 39. Pesticide use in nearby farms f t i i 40' Spreading of manure on agricultural . fields in WINTER 41. Clear-cut logging 42. Naturally occurring arsenic in nearby soil 43. Selective logging ,44. Build up of algae in the water distribution lines ' , > , ; 45. Heavy rainfall mm •^i 3 2 2 3 3 4 4 2 3 2 3 81 SECTION B: THINKING ABOUT DRINKING WATER For the following statements, please indicate your level of agreement by circling the appropriate number for each question. Strongly Disagree Agree Strongly Disagree Agree 1. -1 trust the provincial government's ability to „ manage any risks posed by waterborne ... pathogens. ' ' 2. Local communities can be trusted to properly manage their own water safety programs. 3. The provincial government can be trusted to • properly manage the risks,from technologies,,, suclvas water treatment programs. **. 4. I trust what experts say about the risks of UNCHLORINATED water. 5. People living near a water treatment plant • > * * should beable to vote and close the plant if « ( , they Jhinkwt is not being run safely. = < . 6. There is a safe level of exposure to a chemical that causes cancer. • " » 7. If.a trace.amount of disease-causing ^•microorganisms were found in my tap water, 1, 1 wouldn't drink it.r • ' • ' \ 8. Experts are able to make accurate estimates of health risks from harmful microorganisms. 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 82 Strongly Disagree Agree Strongly Disagree Agree 9 A biological contaniinaritcwa&found in axity,'s supply of drinking water in a concentration of 30 parts per million. The water was filtered by a process that was able to reduce, but not eliminate the contaminant in the water Under most circumstances, this means that • 2 the danqer associated with drinking the water has also been reduced 10. There is a safe level of exposure to disease-causing microorganisms. 11 For disease-causing microorganisms, it's not 12. Low level concentrations of microorganisms in water are okay to drink. 13.7lfarpersbn is exposed toJrarJiation, then that " person will probablyjget cancer some day „~ 14. The more often a person is exposed to a disease-causing microorganism; the more likely he or she is to get sick. 1 . 1 , 0 #2. 3 ^ 15 I don t worry too much about getting sick from water because there are so many other " . thmgslh my life thai I have to deal with. 16. No matter how pristine a watershed seems, the water it provides should be treated before drinking it. 17. 'Exposure to"disease^ausing;rnicroorganisms; 4 , ""would probably make me sick, no matter how*~ -low the dose. ' • " 18. Most microorganisms in the environment are harmful to humans. ^ ^ ^ ^ ^ ^ ^ ^ ^ P U P 1 ^ A 0 4 83 Strongly Disagree Agree Strongly Disagree Agree • 19. If a person is-exposed to a chemical that can cause-cancer then thatpersonjwill probably h 20. Microorganisms are either harmless or dangerous. There really is no "in-between". 1 2 3 4 84 Section C: Water and Your Community People obtain their information on water quality from a number of sources. Using the scale below, please indicate how much information you have received from that source about your local water quality. Almost no A lot of information information 1 TV, radio or other news media •1 m: 2 BfiSlIfi 2. Environment Canada 1 2 3 4 5 3. The Creston Valley Advance H i 2 5 4. Creston town council 1 2 3 4 5 5 Friends and relatives ^^^^^^^ ^ ^ t.^ ^^^^"^ ^^^^f l i n 2 ^ ^ ^ ^ ^ ^ ^ ^ 6. Local medical health officer 2 3 4 5 7 Enckson Improvement ' District water utility office A " >* ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ 8. Health Canada 1 2 3 4 9. Medical doctors ."4 2 5 10. Public interest or 9 c environmental groups L. 11 The provincial Ministry of Health Services »f 2 * ?^ If M ^Uml • „ 5 12. Private industry 1 2 3 4 5 13. University scientists ts? 2 •3'' :.'}^''; 4: 5 85 In considering this list again, please use the new scale to tell me how much confidence you have in these information sources. Almost no A lot of confidence confidence „ 14VJV, radio p'r other,news . % ^media" ' ' 1 2 5 IsSSSIftiaS 15. Environment Canada 1 2 3 4 5 16 The Creston Valley " " Advance , liSii&ftisil 2 A 4 5 17. Creston town council 1 2 3 4 5 1 8 . 'Friends and relatives Si 2 - T 5 19. Local medical health officer 1 2 3 4 5 20 Erickson Improvement District water utility office 2' 5 21. Health Canada 1 2 3 4 5 ;<22:4 Medical doctors^ - - - i 23. Public interest or environmental groups '/24f*The provincial Ministry of Health Services ni 2 • 2 2 3 4 5 5 25. Private industry 2 3 4 5 j~-26. University scientists 2 .-5|^;# 86 For the following questions, please indicate to what extent you agree or disagree with each of the statements. Strongly Disagree Disagree Agree Strongly Agree 27., I.try hard to avoid drinking chlorinated water, even in Creston Ai - n< * 28. Erickson should have some form of water treatment, just not chlorine. 1 2 3 4 29 1 do NOT really mind dnnking chlonnated ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ 30. Adding chlorine to our water would have a very negative effect on the orchards. 1 2 3 4 31. I question the effectiveness of adding chlorine to our water for the protection of Itifiltt •Hi 4 public health 32. 1 prefer to drink untreated water. 1 2 • S B 3 33. Erickson has naturally high quality water. S H I P ! 34. Chlorine adequately disinfects water from a number of contaminants you often find in water. 1 2 3 4 35 Adding chlorine to our system would NOT really affect the orchards and might even be beneficial to the orchardists. 1 M S 36. 1 think it's fine to drink the water in Erickson as is. 1 2 3 4 The following section asks questions about the meaning and importance of Erickson to you. For the following questions, please indicate to what extent you agree or disagree with each of the statements. NOTE: When we say 'place' we mean Erickson and the surrounding area. Strongly Disagree Disagree Agree Strongly Agree 37. 1 feel happiest when I'm in this place. 1 •'. " 2 ; 4 38. 1 enjoy the routine 1 have living in Erickson and the surrounding area. 2 3 4 39. 1 can really be myself in this place. ,r •• Ijjiii^ 4 " * 40. 1 really miss this place when I'm away for even a short length of time. 2 3 4 41. Everything about this place is a reflection of Hi 42. When I'm in this place I feel my most relaxed. 2 3 4 43. Every time I pass particular places in Erickson and the surrounding area, I am struck by the memory of key events in my life that happened there. 2 3 ^ 4 44. This place says very little about who I am. 2 3 4 ,45. As a person who has moved around a lot, it's . places in my memory that matter most to me and not Erickson and the surrounding area... ,'" 2 , -46. This place is the best area for doing the things I enjoy most. 1 2 3 4 88 Strongly Disagree Agree Strongly Disagree Agree 47 Most of the important people in my life live (or have lived) in Erickson and the surrounding join wWfc O t-1 &tf * < j*-™--,.*! 48. For me, personally, many of the stories that make up my life are intimately connected to this place. 1 2 3 4 The following questions ask about certain actions you may have taken in response to the controversy over water treatment in Erickson. Have you ever participated in one or more of the following actions? (Please indicate all that apply). YES NO 49 Participated in the blockade at the water intake area in spring of 1999? 50. Displayed a "No Chlorine" sign on your property? 51 Contnbuted financially to Enckson Water Users Society or the Water Action Group' 52. Boil your water before drinking it? 1 2 1 2 89 SECTION D: B A C K G R O U N D INFORMATION These questions are intended to tell me a little bit about you. I will use this 1 ! information to find out about the way different(pedpl e's spcialbackgrounds influence their views. Please remember that all your answers will be kept -, t , strictly confidential. , , , , t ' 1. What is your date of birth (dd/mm/ yy)? 2. What is your gender? • Female • Male 3. Where were you born? (Indicate name of city/ town, and country) 4. How many years have you lived in your present community? 5. To which ethnic/cultural group(s) do you belong? 6. Please indicate the highest level of education you have completed. • No schooling • Elementary school • Some university • Bachelor's degree • Master's degree • Doctorate or professional degree • Other (please specify) 90 7. Which of the following income categories best describes your household income level from all sources? (Please select only one) • Less than $5,000 • $5,000 to $9,999 • $10,000 to $19,999 • $20,000 to $29,999 • $30,000 to $39,999 • $40,000 to $49,999 • $50,000 to $59,999 • $60,000 to $69,999 • $70,000 to $79,999 • Over $80,000 8. Is your home: • Rented/Leased • Owned • Other (Please specify) 9. If you have any other comments about issues raised by this survey OR Would like to comment on the survey itself, feel free to use the space below: T H A N K Y O U F O R T A K I N G T H E T I M E T O F I L L O U T T H I S Q U E S T I O N N A I R E , Y O U R H E L P IS G R E A T L Y A P P R E C I A T E D . 91 REFERENCES Auslander, BA and PH Langlois. Toronto tap water perception of its quality and use of alternatives. CanadianJournal of'PublicHealth, 84,99-102,1999. Basso, K H and S Feld. Senses cf Place. Santa Fe: School of American Research Press, 1996. Baxter, J and D Lee. Explaining the maintenance of low concern near a hazardous waste treatment facility, fournal cfRisk Research, 6(1), 705-729,2004. British Columbia. Auditor General of British Columbia. Report 5: Protecting Drinking Water Sources. Victoria, BG Officer of the Auditor General of British Columbia, 1998/99. British Columbia. British Columbia Ministry of Health Services. Interior Health Authority. Health Status Vignette, A snapshot of health: Creston local health authority', 2004. Available from http://ww.interiorhealm.ca/NR/rdonlyres/92561836-43Al-46FF-9C19-5664729D4E8E/1028/2004CrestonProfile.doc British Cdurnhia. Office cf the Provincial Health Officer. A Report on the Health of British Columbians. Provincial Health Officer's Annual Report 2002. The Health and Well-being of People in British Columbia. Kzctt^ -M^&y</He^P^m^5 2003. Bickerstaff, K. Risk perception research: socio-cultural perspectives on the public experience of air pollution. Environment International, 30(6) August, 827-840,2004. Brannen, J. Mixing Methods: Qualitative andQuantitative Research. Brookfield: Ashgate Publishing Company, 1992. Canada, Health Canada. Risk/benefit perceptions ofbiotechnologyproducts. Report prepared for Health Canada by Environics Research Group, Ottawa: Health Canada, Jury, 2000. Canada, Statistics Canada. Income statistics for Creston, British Columbia. Retrieved Jan 5, 2005, from Community Profiles Web site: http://wwl2.statcan.ca/english/profil01/Details/details M>SGC=59c^GC=59030045cA=8d^NG=E&Province =59&PlaceName =Creston&CS DNAME =Creston&CMA=&SE ARCH=BE GINS6cDataType=l&TypeNameE =Town& ID=11616.2001. Chawla, L. Childhood place attachments. In I Altman and SM Low (eds.), Place Attachment,. New York: Plenum Press, 1992. Cheng, AS., Kruger, LE and SE Daniels. 'Place' as an integrating concept in natural resource politics: propositions for a social science research agenda. Society and Natural Resources, 16, 87-104,2003. 92 Coughlin, CW., Hoben, M L . , Manskopf, D W and S W Quesada. A SystematicAssessment cf Oflahorative Rresource Management Partnerships. Ann Arbon School of Natural Resources and Environment, University of Michigan, 1999. Available from http://www.umich.edu/ -crpgroup Cozby, P G Methods in Behavioral Research. 5th ed., Mountain View: Mayfield Publishing Company, 1993. Creswell, JW. Research Design' Qualitative, Quantitative, andMixedMethod Approaches. 2nd ed., Thousand Oaks: Sage Publications, 2003. Douglas, M Risk acceptability according to the social sciences. Social Research Perspectives: Occasional Reports on Current Topics, 11,1985. Douglas, M Purity and Danger. London: Routledge, 1966. Douglas, M Risk and Blame: Essays in Cultural Theory. New Y o r k Routledge, 1992. Fessenden-Raden, J., Fitchen, J M and SS Heath. Providing risk information in communities: factors influencing what is heard and accepted. Science, Technology andHuman Values. 12(3/4), surnmer-auturnn, 94-101,1987. Finucane, M L . , Slovic, P., Mertz, C K . , Flynn, J., and T A Satterfield Gender, race and perceived risk the 'white male' effect. Health, Risk and Society, 2(2), 159-172,2000. Fitchen, J M Risk perception in community context: a case study. In BB Johnson and V T Covello (eds.), The Social and Cultural OonstmctioncfRisk. Boston: D Reidel, pp31-54,1987. Griffin, C B . Watershed councils: an emerging form of public participation in natural resource management. Arrrricmfournd of Water Resour^ 35,505-518,1999. Hansen, J., Holm, L . , Frewer, L . , Robinson, P and R Sandoe. Beyond the knowledge deficit: recent research into lay and expert attitudes to food risks. Appetite, 41,111-121,2003. Irwin, A . Citizen Science: A Study cf People, Expertise and Sustainable Development. London: Routledge, 1995. Jardine, C G . , Gibson, N and SE Hrudey. Detection of odour and health risk perception of drinking water. Water Science and Technology, 40(6), 91-98,1999. Jarroll, E L . Sensitivity of protozoa to disinfection. A Intestinal protozoa. In A D Russell, W B Hugo, and G A J Ayliffe (eds.), Principles andPractice cfDisinfktion, Preservation and Sterilization, 3rd ed., Oxford: Blackwell Science, 1999. Jorgensen, BS and R C Stedman. Sense of place as an attitude: lakeshore owners attitudes toward their properties, fournal cfEnuronrnental Psychology, 21,233-248,2001. 93 Kaltenbom, BP. Effects of sense of place on responses to environmental impacts. Applied Geography, 18(2), 169-189,1998. Kasperson, RE., Golding, D and S Tuler. Social distrust as a factor in siting hazardous facilities and communicating risks. Journal of Social Issues, 48(4), 161-187,1992. Kraus, N., Malmfors, T and P Slovic. Intuitive toxicology: expert and lay judgements of chemical risks. Risk Analysis, 12(2), 215-232,1992. Krewski, D., Balbus, J., Butler-Jones, D., Haas, C , Isaac-Renton, J., Roberts, KJ and M Sinclair. Managing health risks from clrinking water - a report to the Walkerton inquiry. Journal qfTaciodogy andEnurmrrental Health, Part A, 65,1635-1823,2002. Kunreuther, H., Slovic, P and D MacGregor. Risk perception and trust: challenges for facility siting. Risk Health, Safety andEnuronrrent, 7,109-118,1996. Low, SM and I Altman. Symbolic ties that bind: place attachment in the plaza. In I Altman & SM Low (eds.), Place A ttaohment, New York: Plenum Press, 1992. Lupton, D. Risk. London: Routledge, 1999. MacGill, S. Risk perception and the public: insights from research around Sellafield. In J Brown (ed.), Enwronmnetal Threats: Perception, A nafysis and Management London: Belhaven Press, pp48-66,1989. MacGregor, DG., Slovic, P and T Malmfors. How exposed is exposed enough? Lay inferences about chemical exposure. (Report No. 94-12), Decision Research, Eugene, Oregon, 1995. Cited in: MacGregor, DG and R Fleming. Risk perception and symptom reporting. Risk A nafysis, 16(6), 773-783,1996. Mc Andrew, FT. The measurement of 'rootedness' and the prediction of attachment to home-towns in college students. Journal cfEnwmnentalPsychology, 18,409-417,1998. McCallum, DB., Hammond, SL., Morris, LA and VT Covello. Public Knowledge and Perceptions cf Cherried Risks in Six Qmrunities. Washington, D.C: US Environmental Protection Agency, 1990. McDaniels, T. Public perceptions regarding water quality and attitudes toward water conservation in the Lower Fraser Basin. Water Resources Research, 34(3), 1299-1306,1998. McGuire, MJ. Off-flavour as the consumer's measure of drinking water safety. Water Science and Technology, 31(11), 1-8,1995. 94 Michitov, AI and SB Rebrik Studies of risk and safety perception in the USSR. In K Borcherding, OI Larichev, and D M Messick (eds.), Conterrporary Issues in Decision Making, North-Holland: Elsevier Science Publishers, pp261-289,1990. Nelkin, D . Oommunicating technological risk the social construction of risk perception. AnnudReuewcfPublic Health, 10,95-113,1989. Otway, H and K Thomas. Beyond acceptable risk: on the social acceptability of technologies. Policy Sciences, 14,247-256,1982. Pallant, J. SPSS Surtivd Manual: A Step-by Step Guide to Data A nafysis Using SPSS for Window (Version 10.0). Buckingham: Open University Press, 2001. Patton, M Q . CjualitatiwEviluatimandR^ 2nd ed., Newbury Park Sage Publications, 1990. Pidgeon, N F . , Hood, C , Jones, D . , Turner, B A and R Gibson. Risk perception. In Royal Society Study Group (eds.), Risk Analysis, PemeptimandMamgenmt, pp89-134, London: Royal Society, 1992. Priest, S H A risky career [Review of the book Risk Penxption], Science and Public Policy, 30(6), December, 463-464,2003. Pynn, L . Province urged to take control of town's waten east kootenay heakh officer calls failure to chlorinate 'shades of Walkerton'. Vancouver Sun, p p A l , Jan. .03,2001. Kleinhesselink, R R and E A Rosa. Cognitive representation of risk perceptions: a comparison of Japan and The United States. Journal cf Crcss-adtural Psychology, 22,11-28,1991. Rubenstein, R L and P A Parmalee. Attachment to place and the representation of the life course by the elderly. In I Altman & S M Low (eds.), Place A ttaobrrent, New York Plenum Press, ppl39-164,1992. Satterfield, T., Mertz, C K and P Slovic. Discrimination, vulnerability, and justice in the face of risk Risk A nafysis, 24(1), 115-124, February, 2004. Slovic, P. The perception of risk Science, 236, April, 280-285,1987. Slovic, P. Perceived risk, trust, and democracy. Risk Analysis, 13,675-682,1993. Slovic, P. Theperceptkn cf risk. London: Earthscan Publications, 2000. Slovic, P., Fischhoff, B and S Lichtenstein. Behavioural decision theory on risk and safety. A aa Psydodogcu, 56,183-203,1984. 95 Slovic, P., Flynn, J., Mertz, CK and L Mullican. Health-risk Perception in Canada. (Report No. 93-EHD-170), Ottawa: Department of National Health and Welfare, 1993. Slovic, P., Malmfors, T., Krewski, D., Mertz, CK., Neil, N and S Bartlett. Intuitive toxicology II: expert and lay judgements of chemical risks in Canada. Risk Analysis, 15(6), 661-675, 1995. Solden, DL. Trust in sources of technical information. Journal qfE rmmnnvntal Educadon, Winter, 26(2), 16-20,1995. Sproull, NL. Handbook cf Research Methods: A GuideforPractioners and Students in the Social Sciences. 2nd ed., Metuchen: The Scarecrow Press Inc., 1995. Stedman, R C Toward a social psychology of place: predicting behaviour from place-based cognitions, attitude and identity. EnmcwrentandBchavlour, 34(5), September, 561-581, 2002, Tashakkori, A. and C Teddlie. Mixed methodology: combining qualitative and quantitative approaches, Applied Soad Research Methods Series, 46, Thousand Oaks: Sage Publications, 1998. Teigen, K H , Brun, Wand P Slovic. Societal risks as seen by the Norwegian public, fournal cf Bchavaural Decision Making, 1,111-130,1988. Tuan, YF. Space and place: humanistic perspective. In S Gale & G Olson (eds.), Philosophy in Geography. Dordrecht: D Reidel, 1979. Tversky, A and D Kahneman. Judgement under uncertainty: heuristics and biases. Science, 185, no. 4157,1124-1131,1974. Tversky, A and D Kahneman. The framing of decisions and the psychology of choice. Science, 211, no. 4481,453-458,1981. United States of America, US Environmental Protection Agency. Total coliform rule. Retrieved Jan. 05,2005, from Drinking Water Standards for Regulated Contaminants Webpage: http://www.epa.gov/safewater/therule.html# Total. n.d. Wakefield, SEL., Elliot, SJ., Cole, D C and JD Eyles. Environmental risk and re(action): air quality, health, and civic involvement in an urban industrial neighbourhood. Health and Place, 7,163-177,2001. Weber, EP. A new vanguard for the environment: grass-roots ecosystem management as a new environmental movement, Society andNaturd Resource Mamgement, 13,237-259,2000. Weyman A K and CJ Kelly. Risk Perception and Risk CcmrmicatiaT A Redewcf the L iterature. 96 Contract Research Report No.248, Health and Safety Executive. Sheffield: Health and Safety Laboratory, 1999. Williams, D R a n d JJ Vaske. The measurement of place attachment: validity and generalizability of a psychometric approach. Forest Science, 49(6), 830-840,2003. Williams, D R and M E Patterson. Environmental meaning and ecosystem management: perspectives from environmental psychology and human geography. Society of Natural Resources, 9,507-521,1996. Wynne, B. Rationality and Ritual: The Windscale Inquiry and Nudear Decisions in Britain. Bucks: The British Society for the History of Science, 1982. Wynne, B. Risk Mamgprent and Hazardous Waste. Laxenberg: International Institute for Applied Systems Analysis, 1987. 97 

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.831.1-0092103/manifest

Comment

Related Items