UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Reconciling Concept and Context : A grounded theory study of implementing school-based health promotion MacDonald, Marjorie Anne 1998

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

Item Metadata


831-ubc_1998-34584x.pdf [ 19.48MB ]
JSON: 831-1.0076932.json
JSON-LD: 831-1.0076932-ld.json
RDF/XML (Pretty): 831-1.0076932-rdf.xml
RDF/JSON: 831-1.0076932-rdf.json
Turtle: 831-1.0076932-turtle.txt
N-Triples: 831-1.0076932-rdf-ntriples.txt
Original Record: 831-1.0076932-source.json
Full Text

Full Text

R E C O N C I L I N G C O N C E P T A N D C O N T E X T : A G R O U N D E D T H E O R Y S T U D Y OF I M P L E M E N T I N G S C H O O L - B A S E D H E A L T H P R O M O T I O N by M A R J O R I E A N N E M A C D O N A L D B . N . , University of Calgary, 1980 M . S c , University of Waterloo, 1983 A THESIS S U B M I T T E D I N P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S F O R T H E D E G R E E OF D O C T O R OF P H I L O S O P H Y in T H E F A C U L T Y OF G R A D U A T E S T U D I E S (Interdisciplinary Studies) We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y OF B R I T I S H C O L U M B I A July 1998 © Marjorie Anne MacDonald, 1998 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia Vancouver, Canada DE-6 (2/88) i i ABSTRACT In response to high prevalence rates of alcohol and other drug use by adolescents in British Columbia (BC), the BC Ministry of Health initiated a pilot project in secondary schools aimed at preventing alcohol and drug misuse among students. The School-Based Prevention Project involved placement of a Prevention Worker (PW) in selected secondary schools. The PW was expected to engage the school community in a collaborative process to develop and implement a comprehensive prevention program in the school using the School-Based Prevention Model (SBPM), an adaptation of the Precede-Proceed Model for health promotion planning. This grounded theory study explored the process by which the PWs implemented a PW role and the SBPM in the schools. Initial data collection involved one hour telephone interviews with 28 of 31 eligible PWs from across the province, conducted 4 to 6 months after the second annual training session. On the basis of the theoretical conceptualizations emerging from analysis of the PW interviews, 6 schools with varied implementation experiences were selected to conduct one week site visits to explore implementation in context. Field notes were" kept on all interactions with PWs and schools over the four . year study period. Over 100 interviews were conducted in the 6 sites with various members of the school community. Data analysis was conducted using the constant comparative method of grounded theory. Analysis yielded the core category of Reconciling Concept and Context. The core category is composed of three sub-categories, Entering the Field, Confronting the Model and "Doing" the Model. Before implementation of the SBPM could begin, PWs had to gain entry to the school by establishing program legitimacy and personal credibility, and by learning the ropes. This was often a challenging and lengthy process. As part of Entering the Field, PWs created a role by finding a focus that was acceptable to all concerned, and by striking a balance between: a) the schools' demand for intervention and the program focus on prevention, and b) the schools' propensity for crisis management and the program focus on pro-active planning. Many schools were not "ready" to implement the prevention program as intended, so PWs also had to enhance school readiness to develop and implement the program in the Ul school. Before PWs could begin "Doing" the Model, they went through a personal Confrontation with the Model, which involved reacting to, learning and contemplating it. When PWs ultimately tried to "Do" the Model, conditions in the school context led them to reinvent, retrofit, reframe, approximate, or abandon the model rather than implement it as intended. Thus, in Reconciling Concept and Context, PWs had to facilitate changes in the school context to accommodate the concept. The concept, in turn, did not fit with the schools' goals and ways of doing things. The challenge for PWs was to modify the concept to fit the context, while retaining the integrity of the concept. In most instances, the concept was modified to a much greater extent than the context and was therefore coopted by the status quo. IV T A B L E OF CONTENTS Abstract ii Table of Contents iv List of Tables xii List of Figures xiii Acknowledgements xiv Dedication xvii Part A: The Research Investigation 1 Chapter One 2 The Problem of Alcohol and Drug Use Among Adolescents 3 The Historical Context of Substance Misuse Prevention . . . . 6 Developing a Study of SBPP Implementation 8 Purpose 10 Research Questions 10 Significance of the Study 11 Overview of the Dissertation 11 Chapter Two - School-Based Health Promotion: Past and Present ; . . . . 14 Schools as the Site for Preventive Interventions 14 An Emerging Consensus 15 Concerns About Health Promotion in Schools 16 Substance Misuse and Education Performance 17 Schools as Both a Setting and a Target for Change 19 The School-Based Prevention Project 21 The School-Based Prevention Model 22 Contractual Arrangements 24 V Current Thinking about Substance Misuse Prevention 29 Community-Specific Approaches to Prevention 29 Community-wide interventions 29 Interventions in community , 30 Community-based programming 30 Community development programming 31 Principles of community health promotion practice 32 The Social Ecological Perspective of Health Promotion 32 Implementation and Evaluation Issues in Community-Specific Approaches 35 Research Dilemmas 36 Resolving the Dilemmas 37 Chapter Three - Implementation 42 What is Implementation? 42 Theoretical Perspectives on Implementation 43 Macro-Implementation versus Micro-Implementation 51 Fidelity versus Adaptation or Reinvention 53 What are the Implementation Issues for this Study? 57 Chapter Four - Theoretical Perspective .61 Symbolic Interactionism, Health Promotion and Grounded Theory .61 Symbolic Interactionism 62 Criticisms of Symbolic Interactionism and Grounded Theory 64 The Conditional Matrix In Grounded Theory 66 The Relationship between Symbolic Interactionism and Health Promotion . . . . 68 Chapter Five - Methodology 72 vi The Study 74 The Data Source 74 Preliminary Participant Observation 75 Telephone Interviews 75 Pilot Case Studies 76 Case Studies 77 Case and Participant Selection 77 The Prevention Workers 77 Pilot Case Study 78 Case Studies • 79 The Nature of the Data 80 PW Interviews 80 Pilot Case Study 81 Case Study Data 84 The Data Base 87 The Analytic Method 88 Theoretical Sampling 88 Data Analysis 89 Trustworthiness 93 Credibility 94 Transferability 96 Confirmability and Dependability 96 Ethics Review and Protection of Human Subjects . 97 Part B: Reconciling Concept and Context. 98 Chapter Six - Gaining Entry as a Part of Entering the Field 108 vii Establishing Program Legitimacy 108 Overcoming Opposition 109 Strategies to Overcome Opposition 113 Clarifying Expectations 117 Establishing Personal Credibility 123 Being Seen 125 Being Known 126 Being Accessible 127 Being Different 129 Doing What We Know . 131 Doing,Everything 132 Doing For 132 Doing Well 134 Keeping Administrators Informed 135 Keeping Secrets 137 Keeping Safe 138 Keeping the Bargain 139 Learning the Ropes 140 Understanding the System 140 Finding the Limits 144 Achieving Peerage and Achieving Acceptance 146 Chapter Seven - Creating A Role as a Part of Entering the Field 150 Finding the Focus 151 What it is and is not? 152 Narrowing the Field 158. viii Taking a Narrow versus a Broader Focus 160 Striking a Balance 165 Prevention versus Intervention 165 Proactive Planning versus Crisis Management 174 Chapter Eight - Enhancing School Readiness as a Part of Entering the Field 181 Selling the Issue 182 Selling the Model 187 Facilitating Participation 188 Steering the Steering Committee 194 Getting Started 195 Maintaining and Working with the Steering Committee 200 Finding Alternatives 202 Summary of Entering the Field 205 Chapter Nine - Confronting the Model 208 Reacting as a Part of Confronting the Model 209 Perceived Learning Needs and Training Expectations 209 PW Background 211 Lack of PW Participation 212 Model Presentation 213 Learning the Model as a Part of Confronting the Model 217 Inadequacies of Training 218 Lack of Knowledgeable Facilitators and Support People 218 Demand of the Work Environment 220 Lack of Access to Learning Resources 221 Contemplating the Model as a Part of Confronting the Model 222 ix Sizing up the Model 224 Strengths 224 Weaknesses 227 Trying it On 229 Assessing the Fit 231 Chapter Ten - "Doing" the Model .235 Planning 236 "Doing" the Model as Project-by-Project versus Comprehensive Planning 237 "Doing" the Model as a Sold or Collaborative Activity 238 Applying the Model 239 "Doing" the Model by the Book 340 "Doing" Approximations of the Model 241 Reinventing the Model 243 Retrofitting the Model 244 The Model as a Mental Map 246 Summary of Confronting and "Doing" the Model 248 PartC: Chapter Eleven - Implications .251 Limitations and Strengths of the Study .251 Implications for the Literature on Model Implementation and School Change 256 Context 257 School Adoption Process 257 School Readiness 258 School Culture 263 School Support 267 Concept 269 The SBPM 269 ADS Environment 271 The PW Background 272 PW Training 273 ADS Agenda .274 ADS and Agency Support 276 Reconciling Concept and Context 276 Implications for Theory 279 Technological Perspective 279 Organizational Development Perspective 280 Bureaucratic Adaptation Perspective 281 Political Perspective 282 Cultural Perspective • 283 Implications for Health Promotion Practice 284 Community Participation 284 Models of Practice: From Service Provision to Building Community Capacity 287 Achieving Peerage versus Challenging the Status Quo 288 Implications for Policy . 290 Implications for Further Research 293 Postscript 296 References 297 Appendix A - SBPP Evaluation Design 316 Appendix B - The Precede-Proceed and School-Based Prevention Models 321 Appendix C - Ministry of Health Organizational Charts 328 xi Appendix D - Generic Service Schedules and Contracts for PWs 331 Appendix E - Policy Guidelines for PWs 339 Appendix F - PW Telephone Interview Schedule 343 Appendix G - Case Description 346 Appendix H - Case Study Interview Schedule 351 Appendix I - Post-interview coding form 354 Appendix J - Document Memo 356 Appendix K - Ethics Approval and Consent 360 xii List of Tables Table 1 Theoretical Perspectives on Implementation 45 Table 2 Comparison of Theoretical Perspectives on Implementation 47 Table 3 Reconciling Concept and Context Conceptual Framework 102 Table 4 A Typology of PW Roles 178 xiii List of Figures Figure 1 Reconciling Concept and Context 100 xiv Acknowledgements Now that this long journey is over, I can take time to acknowledge the many people who supported, encouraged, challenged, and assisted me in so many ways. My heartfelt thanks to my dissertation supervisor, Dr. Lawrence W. Green, whose high standards, wisdom, experience, and kindness helped me strive always to do just a bit better. I learned much in the years I had the privilege of studying with him. His support and his trust in me kept me going. Other members of my supervisory committee were equally supportive. Dr. Judith Ottoson is one of the best teachers I have known. She provided a safe haven for exploring new and radical ideas (at least new and radical to me) yet always pushed me to go just a bit deeper and a little further in my thinking. She has a gift for asking just the right questions and she did not let me take the easy way out. Dr. Garry Grams provided much more than methodological expertise, although his contribution there was also important. He introduced me to Grounded Theory and so this study owes much to him. More than that, however, he listened, and said the right things when I needed to hear them most. His feedback in the final stages of writing was invaluable. Dr. Bonnie Long also held me to high standards and challenged me on some important assumptions. Her thoroughness and attention to detail made my life so much easier when it came time to do final editing. Thanks also to Dr. Allan Best who began as a committee member and saw me through the completion of my comprehensive examination. My pre-doctoral work with him provided a firm foundation for this research. My research benefited enormously from the assistance of the Grounded Theory Club at the University of Victoria. I am deeply grateful to Dr. Rita Schreiber, Dr. Jane Milliken, Tina Jennison, Valerie Watanabe, and Susan Noakes for their "peer debriefing" as I struggled to sort out my core category. They wallowed with me in the data, but kept me from getting mired. Rita, dear friend and colleague, deserves special thanks for getting me "out of my data and into my head". She was always available and no question was too silly or trivial. Without her, I would never have survived the challenges of the last few weeks of writing. Jane too, helped me through the tough times, having just travelled the same path herself. I am deeply grateful to my colleagues at the University of Victoria School of Nursing for their support and patience as I struggled to complete my dissertation. They made it so much easier for me by being flexible about my teaching load and by letting me off the hook for some other responsibilities so I could devote the necessary time to analysis and writing. My students also deserve special mention, particularly those in my Nursing 350 Health Promotion course, because they tolerated my reduced availability over the winter term and were very supportive of me in my final push to finish. This research was made possible by funding from several sources. A research contract from the British Columbia Ministry of Health, Alcohol and Drug Programs Branch to Lawrence W. Green, Dr. PH. at the Institute of Health Promotion Research supported data collection and analysis. Doctoral Fellowships from the Ministry of Health, Alcohol and Drug Programs, and from the National Health Research and Development Program (Award #6610-1987-47) supported me from 1992 to 1995. I am particularly grateful to Wayne Mitic, formerly of the Ministry of Health, for advocating to establish the Ministry Fellowship, and for proposing the idea of conducting my doctoral research in relation to the School-Based Prevention Project. Throughout this project, he was supportive of me and open to hearing whatever findings the study produced. Among the most important of my colleagues in this study were the School-Based Prevention Workers. Without their trust, support, and openness I would have been unable to carry out this research. The work represented in these pages owes as much to them as it does to anyone, although I accept full responsibility for the interpretations offered here. Heartfelt gratitude is extended to Todd Thexton, friend, colleague, and research participant, with whom I spent hours in dialogue about this project and the findings as they emerged. His insightful analysis greatly influenced my own and thus is reflected throughout these pages. Thanks also to the administrators, teachers, counsellors, students, parents, and agency staff affiliated with each of the six case study schools who let me into their worlds, and gave so willingly of their time. xvi I will be forever thankful to Margaret Cargo, dearest friend and sister traveller, for her friendship, support, encouragement, and intellectual companionship. My co-learning experiences with Margaret were among the richest in graduate school. Finally, I offer my deepest and most heartfelt thanks, and my love to my family, Ken, Karlyn, and Hayley Underdahl, who made enormous personal sacrifices so I could return to university to obtain my PhD. To Ken, loving partner and friend, who has patiently supported me in so many ways through three degrees over 19 years, I promise you - this is it! We shall now "have a life". And to my dear children, who have put up with late dinners (or no dinners at all), and an "unavailable" mother, thank you for your patience and your loving support. Their lived experience of adolescence and of being students in secondary schools enriched my own understanding of some of the issues raised in this study. xvii Dedication This dissertation is dedicated to the memory of my father, Ronald Lawrence MacDonald, who died before I could share the joy of my accomplishment with him, and to my mother, Mary Avelle MacDonald, without whose love, faith, support, and encouragement I might never have begun this journey, let alone completed it. PART A THE RESEARCH INVESTIGATION The first part of this document will provide an overview of the problem of alcohol and drug misuse among adolescents, identify the research question and its significance, outline the existing knowledge in school-based health promotion, and describe the School-Based Prevention Project that was established to prevent the problem of alcohol and other drug misuse among secondary school students. Theoretical perspectives on implementation, symbolic interactionism, grounded theory, and health promotion will be discussed, particularly as they relate to this study. The research method used to explore and develop an understanding of the implementation of this program will also be described. 2 CHAPTER ONE In 1992, the Alcohol and Drugs Programs Branch1 (ADP) of the British Columbia Ministry of Health and Ministry Responsible for Seniors (MOH) initiated a three year pilot project entitled the School-Based Prevention Project (SBPP). The aim of the SBPP was to prevent alcohol, tobacco, and other drug misuse2 among secondary school students in British Columbia (BC). The SBPP involved the placement of a Prevention Worker (PW) in selected secondary schools in the province. The PW was expected to use the School-Based Prevention Model (SBPM), a systematic planning process based on the PRECEDE-PROCEED model for health promotion planning (Green & Kreuter, 1991). The intent of the SBPP was that the PW would engage the school community3 in a collaborative and participatory process to identify and define their own health issues related to substance misuse, and to develop, implement, and evaluate strategies to address these concerns. The key features of this innovative project were the implementation of a new position/role in the schools and the implementation of a defined planning process to guide program development. A detailed description of the SBPP is provided in Chapter Two. The Institute of Health Promotion Research (IHPR) at the University of British Columbia (UBC) was contracted by the Ministry of Health to conduct an evaluation of the SBPP. The study reported here is a 1 A subsequent reorganization of the Ministry of Health renamed this branch Alcohol and Drug Services (ADS) and located it in the Prevention and Health Promotion Branch. The acronym ADP will be used in reference to documents, reports or decisions made by that branch prior to their name change in 1994. Otherwise, the acronym ADS will be used. 2 The term 'substance misuse' is used officially by Alcohol and Drug Services in all goal statements related to their funded programs. It is defined in the School-Based Prevention Model Handbook (Alcohol and Drug Programs, 1994) as "any use of a psychoactive (consciousness or mood altering) substance that is harmful to the user, or that has significant risk of causing harm" (p.5). By this definition, use of a drug is not necessarily equated with misuse. The word misuse will be used when referring specifically to SBPP program objectives. Otherwise, the definitions presented in the footnote on the next page apply when the terms drug or substance use, abuse, and dependence are used. Where relevant, specific drugs (such as tobacco and alcohol) will be named. Unless otherwise indicated the terms "drugs" or " substance"are used in the generic sense to include alcohol, tobacco, and illicit drugs. 3 School community' is defined by the SBPP as "all of those who are connected in some way with the school; teachers, students, parents, staff, school nurse, school administrators, and agencies who come into and offer programs in the school. The school itself, is a community in its own right as well as being a member in the larger community in which it resides" (Alcohol and Drug Programs, 1994, p. 132). 3 component of the process evaluation. An overview of the design for the overall SBPP evaluation can be found in Appendix A. The purpose of this study was to explore the process by which PWs implemented both the PW roie and the SBPM in their schools. The Problem of Alcohol and Drug Use4 Among Adolescents The need for the SBPP was premised on the results of two surveys conducted in British Columbia by ADP in 1987 and 1990 (Alcohol and Drug Programs, 1992a; BC Ministry of Health, 1988a). These surveys reported relatively high rates of alcohol and other drug use among BC adolescents, whatever the measure, compared to those in other parts of Canada and the United States, despite an overall reduction in use between 1987 and 1990. Data from a national survey concluded that BC had the highest lifetime prevalence rate of marijuana use by young people of all provinces in Canada. Among those aged 15 to 24 in BC, just over 50% reported having used marijuana at some point in their lives. In contrast, the rate for all of Canada was 34%. BC also had the highest annual prevalence rate for marijuana use by 15 to 24 year olds (27.6%). The next highest provincial rate was in Quebec at 16.9% (Eliany, Wortley, & Adlaf, 1991). At the same time, BC was the only province in Canada that showed increases in the number of reported drug offences among adults and juveniles from 1989 to 1991 (Coordinated Law Enforcement Unit, 1994). Alcohol is the drug of choice across all age groups in Canada. Among adults, 78% of Canadians were current users in 1989; BC had the highest rate of current use at 83% (Health and Welfare Canada, 4 The following definitions of drug use; abuse, and dependence are those put forth by the American Psychiatric Association (1987) in the Diagnostic and Services Manual (DSM-III) and by the Ninth International Classification of Diseases (World Health Organization, 1980). Drug use involves low doses or infrequent use with rare or minor negative consequences. This level of use is sometimes referred to as experimental, social, recreational, or casual, although some might object to these adjectives on the basis that they believe any use in adolescents constitutes abuse. Drug abuse usually involves higher doses and frequencies than those characterizing dependence, and there may be occasional heavy use periods. Generally there are detectable health consequences and some adverse effects on functioning. Drug dependence involves high doses or frequent instances of continuous use over a period of at least one month, and includes elements of compulsion, craving, and withdrawal symptoms. There are generally severe health consequences and serious functional impairment. Injury and violence can often be involved Glantz and Pickins (1991) suggest that there is no clear demarcation in the spectrum of use to abuse, but that there is general consensus on the extreme poles of the continuum. 1990) . Among adolescents, annual prevalence rates of alcohol use and weekly use were highest in British Columbia and Quebec (Health and Welfare Canada, 1996). North American surveys of adolescent drug use indicate that there has been a declining trend in prevalence rates over the past decade or more. Data from Ontario, and the United States suggest that the downward trend has either halted or is reversing (Adlaf, Smart, & Walsh, 1993; Johnston, 1994). Early data from the evaluation of the School-Based Prevention Project support this conclusion (School-Based Prevention Project Evaluation Team, 1994b, 1995c). For example, in 1990 the percentages of students who reported having used alcohol, tobacco, cannabis, and LSD in the previous year were 72%, 27%, 23%, and 9%, respectively. In 1993, the percentages in the SBPP. survey population were 71.4%, 46%, 33.7%, and 15.3%, respectively.(School-Based Prevention Project . Evaluation Team, 1995c). Thus, while alcohol use remained relatively constant (as it has been throughout the 1980s), tobacco,5 marijuana, and LSD use has increased. At the time this project was initiated, however, there was no indication that drug use rates were on the increase. The intent of the SBPP was to support the continued decline in drug use prevalence rates that was apparent in BC rates between 1987 and 1990. The goals of the SBPP, as described in Chapter Two, reflect this intent. ADP viewed this as an important focus in light of the generally higher rates of alcohol and other drug use by BC adolescents, compared to rates of use by youth in the rest of Canada. Addiction, and the problems associated with chronic drug use are complex (Wilkinson & Martin, 1991) , difficult, and costly to treat (Hawkins, Catalano, & Miller, 1992; Single et al., 1996) so it makes sense, both from a human and an economic perspective to prevent these problems before they begin. Because most drug use is initiated in adolescence, the majority of prevention programs target young people prior to initiation to prevent or delay onset or during the early stages of experimentation to prevent progression to regular use. 5 There was a slight difference in the wording of the smoking question between the 1990 ADP survey and the 1993 SBPP survey which may, in part, explain the large differences between 1990 ADP rates and the 1993 SBPP smoking rates. 5 The case for prevention is also supported by data on the patterns of morbidity and mortality in adolescence and adulthood. Unintentional injuries, suicide, and interpersonal violence are the three most common causes of death among adolescents (Blum, 1991). Non-fatal injuries are the major cause of morbidity (Peters & Tonkin, 1994). Drug use, particularly alcohol use, by adolescents can be implicated in all of these (Morisky, McCarthy, & Kite, 1986). Tobacco use is associated with the major causes of morbidity and mortality in adulthood (i.e., cancer and cardiovascular disease), and with respiratory tract symptoms and poorer overall physical health in adolescence (Arday et al., 1995). Tobacco also is a highly addictive substance. Illicit drug use is implicated strongly in the acquisition of HIV. Although few adolescents who use alcohol and illicit drugs during adolescence go on to develop a chemical dependency (Farrow, 1994), those who begin smoking regularly in adolescence do have a high probability of continuing to smoke regularly into adulthood, with all of its attendant health consequences (Kandel, 1988). The health consequences of tobacco and alcohol use have been well documented (Farrow, 1994; MacDonald, 1984; Task Force on Smoking, 1982; U.S. Public Health Service, 1979, 1982). The evidence on the effects of marijuana use on health status is less conclusive. A recent review of this evidence by the World Health Organization (1995) concluded that there are health risks associated with heavy cannabis use (emphasis added), especially when used over a prolonged period of time. The most probable health risks of chronic heavy use include: (a) the development of a dependence syndrome,, (b) an increased risk of involvement in motor vehicle accidents, (c) an increased risk of developing chronic bronchitis, (d) an increased risk of respiratory cancers, and (e) an increased risk of giving birth to low birth weight babies when used during pregnancy. At current levels of use, particularly the levels used by most adolescents in BC, cannabis is a much less serious public health problem than either alcohol or tobacco. There may be some cause for concern, however, given the increased rates of use by BC students between 1993 and 1995 in SBPP schools (School-Based Prevention Project Evaluation Team, 1995d). The risks for motor vehicle accidents, in particular, are 6 likely to be increasing as well. The legal consequences of marijuana use may be one of the most important harmful outcomes because of possible long-term enmeshment in the criminal justice system. From a societal perspective, drug misuse in Canada exacts a heavy toll, both economically and socially. Single and colleagues (1996) have estimated the cost of substance misuse in Canada to be $18.4 billion in 1992, or 2.7% of Gross Domestic Product. The bulk of these costs are related to treatment, law enforcement, and lost productivity. The position of the BC Ministry of Health is that drug misuse is a major health problem with economic and social consequences that extend beyond the health system to the education, criminal justice, and social services systems, and to the private sector. "The cost of drug abuse includes: lost productivity; accidental injuries; deaths; disruption to families and communities; and increased public expenditures in law and order programs, correctional institutions, welfare agencies, and the health care system" (Alcohol and Drug Programs, 1992, p. 4). Thus, the rationale for this particular prevention program is grounded, not only in concerns about the health and educational impact of substance misuse on adolescents, but also on the associated economic and social costs.6 The Historical Context of Substance Misuse Prevention Most of the literature on drug use prevention has involved school-based classroom curricula. Early generations of substance abuse prevention (knowledge-based and affective programs) have been almost universally judged to be ineffective at best, and to increase rates of drug use at worst (Bruvold & Rundall, 1988; DeHaes, 1987; Swisher, Crawford, Goldstein, & Yura, 1971; Tobler, 1986). Other approaches, including risk factor, behaviour, developmental, and community-specific approaches (Gerstein & Green, 1993; Leukefeld & Bukoski, 1993) have had equivocal results (Best, Thompson, Santi, Smith, & Brown, 1988; Gorman, 1992, 1995a, 1995b) or have been inadequately evaluated (Gerstein & Green, 1993). Much of the school-based research has centred on the social influences model, a behavioural approach, in which the emphasis is on developing skills to resist the most salient social influences on a young person's decision to 6 The above statement of the issue represents one particular social construction of the problem of adolescent alcohol, tobacco, and other drug use. It was constructed through a process of sifting, sorting, and selecting "relevant" pieces of data that help to frame the problem in a particular way, in this instance, from a public health perspective. There may be other social constructions of the problem. use or not to use drugs. Results are equivocal. The most conclusive benefits have been demonstrated in the area of smoking prevention in which it can safely be concluded that the social influences model works some of the time for some of the youth (Gerstein & Green, 1993). The latest thinking in alcohol and drug use prevention, stimulated by developments in health promotion, reflects an expanded focus in which single-focus, universal strategies are no longer the preferred option. Rather, preventive interventions must be based variously on the particular local culture, circumstances, traditions, and population needs. Prevention goals are being re-evaluated, and there is more concern now with prevention goals that are relevant for a broader age range in the population, which means extending the focus beyond junior high or middle schools (Gilchrist, 1991). Research on developmental stages in drug use is shifting the attention to other goals, such as preventing transitions in drug use, in addition to the traditional goal of delaying onset (Clayton, 1991; Kandel; 1988; Kandel & Yamaguchi, 1993). Overall, the recent emphasis has been on the development of comprehensive programs targeting multiple levels of intervention. In terms of school programs, the emerging consensus is that drug use prevention is more likely to be successful when implemented in the context of comprehensive school health programs linked with community programs (Carnegie Council on Adolescent Development, 1989). Thus, there has been a shift to a community-specific focus in drug use prevention; one that may be located in a particular setting such as a school but expands from that base to take action in the larger community in ways that support and reinforce school efforts. This recent shift in theorizing about effective substance misuse prevention is the context within which the SBPP was designed by the BC Ministry of Health. As such, the SBPP represents an innovative approach to prevention that has not yet been sufficiently tested. It was a question about the effectiveness of this type of approach that led to the development of the overall evaluation design to assess the impact of the SBPP on drug use prevalence rates in the school population (See Appendix A). To assess impact, however, it is important to know whether a program has been implemented as planned or intended, and ways in which its implementation has been adapted by 8 practitioners to its environment. Therefore, questions about the feasibility and extent of implementation led to this study. Developing a Study of SBPP Implementation The critical importance of studying implementation was recognized in the wake of disappointing results from the "Great Society" era programs in the United States during the 1960s (Palumbo & Callista, 1990; Van Meter & Van Horn, 1975). Many of these programs were found to be ineffective, but on closer examination, the reason for program failure could be traced to a lack of, or inadequate implementation (Pressman & Wildavsky, 1986). The problem was that policy makers and analysts had made the erroneous assumption that once policies were developed the difficult work had been done, and all that was left was simply to implement them (Chase, 1979; Van Meter & Van Horn, 1975). Implementation assessment is important to determine what it was about the program that worked or did not work. Some evaluation theorists (Weiss, 1972; Chen, 1990) point out that many evaluations can be characterized as "black box" evaluations in that they focus primarily on the overall relationship between program inputs and outputs without concern for the transformational processes in the middle. This type of evaluation provides an assessment of whether the program, as implemented, works but does not identify inaccurate assumptions or changes in the program that might aid future program improvement/development or facilitate transfer of the program to a new setting. A black box evaluation does not attend to the political or organizational context of the program, not does it explore the relationships between the program as delivered versus the program as planned. Most importantly, a black box evaluation may have little relevance to policy or practice. Claims of program success on the basis of a black box evaluation may be difficult to apply in other settings or situations because the conditions under which the program works are unknown. If a program is judged to be a failure on the basis of a black box evaluation, this finding may be virtually uninterpretable. A judgment of program failure might be the result of a flawed theory on which the program was based, incomplete or inadequate program implementation, or insensitivity of measurement. Assessment of implementation opens 9 up the black box of the program and allows judgments about the appropriateness of theory, the strength of implementation, and factors that influence the transformation of ideas to action. It also enhances our understanding of the contingencies under which a program may or may not be effective. The SBPP is an innovative idea that is based on health promotion principles, including the principles of participation and community ownership. These principles suggest that, for a program to have the desired impact, community members must participate in defining their own issues and determining and implementing their own solutions. These are important ideas in health promotion. But how relevant are they in the school context? Prior research has demonstrated that schools are reluctant to take on health education and health promotion programs for a variety of reasons (Shamai & Coambs, 1992; Vertinsky, 1989), not the least of which is the lack of time in the curriculum and lack of staff expertise in the area. In recognition of the challenges in implementing health promotion programs in schools because those in schools lack time and expertise, the SBPP was conceptualized to include a PW position in the school. The PW would be responsible for facilitating the development of a comprehensive prevention program. To do this, the PW was expected to implement a health promotion planning process (the SBPM) that would encourage the school to develop and implement primary prevention strategies in the school setting. This conceptualization of the SBPP raises a number of unanswered questions that have not been addressed in the literature. Is the PW role able to be integrated in the school context in a way that facilitates the development of a comprehensive and potentially effective prevention program? How well is the PW able to implement the SBPM to develop such a program and what are the factors that influence the implementation of that process? Is the PW able to implement a new role and the SBPM in the way that was intended by the developers of the program? How well did PWs understand the SBPM? What did it mean to them? I came to this study with an interest in implementation issues and a curiosity about the answers to the questions raised above. Ottoson's (1993) findings that participants in her study used the Precede-Proceed model (on which the SBPM is based) in a wide variety of ways, intrigued me. Participants in her study were not required to use the model for their jobs. Was there a difference in application by PWs who were required 10 to use the model? My original research question was therefore related to whether PWs were able to implement the model in the school context and if so, how did they actually apply it? As I began to interact with PWs as they started to work in schools, however, it became clear that to implement the model, PWs first had to implement the PW role in a challenging context. Implementing the model might only be a small part of the story. I struggled with the need to focus my research question to keep the study manageable, but it became impossible for me to explore the SBPM implementation without understanding the context within which implementation evolved in its various ways. Thus, my interest shifted from a focus on how the model was implemented to a focus on how PWs implemented both a PW role and the model and the contextual influences on that process. Purpose The purpose of this study was to explore the process by which PWs managed the implemention of the SBPM and the PW role in a sample of secondary schools in BC, and to identify and describe the influences on that process. I used the grounded theory method as outlined by Glaser and Strauss (1967), Glaser (1978), Strauss (1987), and Strauss and Corbin (1990) to explore the experiences of PWs and other members of the school communities participating in the project. Research Questions Grounded theory begins with a broad research problem that identifies the phenomenon to be studied (Glaser, 1978). The phenomenon, in this instance, was the implementation of the SBPP, specifically the implementation of the PW role/position, and the implementation of the SBPM, a defined planning process. Strauss and Corbin (1990) suggest that a novice grounded theory researcher should delimit the broad problem into a more specific question or questions, but these should be stated in a way that remains broad enough to allow emerging possibilities. In relation to this, the research questions were: * What is the basic social problem7 experienced by the participants in implementing the SBPM and the PW role? 7 The "basic social problem" is a grounded theory term to describe the basic concern or problem that participants in the action scene must handle or manage. 11 » How do schools and PWs manage the implementation process? • What are the major influences on implementation of the PW role and the SBPM? Significance of the Study The SBPP was an innovative pilot project aimed at preventing substance misuse among adolescents in the secondary school system. It was developed and funded by the Ministry of Health and implemented in collaboration with the Ministry of Education. The project is therefore an example of interministerial collaboration and so the findings have significance for developing future collaborations. The organizational challenges of implementing a complex project, involving multiple levels of two distinct systems provides an excellent opportunity for "policy learning" (Springer & Phillips, 1994), particularly in relation to how projects of this nature can be implemented in the school system and how the two systems can work together to achieve complementary goals. As mentioned above, the SBPP is based on health promotion principles and embodies a set of values that may differ from those of the education system into which it was being introduced. The study will therefore provide a theoretical understanding of the contextual conditions that need to be in place for a program of this nature to be implemented successfully and ultimately, to succeed in its goals. Alternatively, it may provide an understanding of the ways in which the program needs to be adapted in design to fit the contextual conditions in most schools. In focussing on the implementation experiences of the key participants in the process, this study identifies the challenges of translating basic principles of health promotion, embodied in the SBPM, into practice at the local level. In doing this, the study may also provide useful information to policy makers and program managers about how the process can be improved in future sites or similar programs. Overview of the Dissertation This dissertation is organized into three major sections: Part A , the Research Investigation, Part B, Reconciling Concept and Context, and Part C, Implications. Part A , The Research Investigation, is composed of five chapters. Chapter One provides the introduction and overview of the dissertation. Chapter Two 12 reviews the literature8 on the history of school-based health education and alcohol and drug misuse prevention, describes the SBPP, the implementation of which is explored in this study, and then provides an overview of current thiriking in substance misuse prevention. Chapter Three discusses what is meant by implementation and how the understanding of implementation is influenced by the theoretical perspectives within which an implementation issue is framed. The chapter closes with a review of the implementation issues that shape the focus of this study. Chapter Four reviews symbolic interactionism, the theoretical and philosophical perspective that forms the basis for grounded theory methodology. Criticisms of grounded theory and symbolic interactionism are identified and challenged. Symbolic interactionism and grounded theory are explored for their relevance to the study of health promotion concerns. The final chapter in Part A , Chapter Five, describes the methodology of this particular study. It begins with a rationale for using grounded theory, then describes the data source, research participants, case selection, and the analytic methods. The chapter ends with a discussion of rigour in qualitative research. 8 Grounded theory method (Glaser & Strauss, 1967) attempts to avoid the possibility that prior theoretical conceptions force data collection and analysis into preconceived categories. Glaser, in Theoretical Sensitivity (1978) cautions against reading too much in the substantive area for fear of contaminating the ability to generate concepts from the data by gathering data according to preconceived concepts identified in the literature. This is an impossible methodological dictate to adhere to as a graduate student. In preparation for the dissertation research, course work in the subject area(s), and wide reading in the substantive and theoretical literature of the discipline(s) was necessary. In this case, it is impossible to take a "pure" atheoretical stance by avoiding the literature altogether. There is a balance to be struck between cultivating the necessary theoretical sensitivity and avoiding the imposition of prior theoretical conceptualizations on the data, as the authorities below suggest. Given that it is often impossible to avoid exposure to prior theoretical conceptualizations, it is important to put these on the table for exposure, critique, and deconstruction. In fact, Denzin (1989) talks about the process of "deconstruction" as a first step in interpretive research in the symbolic interactionist tradition. Deconstruction is a way of laying bare prior conceptions of the phenomenon in question so that biases surrounding existing understandings of the phenomenon can be presented, challenged, and critically examined. In doing this, the researcher is "coming clean" (Locke, Spirduso, & Silverman, 1993), and the reader will be able to make a judgment about the extent to which preconceptions and biases may have influenced data collection and analysis. In part, this is what I have attempted to do in my review of the literature. Strauss and Corbin (1990) have expanded their conception of theoretical sensitivity, and sensitizing concepts to encompass this type of situation. They view the systematic analytic techniques of grounded theory as safeguard against preconceiving the theory. 13 Part B of this dissertation presents the findings of the research investigation. The analysis of the data yielded the core category of Reconciling Concept and Context, which is comprised of three major categories: Entering the Field, Confronting the Model, and "Doing" the Model. Entering the Field is a large and complex category, so it is presented in three separate chapters, one for each of the three processes that make up Entering the Field. Chapter Six presents gaining entry, which describes the process PWs go through to establish the legitimacy of the program in the school, establish personal credibility and learn the ropes. In Chapter Seven, the process of creating a role is presented, which describes the PW's struggle to manage find the focus of the role and to strike a balance in managing several dilemmas they encounter in trying to implement the PW role in the school. Enhancing school readiness is presented in Chapter Eight and explains how PWs met the challenges of selling the issue of alcohol and drug prevention, selling the model, facilitating participation and steering the steering committee. The second major category, Confronting the Model, is presented in a single chapter because it is a smaller and less complex category than Entering the Field. Chapter Nine, Confronting the Model is about the process by which PWs find themselves reacting to the model and the context within which it was presented to them at the SBPP training session. PWs' experiences in learning the model are then explored followed by a description of the process of contemplating the model, in which the PWs sized up the model, assessed its fit with their own practice and with the school's way of doing things, and tried it on for size. The tenth and final chapter in Part B, "Doing" the Model, describes how PWs actually implemented the SBPM or adapted it to fit the school context. . Part C of the dissertation is the Implications chapter in which the limitations and strengths of the study are presented. The findings of this study are then examined in relation to previous research and the implications for policy, practice, and future research in health promotion are discussed. 14 CHAPTER TWO SCHOOL-BASED H E A L T H PROMOTION: PAST A N D PRESENT The literature reviewed here explores briefly issues related to using schools as the site for preventive interventions. The SBPP is described, including its key elements and organizational structure. Following a description of the project, current thinking in substance misuse prevention is discussed, and the implications of this thinking for exploring implementation issues using a qualitative approach are presented. The remainder of the review focusses on implementation issues and theoretical perspectives on implementation drawing from the literature of diverse disciplines. The intent is to identify the gaps in the literature, to establish the significance of this research, and to situate it in the larger body of theoretical work on implementation, particularly in relation to health promotion. Schools as the Site for Preventive Interventions "...the school should because the school can and, therefore, the school must" Seffrin, 1990 p. 152 The school health education and health promotion literature generally reflects an agreement that schools are important settings for action to improve the health status of the population (Allensworth & Wolford, 1988; Basch, 1984; Cleary, 1991; Green & Iverson, 1982; Green, 1984; Kolbeet al., 1985; Mason & McGinnis, 1985; Mutter, Ashworth, & Cameron, 1990; Stone, 1990; Vertinsky, 1989). In the United States, schools were viewed as "powerful and effective agents to facilitate attainment of the 1990 Health Objectives for the Nation" (Allensworth & Wolford, 1988, p. 3). Approximately one third of these 227 health objectives were identified as those that could be achieved either directly or indirectly through the school setting (Iverson & Kolbe, 1983). The Carnegie Council on Adolescent Development (1989) also stated that "orchestrated drug abuse prevention in schools might constitute society's most cost effective prevention strategy" (Gerstein & Green, 1993, p. 131). The Canadian literature has also been supportive of, or implicitly assumed the appropriateness of, the use of schools for achieving health objectives (Best, Thomson, Santi, Smith, & Brown, 1988; Best, 15 Brown, Cameron, Smith, & MacDonald, 1989; Cameron, Mutter, & Hamilton, 1991; Cogdon & Belzer, 1991; Mutter et al, 1990). In 1988, seven national Canadian education organizations sought funding from Health and Welfare Canada9 to bring together a multi-disciplinary group representing government officials, teachers, school administrators, and health professionals to discuss strategies to address health issues, particularly in high schools. This 'Exchange 88' conference reached a consensus that children's health was a prerequisite for learning, and that a comprehensive approach was both necessary and preferable to the traditional fragmented, issue- and crisis-oriented approach that had characterized health teaching in Canadian schools (Anonymous, 1991). An Emerging Consensus In 1990, the Harvard School Health Education Project (Lavin, Shapiro, & Weill, 1992) initiated a national policy analysis in the United States on school-based health promotion. The authors reviewed 25 reports, published between 1989 and 1991, by diverse individuals and groups representing government, private, and professional organizations, and experts in the field. Some of these reports reflect an education perspective, others a public health perspective. The authors of this article conclude that these reports reflect a growing consensus on the issues and strategies for action. Five main conclusions emerged from that analysis: (a) education and health are interrelated in that each can contribute to improving the other, (b) a more comprehensive, integrated approach is needed because most efforts have been fragmented, targeting categorical symptoms rather than the underlying causes of the problems, (c) the biggest threats to health of school-aged children are social morbidities, many of which are preventable, (d) health promotion and education efforts should be centred in and around schools because schools are community institutions, and need to play a larger role in addressing the health and social problems that limit academic achievement, public health, and economic productivity, and (e) prevention efforts are cost-effective, and the social costs of inaction are too high. 9 Now Health Canada, following a federal government reorganization. 16 Concerns About Health Promotion in Schools The reports reviewed by the Harvard School Health Education Project (Lavin et al., 1992) make it clear that many educators and educational policy makers believe that the school is an appropriate place for education and action on health concerns. Despite this emerging consensus, the practice-based educational literature has been equivocal in its support of health education and health promotion efforts in schools (Shamai & Coambs, 1992; Walker, 1992). Many have expressed concerns that the demand for curriculum time to address a broad range of categorical health problems will dilute the educational mandate of schools (Nader, 1990). Certainly, the experience of school-based health professionals suggests that whatever the rhetoric about support for school health at the policy, government, and even school district levels, when it comes to implementation in schools, there is often a gap between the rhetoric of support, and the actual implementation of health education and health promotion programs (School-Based Prevention Project Evaluation Team, 1993). Gerstein and Green (1993) observed that "those concerned about drug abuse sometimes promote the health or social objectives of prevention without much apparent attention to the priorities of cooperating organizations" (p. 131). For example, in their article on the role of the schools in achieving the U.S. 1990 Health Objectives for the Nation, Allensworth and Wolford (1988) cited the views of those primarily from the health sector. The views of the education sector were not represented in the literature cited in support of their conclusions. For some, the school is an important arena for health promotion efforts simply because it is a convenient way of reaching a large number of adolescents (Millstein, 1993) or, as Seffrin (1990) suggested, "...the school should because the school can and, therefore, the school must" (p. 152). Werner (1991b), however, questions the extent to which schools can be accountable for solving all of society's social ills. He said that Canadians expect a lot of schools, and sometimes these expectations are unrealistic. It may be that the expectations of health professionals concerning the role and impact of school health education is based on an invalid interpretation of the function of education in general (Kolbe et al., 1985). Most school health education models aim to improve cognitive, affective, and skill development to 17 achieve positive changes in unhealthy behaviour thus resulting in improved health outcomes. A focus on health outcomes may not be in keeping with the primary concerns of educators. A more appropriate model may be the comprehensive school health model that incorporates school health services, school health education, and school environment interventions to increase educational performance directly through preventive health services and the environment, or indirectly through changes in behaviour, and thus health (Green & Iverson, 1982). Substance Misuse and Educational Performance For many, the justification for school health programs lies in their potential to improve school performance, and to meet educational objectives (Kolbe et al., 1985). Substance misuse in the developing adolescent may undermine motivation and interfere with cognitive processes (Hawkins, Catalano, & Miller, 1992) including memory and concentration (Farrow, 1994). This has obvious implications for educational performance. Use of drugs and alcohol also affects school functioning because it is often associated with disciplinary problems in the classroom and at school functions (Perry, Kelder, & Komro, 1993). Perhaps the most serious educational consequence of drug abuse may be dropping out of school. Although we know that youth who drop out of school are more likely to abuse drugs after they leave school (Clayton, 1991), there is also evidence that young people who drop out are more likely to have previously used cigarettes, marijuana, and other illicit drugs (Mensch & Kandel, 1988). If drug use has resulted in problems with school performance, then it may have contributed to dropping out, either directly or indirectly. The paradox for schools is that, on the one hand, they are being asked to solve a growing list of health and social problems, including substance use/misuse, while on the other they are under increasing pressure to get "back to the basics," all while resources are diminishing (Elias, 1990). As schools are confronted with increasing demands and reduced resources, some authors suggest that decisions about including health programs in an already crowded curriculum are likely to be based on its value in meeting the school's educational objectives (Green & Iverson, 1982b; MacDonald & Green, 1994b). Educational studies, however, have found that school district administrators and principals often make decisions about adopting 18 innovations on the basis of factors other than their potential to improve academic achievement, and other student outcomes (Huberman & Miles, 1984). Because schools do not exist in isolation from the community, which sanctions and supports their existence through tax dollars, many believe that schools do have an obligation to address their communities' needs, concerns, and interests (Killip, Lovick, Goldman, & Allensworth, 1987). There is some evidence that drug abuse is viewed by the general public as one of the most important problems facing the educational system (Vertinsky, 1989). On the other hand, this is only one of many issues being promoted by a range of interest groups. Werner (1991a) reports on a study by Anderson (1983) in which at least 66 government departments, and other agencies, had specific interests to advance in schools and were backed by the financial resources to make their representations. It is clear that a lack of recognition of the concerns and dilemmas of schools in relation to health promotion, and other interests, will contribute to problems in implementation and may be wasteful of precious resources. The problem has been conceptualized at times by health professionals and researchers as a diffusion or dissemination problem (Anderson & Portnoy, 1989; Basch, 1984). That is, experts outside of schools know what 'works' in schools. A l l that is needed is to get people inside of schools to understand how important it is, and to agree to adopt and implement the curricula developed by the outside experts. Thus, 'we' want the schools to come around to 'our' view of what-is important for 'them' to do. This view is exemplified in the definition of a "health promotion gap" as the gap between what health professionals think people should do, and what they actually do. The role of health education is to close this gap (Orlandi, Landers, Weston, & Haley, 1990). This view appears to violate an historical principle of health education and philosophical tenet of some definitions of health promotion; that is, those affected by a problem must be involved in naming it and seeking solutions. What rarely seems to be acknowledged in the school health education literature are the inevitable trade-offs and value conflicts that must be addressed at the school level. Also, what works is not always so clear cut. 19 An emerging rationale for locating prevention programs in schools is that schools themselves may be sources of risk for the initiation and maintenance of drug use. One line of prevention research suggests that academic failure, a low degree of commitment to education, and low attachment or bonding to teachers, and schools increase the risk for adolescent drug abuse (Hawkins, Doueck & Lishner, 1988; Hawkins & Catalano, 1990). Data from the World Health Organization multi-national survey on youth health behaviour support this conclusion (King & Coles, 1992; Nutbeam & Aaro, 1991). There is strong association between regular smoking and alienation from school. This finding is consistent across gender, culture, and the social organizations represented by different countries. Youth who are alienated from school because of low levels of achievement, and who experience difficulties in relationships with their parents are more likely to engage in health risk behaviours, including substance use. Schools as Both a Setting and a Target for Intervention Trickett and Doherty Schmid (1993) suggest that adolescents may bring into the school their problems that originate in other contexts, but the way the school manages these difficulties can exacerbate existing problems, and even create new ones. In reviewing a range of studies that examined the impact of school structures and processes on student outcomes, they concluded that schools do affect adolescents in a variety of important ways. Thus, the school becomes an important social context that may contribute to the development and/or maintenance of problems such as substance misuse. This makes it an important setting for prevention but also an important target for intervention as well. Walker (1992) even questions whether schools, as they are currently organized and structured, can bring about significant changes in drug use behaviour through curriculum efforts because schools themselves are settings that actually encourage drug use. He argues that drug use is prohibited by adult authority yet school communities informally sanction use among adults. Schools also provide a "venue for close peer relationships and social networks and an implicit culture of deception and risk taking that appears designed to support a culture of resistance" (Walker, 1992, p. 47). In other words, the adolescent culture in schools promotes drug use and risk taking in opposition to adult authority. 20 Thus, the problem of drug use may not lie in the curriculum but in the institution of schooling itself. However, the analysis supporting Walker's conclusion that schools are inappropriate settings for prevention programs might equally be used in support of a conclusion, or at least a hypothesis, that effective prevention requires simultaneous change on the structure and organization of the school itself in addition to health education in the classroom. The BC Ministry of Health, in implementing the SBPP, makes the basic assumption that the school is the appropriate locus for substance misuse prevention efforts. In an official background and rationale document for the SBPP (Alcohol and Drug Programs, 1992b) the following is stated: The school setting is a natural place for learning. There are many who believe that schools are the best place for teaching children about the complex risks to health and about how both individuals and society can attempt to control those risks (Vertinsky, 1989). Efforts to prevent drug use and misuse by adolescents and to effectively intervene when problems do arise requires [emphasis added] a focus on the school setting. Research has demonstrated . that overall rates of treatment effectiveness can be substantially improved i f alcohol and other drug problems are identified early. Traditional approaches to alcohol and other drug intervention typically do not touch adolescents until a crisis is reached. It is the school setting that offers an opportunity to promote and encourage healthy behaviours, and to identify and intervene with adolescents to ultimately reduce the social and economic costs associated with drug misuse. The School-Based Prevention Project (SBPP) is based on the concept that it is within the school environment that the greatest potential exists to: capitalize on the energy of adolescents; reach adolescents during their formative stages; and develop services that are comprehensive, continuous and age appropriate (p. 15). [emphasis in the original] The School-Based Prevention Model Handbook (Alcohol and Drug Programs, 1994) extends this rationale for substance misuse prevention in schools by making the argument that it is in keeping with a comprehensive school health model, and that PWs are an important adjunct to the Learning for Living curriculum. The School-Based Prevention Project The School-Based Prevention Project (SBPP) was developed and funded by the Alcohol and Drug Programs (ADP) Branch, originally of the Ministry of Labour and Consumer Services in British Columbia (BC). The branch was subsequently relocated to the Ministry of Health and Ministry Responsible for Seniors and renamed Alcohol and Drug Services (ADS) in the Prevention and Health Promotion Branch where it 21 remained during the conduct of the field work for this study. As of January 31, 1997, the part of the branch responsible for youth programs and services was moved to the newly created Ministry for Children and Families. The SBPP was initiated as a 3 year pilot project that began September 1,1992, and ended on August 31, 1995. The pilot phase began several months after the first PWs were hired and had begun to work in the schools. In September of 1995, the SBPP achieved official program status with the Ministry of Health, and was renamed the School-Based Prevention Program. The aim of the SBPP was to involve youth, parents, and other members of the school community in activities that promote health by preventing and reducing substance use/misuse by BC youth in selected secondary schools in the province. The SBPP was based on a community participation philosophy. The Ministry of Health provided the basic structure and personnel for the programs, but the intent was that schools and communities would plan and direct the specific activities of the SBPP. The intended role of the PW was to facilitate collaboration and consultation among students, the school, parents, and the community related to substance use specifically, and health promotion more generally. In 1991, a request for proposals (RFP) went out to alcohol and drug agencies and school districts across the province requesting that agencies and school districts form partnerships to develop a school-based prevention program, and submit a proposal for funding. The central feature of the program was the provision of contract funding to hire a school-based PW. A small amount of money was also provided to support the school planning process, and to fund local strategy development. The specifics of program development and implementation were left up to the local community partners but the emphasis was to be on preventing alcohol and other drug misuse by adolescents aged 12 to 18 in the secondary school system. The objectives of the SBPP, as outlined in the program documentation, included both process and outcome objectives. The outcome objectives were: (a) to increase the mean age of onset of substance use, (b) to decrease substance use by youth, (c) to reduce the proportion of youth at high risk for substance misuse, (d) to decrease the negative consequences of substance use in youth, and e) to increase the proportion of youth abstaining from substance use. The process objectives of the project were: (a) to develop activities for 22 students who do not have a substance use problem, with the purpose of avoiding or reducing the likelihood of a problem occurring or improving or reinforcing healthy attitudes and/or behaviours, (b) to plan activities intended to develop the capacity of school and community personnel to be effective partners in the prevention of substance use, (c) to assist school personnel in developing programs for students at risk of developing substance misuse, and (d) to assist teachers with the implementation of the Learning for Living 1 0 curriculum (School-Based Prevention Project Evaluation Team, 1993). The SBPP began in the fall of 1991 when 18 PWs were hired to work in selected middle and secondary schools across the province. Most of the original contracts (N=15) were held by a local alcohol and drug agency that was responsible for hiring the PW and overseeing the project. A few of the contracts (N=3) were established directly with school districts. The School-Based Prevention Model Shortly after the first wave of PWs were hired, a new Director was hired in the ADP branch. This person introduced the Precede-Proceed model (Green & Kreuter, 1991) as the generic planning framework for the project and contracted an evaluation with the Institute of Health Promotion Research at the University of British Columbia. Many PWs believed that the new Director was the driving force behind the commitment of ADP to the Precede-Proceed model and then the SBPM. See Appendix B for an overview of the original model and subsequent revisions made to the model by ADP. The new Director of the Prevention Branch in ADP had been a university professor involved in training health education specialists using the model. Several people in the ADS service system, including PWs, believed that Precede-Proceed had been instituted to ensure a sound theoretical basis for the program although a Ministry representative suggested that this was not the primary reason. "The concern from the ADS perspective was to help the programs focus and remain accountable to definable prevention goals. The theoretical basis was secondary, and viewed as a support to 1 0 learning for Living was the comprehensive school health program, formerly mandated by the Ministry of Education in British Columbia. Learning for Living had three components: a health education curriculum, services for students in the school setting, and a healthy school environment (Ministry of Education, 1990). In 1995, Learning for Living was subsumed (in a diluted form) by the new Career and Personal Planning (CAPP) program, which was not based on a comprehensive school health framework but included only the curriculum component (Province of British Columbia, 1995). 23 the issue of accountability and focus. The Precede-Proceed model helped in this regard" (personal communication, Michael Egilson, Ministry of Health, December 17, 1996). No training had been provided for the original 18 workers prior to their entry to the school system, and the majority of these PWs had not worked previously in schools. This was not unusual in the Ministry because it was rare for Ministry-wide training to be held for this type of program. The PWs were assumed to have been hired with the requisite skills to work in the area of alcohol and drug prevention. Once the Precede-Proceed model was introduced, however, ADS believed that training was necessary for PWs to learn the model. The first training session was held in February of 1992 in Vancouver for PWs, and their agency and school supervisors. The main focus of the training session was to introduce the Precede-Proceed Model, to provide an opportunity for PWs and their supervisors to learn the model, and to help them develop a clear understanding of what was expected from the program in each local setting. The majority of Wave One PWs had been working in their schools for up to 6 months before the first training session. In the summer and fall of 1992, several new PWs were hired11, and an orientation/training session was held in October for the existing, and the recently hired PWs. Based on comments about the Model from the previous training session, several modifications had been made to simplify it prior to this second training session. A diagram of this revision is presented in Appendix B. The major changes involved turning the diagram around so that it began on the left, and moved to the right rather than vice versa, revising the titles of each phase in the model, and changing the names of the predisposing and reinforcing factors to motivating and rewarding factors. The model was renamed the School-Based Prevention Model (SBPM). Although the model direction was reversed in this first revision, the text in the training manual (Alcohol and Drug Programs, 1992) still stated that the process begins with the final desired goal, which was now on the left of the page instead of the right, and works backwards to identify what preceded it. Thus, there was no real change to the core elements of the model. 1 1 By 1993, there were 43 PWs working in 56 schools, and 8 contracts were held by school districts. By February of 1995, there were 42 PWs working in 53 schools, and 10 of the contracts were held by school districts. The program was dropped from 3 schools for a number of reasons, primarily related to difficulties in getting the program established satisfactorily in those schools. 24' At this point, the policy and regulatory components of the Proceed portion of the model remained in the SBPM. A handbook was developed for PWs that was intended to provide a user-friendly, step-by-step set of guidelines for using the model. A draft of the handbook was introduced to PWs at the second training session for feedback and discussion (Wharf Higgins & MacDonald, 1992). Feedback from this second training session, and subsequent discussions with PWs led to further revisions to the model (see Appendix B). The model was simplified in that stages were merged, the Proceed portion of the model was omitted,12 and a preliminary "Getting Ready" stage was added as the first stage. The diagram was now represented as a circle, rather than as a linear model with boxes and arrows. The core of the Precede model was retained (i.e., emphasis on the predisposing, reinforcing, and enabling factors) as the central element of the planning process. Recall that the predisposing and reinforcing factors had been renamed motivating and rewarding factors in the first revision to the model. In the second revision, the motivating and rewarding factors reverted to their original names, but the enabling factors were renamed facilitating factors. Many of the PWs had been addictions counsellors, and the term "enabling" had a negative meaning for them. The first draft of the model handbook (Wharf Higgins & MacDonald, 1993) was revised in line with the changes to the model (Alcohol and Drug Programs, 1994) and distributed to PWs several months following the second training session, just prior to the final year of the pilot phase of the project. A description and diagram of the final revised model (i.e., the SBPM) is outlined in Appendix B. Contractual Arrangements At the beginning of the SBPP, Alcohol and Drug Programs was organized into five service regions. In each region, Area Managers were responsible for managing SBPP contracts within a defined area. This changed with the restructuring that followed the formation of the new Ministry for Children and Families. For the purposes of this study, however, the relevant organizational structures are those that were in place. 1 2 The Proceed portion of the model did not appear to be deliberately omitted but in the attempt to make the entire process more user friendly, the policy and regulatory components were somehow left out. However, there was an explicit requirement in the PW contract that specified that the PW was to develop school policies related to alcohol and other drug use. Thus, despite the elimination of the Proceed portion of the model in the revision of the SBPM, there remained an expectation that school policy would be an important prevention strategy in the SBPP. 25 prior to 1994, and from 1994 to January 31, 1997. Appendix C presents the organizational charts for these two periods of time. In each region, an ADS area manager was assigned to manage the contract with either the alcohol and drug agency or the school district. The reporting relationship for the PW varied depending on which organization held the contract. If a school district held the contract, the lines of authority, responsibility, and accountability were straightforward. In this case the PW was responsible to the school district through the local school administration. The school district was accountable to the local ADS area manager for meeting the terms of the contract. Schools did not have any other contractual relationships with ADS, and for the most part, there was no history of a prior relationship between the two organizations. If a community agency held the contract, the PW was accountable to the agency who paid her or his salary, but was also accountable to the school in which he or she worked, usually through an administrator or head counsellor. Thus, there was one more organizational level to which the PW was accountable, and two service delivery systems to accommodate. The agency, as for the school district, was accountable to ADS for the management of funds, and completion of the contract terms. The majority of these community alcohol and drug agencies had long-standing relationships with ADS and relied on them for a large portion of their funding. Thus, most agencies had more than one contractual relationship with ADS. The contract specified the terms and conditions of payment to the organization and outlined a service schedule for the PWs. The terms of the contract were revised at the end of the second year of the 3 year pilot phase to reflect more closely the stages in the SBPM. Copies of the original and revised generic service schedules, and their associated Schedule A, which outlines specific contract requirements, are presented in Appendix D. The original service schedule outlined the requirement for the SBPP to provide prevention services within the context of a comprehensive school health program. In addition, the PWs was responsible for carrying out the following: (a) developing and maintaining a prevention steering committee comprised of students, teachers, administrators, and parents, (b) development of comprehensive school health policies, (c) 26 developing and implementing health promotion programs specifically in substance misuse prevention, (d) providing short term early intervention services including screening, referral and support groups; and (e) providing teacher in-service training and support on strategies relevant to substance misuse prevention. During the first 2 years of the project, PWs were expected to engage in activities defined in terms of primary prevention (30%), secondary prevention (25%), and community development (45%). The original contract did not specify that the SBPM was a required element of the program but this requirement was made explicit to PWs and their agency and school supervisors at the training sessions. The revised contract service schedule (implemented at the end of Year 2 of the program) continued to be framed within a comprehensive school health model but explicitly required the use of the SBPM to plan programs to meet three stated goals: preventing the onset of alcohol and other drug use, encouraging students who use alcohol or other drugs to decrease or cease use, and reducing the consequences associated with use. The service requirements were now specified in the language of each phase of the model, and the contract required the submission of a program proposal developed according to each stage of the SBPM. There was no time allocation specified for any single component of the model but 75% of the PW's time was to be spent in prevention, and 25% in screening and referral. The community development requirement was subsumed in the early stage of the revised SBPM, and in the 75% prevention component. Thus, unlike the original contract, the revised contract made the requirement for using the SBPM explicit. PWs did not have input into the revisions to the contract. The major documentation provided to the first group of PWs to guide their work in schools included Schedule A of the contract between ADS and the agency/school district, and a binder providing an overview of the Precede-Proceed model. The second wave of PWs received Schedule A of the contract, the first draft of the School-Based Prevention Model Handbook (Wharf Higgins & MacDonald, 1993), a statement outlining the Prevention Worker role, a statement of policy guidelines concerning reporting, confidentiality, and classroom presentations, and a statement outlining guidelines for standards of ethical conduct. Copies of these last three documents can be found in Appendix E. The first wave of PWs also received these 27 documents at the same time as the second wave. The final version of the SBPM handbook was sent out to all PWs at the end of the second year of the pilot phase. This handbook (Alcohol and Drug Programs, 1994) outlined the "final" components of the SBPP, which included the following: 1. Placement of a full or part-time13 PW in a secondary school. The PW was responsible for facilitating a collaborative planning process in which members of the school community defined their major health concerns, specifically related to substance misuse, and determined a course of action. The key tasks of the PW, as defined in the SBPM handbook, included: (a) building consensus on the nature of the underlying problems, (b) building consensus on possible solutions, (c) building a healthier community with the assistance of as many stakeholders as possible, and (d) displacing simplistic notions of behaviour change, and the causes of lifestyle problems. The handbook states that the SBPM was considered by ADS to be the mechanism for accomplishing these tasks. 2. The use of a defined and systematic planning process known as the SBPM, based on the Precede-Proceed model for Health Promotion Planning (Green & Kreuter, 1991). The model begins with the ultimate goal desired by the school community, and works backward to determine what must "precede" it. It includes gathering relevant data that are expected to help the community make sound programming decisions14. A description of the model and its stages can be found in Appendix B. 3. The formation of a steering committee comprised of representatives from the school community including students, teachers, parents, administrators, and sometimes members of related or concerned community agencies. The steering committee was viewed by ADS as the central mechanism for participation of the school community in decision making in the SBPP, and was expected to result from a community networking process engaged in by the PW early in the process. 1 3 This was a local decision. In many instances, school size determined whether a PW would be hired full- or part-time, but in other instances, school districts and community agencies sought to maximize the PW resource by utilizing the service in more than one school. 1 4 Earlier in this chapter, a Ministry official was quoted as saying that the model had been introduced to help the programs focus and remain accountable to definable prevention goals. This purpose might not be the same thing as using the model to make sound programming decisions, which was the purpose of the model as stated in the SBPM handbook. 28 4. Staff development activities by ADS intended to provide PWs, agency personnel, and regional ADS staff with an understanding of the purpose and the process of the SBPM, and skill development related to working in schools with young people. Staff development included an annual training conference held in Vancouver for PWs, school, and agency personnel, and workshops, in the regions for ADS staff. In the third year of the project, ADS meetings and workshops were held in some regions for all workers in the ADS continuum of care,15 which included PWs. Funds were also provided in the contract for PW staff development. Many PWs availed themselves of a variety of staff development opportunities depending on their personal continuing education needs. In summary, as the above description of the SBPP indicates, there are a number of features of the program that reflect an attempt to put health promotion principles into action in the school setting. Several features of the program represent real innovations for schools. In many ways, the ideas and principles underlying the SBPP reflect current and emerging thinking in health education and health promotion. For example the intent of the SBPP was that the PW would facilitate a collaborative process by which the school would determine the direction of their own prevention program and be involved in its implementation. The PW's role was to help build the capacity of the school to take on its own health issues and concerns. The SBPP does not involve a defined or specified intervention or program. Rather, it involves the implementation of a systematic planning process that is expected to result in the development of strategies that suit the unique needs and circumstances of every school. No two programs will look alike. In the next section of this chapter, current minking in substance misuse prevention is reviewed. This helps to place the development of the SBPP into a context that helps us to understand the elements of the SBPP in relation to emerging theoretical perspectives in substance misuse prevention. 1 5 The ADS "continuum of care" refers to the full range of alcohol and drug services from primary prevention through early intervention, and brief therapy to more intensive, longer term treatment. 29 Current Thinking in Substance Misuse Prevention Historically, most interventions aimed at preventing alcohol and drug misuse have involved school-based classroom curricula, implemented by teachers or health care providers. A variety of theoretical frameworks have undergirded these approaches. Overall, however, the results of these studies have been disappointing. In recent years, a number of reviews have concluded that, to be successful, prevention programs must go beyond the classroom to make changes in the larger school environment (Parcel, Simons-Morton, & Kolbe, 1989; Gerstein & Green, 1993). Community-Specific Approaches to Prevention The vast differences between communities in terms of population make-up (e.g., ethnicity and culture), socio-political environment, and the scope and extent of drug abuse problems, suggest that a community-specific approach may be most relevant. There is a large body of literature on community-specific approaches to a wide variety of health and social issues. Until recently, however, there have been few published studies of community-specific drug and alcohol prevention interventions. Therefore the relevance of community-specific approaches to drug use prevention has not been established (Gerstein & Green, 1993). There is considerable variability in what constitutes a community-specific approach. Many approaches are referred to as "community-based" although they differ significantly in how "community" is defined, on the theoretical and philosophical underpinnings, and on the key features of the intervention. Green and Kreuter (1991) identify two types of community-specific interventions: community-wide interventions and interventions-in-community. Community-wide interventions Community-wide interventions attempt to obtain small changes at the individual level, but among a large proportion of the population thereby generating a significant population impact. These interventions are based on the assumption that risk factors for a particular condition or behaviour are normally distributed in populations. Examples of this approach include the large-scale cardiovascular and cancer risk-reduction trials such as the North Karelia project in Finland (Puska et al., 1983), the Stanford Three- and Five-City 30 projects (Farquhar et al., 1982; Flora, Maccoby, & Farquhar, 1989), the Minnesota Heart Health Project (Lando et al., 1995; Luepker et al., 1994), the Pawtucket Hearth Health Program (Carleton, Lasater, Assaf, Lefebvre, & McKinlay, 1987), the Community Intervention Trial for Smoking Cessation (COMMIT) (Lichtenstein, Wallack, Pechacek, 1991; The COMMIT Research Group, 1995a & b) and the CART project in Australia (Hancock et al., 1997). Interventions in community The second category of community-specific programs are "interventions-in-community." These tend to operate from a specific site or institution within the community, such as a school, a work site, a clinic, or a hospital. In contrast to community-wide interventions that seek small changes across an entire population, these approaches seek more intensive or extensive moderate change in a sub-population. Most of the drug prevention research has been in school settings, and thus would be classified in this category. As discussed in the introduction, however, there is an emerging consensus that drug use prevention is more likely to be successful when implemented within the context of comprehensive school health programs linked with community programs. This could be said of the institutionally-based programs in other settings as well. The types of interventions characterized as interventions-in-community vary widely. Rothman and Tropman (1987) distinguished three models of community organization: locality development, social planning and social action. Locality development and social planning are what others have characterized as community development and community-based programming (Chavis & Florin, 1990; Labonte, 1994a; Minkler, 1990). Community-Based Programming In community-based planning, health agencies or professionals define the problem, develop strategies to remedy the problem, and involve local community members or groups in solving the problem. Ultimately, the intent is to transfer responsibility for the on-going program to local community members and groups. Generally, there are clearly defined program timelines, and decision making power is in the hands of the institution or agency although community members may be involved in advisory groups to the decision ' 31 makers (Labonte, 1994a). Overall, these tend to be more "top down" initiatives directed by experts/professionals. Community Development Programming By contrast, community development programming is "the process of supporting community groups in their identification of important concerns and issues, and in their ability to plan and implement strategies to mitigate their concerns and resolve their issues" (Labonte, 1994a, p. 37). Thus, the problem is locally defined, the time-lines may not be set, and power relations are continually negotiated within the process. Ultimately, the intent is that the group will have improved its capacity to solve its own problems effectively. This approach is more "bottom up," directed by community members. Thus, it appears that the first classification scheme distinguishes community-specific approaches in terms of the size and scope of the program, and the number of agencies and levels of organization involved. The second scheme has more to do with the nature of the relationships and transactions between health professionals and community members, and the types of processes involved, although it is clear that each type in the second scheme could easily fit within each type of the first. In other words, community-based and community development programming could characterize both community-wide interventions and interventions-in-community. The difficulty with these classifications is that, in practice, the boundaries between them are somewhat blurred. Characteristics of each can be found in any one project. For example, the SBPP is situated in a school but it operates in partnership with at least one other community agency. Many of the PWs have established collaborative relationships with other community agencies around the issue of drug use prevention in particular and adolescent health promotion more generally. In some schools, there is a steering committee in which members of the community participate. Some PWs also sit on advisory committees for other community agencies. Thus, although the program is clearly not a community-wide intervention, it is intended to be more than just a school-based program. In the SBPP, the problem of alcohol and other drug use by adolescents has been named by the sponsoring agency. But the approach that is advocated, at least in 32 its official policy documents and conceptualization, has some features in common with the community development model outlined above. Principles of Community Health Promotion Practice In relation to community health promotion practice, Gerstein and Green (1993) have identified a set of fundamental propositions that have come to be widely accepted by a variety of practitioners as basic principles of practice in community settings. These are: (a) begin from a base of community ownership of the problems, and the solutions; (b) plan thoroughly using relevant theory, data, and local experience as bases for program decisions; (c) know what types of interventions are most acceptable and feasible to implement for specific populations and circumstances; (d) have an organizational and advocacy plan to orchestrate multiple intervention strategies into a complementary, cohesive program; and (e) obtain feedback and evaluation of progress as the program proceeds (Gerstein & Green, 1993, p. 119). They question, however, whether these broad generalizations apply as well to drug abuse prevention as to the other fields of research and practice from which they emerged (e.g., education, public health, community psychology) since they have not been widely applied in a systematic way in this field. The intent of the SBPP is to apply these principles in the school setting, guided by a defined planning process based on the Precede-Proceed model for health promotion planning. The Social Ecological Perspective of Health Promotion The emergence of the community-specific approach to drug use prevention has parallelled (and been influenced by) developments in health promotion. A central feature of the post-Ottawa Charter (World Health Organization, 1986) health promotion is its ecological perspective which sees health as "the product of the continuous interaction and interdependence of the individual within his or her ecosphere: that is, the family, community, culture, societal structure, and physical environment" (Green & Raeburn, 1988, p. 154). Stokols (1992) describes 'ecology', which had its earliest roots in biology, as referring broadly to the interrelations between organisms and their environments. It has evolved in several disciplines into a general framework for understanding the nature of people's transactions with their physical and social environments. 33 Social ecology is concerned with social, institutional, and cultural contexts of people-environment relations. Stokols outlines several assumptions underlying this perspective: (a) the healtirfulness of a situation and the wellbeing of its participants are influenced by multiple facets of both the physical and the social environment, (b) analyses of health and health promotion should address the multidimensional and complex nature of human environments, (c) the socioecological perspective incorporates multiple levels of analysis and diverse methodologies for assessing the healthfulness of settings and wellbeing of individuals and groups, (d) the effectiveness of health promotion programs can be enhanced significantly through the coordination of individuals and groups acting at different levels, and e) people-environment interactions are characterized by cycles of mutual influence. . The ecological approach is transactional in nature, with reciprocal effects between the individual/family/group/community and the environment. To exploit this theoretical understanding fully, actions based on this approach are directed at influences in each of the levels, and consideration is given to outcomes at levels other than the individual (Green, Richard, & Potvin, 1996; Kickbush, 1989; McLeroy, Bibeau, Steckler & Glanz, 1989; Richard, Potvin, Kishchuk, Prlic, & Green, 1995; Stokols, 1992). Perry, Kelder, and Komro (1993) have described an ecological framework for examining the health. of adolescents. They use Bronfenbrenner's (1979) nested spheres of micro-, meso-, exo-, and macro-systems. The rationale for using such a framework is that "...explanations of behaviour and developmental patterns cannot be sought at only one level, but need to include several layers of context" (Perry, Kelder, & Komro, 1993, p. 75). Thus, the other spheres are assessed as a source of'explanations' for behaviour but not necessarily targeted for change. Behaviour remains precisely at the heart of the model as the primary focus and concern. This appears to be a unidirectional model. There is no analysis of the influence of adolescents' behaviour in affecting their own social or environmental context or, in health promotion terms, of taking action on their environments to gain control over the factors influencing their health or lives. The absence of this analysis is surprising, given the notion of reciprocal determinism that lies at the foundation of Social Learning Theory (Bandura, 1977), which undergirds virtually all of Perry's work. 34 In contrast, Leventhal and Keeshan (1993) have a different understanding of the concept of interdependence and mutual influence in relation to adolescent health generally, and substance use in particular. They identify that there is a great need in drug prevention research "to improve the conceptualization of the individual in the social context" (Leventhal & Keeshan, 1993, p. 265). For example, they argue that treating peer influences as solely an environmental factor ignores the adolescent's representation of the relationship between peer and self. Adolescents are not passive recipients of social influences from the external context, but actively construct and regulate the world around them in relation to the meanings these influences hold for them.16 Thus, "young people, in action with their context, create the basis for their own development" (Leventhal & Keeshan, 1993, p. 268). Green and colleagues (1996) identify some limitations of the ecological approach. They argue that health promotion is drawn to ecology because it enlarges the focus from behaviour17 to the environment. They contend, however, that we are forced to "retreat to behaviour at some level" because "we do not manage ecosystems, we manage our interactions with them" (Green et al., 1996, p. 273). What they mean by a retreat to behavior is that we cannot abandon a concern with human behaviour in a truly ecological perspective. The term "retreat to behavior" is also meant to convey the understanding that the "management of interaction" is, in fact, human behaviour (personal communication, L.W. Green, October 15, 1996). This notion of managing interactions with the environment was exactly the point I was making above. Adolescents actively manage their transactions with their environments rather than responding solely to external influences. Thus, 1 6 Although the authors do not specifically refer to symbolic interactionism, the notion that adolescents actively construct the world around them in relation to the meanings they make of that world, is a key precept of symbolic interactionism (Blumer, 1969). This theoretical perspective will be discussed in more detail in Chapter Four. 1 7 1 have used both the British and American spellings in this document. I use the British spelling as a matter of course; the American spelling is used in direct quotations where that particular spelling was used in the original. 35 the centrality of human action is acknowledged but there is also a recognition of the reciprocal and interdependent relationship between human beings and their social/physical worlds.18 Implementation and Evaluation Issues in Community-Specific Approaches In part, the limited evidence for the effectiveness of community-specific approaches to drug prevention may be due to the inherent difficulties in evaluating community interventions that are based on ecological principles. The move to recognize the importance of community context in prevention programs has added enormous complexity to the issue. It is difficult to evaluate interventions that consider the multiple levels of influence across multiple settings and to track the iterative and ongoing process of reciprocal influence between human action and environmental conditions. The applicability of traditional experimental and quasi-experimental approaches to evaluating such interventions increasingly is being questioned (Potvin, 1995; Green, Richard, & Potvin, 1996). The complexity of these interventions also raises questions about what is actually being implemented, and how that implementation is taking place. The challenges of implementation have been demonstrated in situations in which the intervention is well defined and clearly circumscribed. Implementation is surely a greater challenge when the intervention is more complex. In recent community-based substance misuse prevention interventions, in which there was a standardized and controlled intervention, the researchers reported a number of implementation challenges. One of the challenges was: ....to maintain a structure for the study that encourages collaboration between scientist and citizen, that is sufficiently rigorous for the scientific enterprise and yet flexible enough to incorporate ideas from the community on how to proceed, and thereby to embrace the idiosyncrasies and local community problems that are expected to occur when the community is the 'subject' of research (Perry et al., 1993, p. 135). An international symposium on Experiences with Community Action Projects for the Prevention of Alcohol and other Drug Problems (Greenfield & Zimmerman, 1993) focussed much of their discussion on 1 8 The concept of mutual interdependence is central to both social ecology and symbolic interactionism. As discussed in Chapter Four, this observation has implications for a choice of methodology in a study of health v promotion. 36 the issues associated with melding the differing agendas of researchers and community members around issues of alcohol and drug prevention. The issue raised above is only one of many that researchers and communities are grappling with in the wake of sustained interest in community approaches to prevention. The same question has been raised by the various traditions of participatory research (Green et al., 1995). Research Dilemmas Several dilemmas were identified throughout that symposium in relation to the issue of research versus community-driven prevention programming. Wagenaar and Wolfson (1993) identify six tensions between scientific research and community action. Three of these are particularly relevant to this discussion. The first is the issue raised above about the tradeoffs between scientific rigour and community relevance. The value-free assumptions of the traditional scientific method often conflict with the explicit values inherent in community action. The value-free stance has been widely called into question and few researchers today would argue that the scientific method is itself value-free. Wagenaar and Wolfson argue that the conflict is not between value-free objectivity (since science cannot be value-free) and community advocacy, but between competing values. The second dilemma has to do with the power of the research design versus the power of the intervention. A powerful research design, in the traditional scientific sense, is likely to minimize the impact of the intervention because of the need for standardized protocols and researcher control. On the other hand, a powerful intervention may require extensive community control, and therefore is antithetical to the scientific method. Thus, to the extent that research values take precedence, the success of the intervention may be jeopardized. Some researchers address this dilemma by attempting to strike a balance between research and community needs; that is, attempting to maximize community participation while maintaining scientific rigour. The third dilemma is related to the first two; that is, the issue of community empowerment versus researcher control. Community organization theory suggests that problems and solutions that are community-defined and owned are likely to be more effective than those imposed from outside (Minkler, ' 37 1990; Rothman & Tropman, 1987). The dilemma here is that generally, local autonomy leads to treatment heterogeneity. The scientific method, of course, relies on stringent researcher control over the intervention (i.e., the 'independent variable') and desires, if not requires, standardized interventions across sites. These dilemmas have also been described in relation to health education research and reflect a long-standing tension in the evaluation of health education interventions (Green, 1977). Resolving the Dilemmas Fisher (1995) comments on this difficulty posed by the dilemma noted above, and proposes an alternative approach that may help researchers and practitioners to get around this dilemma. He suggests that it is important to distinguish between "defined, community-based" interventions with standardized protocols and "community organization" approaches that are truly community-driven, and that place greater emphasis on intended audiences' active participation in program development. Instead the intervention might focus, not on specifying the intervention components and governance structures, but on defining the process by which an intervention is developed. In the SBPP, a key component of the intervention is the use of a defined planning process, the SBPM. Process as intervention. Unfortunately, few studies have been published in which a "process" such as the SBPM is the intervention. For example, the Precede-Proceed model, on which the SBPM is based, has been applied in over 700 published applications.19 Although I have not reviewed all of these, a search of close to 400 of these has not turned up an instance in which the model itself is the intervention. In most applications, the model is used to develop a defined intervention which is then tested and evaluated. Some researchers have used the model as a framework for assessing the comprehensiveness of the program, as a conceptual framework for deriving variables to study in relation to a specific research question, or to define variables for evaluation of a program. 1 9 A bibliography of this literature can be obtained through the internet by browsing the web site of the Institute of Health Promotion Research, University of British Columbia at http://www. ihpr.ubc.ca 38 A recent project in Minnesota (The Communities Mobilizing for Change on Alcohol - C M C A ) is testing a similar proposition. That is, it is evaluating a defined approach to community organization for alcohol use prevention, not a defined intervention (Wagenaar, Murray, Wolfson, Forster, & Finnegan, 1994). C M C A is an 18-community trial that targets community-level influences on alcohol use. This program differs from those previously cited in that it does not involve a standardized program developed by the research team. Rather, the communities themselves develop and devise the strategies they feel are most appropriate to meet community-defined needs. Researchers serve as consultants only, providing information and materials as required related to such issues as the nature and extent of alcohol use among youth, and the effectiveness of various strategies. The project aims to test a theory-based process of community organization, and "the objective is to change community policies and practices, not simply the behaviour of an aggregate of individuals in a community" (Wagenaar et al, 1994, p. 80). The intervention model utilizes a community organization process to influence formal community policies, and the practices of law enforcement agencies, parents, schools, alcohol merchants, youth, and other major players in the community. The specific policies and practices targeted are those influencing youth access to alcohol, including institutional policies, local ordinances, and enforcement of existing laws. Youth access is defined as the relative ease with which youth can acquire alcohol, and is assumed to be strongly related to the actual level of consumption in the community. The ultimate goal is the reduction of youth alcohol problems. Unfortunately, the results are not yet available. Alternative means of knowledge construction. Another response to this dilemma is to examine alternate means of knowledge construction that are not based on the scientific method, and its associated definition of research rigour. Virtually all of the substance use/abuse evaluations cited above have involved quantitative designs based on survey data collection of self-reported, individual-level behavioural outcomes. Many questions have been raised about the appropriateness of using experimental and quasi-experimental designs in these types of community interventions. The complexity of the programs now being advocated make it very difficult to implement these more traditional designs, and to model program processes and 39 outcomes. Even the most intransigent of positivists are beginning to acknowledge that qualitative designs, serve an important function in the evaluation of the new generation of community interventions for alcohol, tobacco, and other drug use prevention, particularly in terms of describing and theorizing implementation processes. A new development in the generation of knowledge about effective substance abuse prevention may be the findings beginning to emerge from a series of community partnership grants programs. These programs were funded and implemented by such organizations as the Kaiser Family Foundation (Henry J. Kaiser Family Foundation, 1989), the Robert Wood Johnson Foundation (Green & Kreuter, 1991), and the Centre for Substance Abuse Prevention (CSAP) 2 0 of the U.S. Department of Health and Human Services (Kaftarian & Hansen, 1994). Because there are similarities among these community grants programs, I will use the CSAP partnership project as an example to illustrate my point about the importance of new approaches to knowledge development that do not rely on traditional quantitative methods. The CSAP partnership project is a demonstration project that has a number of features in common with the SBPP. First, the CSAP project emphasizes the importance of local control over program design, delivery and evaluation. In this regard, it is similar to the SBPP. Second, in the CSAP project, funding is channelled directly to community agencies in the expectation that coalitions will be formed to address the issue of alcohol and other drug use. In the SBPP the partnership is not a community coalition as such, but a relationship between at least the school/school district and a community alcohol and drug agency. The local steering committee in the SBPP is the central mechanism for partnership while the coalition is the central mechanism in the CSAP project. The SBPP specifically addresses alcohol and other drug use prevention among adolescents rather than drug use more generally among the community at large, as in the CSAP projects. 2 0 Formerly, Office of Substance Abuse Prevention (OSAP). The name change to CSAP followed a federal reorganization of the parent organization - Alcohol, Drug Abuse, and Mental Health Administration (ADAMHA) which became the Substance Abuse and Mental Health Services Administration (SAMHSA) on October 1, 1992. 40 Third, in both CSAP projects and SBPP, the sponsors were not disseminating a defined program. Each community is expected to engage in their own planning process to develop a credible approach to alcohol and other drug use that is grounded in the local context, and includes a local understanding of the issues and appropriate solutions. Local control over the planning process, the solutions implemented, and the evaluation is considered to be a critical element in both the SBPP and the CSAP. This could be said of most community grant programs of federal, state, and provincial agencies, and of foundations (e.g., BC Health Research Foundation). Finally, community level participation is a key feature of both innovations. What this actually means at the local level is likely to vary across communities. The intent, however, is that local participation will engender community concern and ownership of the issue and the solutions. It will also develop the capacity of the community to deal with these and similar issues in the future. As a demonstration project, the CSAP projects are concerned primarily with "policy learning" (Springer & Phillips, 1994). That is, grants are provided to communities to encourage local partnerships to develop new and innovative approaches to substance abuse prevention that will provide important lessons for future programming and policy development. For these purposes, it is as important to understand the local context and its interaction with the local program as it is to understand the program's impact. Policy learning, in this sense, is also an important focus in the SBPP evaluation. To date, limited outcome data have been published from any of the CSAP partnership programs implemented in 251 communities across the U.S. There have been, however, numerous publications that report on formative and process evaluation data with an important emphasis on describing the implementation issues that emerge in these complex and comprehensive community interventions. There are several other reasons why traditional evaluation designs in substance abuse prevention are a poor fit with the types of tasks faced in evaluating programs such as the Community Partnership Programs, or the SBPP for that matter. First, there is an increasing recognition that the possibility of identifying and using appropriate control communities is rapidly becoming an inviable option in evaluation. This concern 41 was echoed throughout a recent CSAP monograph (Greenfield & Zimmerman, 1993). This was an important issue in the SBPP evaluation, in which an inadequate number of control schools were recruited, and those that did agree to participate were so unique that any comparisons between them and the experimental schools are suspect (School-Based Prevention Project Evaluation Team, 1995c). Second, classic experimental and quasi-experimental designs aim primarily to reduce Type I errors, that is, to avoid reporting an effect that does not exist. Hansen and Kaftarian (1994) argue that Type II errors (i.e., failing to detect effects that do occur) are potentially more limiting to an emerging field than Type I errors because a field cannot progress without hypotheses. Experimental and quasi-experimental designs are only appropriate when a body of knowledge already exists about a topic. In new or emerging fields, there is often insufficient theory development to allow the deduction of sophisticated hypotheses that can later be tested. The testing of ill-formed hypotheses can be wasteful and an inefficient trial-and-error process. In the case of the CSAP partnerships, there is little information currently available about what types of partnerships exist or the contexts in which they function effectively. Similarly, for the SBPP, there is little information on drug use prevention at the high school level that does not centre around classroom curricula as the primary strategy. We do not know what types of SBPP programs are being implemented successfully in schools nor do we know anything about the school contexts in which they function effectively. Most importantly, we do not know whether and how the intended planning process is actually implemented in the schools. Thus, generating this understanding through exploratory research may be a more important task for evaluation, at this stage of SBPP development, than demonstrating effectiveness. Thus, it seems that a focus on the implementation process in the SBPP represents an important area for exploration and theoretical development that will make an important contribution to knowledge in the field of drug use prevention as well as in the field of health promotion. The next chapter reviews the rationale for studying implementation, defines it, and provides an overview of various theoretical perspectives on implementation. The specific implementation issues for this study are then discussed. 42 CHAPTER THREE IMPLEMENTATION "The process unwinds in its own terms, mocking standard frameworks, and challenging the researcher to make a coherent summary of the welter of observed and reported events" (Huberman &Miles, 1984, p. 1). Much of the literature on implementation does not define the term, and assumes the meaning to be self-evident (Hasenfeld & Brock, 1991; Monahan & Scheirer, 1988). On the other hand, multiple definitions of implementation have been offered by various theorists, and although there are similarities across definitions, each reflects a particular theoretical, disciplinary, or philosophical perspective on implementation. This chapter therefore addresses two questions in relation to implementation: What is implementation, and what are the implementation issues for this study? To answer the first question, the literature on various theoretical perspectives on implementation is summarized, followed by a discussion of two major issues identified in the implementation literature: the macroimplementation versus microimplementation debate, and the fidelity versus reinvention debate. The chapter closes with a discussion on the implementation issues for this study. What is Implementation? Implementation is usually studied as a stage or component of a larger change process. The term implementation implies that "something" is to be implemented; that is, carried out, put into place, or acted upon. It further implies that this "something" is new and different. Thus, implementation involves a change from the current situation or practice. In general, implementation refers to a process by which intentions are translated into action. These intentions are reflected in various instruments including policies, plans, technologies, programs, and innovations. It seems likely that the implementation process will be affected by the nature of the "thing" being implemented (Ottoson & Green, 1987). Implementing a broad policy may be quite different than implementing a clearly specified program. How implementation is defined will also be influenced by the disciplinary and/or theoretical lens one adopts in viewing implementation. The following discussion identifies and describes five theoretical lenses on implementation. 43 Theoretical Perspectives on Implementation Much has been written about different theoretical perspectives on implementation, although each author seems to use a different label when talking about essentially the same thing. The social policy literature described two general perspectives on implementation: the "top down" (Sabatier & Mazmanian, 1979; Van Meter & Van Horn, 1975) and the "bottom up" approach (Lipsky, 1970; Moore, 1987; Weatherley & Lipsky, 1977). These perspectives have also been termed "forward mapping" and "backward mapping" (Elmore, 1985) or "implementation as control" and "implementation as interaction" models (Majone & Wildavsky, 1979). They have also been closely associated with macro- and micro-implementation. A third perspective is the iterative or evolutionary view of implementation (Berman & McLaughlin, 1975; Majone & Wildavsky, 1979; Ottoson & Green, 1987; Palumbo & Oliverio, 1989; Van Meter & Van Horn, 1975). Also in the field of social policy, Hasenfeld and Brock (1991) describe five theoretical perspectives that explain implementation; the "pursuit of rationality," the "organization-policy environment fit," the "bureaucratic discretion and adaptation," the "power relations," and the "leadership skills" models. The organizational development literature describes four general perspectives; "implementation as systems management," "implementation as bureaucratic process," "implementation as organizational development," and "implementation as conflict and bargaining" (Elmore, 1978). In the education field, House (1981) suggested that three major theoretical perspectives have been used to analyse and interpret the implementation of change in education systems and schools; the "classical or technological model", the "political" model and the "cultural" model. This classification has been used widely by educational researchers to guide research and analyse the process of implementation (Larson, 1992; Musella, 1989; Rossman, Corbett, & Firestone, 1988; Williams & Smith, 1993). Although a "bureaucratic adaptation" model was not included in House's classification, it has, in fact, been used to explain implementation in educational studies (e.g., Lipsky, 1980; Radin, 1977; Weatherley & Lipsky, 1977). 44 In the area of public administration, Yanow (1987, 1990) identifies five conceptual lenses for analyzing and interpreting implementation; the "human relations," the "political," the "structural," the "systems," and the "cultural" lenses. These overlap considerably with the models already identified. In contrast to other reviews of the various perspectives, however, Yanow explicitly identifies different levels of analysis as the focus of implementation in each of these perspectives. In health education, the research and development model (Basch, 1984; Williams & Smith, 1993) has guided much of the implementation research in that field. It has been closely tied to diffusion of innovation theory (Lorig, 1986; Goodman, Tenney, Smith, & Steckler, 1992; McCormick, Steckler, & McLeroy, 1995; Orlandi, Landers, Weston, & Haley, 1990; Parcel et al., 1989; Rogers, 1983; Schinke & Orlandi, 1991;). Classical diffusion theory is closely related to what others have called the technological model; however, Basch (1984) identifies four different diffusion perspectives that have guided implementation research in this area.. These are the "research and development" model (R&D), the "social diffusion" model, the "innovative organization" model, and the "organizational development" model. A careful review of all of the above perspectives reveals that there is considerable overlap in these, with five more or less distinct perspectives emerging. Table 1 identifies the five major perspectives found in the literature (technological, bureaucratic adaptation, organizational development, political, and cultural) and lists their various synonyms with supporting references. 45 Table 1 Theoretical Perspectives on Implementation: Synonyms and Sources Synonyms Sources TECHNOLOGICAL Top Down Van Meter & Van Horn, 1975; Sabatier and Mazmanian, 1979; Palumbo & Callista, 1987; Palumbo & Oliverio, 1989 Implementation A s Control Majone & Wildavsky, 1979 Pursuit of Rationality Hasenfeld & Brock, 1991 Systems Management Elmore, 1978 Foreward Mapping Elmore, 1985 Classical Firestone & Corbett, 1988; Larsen, 1992; Musel la, 1989; House, 1981 Research and Development Clark & Guba, 1967; Basch, 1984; Rogers, 1983 Diffusion of Innovations Rogers, 1983; Basch, 1984; Orlandi et a l . , 1990 BUREAUCRATIC ADAPTATION Bottom Up Palumbo & Oliverio, 1989; Palumbo & Callista, 1990 Backward Mapping Elmore, 1985 Street-Level Bureaucracy Wea