HOSPITAL SOCIAL WORKERS' PERCEPTIONS OF ETHICAL ENVIRONMENT: RESULTS FROM A NATIONAL SAMPLE By Gregory Lloyd Pugh A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Social Work 2011 i ABSTRACT HOSPITAL SOCIAL WORKERS' PERCEPTIONS OF ETHICAL ENVIRONMENT: RESULTS FROM A NATIONAL SAMPLE By Gregory L. Pugh Hospital social workers are in the unique position of being one of the only non-medical professions working in the hospital setting. This raises some interesting challenges in hospital social work practice, especially when it comes to professional ethics. The medical model of ethical reasoning is different from social work models, and the medical model contributes a great deal to the overall Ethical Environment in a hospital. Ethical Environment is a new concept for hospitals to consider and has only recently been explored, almost exclusively with samples of nurses. Ethical Environment includes the hospital practices and resources directed at ethical issues and is thought to affect ethical practices and behaviors. The research suggests that a number of personal, professional, and organizational factors may influence perceptions of Ethical Environment. The hospital social work and ethics literature also suggests relationships between some of the same factors and the ethical practices and behaviors of social workers, but has not considered Ethical Environment. Both the nursing and social work literature suffer from a lack of large, representative samples. This study was designed to explore the relationships between personal, professional, and organizational variables and perceptions of hospital Ethical Environment among social workers in a large, nationwide, representative sample. The study succeeded in collecting a sample of 973 social workers from 290 hospitals in a random closed population cluster sample across 40 States. The participants completed an online survey questionnaire about the Ethical Environment of ii their hospitals, and resulted in hospital social workers rating the Ethical Environment significantly higher than nurses. Significant predictors of the rating of Ethical Environment are based primarily on job satisfaction, with some contributions from years of service, attendance at hospital-based ethics education programs, and a centralized social work department configuration. Social workers also rate the environment of for-profit hospitals greater than nonprofit hospitals. Results indicate that larger hospitals with greater resources directed at ethics provide the best Ethical Environment for social workers. Professional social work ethics education and training was not predictive, which raises questions about the content and methods of teaching social work ethics. iii Copyright by GREGORY LLOYD PUGH 2011 iv ACKNOWLEDGEMENTS This dissertation is the product of a long process involving a great deal of work, none of which would have been possible without the kind and generous support of numerous amazing people, and the Michigan State University School of Social work. My dissertation committee chair, Dr. Rena Harold, was incredibly patient, supportive, and challenging in all the right ways, and at all the right times. Committee members Dr. Gary Anderson, JoAnn McFall, and Dr. Thomas Conner from Sociology all contributed support, encouragement, and expertise. As a committee, they all made our meetings fun and engaging. I owe debts of gratitude to a number of doctoral student colleagues in the school of social work, who shared and listened as we all struggled through our own process. Especially helpful and fun were Karen Newman, Jessica Gladden, and Louanne Bakk. Several faculty and staff were also instrumental in my development as both a researcher and teacher, among them Dr. Mary Barron, Linda Campbell, Angelique Day, Julie Farman, Dr. Victoria Fitton, Dr. Sheryl Kubiak, Dr. Marcia Lampen, Dr. John Seita, Dr. Marya Sosulski, and Dr. Amanda Woodward. A special mention for those who are Graduate Office team members, including the incomparable Nancy Gray, Julie Navarre, and Dr. Sally Rypkema. My friends and family were always supportive, encouraging, and patient with me. I am especially grateful to my Mom and Dad, Nancy Lewis, Damian Pisanelli, and Tim Young, all of whom listened to me blather on about my research and dissertation as if it was important to anyone other than me. I have had the pleasure of being part of the MSU School of Social Work, one of the finest academic units I have ever known, with some of the most impressive students I have ever had the pleasure to teach. My students have taught me much more about teaching than they realize. v TABLE OF CONENTS LIST OF TABLES viii LIST OF FIGURES ix CHAPTER I: INTRODUCTION Research Questions Contributions to the Field 1 5 7 CHAPTER II: LITERATURE REVIEW A. Organizational Culture, Climate, and Ethical Environment The Relationship between Culture and Climate Ethical Culture, Climate, and Environment Ethical Environment Summary and Theoretical Model B. Ethical Development and Reasoning in Medicine and Social Work General Moral Theory Ethical Theory Ethical Dilemmas The Process of Ethical Reasoning Reasoning in Medical Ethics The Medical Model of Ethical Reasoning: Principlism Reasoning in Social Work Ethics Sequential Models The NASW Code of Ethics as Ethical Reasoning Ethical Principles in Social Work C. The Context and Practice of Hospital Social Work The History of Social Work in the Hospital Setting Practice in a Host Setting The Hospital Ethics Committee Discharge Planning and Psychosocial Care The Changing Nature of Practice Ethics in Discharge Planning Social Work Involvement in Hospital Ethics D. Research on Social Work Ethics in the Hospital Setting The Need for Social Work Ethics Research Barriers to Social Work Ethics Research in Hospitals Social Work Ethics Research in Hospitals Social Work Ethics Research Literature Reviews Research Gaps in Hospital Social Work Ethics Summary and Theoretical Model 9 12 14 15 19 20 21 23 24 25 25 26 28 29 29 30 32 32 34 35 37 38 40 41 42 42 43 44 44 54 56 CHAPTER III: METHODOLOGY Sampling Frame Design and Procedures 61 61 63 vi Measures Dependent Variable and Instrument: Ethical Environment Independent Variables Process and Data Collection Data Analysis 68 68 70 73 75 CHAPTER IV. RESULTS Hospital Sample and Response Rates Social Worker Sample and Response Rates Sample Characteristics Dependent Variable: Ethical Environment Independent Variables Relationships with Ethical Environment Regression Model Testing Summary and Results Model 79 79 82 83 85 86 89 94 97 CHAPTER V. DISCUSSION Discussion Personal Variables Professional Variables Organizational Variables Strengths and Limitations Implications Conclusion 99 100 103 105 107 109 115 APPENDICIES A. Evite Text and Reminder Texts B. Online Consent Form Letter C. Survey Questionnaire 118 120 121 REFERENCES 126 vii LIST OF TABLES Table 1. Ethical Climate and Environment Studies in US Hospitals 18 Table 2. Theoretical Literature on Hospital Social Work Ethics 45 Table 3. Descriptive Research Literature on Hospital Social Work Ethics 46 Table 4. Hospital Participation by Profit Status Contingency Table 81 Table 5. Hospital Participation by Census Region Contingency Table 82 Table 6. Sample Demographic and Descriptive Elements 83 Table 7. Sample Distributions by Census Region and Census Division 85 Table 8. Distribution of EEQ Scores 85 Table 9. Means, Standard Deviations, and Pearson's Correlations for all Variables 87-88 Table 10. Group Differences in Ethical Environment Questionnaire Scores 91 Table 11. Sequential Regression Model Predictors of Ethical Environment Questionnaire Scores 95 viii LIST OF FIGURES Figure 1. A Theoretical Research Model 6 Figure 2. A Theoretical Model of Hospital Ethical Environment Literature 19 Figure 3. A Theoretical Model of Social Work Hospital Ethics Literature 58 Figure 4. A Combined Theoretical Model 59 Figure 5. Ethics Preparation Frequency Distribution 89 Figure 6. Job Satisfaction Frequency Distribution 89 Figure 7. Religiosity as Measured by Religious Service Attendance 89 Figure 8. Research Results Model 98 ix I. Introduction This dissertation will explore some of the challenges of the hospital environment to social work practice, with a focus on professional ethics. Based on the literature reviews that follow, one potentially influential factor in the ethical practices in hospitals is the concept of the organizational Ethical Environment. The Ethical Environment is essentially the perception among employees, including social workers, as to how the hospital implicitly and explicitly deals with ethical issues. Ethical Environment is strongly influenced by the ethical reasoning of the medical profession, which is different from the ethical approaches of social work. To understand this contrast, the development and current status of ethical reasoning in both social work and medicine is explained. An understanding of the development, context, and current practice of hospital social work follow this, with an emphasis on how this might impact ethical issues, practices, and the Ethical Environment. The social work literature suggests a series of personal, professional, and organizational variables that influence ethical actions. Ethical actions may also be influenced by the hospital Ethical Environment, and the Ethical Environment literature argues that many of the same personal, professional, and organizational variables affect perceptions of Ethical Environment. Ethical Environment has not been adequately investigated among hospital social workers, and this dissertation intends to fill that gap in the existing research. Ethical Environment is a concept that has only been touched upon in the social work literature, but is becoming an important organizational factor to consider given its apparent influence on ethical (and unethical) behavior. There is a body of knowledge in industrial and organizational psychology about the concepts of organizational culture, climate, and environment, and more specifically the concept of Ethical Environment. The Ethical Environment of a hospital communicates to employees what the organizational practices, 1 attitudes, conditions, and resources are pertaining to ethical issues (Hamric & Blackhall, 2007; Lyon & Ivancevich, 1974). The ethical environment also signals to employees how they will be treated when ethical issues arise, and indicates who holds power and authority in ethical decision-making (Hamric & Blackhall, 2007; Hart, 2005). Some authors suggest that the Ethical Environment of hospitals may leave social workers feeling unsafe and unwelcome in participating in ethical discussions and deliberations (Jansson & Dodd, 2002; Walsh-Bowers, Rossiter, & Prilleltensky, 1996). Compliance with the explicit and implicit rules of this environment indicate to social workers the options they have when it comes to involvement in ethical issues (Kugelman, 1992; Landau, 2000a; Walden, Wolock, & Demone, 1990). Ethical Environment and other organizational variables (such as the size of the hospital and profit status) may be barriers and sources of conflict for social workers in hospital ethics (Csikai & Sales, 1998; Jansson & Dodd, 1998, 2000; Landau, 2000b; Walsh-Bowers, et al., 1996). Within the environment literature, instruments to measure the Ethical Environment have been developed (McDaniel, 1997; Olsen, 1998; Victor & Cullen, 1988), and have been applied to hospitals with samples of nurses (Deshpande & Joseph, 2008; Joseph & Deshpande, 1997; Rathert & Flemming, 2008). One of these instruments was developed specifically for hospitals (Olson, 1998), and another was initially validated in hospitals (McDaniel, 1997). All of these have been utilized with samples of nurses (Corley, Minick, Elswick, & Jacobs, 2005; Hart, 2005) with two small inclusions of hospital social workers (O'Donnell, et al., 2008; Ulrich, et al., 2007). There have been no studies exclusively done with hospital social workers, and rarely are the studies based on large representative samples. The similar, and in some cases identical, personal, professional, and organizational factors that are related to both ethical environment and the ethical practice of hospital social work represent an opportunity for interconnected research 2 in this area. Given that little is known about how hospital social workers perceive the ethical environment of their work setting, an exploratory study is indicated. To understand Ethical Environment in the case of hospitals and social workers, it is necessary to explain how medicine and social work have developed their own separate professional ethics to deal with ethical issues in their practice. These ethical approaches contribute to the hospital Ethical Environment as the methods by which ethical issues are addressed. The ethics of both professions are based on the same moral philosophy and ethical theory, and include some of the same ethical principles. Both have developed and modified their own professional ethics and models of ethical reasoning in response to developments in ethical theory and societal changes. Social work and medical professionals may work together in some aspects of hospital care, such as on ethics committees (Csikai, 1997; Guo & Schick, 2003; Hoffmann, Tarizan, & O’Neil, 2007), but overall, the medical approach to ethics is the dominant framework. There is only a small body of research representing the normative applications of social work ethics in the hospital, leaving many unanswered questions. The medical model of ethical reasoning is a major contributing factor to the hospital Ethical Environment. The hospitals within which social workers practice present some interesting contextual issues as well. Hospital-based social workers are in the unique position of being one of the only non-medical professions working in the medical setting (Caputi, 1978; Cowles, 2003; Greene & Kulper, 1990; Nacman, 1976). They typically provide a range of services to hospitalized patients and families that generally center on the two primary functions of discharge planning and psychosocial support (M. Abramson, 1981; Beder, 2006; Cowles, 2003; Holliman, Dziegielewski, & Teare, 2003). Although social workers have been providing these services in hospitals for over 100 years (Bartlett, 1975; Beder, 2006), the function of the hospital is 3 primarily medical care, not social services (Caputi, 1978; Green & Kulper, 1990; Nacman, 1976). As such, social workers are often perceived as secondary professionals in this host setting, where professional power, authority, and prestige are centered on physicians (Abramson & Mizrahi, 1996; Davidson, 1999; Gregorian, 2005; Landau, 2000b). This may lead to conflicted relationships between social workers and medical personnel, especially in the area of ethical issues, practices, deliberations, and resolutions. This potential for conflict is another element contributing to the hospital Ethical Environment. The existing social work research on ethics in the hospital setting explores the ethical issues in which social workers may be involved, the types of dilemmas they encounter, and the process of ethical reasoning in which they may engage (Blumenfield & Lowe, 1987; Doyle, Miller, & Mizra, 2009; Foster, Sharp, Scesny, McLellan & Cotman, 1993; Walden, et al., 1990). In most cases, the research examines how and why hospital social workers make ethical decisions, what involvement, influence, or competence they have in ethical issues, and occasionally includes considerations for the contextual elements of the hospital setting (Foster, et al., 1993; Holland & Kilpatrick, 1991; Proctor, Morrow-Howell, & Lott, 1993). It is generally found that social workers are limited in their involvement and influence in ethical decisionmaking in the hospital (Csikai, 1997; Csikai & Sales, 1998; Joseph & Conrad, 1989; Landau, 2000a, 2000b; Walsh-Bowers, et al., 1996). The most common conclusion in the research is to advocate for more ethics education and training for social workers (Boland, 2006; Davis, 2004; Foster, et al., 1993; Grady, et al., 2008; Jansson & Dodd, 1998, 2002). At the same time, a consistent theme in the research is that there are organizational level factors that greatly influence the desire and ability of social workers to be involved in hospital ethical issues (Proctor, et al., 1993; Doyle, et al., 2009). Among other variables, this indicates the need to 4 explore Ethical Environment. The social work research indicates that there are a host of potentially influential factors when it comes to ethical social work practice in the hospital, but the findings are inconsistent. Personal variables such as age, race, gender, religiosity, and years of service may be related to social workers' involvement and influence in hospital ethical issues (Boland, 2006; Csikai, 1999; Doyle, et al., 2009; Grady, et al., 2008; Jansson & Dodd, 2002). Professional variables such as ethics education and training may also affect the ethical activities of hospital social workers (Boland, 2006; Csikai, 1999; Doyle, et al., 2009; Foster, et al., 1993; Grady, et al., 2008; Jansson & Dodd, 2002; Manetta & Wells, 2001). In addition, organizational level factors, such as the size and structure of the social work services within the hospital, are often mentioned as areas that may influence the ethical actions of hospital social workers (Abramson & Mizrahi, 1996; Csikai & Sales, 1998; Jansson & Dodd, 1998, 2002; Mizrahi & Berger, 2005; Walsh-Bowers, et al., 1996), but this is the least investigated area. Based on the literature reviews that follow, a theoretical model is proposed for the potential relationships between personal, professional, and organizational variables and ethical environment (Figure 1). The literature suggests the theoretical model, but the nature of the relationships is uncertain. Therefore, the study conducted for this dissertation is exploratory and does not necessarily indicate directional research hypotheses. Based on the literature and model, this study will explore the following research questions. Research Questions 1. How do social workers in the US perceive the Ethical Environment of hospitals? a. Do the Ethical Environment ratings of social workers differ from the results of previous studies with nurses? 5 1. Personal Variables: Race, Age, Gender, Religiosity, Social Work Tenure, Hospital Tenure, Job Satisfaction 2. Professional Variables: Ethics Education, Training, and Preparation; Degree Perceptions of Ethical Environment 3. Nature of Organization: Profit Status, Hospital Size, Department Organization, Department Size, Supervision, Hospital Location, Job Functions Figure 1. A Theoretical Research Model 2. Is there any relationship between perceptions of Ethical Environment and personal variables such as: a. Age, race, religiosity, and gender? b. Number of years of hospital social work practice or number of years employed in current hospital? c. Job satisfaction? 3. Is there a relationship between perceptions of Ethical Environment and professional variables such as: a. Ethics education, both in professional schooling (ethics classes and/or ethics content) and in continuing education (both internal and external to the hospital) contexts? b. Level of social work education (BSW-MSW)? c. Having served on an ethics committee? 6 4. Is there a relationship between perceptions of Ethical Environment and organizational variables such as: a. Hospital size and social work department size? b. Social work department configuration (centralized or decentralized)? c. Profit or non-profit status? d. Hospital social worker job functions such as full time or part time status, working in inpatient or outpatient settings, or discharge planning or psychosocial support duties? e. Hospital location (regional differences)? Contributions to the Field This exploratory study has the potential to contribute to the growing body of knowledge on Ethical Environment. Although a common research topic in business ethics, Ethical Environment has only recently become a research topic for hospitals, and has only once appeared in any social work research. The impact of Ethical Environment on social workers in multidisciplinary host settings could provide information for those host organizations to improve ethical behavior and engagement in ethical processes among employees. There are limited studies of Ethical Environment with nurses, but some comparisons with social workers are possible. Hospital and social work department managers can utilize the information gained about how social workers perceive the Ethical Environment of their hospitals to potentially alter and improve the Ethical Environment and subsequent ethical behavior. Outside of modifying hospital policies and procedures, this research might have larger policy implications in the broader realm of hospital and medical care in the United States. This research may also indicate 7 the impact of social work ethics education and training on perceptions of Ethical Environment, and these relationships may suggest future directions for that education. Correlates of Ethical Environment could inform hospital social work practice by bringing to light how the environment interacts with practice dynamics in the hospital setting, such as the need for interdisciplinary collaboration. Given the exploratory nature of this dissertation project, it will likely raise a number of additional questions to guide future research as well. As there are a number of complex variables that interact with Ethical Environment, this study must necessarily restrict its focus to specific aspects that are attainable and measurable within time, funding, and methodological constraints. As such, only the correlational and predictive factors related to Ethical Environment are explored, leaving the impact of Ethical Environment on the ethical actions of social workers for future research. 8 II. Literature Review A. Organizational Culture, Climate, and Ethical Environment The dependent variable of interest in this dissertation is the Ethical Environment of hospitals. As such, the term needs to be defined and placed within the context of the hospital. This will first require an exploration of the industrial and organizational psychology literature, which frames Ethical Environment within organizational culture, and has only recently begun to specify differences between organizational climates and environments. Measures of Ethical Environment have been developed and a small body of literature has applied these instruments to the hospital setting with samples of nurses. Organizational Culture and Climate In essence, an organization's culture is its personality: a fundamental way of being in the world that allows the organization to interact and behave in its own unique interaction with the environment around it in productive ways (Ott, 1989). Organizational culture is rooted in deep, unconscious social constructions that may not be apparent to organizational members, but are the basis for the beliefs and behaviors within the organization (Alvesson, 2002; Ott, 1989; Pettigrew, 1990). It manifests itself in a wide range of features of organizational life, including the way employees are treated (Pettigrew, 1990), the way employees behave, and the implicit organizational limits on behaviors (Ott, 1989). Organizational culture is an important aspect of hospitals in that it is a unique, enduring, and powerful phenomenon within each hospital where leaders, leadership styles and practices, and patterns of decisions may be created and shaped more by culture than by the leaders themselves (Ott, 1989). Shared beliefs, values, moral and ethical codes are central to organizational culture as they provide the justification for why organizations behave as they do 9 (Ott, 1989). The unstated and perhaps unrecognized values and assumptions about the social reality of an organization are key to understanding organizational culture (Alvesson, 2002), as these are the perceptions utilized to measure the concept. Still, the concept of organizational culture is incredibly broad, multidimensional, and difficult to ascertain, thus the focus of research on organizational climate. Organizational climate definitions have varied over the years, the term having first been mentioned in quotation marks without any definition in 1939 by Lewin, Lippit, and White (as cited in Reichers & Schneider, 1990, p. 14). Given the date, it seems that climate research came first, before culture research; however, culture was defined more explicitly before a specific conceptualization of climate was offered. In 1966, Litwin and Stinger offered that climate was dimensions of “structure, reward, and warmth and support – as reported by member perceptions (as cited in Reichers & Schneider, 1990, p. 18).” There are a number of variants to the definition of the climate concept, but it seems to be commonly understood to be a set of organizational attributes that are perceived by organizational members based on organizational behavior that indicate to those members what is important in the organization, inclusive of practices, procedures, and the type of behaviors that are expected of organizational members (Butcher, 1994; Gray-Toft & Anderson, 1985; Hellriegel & Slocum, 1974; Lyon & Ivancevich, 1974; Schneider, 1990). Most commonly, the various climate study authors have agreed that there is no singular, overarching climate within an organization, but a multitude of climates for specific organizational behaviors and practices (Anderson & West, 1998; Gray-Toft & Anderson, 1985; James, James, & Ashe, 1990; Parker, et al., 2003; Schneider, 1990; Wallace, Ivancevich, & Lyon, 1975). Instead, they purport that climate is more specific in that it has a particular 10 referent, as in the 'climate for safety,' or the 'climate for ethics' (Reichers & Schneider, 1990). Some have gone so far as to say that without a particular referent, the concept of climate is meaningless (Anderson & West, 1998). For the hospital, this means there may be multiple climates that characterize specific work groups, departments, or professions, among other possible referents (such as ethics). The complex organization of a hospital across units, specialties, work hours and shifts, and varying occupational and professional hierarchies seems to preclude the formation of a uniform organizational climate (Anderson & West, 1998; Jansen & Chandler, 1994; Schneider, 1990). Despite the amorphous nature of climate, research has shown that it is still a very important aspect of the work environment, particularly as perceived by employees. Negative perceptions of climates that are inconsistently punitive or restrictive have been found to decrease organizational performance (Hellriegel & Slocum, 1974), where as climates that are perceived as open and innovative result in increased staff satisfaction and performance (Gray-Toft & Anderson, 1985). As with culture, the key to climate is organizational members’ perceptions of that climate (Butcher, 1994). To some degree, shared climate perceptions are created through organizational socialization, such as at the beginning of an employee's tenure, while also potentially subject to creation and modification through management communications (Anderson & West, 1998). Despite attempts at modification and management of organizational climate, it is believed to be a relatively enduring characteristic that can change only over long periods of time (Gray-Toft & Anderson, 1985). However, there is still the issue of defining the climate construct in a way that facilitates assessment (Schneider, 1990), and in addition to the referent, the unique role, function, and behaviors of the hospital as an organization matter. Previous climate measurement instruments 11 that appeared effective in one setting (such as business and education), did not generalize well to the hospital setting (Wallace, et al., 1975). In addition, although it may seem that climate now refers to attributes of people, it is really a set of attributes of the organization, merely measured by shared perceptions among individuals (Hellriegel & Slocum, 1974). Measuring climates within organizations has resulted in two main approaches with substantial support (Anderson & West, 1998; Reichers & Schneider, 1990). The Cognitive Schema approach determines climate by ascertaining the psychological meaning and significance that individuals’ constructive representations of their work environment holds for them (Anderson & West, 1998; Parker, et al., 2003). The Shared Perceptions approach extends this sense-making exercise to determine how and to what extent individuals share their assessments of the work environment, particularly shared perceptions of organizational policies, practices, and procedures (Reichers & Schneider, 1990). Climate research focuses on organization members’ perceptions of the way things are, but perception also includes the idea that meaning is attached to the perceived event or thing. The Shared Perceptions approach emphasizes the importance of shared perceptions, but also accounts for the psychological meaning of those perceptions (Parker, et al., 2003). The Relationship between Culture and Climate Originally, several authors interchanged the terms climate and culture (Tagiuri & Litwin, 1968; Miles & Schmuck, 1971; Lippit, Langseth, & Mossop, 1985; as cited in Ott, 1989). Ott (1989) argues that climate is similar to, but not as general as culture; that it is a separate but related phenomenon; but insists that climate is not a component element of organizational culture. Others have argued that climate is indeed a manifestation of culture (Reichers & Schneider, 1990). Clearly, there is substantial overlap between the two concepts. Both climate 12 and culture are learned, largely through the socialization process and through symbolic interaction among group members. They are monolithic, complex, and multidimensional constructs, with culture at a higher level of abstraction, and climate as a manifestation of that culture (Pettigrew, 1990; Reichers & Schneider, 1990). Both climate and culture indicate to organizational members how they are to make sense of their organizational environment. These sense-making attempts manifest themselves as shared meanings that form the basis for action. Although these organizational features are established prior to the entry of a given employee, once an organizational member, that employee participates in constructing, maintaining, and communicating culture and climate to others, as well as themselves (Reichers & Schneider, 1990; Ott, 1989). Both climate and culture are operationalized and measured in similar ways. Parker, et al. (2003) argue that both climate and culture are group-level constructs that may be measured by aggregating psychological perceptions. For climate, the researchers tend to be theorists with an interest in how and why organizations do what they do and how this might be improved or made more effective (Reichers & Schneider, 1990). For culture, it seems the interest lies more in description and comparison, without ascription of any value judgments about effectiveness (Reichers & Schneider, 1990). More so than culture, climate measures tend to focus on shared meanings, especially regarding organizational attitudes toward specific routines and rewards for specific groups and activities (such as social workers and ethical actions). The values and beliefs communicated to organizational members via culture, the routines and rewards communicated via climate, and perhaps even the values and beliefs of individuals, may all converge or be the same (James, et al., 1990; Vandenberghe, 1999). 13 Ethical Culture, Climate, and Environment In general, it appears that ethical climate is a specific concept with multiple dimensions and a subordinate concept of the construct of organizational ethical culture (Agarwal & Malloy, 1999; Vardi, 2001). The two concepts are closely related and overlap, however the ethical climate is considered a more discernable and measurable attribute of an organization (Trevino, Butterfield, & McCabe, 1998; Vardi, 2001). Ethical climate is also more specific in guiding organizational members in identifying ethical issues and determining how they are to be addressed (Weber & Seger, 2002). Although it seems the introduction of the term ethical environment was merely a semantic choice in reference to ethical climate, there is some evidence now that it may well be a different aspect of organizational ethics. In their investigation of ethical culture and climate, Trevino, et al. (1998) found a common factor within ethical culture measures, different from and yet related to several specific dimensions of ethical climate. This mid-range concept they termed Ethical Environment. The understanding of the term environment presented here is also supported by an examination of established measures of ethical climate and environment. Questions in the measures of ethical climate assess more detailed and specific dimensions as applicable to specific work units (the Ethical Climate Questionnaire as established by Victor & Cullen, 1988 and the Hospital Ethical Climate Scale as established by Olsen, 1998). Questions in the measure of Ethical Environment, quite similar to those of Trevino, et al. (1998), assess the larger environment across an organization (the Ethical Environment Questionnaire, as established by McDaniel, 1997). Even so, the overlap in content and concept between climate and environment is substantial, so comparisons between results based on both the climate and environment research and measures are appropriate. For clarity, the term Ethical Environment is used 14 throughout the rest of this dissertation, unless the term climate is necessary to specify findings that may not translate to the environment concept. Ethical Environment Ethical Environment is defined as the organizational practices, attitudes, conditions, and resources that employees perceive and take as an indication of how ethical issues and concerns are dealt with within the organization (Hamric & Blackhall, 2007; Lyon & Ivancevich, 1974). An organization’s Ethical Environment determines how difficult ethical problems are addressed within the organization (Hart, 2005). Ethical Environment also directly influences the ethical attitudes and behaviors of employees (Appelbaum, Deguire, & Lay, 2005; Deshpande & Joseph, 2009; Olson, 1998; Peterson, 2002; Schluter, Winch, Holzhauser, & Henderson, 2008; Sims & Keon, 1999; Trevino, et al., 1998; Vardi, 2001). The Ethical Environment of an organization is communicated to employees by the way the organization deals with ethical behavior of employees and the ethical issues arising from the environments (e.g., social, legal, and economic) in which the organization functions (Hellriegel & Slocum, 1974). If the Ethical Environment of an organization is perceived as poor (in which ethical behavior is not discussed, encouraged, or actively supported by the organization) then employees feel no impetus to maintain ethical behavior for themselves, nor hold others accountable for ethically questionable actions (Appelbaum, et al., 2005; Deshpande & Joseph, 2009; Peterson, 2002; Sims & Keon, 1999; Trevino, et al., 1998; Vardi, 2001). Research into the Ethical Environment of organizations, including hospitals, has found a number of areas affected by a poor environment, such as organizational functions, employee behavior, and job satisfaction (Ambrose, Arnaud, & Schminke, 2008; Goldman & Tabak, 2010; Joseph & Deshpande, 1997; O'Donnell, et al., 2008; Schwepker, 2001). A poor Ethical 15 Environment can increase the rate of employee turnover (Ambrose, et al., 2008; Cullen, Parboteeah, & Victor, 2003; Hart, 2005; Schwepker, 2001; Trevino, et al., 1998), and in fact, Ethical Environment was the most important factor in turnover intent among hospital nurses (Ulrich, et al., 2007). A poor Ethical Environment can also have negative consequences for interdisciplinary teamwork (Rathert & Fleming, 2008), as well as increase the frequency and intensity of ethical problems within the organization (Gray-Toft & Anderson, 1985; McDaniel, 1998; Ulrich & Soeken, 2005). Employees who perceive a poor Ethical Environment, with inconsistent or lacking ethics policies and procedures, are reluctant to engage in ethical deliberations and problem solving. Employees are found to experience moral distress, the stress caused not by ethical dilemmas themselves, but by the constraints of the organizational environment in preventing action on those dilemmas (Corley, et al., 2005). The association between Ethical Environment and unethical behavior is greater in organizations without codes of ethics or other visible institutional ethical statements (Peterson, 2002; Trevino, et al., 1998). In fact, the research indicates that a poor Ethical Environment is more important than the presence of a code of ethics when it comes to the behavior of employees. However, others have stated that the effect of ethical environment on behavior is still being explored and may be questionable (Deshpande, Joseph, & Prasad, 2006). Within the body of research on Ethical Environment, there are actually a number of variables and relationships that are contradictory. Some have noted that Ethical Environment varies between different organizations, depending on the business engaged in and on for- and not-for-profit status (Agarwal & Malloy, 1999; Brower & Shrader, 2000; Gray-Toft & Anderson, 1985; Lane & Schaupp, 1989), but the findings have been weak and difficult to replicate. 16 Differences in Ethical Environment perceptions by gender have been noted in some cases, with women tending to be more critical of Ethical Environment with ratings lower than men (McDaniel, Schoeps, & Lincourt, 2001), and some have stated that the female ratings are more accurate (Tyson, 1990). Others have found no difference in Ethical Environment perceptions by gender (Deshpande, et al., 2006; O'Donnell, et al., 2008). Another equivocal relationship is between Ethical Environment and years of experience on the job or in the field (or age of employee), where those with more experience seem to improve their abilities to engage in ethical deliberations and tend to rate the environment higher (Goldman & Tabak, 2010; McDaniel, 1998). Others have found this untrue, if not the opposite, where tenure indicates a more critical evaluation of environment (Malloy & Agarwal, 2003; McDaniel, 1998). Employee race has also been equivocal with some noting lower environment ratings by African American employees (McDaniel, 1998; Ulrich et al., 2007), while others have noted no differences by race (Deshpande, et al., 2006; O'Donnell, 2008). Finally, there is an ongoing debate over the effect of ethics education on environment ratings, where some have found university courses during professional training improve ratings (Deshpande, et al., 2006; Malloy & Agarwal, 2003; Schluter, et al., 2008), while others have found that continuing education after university training is more important (Ulrich & Soeken, 2005). Although there is a significant body of literature on the ethical climate of businesses and corporations dating back to the 1960’s, there is only recently (beginning in the late 1990’s) an interest in the ethical environments of hospitals and the perceptions of medical staff. After excluding work outside the US, and work not hospital-based, there are only eleven studies on the perceptions of ethical environments in hospitals, all conducted with samples of nurses. Three of the eleven studies (Corley, et al., 2005, and McDaniel, 1997, 1998) used the Ethical Environment 17 Questionnaire (EEQ) designed by McDaniel (1997). Four used the Hospital Ethical Climate Scale (HECS) (Hart, 2005; O'Donnell, 2008; Olson, 1998, and Ulrich, et al., 2007) designed by Olson (1998). One study (Hamric & Blackhall, 2007) used the EEQ in part of their study and then the HECS in another. The final three used the Ethical Climate Questionnaire (ECQ) (Deshpande & Joseph, 2009, Joseph & Deshpande, 1997, and Rathert & Fleming, 2008) designed by Victor and Cullen (1988). Table 1 lists the relevant literature for each measure, with special attention to sample limitations and the dependent variables in each study. Included in Table 1. Ethical climate and environment studies in US hospitals. Instrument Ethical Environment Questionnaire (EEQ) (McDaniel, 1997) Hospital Ethical Climate Survey (HECS) (Olson, 1998) Ethical Climate Questionnaire (ECQ) (Victor & Cullen 1988) EEQ, HECS, ECQ Study (by date) Nurse Sample Hamric & Blackhall (2007) EEQ & HECS Corley, et al. (2005) 196 at 2 hospitals in 1 State 106 at 2 hospitals McDaniel (1998) McDaniel (1997) O'Donnell, et al. (2008) 450 in 2 States 450 in 4 cities 78 hospital social workers (same as Ulrich, et al. (2007) 793 in 4 States (+78 social workers) 463 in 1 State 360 at 2 hospitals in 1 city 103 at 3 hospitals Ulrich, et al. (2007) Hart (2005) Olson (1998) Deshpande & Joseph (2009) Rathert & Fleming (2008) Peterson (2002) Joseph & Deshpande (1997) Schluter, Winch, Holzhauser, & Henderson (2008) 220 at 1 hospital Meta analysis 114 at 1 hospital Systematic Review of Ethical Climate Literature Dependent Variables Perspectives on endof-life care Moral Distress, Ethical Involvement Ethical Environment Ethical Environment Ethical Stress, Moral Action, Job Satisfaction Moral Stress, Job Satisfaction Turnover Intent Ethical Climate Ethical Behavior Teamwork & Leadership Ethical Behavior Job Satisfaction Turnover Table 1 under the ECQ instrument is the meta analysis conducted by Peterson (2002). Although it is not specific to hospitals or nurses, it provides some of the strongest indications that ethical climate (environment) affects the ethical behavior of employees within an organization. Also 18 included in Table 1 is a systematic review by Schluter, et al. (2008) that included all three instruments in their attempt to explain the effects of moral distress and poor ethical climate on nurse turnover. Summary and Theoretical Model The Ethical Environment literature suggests throughout that some personal, professional, and organizational variables are relevant, but poorly understood and not systematically investigated in the case of hospitals. Only O'Donnell, et al. (2008) is exclusive to hospital social workers, but like the samples of nurses, it is small and not representative or generalizable (Table 1). O'Donnell's (2008) sample was drawn from State licensing lists in four states, based on the availability of those licensing lists. There are inconsistent dependent variables (Table 1), and conflicting findings about independent personal and professional variables that leave doubt as to what elements are important. Organizational variables are not well specified or measured. Nevertheless, the available literature suggests a pattern of relationships. Based on the literature a theoretical model representing the potential relationships is presented (Figure 2). Some of the 1. Personal Variables: Race, Age, Gender, Religiosity, Tenure 2. Professional Variables: Ethics Education and Training Perceptions of Ethical Environment 3. Organizational Variables: Type and function of organization, Profit Status, Communication and Collaboration, Nature of Practice Dependent Variables: Ethical Involvement, Teamwork, Job Satisfaction, Turnover Intent, Moral Distress, Ethical Behavior Figure 2. A theoretical model of hospital ethical environment literature Ethical Environment literature would suggest another conception of this model, where personal and professional variables function as intervening between the organizational variables and 19 Ethical Environment (Vidaver-Cohen, 1998). Figure 2 represents a more common theoretical model as derived from the literature that focuses on the hospital setting. The literature review of organizational culture, climate, and environment has established the key variable of Ethical Environment for this dissertation. In considering Ethical Environment, there are a number of ideas from the literature about what contributes to that environment. In the case of hospitals, how ethical issues and conflicts are addressed within the organization are strongly informed by the attitudes toward those issues, and the routines the hospital utilizes and sanctions for use by employees. In the case of hospitals in the US, the standard approach to ethics and ethical reasoning is a medical model, which is different from a social work model. Both of these professional approaches need to be explored in order to understand their contributions to the Ethical Environment of hospitals. The next section explores how these ethical models developed, and the process of ethical reasoning for both professions. B. Ethical Development and Reasoning in Medicine and Social Work To begin, medicine has a much longer ethical history than social work, beginning with the admonition to physicians to “bring benefit and do no harm” (Jonsen, Siegler, & Winslade, 2006, p. 18), often credited to Hippocrates around 400 BCE (Jonsen, 1993; Loewenberg & Dolgoff, 1992; Ponton & Duba, 2009; Strom-Gottfried, 2007; Zussman, 1997), although its actual origin is unknown (Jonsen, 2000). Over time, medical ethics have changed in reaction to changing social institutions, legal precedents, and economic forces (Abbott, 1983; Backof & Martin, 1991; Berlant, 1978), as well as more recent developments and expansions in ethical theory (Jonsen, 2000). Much like the history of medical ethics, the history of social work ethics demonstrates that ethical standards, and in particular professional ethics, are drawn from the social and 20 historical contexts of the time (Doyle, et al., 2009; Reamer, 1998, 2006a, 2006b; Todd, 1930). This history begins with the founding of one of the first national social work organizations, The American Association of Social Workers (AASW) in 1921 (Hurlbutt, 1934). Writers at the time were calling for attention to social work ethics, noting an “utter lack” of ethical standards in social work education or in the AASW (Todd, 1930, p. 553). In 1955, one of the first tasks of the newly formed National Association of Social Workers (NASW) was to develop a code of ethics (Congress, 1999; Dolgoff, Loewenberg, & Harrington, 2009; Freud & Krug, 2002b; Linzer, 1999; Reamer, 1998, 2006a, 2006b; Sparks, 2006). As with the medical ethics, social work ethics have been revised in response to social pressures, as well as the a growing body of ethical theory (Congress, 1999; Freud & Krug 2002b; Linzer, 1999; Reamer, 1998, 2006a, 2006b; Sparks, 2006). General Moral Theory As noted above, both medicine and social work draw from the same moral and ethical theory to establish their professional ethics, as well as their process of ethical reasoning. Ethics has historically been a branch of moral philosophy, and it “involves systematizing, defending, and recommending concepts of right and wrong behavior” (Dabby, Faisal, Holliman, Karliner, & Silverman, 2008, section 2.1). Prominent ethical theories are based on classic perspectives in moral philosophy (Reamer, 1993), and the two most prominent among these are deontology (that decisions should be made on the basis of principles themselves), and teleology (that decisions should be made on the basis of the likely consequences of those decisions) (M. Abramson, 1996a, 1996b; Beckett & Maynard, 2005; Linzer, 1999; Loewy & Loewy, 2004; Reamer, 1979, 1987, 1993, 2006b; Rhodes, 1991; Sommers-Flanagan & Sommers-Flanagan, 2007; Sparks, 2006; Strom-Gottfried, 2007; Wenston, 1987). Many ethics texts begin with a review of these 21 general ethical theories, and more recently tend to include theories of justice that are concerned with fairness and the distribution of the benefits and burdens of the goods and services necessary for a functional society (Beauchamp & Childress, 2009; Blumenfield & Lowe, 1987; Galambos, 1999; Post, Blustein, & Dubler, 2007; Zussman, 1997). The theories of deontology, teleology, and justice are the basis for both medical and social work ethics. Deontology is credited to Immanuel Kant, an 18th century German philosopher (Beauchamp & Childress, 2009; Reamer, 1993) and proposes that there are certain actions that are inherently good or bad, right or wrong, without concern for outcomes (Congress, 1999; Furman, 2003; Rhodes, 1991). The deontological perspective holds that there are rigid duties all people are obligated to and that must always be fulfilled (Blumenfield & Lowe, 1987; Dolgoff, et al., 2009; Meacham, 2007). The weakness of deontology is that strict duties based on good principles can lead to bad outcomes. A classic example is telling the truth to someone about the location of another for whom they are in search of for the express purpose of murdering that individual (Hartsell, 2006). Teleology is sometimes referred to as consequentialism or by its most recognized theory, utilitarianism (Furman, 2003; Rhodes, 1991). Utilitarianism was initially proposed by Britain David Hume, and developed in detail by Jeremy Bentham, an 18th century British philosopher, and John Stuart Mill, a 19th century American (Beauchamp & Childress, 2009; Rachels, 2007; Reamer, 1993). The utilitarian view argues that we should act for what is the best for the most people, aiming for that which will produce the greatest balance of good for the greatest number of persons possible, and that acts are right or wrong based on these consequences (Blumenfield & Lowe, 1987; Congress, 1999; Furman, 2003; Meacham, 2007; Rhodes, 1991). The weakness of teleology is that it may result in grave harms to a few, for the benefit of the greater good 22 (Hartsell, 2006). A classic example is the killing of one person of little social worth (single, orphaned, and homeless) in order to harvest their organs to benefit a number of other highly valued people, their families, and society. Justice is an elusive concept and principle, and no widely accepted unified theory of justice exists (Beauchamp & Childress, 2009). The 20th century American philosopher, John Rawls is most widely known for his theory of justice that posits that the distribution of goods and benefits should be arranged from the point of view of the least advantaged in society (Blumenfield & Lowe, 1987; Hinman, 2008; Reamer, 1993). Rawls’ theory is considered an egalitarian theory that proposes that goods should be redistributed based on need, that those who are the least well off and in the most need are the first consideration in any formulation of distributive justice (Beauchamp & Childress, 2009; Hinman, 2008; Post, et al., 2007). Rawls' may be one of the most prominent theories, but there are others such as the libertarian theory that proposes individual free liberty and rights as the basis for just distributions based on merit (Beauchamp & Childress, 2009; Hinman, 2008; Post, et al., 2007; Rhodes, 1991). Utilitarianism is also considered a theory of justice, as it addresses the distribution of goods based on the greatest good for the greatest number (Beauchamp & Childress, 2009; Hinman, 2008; Post, et al., 2007). In general, justice is concerned with fairness and the equitable distribution of both benefits and burdens in society (Beauchamp & Childress, 2009; Post, et al., 2007; Reamer, 2006a), often with some conception of distribution according to need (Galambos, 1999). Justice and the other moral theories are the basis of ethical theory. Ethical Theory Ethical theory follows from the general (moral) deontological, teleological, and justice theories, but with more emphasis on choosing and defining principles, and on their actual 23 application to moral problems of ethical dilemmas (Beauchamp & Childress, 2009; Reamer, 1979, 1987, 1993, 2006a; Sparks, 2006). The goal of ethics is “the systematic exploration of questions about how we should act in relation to others (Rhodes, 1991, p. 21),” and how we go about establishing a “habitable moral world” (Holstein, 1995, p. 179) in which to live. The processes of determining which ethical principles should be put into use in the first place, how terms are defined, and the methods by which these determinations are made are metaethics. The actual applications of those principles once chosen are normative ethics. Normative ethical theories do not resolve ethical issues or difficulties, but provide ways of analyzing and clarifying them (M. Abramson, 1981, 1983; Blumenfield & Lowe, 1987; Congress, 1999; Rhodes, 1991). Both deontology and teleology are normative ethics (Reamer, 2006a), as are professional ethics for both social work and medicine. No matter which moral or ethical theory is utilized, the point in their application to real world situations is to assist people in resolving ethical dilemmas by engaging in a process of ethical reasoning. Ethical Dilemmas Regardless of which ethical theory or principles are chosen, they are all meant to address the ethical dilemma. Ethical dilemmas are usually classified into two types; one where an action is both right and wrong (or the evidence is inconclusive), and one where two courses of action are equally right but only one can be done, thus violating the other (Beauchamp & Childress, 2009; Linzer, 1999; Post, et al., 2007). Dilemmas are embedded in complex, real-life situations, where they can be constructed (perceived and interpreted) in numerous ways (Fleck-Henderson, 1991; Walden, et al., 1990). In any given ethical quandary, our feelings are engaged first and far before abstract thinking about morality and ethics in a broad philosophical sense (Doyle, et al., 2009; Fletcher, 1973; Holland & Kilpatrick, 1991). The personal beliefs and strong moral 24 convictions of individuals can interfere with moral reasoning (M. Abramson, 1996a; FleckHenderson, 1991; Goldstein, 1998; Meacham, 2007). To address potential confounding factors in ethical reasoning about ethical dilemmas, attention to an ethical reasoning process is required. The Process of Ethical Reasoning The goal of good ethical reasoning is not a resolution to an ethical dilemma, but the actual process of thoughtful reflection and carefully considered judgments, keeping in mind that responsible, thoughtful, diligent, and reasonable people may disagree about the most appropriate course of action (Reamer, 1993). At the core of ethical reasoning is deliberation among people with differing perspectives (Gracia, 2003). As the moral training and professional socialization for medicine and social work is different (Clark, 1997; Congress, 1999; Csikai, 1999; Foster, et al., 1993), there is no one set of knowledge or process skills that both will share in facing an ethical dilemma in the hospital setting. Strong communication and listening skills are necessary to promote dialogue (M. Abramson, 1996a), and moral agents will have their own underlying moral values, and will believe that they know, personally, deeply and without question, what is right and wrong (Goldstein, 1998). Although general ethical reasoning requires a deliberative process, how that process is actualized also differs among the medical and social work disciplines. Reasoning in Medical Ethics Reasoning in medical ethics tends to be much more deontological, evaluating acts as either right or wrong, apart from any consequences (Zussman, 1997). As such, medical ethics have become quite institutionalized as principles and procedures. Although no one profession or discipline has exclusive rights over medical ethics (Fletcher, 1973), the fact is that physicians dominate the practice of reasoning over ethics in medicine. This is likely tied to the social and 25 professional prestige of modern medicine, physician authority in the hierarchy of medical care, and may be due to the professional socialization of physicians. Physician dominance in medical ethics raises two main concerns as noted in the extant literature. First, “research on moral development in medical students indicates that their training inhibits, rather than enhances, the development of moral reasoning” (Clark, 1997, p. 444). It is believed that this is due to an emphasis on 'clinical detachment' in order to focus on the cognitive processing required for medical practice. Second, physicians tend to concentrate on the individual case, and have difficulty understanding the larger social structures involved and the collective consequences of their decisions (Zussman, 1997). Nevertheless, medical ethics has strongly established its own model for reasoning through an ethical dilemma. The medical model of ethical reasoning: Principlism. ‘Principlism’ is the established model for normative reasoning in ethical dilemmas and has become the most popular, influential, and dominant approach in Western bioethics and medical practice (R. Davis, 1995; Levi, 1996; McCarthy, 2003). There are other models of reasoning in medical ethics (e.g., Jonsen, et al., 2006); however, they tend to reflect a reorganized form of the same content as principlism. Principlism is drawn from the Belmont Report – the findings of the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research (Antle & Regehr, 2003; Evans, 2000; McCarthy, 2003). The Belmont Report and the subsequent medical ethics text based on that report (Beauchamp & Childress, 2009) both identify four cardinal principles for consideration in research and medical practice with human participants: autonomy, beneficence, nonmaleficence, and justice (Antle & Regehr, 2003; Beauchamp & Childress, 2009). A. Autonomy is the individual’s right to choose intentionally and knowingly for himself or herself, free from controlling interference by others (Beauchamp & Childress, 2009). 26 B. Beneficence is promoting good, preventing harm, and removing harm (Beauchamp & Childress, 2009). Beneficence requires taking action or positive steps to help others and includes all forms of action intended to benefit others. C. Nonmaleficence is “the obligation not to inflict harm on others” (Beauchamp & Childress, 2009, p. 149). It is limited to just not inflicting harm, and may be considered the converse of beneficence (Jonsen, et al., 2006). D. Justice is concerned with “fair, equitable and appropriate treatment in light of what is due or owed to persons (Beauchamp & Childress, 2009, p. 241).” Due to persons means that any one individual is only asked to bear the burdens that are fair in consideration for their benefits, as well as the burdens and benefits of others (Beauchamp & Childress, 2009; Post, et al., 2007; Reamer, 1993, 2006a). Justice is further classified into two types: 1. Comparative Justice: fair and equitable treatment between individuals when their interests compete and are compared with the interests of others (Wenston, 1987). 2. Distributive Justice: fair and equitable treatment between individuals and larger groups, such as communities and societies, and how benefits and burdens are distributed in society, among all members (Post, et al., 2007; Wenston, 1987). The framework for moral reasoning offered by principlism is intended as a process, not a product with which to arrive at moral certainty (McCarthy, 2003). The aim of this process is not really to make decisions, but to explore and investigate what the ethical issues are (Gracia, 2003). Additionally, the principles are not meant to be applied in a mechanical or sequential fashion. They need to be used with judgment and sensitivity to inform ethical reasoning (Holstein, 1995; Osman & Perlin, 1994; Post, et al., 2007). In this process, each principle must be carefully and fully considered, weighed, and balanced, with none having preeminence over 27 another in order to determine which imposes the overriding obligation (Beauchamp & Childress, 2009; R. Davis, 1995; McCarthy, 2003). The outcome of this process is meant to be considered judgments, which are the moral convictions about which we have the highest degree of confidence (Jecker, 2008), and which we believe are the least biased (Beauchamp & Childress, 2009). Reflective equilibrium is the process of reviewing, testing, and revising these considered judgments to make them as coherent as possible, and representative of our most general (universal) moral commitments about right and wrong (Beauchamp & Childress, 2009; R. Davis, 1995; McCarthy, 2003). Although it is argued that each principle is of equal consideration, the practical reality in the practice of medical ethics is that autonomy always takes a central and overriding role (Callahan, 2003; Walker, 2008; Wenston, 1987). Reasoning in Social Work Ethics Social work practice is inextricably and inevitably involved in moral and ethical issues (Csikai, 1999; Fleck-Henderson, 1991; Jansson & Dodd, 1998), and values and ethics have always had a central place in social work (Reamer, 1993). At the same time, as with all historically situated ethics, social work's professional values and ethics have been reactive to political, economic, social, and cultural influences (Freud & Krug, 2002a). This suggests that in order to advance the profession, social workers need to take a proactive stance in ethical practice, rather than reacting to social, cultural, and legal events (Dabby, et al., 2008; Reamer, 1985; Rhodes, 1991). It also means that socialization into the social work profession must continue its unique emphasis on personal, professional, and social values and value conflicts (Clark, 1997; Silverman, 1992). The social work literature suggests that some form of moral reasoning is pervasive in social work practice (Fleck-Henderson, 1991), and there is a body of literature on methodologies for identifying, clarifying, and analyzing these ethical conflicts (M. Abramson, 28 1996a). The available literature suggests that when ethical conflicts arise, social work can draw from three forms of ethical reasoning; (1) sequential models of ethical analysis; (2) the rules and guidelines for professional relationships, most prominent as the code of ethics; and (3) the general ethical principles representing the professions’ values. Sequential models. The first step in any system of ethical reasoning is to recognize a dilemma as a dilemma and then to take responsibility for the need to take action to address the dilemma (Fleck-Henderson, 1991). After this, there are a number of systematic and sequential models that have been proposed, however they all seem to have similar steps (modified from Reamer, 1998): 1. Identify conflicts between values, duties, laws, and/or ethics. 2. Identify those involved in, and those affected by, the decision. 3. Tentatively identify all possible courses of action and the risks and benefits of each course. 4. Examine reasons for and against each possible course of action by applying ethical theory, the code of ethics, and applicable laws. 5. Consult with colleagues and experts. 6. Make the decision and document the process. 7. Monitor and evaluate the decision. The problem with the sequential models is that they are rarely, if ever, utilized by practicing professionals (Doyle, et al, 2009; Holland & Kilpatrick, 1991; Joseph & Conrad, 1989; Kugelman, 1992; Mattison, 2000; Walden, et al., 1990). The NASW code of ethics as ethical reasoning. The unequal power in professional helping relationships (medicine or social work) requires a special type of social control within 29 the profession, most often represented in codes of ethics (Abbott, 1993; Kultgen, 1988; Manning, 1997; Rhodes, 1991; Strom-Gottfried, 2007). Codes of ethics have three main functions; (1) they describe the highest moral ideal for practitioners; (2) they educate the profession and the public as to the nature, limits, and expectations of the profession; and (3) they have a proscriptive function in clearly stating what may not be done within the profession (Congress, 1999; Sommers-Flanagan & Sommers-Flanagan, 2007). The intention of the NASW code of ethics (National Association of Social Workers [NASW], 2006) may be interpreted as a resource for social workers to consult in an ethical quandary to help guide their deliberations and inform their decision-making. Unfortunately, the NASW code of ethics is often considered too vague to be of use by practitioners in resolving ethical dilemmas or assisting with ethical reasoning (Dabby, et al., 2008; Doyle, et al., 2009; Fleck-Henderson, 1991; Peterson, 1987; Reamer, 1987). Like the sequential models, professionals in practice rarely use the NASW code of ethics (Congress, 1999; Hartsell, 2006; Rawwas, Strutton, & Pelton, 1994; Rhodes, 1991). Ethical principles in social work. Some social work literature suggests that ethical issues can be resolved by applying the ethical principles drawn from the professional values as implied in the social work code of ethics (Fleck-Henderson, 1991). Strom-Gottfried (2007) states that there are five principles that underlie social work ethics; the same cardinal principles of autonomy, beneficence, nonmaleficence, and justice as in principlism, and the additional principle of fidelity – defined as keeping promises and upholding trust. In comparing codes of ethics between helping professions (physicians, counselors, psychologists, marriage and family therapists, and social work), once again the same four cardinal principles are identified (Hadjistavropoulos, 1996; Parsons, 2001; Reamer, 2006a; Strom-Gottfried, 2007). The difference between the codes of ethics is one of emphasis on, rather than omission of, any of the 30 four cardinal principles. Although some contend that the principlist approach is not appropriate for social work (M. Abramson, 1996b), it nonetheless has broad applications and implications for social work practice (Reamer, 2006a) and the fact is that social work has focused on principlism of a sort, by relying on the core values of the profession and the principles they represent (M. Abramson, 1996a). The NASW code of ethics heavily emphasizes the traditional and enduring central social work principle of self-determination; that the client's right to choose for themselves how to live should supersede other professional social work responsibilities (Antle & Regehr, 2003; Clark, 1997; Congress, 1999; Furman, 2003; Galambos, 1999; Reamer, 1985, 2006a, 2006b; StromGottfried, 2007). Self-determination is essentially autonomy by another name, so, as with medical principlism, this principle also takes precedence in social work practice. The vigorous focus on self-determination at the expense of all other social work values and principles is a notable concern for the profession (Manetta & Wells, 2001), especially in light of the fundamental social work mission to ensure social justice (Freud & Krug, 2002b). The NASW code is considered to have representations of beneficence and nonmaleficence, but it is the weakness on issues of justice, and especially on distributive justice, that concerns social work scholars (Furman, 2003; Hadjistavropoulos, 1996). Although the code can be interpreted as emphasizing justice as a core value (Galambos, 1999), it is a different conception of the principle that focuses on social justice in the form of advocacy for the disadvantaged and oppressed (NASW, 2006; Reamer, 2006a, 2006b). This is a formulation once removed from the broader justice concern for the equitable distribution of social goods in both the comparative and distributive senses. 31 Reviewing the literature on the development of ethics and ethical reasoning in social work and medicine raises some interesting connections. Both professions have built up ethical understandings, ideas, and principles from a common foundation of ethical theory. Each profession has establish ethical reasoning processes and ethical models that would seem to meet their professional needs, as well as the needs of the social institutions in which they operate, and the needs of society. However, they have certainly diverged in the most recent formulations of principlism vs. the sequential models or the NASW code of ethics. Since the approach to ethics in the hospital setting is based on the medical ethics model, it is reasonable to assume that this is a strong contributor to the Ethical Environment of hospitals. The influence this might have on social workers seems compounded given that the social work systems for addressing ethics are not being utilized. But the measurement of Ethical Environment is not entirely dependent on the organizational routines and processes of ethical reasoning and deliberations based in principlism. The concept of Ethical Environment also includes the dynamics inherent in the practice setting, the roles organizational members play, and the impacts these have on ethical practice. C. The Context and Practice of Hospital Social Work A Brief History of Social Work in the Hospital Setting The first consideration when it comes to the context and practice of hospital social work is a brief explanation of how social workers first entered into hospitals in the US. The history of social work in the medical setting began in 1905 when Richard Cabot, a physician at Massachusetts General Hospital (MGH) in Boston, hired a nurse to help address the social issues of his patients that he felt were interfering with their medical care and compliance (Bartlett, 1975; Beder, 2006; Cannon, 1923, 1952; Caputi, 1978; Cowles, 2003; Gehlert, 2006; Hurlbutt, 32 1934; Luptak, 2004; Nacman, 1976; Stuart, 2004). In 1907, Dr. Cabot hired Ida Maud Cannon, a nurse who had just completed social work training at the Boston School for Social Workers, later Simmons College (Bartlett, 1975; Beder, 2006; Cowles, 2003; Gehlert, 2006; Luptak, 2004; Nacman, 1976). Ida Cannon and other early social work pioneers were charged with investigating the social situations of Dr. Cabot’s patients in the dispensary, or outpatient clinic, and coordinating with community resources that could improve the patient’s situation and compliance with medical care (Bartlett, 1975; Cannon, 1923, 1952; Caputi, 1978; Cowles, 2003; Nacman, 1976). In 1914, social work was expanded into the inpatient wards of MGH, establishing hospital social work (Nacman, 1976). Cannon then hired Ruth Emerson as the first inpatient ward social worker (Cannon, 1952). Hospital social work spread across the nation, and a large number of the new directors of social services departments were employed or trained at MGH under Cannon and Dr. Cabot (Bartlett, 1975). The growing number of hospital social workers resulted in the founding of the American Association of Hospital Social Workers in 1918 (Caputi, 1978; Nacman, 1976). The social services department was officially recognized and fully financially supported by MGH in 1919 (Nacman, 1976). Ida Cannon remained the Chief of Social Services until her retirement in 1945 (Bartlett, 1975). There have been numerous social and cultural changes since Ida Cannon laid the foundations of hospital social work. The prominence and integration of social work in the hospital has waxed and waned, and yet modern medical social work remains very much the same. The primary function of the hospital social worker was and still is discharge planning and the psychosocial care of the patient and family (M. Abramson, 1981; Beder, 2006; Cannon, 1923, 1952; Cowles, 2003; Gregorian, 2005; Holliman, et al., 2003). Almost a century of social 33 work in the inpatient hospital setting has passed, and throughout this time, regardless of the presence of social workers in the inpatient hospital and the prominence social work enjoyed at MGH, the primary function of the hospital has been medical care, not social work (Caputi, 1978; Cowles, 2003; Greene & Kulper, 1990; Nacman, 1976). This is expected and appropriate for hospital care, but perhaps too exclusive of the medical perspective. This leads to the position of social work in the hospital as a sort of guest in a host setting, which could certainly have an impact on social workers' perceptions of Ethical Environment. Practice in a Host Setting Given that social work is operating in the host setting of the hospital, where the primary function of the agency is medical care, not social work (Caputi, 1978; Cowles, 2003; Greene & Kulper, 1990; Nacman, 1976), social workers are often perceived as secondary professionals. In the host setting, authority and prestige are centered on the physicians who head a hierarchy of disciplines with lower status (Abramson & Mizrahi, 1996; Davidson, 1990; Gregorian, 2005; Landau, 2000b). This kind of power differential allows physicians to direct ethical practices and deliberations in the hospital and to apply the principlism model. Since the host setting concept is not applied to medical personnel, there is no known research about how it might affect hospital Ethical Environment. Although not explored in the social work literature specifically for hospital social work, the host setting aspect of social work practice has been touched upon. DiFranks (2008) conducted a nationwide study of NASW members and found that working in a host setting (loosely defined to include hospitals, prisons, and "other") made no difference at all in aspects of ethical practice. Carlton (1989) suggests that the host setting concept for hospital social work should be rejected because it “puts social workers in the position of perpetual guests – a self- 34 assumed, subservient position that is a serious disadvantage from the outset” which is erroneous given that social workers are in hospitals “because their presence is sanctioned by the public” and is necessary for hospitals and society to achieve the goal of good, effective health care (Carlton, 1989, p. 148). Even so, the research explored below suggests that some dynamics of the host setting do affect both ethics and social work practice. Before looking at the contrasts however, there is one space within the hospital where the social work perspective might be invited and complimentary: the hospital ethics committee. The Hospital Ethics Committee A hospital bioethics committee is a multidisciplinary group of health care professionals that has been established specifically to address the ethical dilemmas that occur within a health care institution (Csikai, 1997; Silverman, 1992). Ethics committees address the need for a more systematic and principled approach to ethical decision-making (Silverman, 1992), and the need to protect both the interests of the patient, and of the hospital providing care (Csikai, 1997). In 1992, The Joint Commission on Accreditation of Hospital Organizations (JCAHO), the accrediting body over hospitals in the US, mandated that hospitals put in place a means for addressing ethical concerns (Csikai, 1997; Jonsen, et al., 2006; McGee, Spanogle, Caplan, Penny & Asch, 2002), and this is commonly in the form of an ethics committee. The hospital ethics committee is generally ascribed with four main functions: (1) the education of hospital staff, management, and patients and families on ethical issues and their resolution; (2) the review and development of institutional policy to prevent or address ethical issues in medical care; (3) consultation on ethically complex cases; and (4) retrospective review of ethically complex cases (Congress, 1999; Csikai, 1997; Furlong, 1986; Jonsen, et al., 2006; Kanoti & Vinicky, 1987; Levine, 1984; Reamer, 1987, 2006b; Silverman, 1992). Ethics 35 committees generally spend most of their time on the active and retrospective case consultations (Csikai, 1997; Reamer, 2006b), although recently it seems that this has diminished and the focus is now more on policy review. Within the social work literature on ethics committees, a number of authors comment on how social workers are in an excellent position, by way of professional values, training, skills, and practice, to serve on ethics committees (Csikai, 1997; Csikai & Sales, 1998; Furlong, 1986; Silverman, 1992). Silverman (1992) notes, in particular, that the examination of committee members’ own biases may not have been part of their professional training, but is a crucial part of social work education. An often-cited 1991 national study found 60% of hospitals had an ethics committee, and that a social worker served on 76% of them (Skinner, 1992 as cited in Csikai & Sales, 1998), which Csikai (1997) confirmed in a sample of hospitals in Pennsylvania. In Maryland, where ethics committees are mandated for all hospitals by state law, a survey found that 97.5% of committees had a social worker as a member (Hoffmann, et al., 2000). However, Foster, et al. (1993) found that across seven states only 11.5% of hospital social workers (in a sample of 255) had ever served on a hospital ethics committee. In a 2003 nationwide sample, only 3.5% of 517 ethics committees reported having a social worker as a member (Guo & Schick, 2003). It seems that over time, both the prominence of ethics committees and active involvement by social workers has diminished. Social work continues its tenuous and yet consistent presence within the hospital, even if not on the ethics committee. The role of social workers in hospital ethical issues and events may be minimal, but the future role of the profession is yet to be determined. As such, the importance of the enduring roles of hospital social workers may contribute more to the understanding of the context and practice of their work, and how this 36 might relate to Ethical Environment. The two key roles or functions of hospital social workers are discharge planning and psychosocial care, and these are critical to the understanding of both historical and modern hospital social work practice. Discharge Planning and Psychosocial Care Discharge planning and the psychosocial care of hospitalized patients and their families are the core functions of social workers in the hospital (Abramson, Donnelly, King & Mailick, 1993; Beder, 2006; Cowles & Lefcowitz, 1995; Cowles, 2003; Kadushin & Kulys, 1993). As noted in the brief history above, these have been the traditional responsibilities of social work for more than a century (Holliman, et al., 2003). In some hospitals, social workers hold primary responsibility for discharge planning, while in others they may be focused almost exclusively on psychosocial care, and in many they are responsible for both (Cowles, 2003; Gregorian, 2005). Discharge planning is a continuum of activities intended to engage patients and their family members in considering various options to facilitate a patient’s transition out of the hospital and to arrange for continuing care and services after hospitalization (J. Abramson, 1988; Blumenfield & Lowe, 1987; Boland, 2006; Cummings & Cockerham, 1997; Kadushin & Kulys, 1993; Wimberley, 1988). Psychosocial care and support is a function of the assessment and management of psychosocial issues that occur in and complicate discharge planning. It may stand separately when discharge planning needs are minimal and the focus remains on psychosocial adjustment and coping with medical illness and its impact on patients and their families (Beder, 2006; Cowles, 2003). Both discharge planning and psychosocial care occur within, and are influenced by, the hospital setting, and in both arenas, ethical issues can easily arise (Cummings & Cockerham, 1997; Holliman, et al., 2003). 37 The Changing Nature of Practice Even if discharge planning and psychosocial care remain the primary functions of hospital social workers, the context of practice has certainly changed over time. The major changes have been in the discharge planning role, where there is a new level of interdependence between the professional decision makers involved in the provision of modern medical care (Abramson, et al., 1993; Boutin-Foster, et al., 2005; Cummings, 1999; Holden, Cuzzi, Grob, & Bazer, 1995). The social work role of discharge planning must also attend to and negotiate the varying needs and competing interests of patients, families, medical staff, and community agencies (Cummings & Cockerham, 1997). As discharge planning cases become more challenging, especially with additional psychosocial issues, social workers are often asked to be involved, which also increases the likelihood of encountering ethical issues (Cummings & Cockerham, 1997; Holliman, et al., 2003). Social workers need to be clear on their role and contributions, and they must have the capacity to exchange information with and educate others. However, even when social workers have a clear concept of their roles in the hospital and can communicate them well – those others (physicians and nurses) may not agree with the social work conception (Feather, 1993; Gregorian, 2005; Keigher, 1997; Landau, 2000b; Nelson & Merighi, 2003). Interdisciplinary collaboration. The increased interdependence of hospital providers demands interdisciplinary collaboration. Given the pressures on social workers in discharge planning, the medical setting, and the involvement of differing professions, effectiveness in the job is highly influenced by the cooperation (or lack thereof) of physicians (Feather, 1993; Kadushin & Kulys, 1993, 1995). Both the power and authority of physicians, and the feeling that their own autonomy, professional decision-making, and expertise is being encroached upon 38 by the pressures of cost-containment, can seriously complicate discharge planning efforts (Feather, 1993; Kadushin & Kulys, 1993). The need for collaboration and interdisciplinary work is vital in modern hospital social work (Berkman, et al., 1996) and both social work and the medical professions find themselves sacrificing some of their autonomy to facilitate collaborative problem solving, both in practice and in ethics (Abramson & Mizrahi, 1996). For social workers, these challenges may include outright resistance from physicians who simply do not understand or appreciate the social work role (Feather, 1993; Kadushin & Kulys, 1993, 1995). Social worker involvement in ethics issues and deliberations hinges on the behavior and attitudes of physicians in the hospital, and especially within a given physician’s specific hospital unit or service area (Abramson & Mizrahi, 1996; Landau, 2000a). The mention of specific hospital units alludes quite nicely to the concept of ethical climate, and thus the overall Ethical Environment of the hospital. The difficulties in interdisciplinary communication are partly due to the differing professional socialization, training, and education of each discipline that imparts distinct values to each (Abramson, et al., 1993; Abramson & Mizrahi, 1996). In this environment, social workers need to work actively to build and maintain relationships with the medical staff (Gregorian, 2005), increase their involvement and visibility in ethical decision-making in the hospital, and improve their communication skills with the other professions (Kadushin & Kulys, 1995; Landau, 2000b). Social workers offer a unique knowledge base that can be very beneficial in the process of ethical decision-making (Landau, 2000b). With their particular skill set and unique perspective, and the changes in the nature of hospital practice, there is some evidence of expanding roles for social work, as well as more recognition and a reduction in inter-professional conflicts (Abramson & Mizrahi, 1996). 39 Ethics in Discharge Planning Whatever the interdisciplinary complications, there are additional forces to consider when it comes to ethical issues in discharge planning. There are ethical tensions in hospital social work between advocacy for the patient and family, the need for collaboration with other medical professionals, and the responsibility the social worker has to the employing organization (Auerbach, Mason, & LaPorte, 2007; Cummings & Cockerham, 1997; Gregorian, 2005). Out of necessity for survival during the recent period of cost-containment, social workers in health care have become more attuned to organizational factors (Mizrahi & Berger, 2005), including the organizational Ethical Environment. Hospital social workers tend to rely on external laws, rules, and authorities (a deontological approach), with a focus on resolving ethical dilemmas on a case-by-case basis (with a utilitarian perspective on outcomes) without any systematic application of ethical principles or models (Landau, 2000a). A lack of social work preparation in ethical analysis and decision-making increases the likelihood of ethically questionable practices in discharge planning (Cummings & Cockerham, 1997). If the deontological approach fails to resolve the ethical issues for the social worker, they then seek consultations from multiple sources, including social work and non-social work colleagues. Nevertheless, the level of influence the social worker discharge planners have is limited, constrained, or undermined by the authority and role status of physicians and nurses (Clemens, 1995). The literature suggests that organizational variables can have an impact, such as profit status and the economic pressures this may bring to bear on ethics in discharge planning, but this has not been explored in detail. 40 Social Work Involvement in Ethics It is clear that social workers need to be involved in the ethical deliberations of the hospital, especially as these arise in and involve discharge planning and interdisciplinary collaboration. Social workers can take steps to take part in and improve their role in ethical deliberations and decision-making in the hospital (Landau, 2000b). The available research raises doubts about this involvement. In some instances, depending on the hospital, and more so the physicians, social workers seem to be discouraged from raising ethical issues or involvement in their resolution. However, this seems to hold only insofar as the issues involve a singular patient; as the complexities of a given case increase, and there is more involvement of family and external agencies and resources, social worker participation is expected and welcomed (Landau, 2000b), although this may be the case only so long as the social workers acquiesce to the authority of physicians and the financial needs of the hospital. The context, practice, roles, and functions of hospital social workers are strongly related to the ethical dilemmas they encounter, and the previous review of ethical reasoning suggests that the professional social work reasoning models are poorly utilized. The medical model takes precedence in this host setting, both in clinical practice, and within the ethics committee. The changing nature of hospital social work practice may still center on discharge planning and psychosocial care, but the need for greater collaboration and attention to ethics suggests the increasing importance of the Ethical Environment, and its impact on ethical actions. Research on social work ethics in the hospital setting is beginning to explore these dynamics. 41 D. Research on Social Work Ethics in the Hospital Setting The Need for Social Work Ethics Research There has been a call for more basic research in all fields of social work practice for some time (Wade & Neuman, 2007). This is true for health care as well, where empirical and practice outcomes research is needed to affirm the efficacy, value, and contribution of social work in the hospital setting (Auerbach, et al., 2007; C. Davis, 2004; Keigher, 1997; Kossman, Lamb, O’Brien, Predmore, & Prescher, 2005; Neuman, 2003). Most studies of social workers in the hospital simply describe what hospital social workers are doing, or how other professionals view hospital social workers (C. Davis, 2004; Neuman, 2003; Simon, 1991). There has been very little in this research agenda on ethics. In a review of ten years of hospital-based research (19912001), only two were found that addressed ethics (C. Davis, 2004). This lack of information belies the fact that investigations and knowledge development of the social work profession’s ethics are just as critical as other knowledge and theory (Walz & Ritchie, 2000). The need to demonstrate social work value includes the need to explore and show that social worker participation in ethical situations improves patient well-being or patient outcomes (Jansson & Dodd, 2002). Social work practice roles in health care are being profoundly affected by the more frequent occurrence of ethical issues (Keigher, 1997), and this represents an opportunity for social work to be more involved (Csikai, 1997). Multiple authors in the area of hospital social work ethics comment on how the purposes, knowledge, skills, training, and values of social workers can contribute significantly to ethical decision-making (Csikai, 1997, 1999; Csikai & Sales, 1998; Furlong, 1986; Joseph & Conrad, 1989; Silverman, 1992). They also note that professional socialization during social work education provides a different and contributory perspective than the training of other professions 42 working in medicine. In particular, social work training typically includes the examination of social workers’ own biases, which may not have been a part of the training of other medical staff (Silverman, 1992). On the other hand, many note that the marked differences between the values and professional socialization of the various health professions can make communication between them difficult (Clark, 1997; Congress, 1999; Keigher, 1997). Barriers to Social Work Ethics Research in Hospitals Interdisciplinary communication, collaboration, and the complexities of discharge planning are issues, if not barriers, that would make research in this area difficult. Some social workers feel that they lack the resources, expertise, and time to engage in research (Jansson & Dodd, 2002; Neuman, 2003). Others are even opposed to social work involvement in research, feeling that it detracts from their primary clinical practice (Christ, Siegel, & Weinstein, 1995) and responsibilities for the care of their clients (Simon, 1991). In addition, given that social workers in the hospital setting have rarely engaged in research activities, some of the initial efforts have met with resistance from other professions (Christ, et al., 1995; Neuman, 2003; Simon, 1991). The ongoing social work struggle to attain recognition in the hospital setting where physicians and other medical staff hold much greater authority and prestige (Joseph & Conrad, 1989) presents a significant barrier to engaging in ethics, let alone research. Social work has had a peripheral presence in the hospital, perhaps not by choice given the host setting, and this has contributed to a history of tenuous relationships with colleagues in medicine and nursing (Keigher, 1997). Some of the ethics research has indeed found that some hospitals and some physicians are not highly receptive to social work participation in ethical deliberations (Jansson 43 & Dodd, 2002). Social workers may assume submissive roles due to physician power and authority, and because they may feel inadequately prepared for ethical deliberations. Social Work Ethics Research in Hospitals Regardless of barriers, research does progress, usually beginning with the theoretical literature. In the hospital social work area, this literature comments on the need for ethical practice and proposes models of ethical reasoning, but very little speaks to empirical research on ethical issues, decision-making, or the value of social work to these processes and outcomes (Csikai, 1999; Doyle, et al., 2009; Foster, et al., 1993; Holland & Kilpatrick, 1991; Jansson & Dodd, 1998; Proctor, et al., 1993; Walden, et al., 1990). The outcome of most of the research on ethics has been the numerous, remarkably similar, and rarely utilized sequential models of analysis to help in decision-making (Blumenfield & Lowe, 1987; Congress, 1999; Linzer, 1999; Loewenberg & Dolgoff, 1992; Reamer, 2006a; Strom-Gottfried, 2007). In addition to model development, there is a growing body of descriptive research that speaks to what social workers do, what ethical dilemmas they encounter, how social workers perceive themselves and how they are perceived by others, and to some extent their process of ethical reasoning (Blumenfield & Lowe, 1987; Doyle, et al., 2009; Foster, et al., 1993; Walden, et al., 1990). A detailed review of this literature will provide a number of independent variables for this dissertation research, many quite similar to those in the Ethical Environment literature. Social Work Ethics Research Literature Reviews Jansson and Dodd (1998) reviewed the social work literature on ethics in health care (not specific to hospitals) from 1980 to 1996. Their review was ostensibly to propose a heuristic framework to help organize and generate ethics research. Only five articles from their review that are pertinent to hospital-based social work are included in the following analysis. The 44 majority of studies analyzed the ethical choices of individual professionals (Foster, et al., 1993; Holland & Kilpatrick, 1991; Proctor, et al., 1993). These same studies also include other relevant factors, such as descriptions of ethical dilemmas, the ethical reasoning process of social workers, and rarely, contextual elements of the hospital setting. Only Kugelman (1992) looked at whether social workers recognized specific ethical dilemmas, and only Joseph and Conrad (1989) analyzed the ethical decision making of multidisciplinary teams. C. Davis (2004) reviewed the social work literature on hospital social work (not limited to ethics) from 1991 to 2001. That review did find two articles that addressed ethics (Silverman, 1992; a theoretical article on the importance of social workers on hospital ethics committees, and Proctor, et al., 1993; an analysis of discharge planning factors that correlated with ethical complications during hospitalization). Since the Jansson and Dodd (1998) and C. Davis (2004) reviews of literature through 2001, there appear to have been only four additional articles specific to hospital social work and ethics in the U.S.: Jansson and Dodd (2002), Boland (2006), Grady, et al. (2008), and O'Donnell, et al. (2008). Overall, the research literature on ethics within the hospital setting falls into two broad and nonexclusive categories. First are the theoretical research literature comprised of various authors’ efforts either to demonstrate the use of ethical decision-making models, or to espouse the value and skills that social workers bring to ethical deliberations. These are few and are summarized in Table 2. Table 2. Theoretical literature on hospital social work ethics. Author (by date) Cummings & Cockerham (1997) Silverman (1992) Blumenfield & Lowe (1987) M. Abramson (1983, 1981) Summary Proposes a decision-making model. Discusses the importance of social workers on ethics committees. Proposes a decision-making model. Applies ethical principles to the discharge planning process. 45 Second are the descriptive research studies. These are summarized in Table 3 with particular attention to the sample of hospital social workers and the dependent variable of interest in the study. Within all the literature, there are four main themes when it comes to hospital social work: 1. The nature of ethical dilemmas encountered. 2. The decision-making processes employed by hospital social workers. 3. The perceptions of social workers and other medical personnel as to the involvement, influence, and competence of medical social workers in ethical issues. 4. The adequacy of and need for ethics training and education for social workers. Table 3. Descriptive research literature on hospital social work ethics. Study (by date) Grady, et al. (2008) O'Donnell, et al. (2008) Boland (2006) Guo & Schick (2003) Jansson & Dodd (2002) Landau (2000a, 2000b) Hoffmann, Tarzian, & O’Neil (2000). Csikai & Sales (1998) Csikai (1997) Walsh-Bowers, Rossiter, & Prilleltensky (1996) Clemens (1995) Foster, et al., (1993) Proctor, Morrow-Howell, & Lott (1993) Kugelman (1992) Walden, Wolock, & Demone (1990) Joseph, & Conrad (1989) Hospital Social Worker Sample 83 in 4 States 74 in 4 States (same sample as Grady, et al., 2008) 239 nationwide 13 on ethics committees nationwide 162 at 37 hospitals in 1 State 18 at 14 hospitals in Israel 192 on ethics committees in 58 hospitals in 1 State 85 (mostly dept. directors) in 1 State 65 on ethics committees at 143 hospitals in 1 State 14 at 1 hospital in Canada Dependent Variables Ethical Competence Ethical Stress, Moral Action, Job Satisfaction Ethical Reasoning Success of Ethics Committees Ethical Activism Ethical Decision-making Ethics Committee Competence Ethics Committee Involvement Ethics Committee Participation Ethical Climate Perceptions 4 at 1 hospital in 1 State 255 in 7 States 16 in 1 State Ethics in Discharge Planning Ethics Education and Training Ethical Dilemmas 10 in 1 State 75 at 9 hospitals in 1 State Ethical Decision-making Ethical Decision-making 123 at 15 hospitals in 3 States Ethical Influence 46 The nature of ethical dilemmas. Among hospital social workers, the most frequently encountered ethical conflicts arise between the social workers’ pursuit of what they deem as in their clients’ best interests (beneficence and paternalism) and fostering maximum selfdetermination (M. Abramson, 1983; Linzer, 1999; Proctor, et al., 1993). Proctor, et al. (1993) also found that dilemmas may arise from the social workers’ multiple loyalties to other professions (who had greater authority to make key decisions that affect client well-being and self-determination), and to a lesser extent to the employing organization when faced with rapidly discharging a patient in line with the fiscal interest of the hospital when more time is needed. Hospital social workers are on some level engaged in trying to balance the demands of both clients and their employing organizations (Cummings & Cockerham, 1997; Gregorian, 2005; Walden, et al., 1990). Other factors mentioned that complicate ethical dilemmas in discharge planning are the numbers of people involved (from providers to family), which may include dilemmas about the way patients are treated by physicians and nurses (Proctor, et al., 1993). Authors on the nature of ethical dilemmas suggest that social workers need more ethics education, especially as it relates to personal, professional, and organizational dynamics. The ethical decision-making process. Some research has investigated the ethical decision-making process of social workers, including those based in hospitals, typically using questionnaires or interviews to explore how social workers contemplate ethical issues and attempt to arrive at resolutions. Two studies were not specific to hospital social workers but seem to have included them in the sampling frame (Doyle, et al., 2009; Holland & Kilpatrick, 1991), while others have been specific to hospital social work either in the U.S. (Walden, et al., 1990) or in Israel (Landau, 2000a, 2000b). Several studies use a vignettes methodology where participants are given descriptions or case studies of ethical issues or situations, and then 47 complete a questionnaire, interview, or both, asking about what they would do in that situation and why (Boland, 2006; Doyle, et al., 2009; Kugelman, 1992). Most authors readily acknowledge that this is a limitation to the study design, as the participant responses may reflect what they should do more so than what they would do in a practice situation. Doyle, et al. (2009) tried having participants specify between what they should do, versus what they actually would do in a given situation. They found discrepancies between what social workers say they should and would do, but could not distinguish a clear pattern in those discrepancies. In some cases, the differences are related to professional factors such as a level of commitment to social work values, while in others the discrepancies are related to personal factors such as ethnicity and religion (Doyle, et al., 2009). The investigations into the decision-making process of social workers have several dimensions that both inform and raise questions about ethical practice. Holland and Kilpatrick (1991) found that social workers vary in the locus of authority they rely on in ethical decisionmaking, ranging from internal (relying on individual, personal judgment) to compliance with external rules. This held true for hospital-based social workers as well, wherein the personal orientations of the practitioner and compliance with the rules and authority-bases of the hospital always came first (Kugelman, 1992; Landau, 2000a; Walden, et al., 1990). This does not mean that social workers are avoiding ethical dilemmas, as they are quite willing to take action and engage in ethical deliberations (Walden, et al., 1990). The issue seems to be more that hospital social workers avoid any extreme position or absolute dedication to any personal or professional ethics, instead preferring a compromise, middle-range option that allows them to be ethically responsive, but within the context of conflicting personal, professional, organizational, and societal demands (Kugelman, 1992; Landau, 2000a; Walden, et al., 1990). The social workers 48 themselves may not be able to identify these conflicts clearly, but they are keenly aware of power differentials in the hospital, and know they can rely on the rules and regulations (Doyle, et al., 2009; Kugelman, 1992). The interdisciplinary nature of ethical decision-making in the hospital setting is also evident in the work of Landau (2000b) in Israel. Like Joseph and Conrad (1989), Landau (2000b) found that ethical decision-making in the hospital is an interdisciplinary process, but how much social work is involved in and contributes to this depends on a number of organizational and professional environmental factors: primarily the cooperation of physicians, and the lack of clarity among the other professions as to what social work does and can do. Landau and others (Csikai, 1997; Csikai & Sales, 1998; Furlong, 1986; Silverman, 1992) comment on how the social work contribution to ethical deliberations and decision-making can be unique and especially beneficial given the professional perspective, knowledge, and skills of social workers. In addition to the biopsychosocial (person-in-environment) perspective, Grady, et al. (2008) also found that social workers typically have more training and education in ethics than nurses, and tend to be more directly involved in ethics issues in the hospital. Social workers report that they do feel prepared to deal with ethical dilemmas because of their training and education (Foster, et al., 1993); however, it seems there is still a need to make the unique contributions of social work explicit and improve communication with the medical professions (Landau, 2000b). Both the general ethics research and that specific to hospital social workers come to the same conclusions about social workers and ethical decision-making. First, ethical decisionmaking begins with the individual personal feelings of the practitioner, which is followed immediately by compliance with rules and regulations (Boland, 2006; Doyle, et al., 2009; Grady, 49 et al., 2008; Holland & Kilpatrick, 1991; Jansson & Dodd, 2002; Mattison, 2000). In this context, social workers are less able to analyze ethical dilemmas, are more vulnerable to personal preference, bias, and prejudice, and are influenced by organizational rules and the power of others (Joseph & Conrad, 1989; Kugelman, 1992). The research suggests that there are internalized professional values at work in the personal ethics of social workers, and that the more internalized these are, the more likely practitioners are to utilize a systematic process in dealing with ethical dilemmas (Doyle, et al., 2009; Holland & Kilpatrick, 1991). Still, the ethical decision making process of hospital social work seems prone to the effects of personal, professional, and organizational factors, and the proposed solution throughout the literature is more ethics education. Perceptions research. Some of the literature under the category of perceptions research focuses on the perceptions of hospital social workers and department directors as to the level of involvement and influence social workers feel they have in ethical decision-making (Joseph & Conrad, 1989; Landau, 2000a, 2000b; Walsh-Bowers, et al., 1996). Research here often includes additional variables of interest, such as correlations between ethical decision-making and personal values, social work values, job satisfaction, and interdisciplinary collaboration (Joseph & Conrad, 1989; O'Donnell, et al., 2008). Walsh-Bowers, et al. (1996) also delve into the Ethical Environment of the hospital (interdisciplinary and organizational factors) in their Canadian hospital study. Involvement and influence in ethical decision-making and deliberations varies along a number of factors. Joseph and Conrad (1989) explored several and found that if an individual social worker’s personal values and understanding of social work roles are congruent with the professional social work values and roles, they tend to have more influence in ethical decision-making. Social worker job satisfaction and satisfaction with their 50 interdisciplinary collaborations are also indicative of higher levels of influence. Joseph and Conrad (1989) also found it helpful if the other professions in the hospital, namely physicians and nurses, seemed to have a good and accurate understanding of social work roles and contributions. An interesting element in the perceptions research is the work of Hoffman, et al. (2000) on perceptions of competence among ethics committee members. They report that assessment, communication, support, and empathy are critical in the functions of hospital ethics committees, although these are often lacking in physicians. They also report that a foundational ethics education, including philosophical principles, morality, and decision-making schemes is necessary. However, this type of education was lacking in nearly every person and profession involved in the ethics committee. Several committees in their study required no training or education whatsoever as a prerequisite of committee membership. Although a study conducted in Canada, the work of Walsh-Bowers, et al. (1996) deserves some extra attention. Theirs is one study that truly explored the depth and breadth of the complicated and contradictory cross-disciplinary and staff-management perceptions in one specific hospital. While the social work department management felt that ethical conflicts and issues are handled well, the staff described a hostile ethical environment where social work is marginalized by the unquestioned authority and values perspective of physicians. The social workers in the study felt they were compromising their professional values, and that it was unsafe to raise questions and concerns. As to the ethics committee, although management believes it is a great resource, the line staff reports that the committee is clearly under the control and direction of physicians, and is thus underutilized. 51 The concern over an Ethical Environment potentially hostile to social work involvement in ethical issues in the hospital is touched on by other US-based research as well. Jansson and Dodd (2002) report that social workers feel the hospital atmosphere is relatively positive in that they are allowed to question ethical resolutions or dilemmas with specific patients, and that sufficient discussion is devoted to ethical issues. The social workers do not feel that they are systematically excluded, overall, but perhaps on a case-by-case basis, and depending on a host of other factors, such as how difficult the case is, how large and visible the social work department is, and how supportive the organization is of social work (Jansson & Dodd, 1998, 2002). The bottom line is that social workers are the most active and involved in ethical issues in hospitals where there is the least opposition (or most receptivity) to begin with (Jansson & Dodd, 2002; O'Donnell, et al., 2008). The perception research arena again highlights personal, professional, and organizational influences, and suggests social workers increase their visibility and selfadvocacy, as well as the need for excellent social work ethics education. Ethics education and training. Research in this area focuses on social workers’ and researchers’ perceptions of the adequacy and impact of social work training and education in ethics, some exclusive to hospital social workers (Foster, et al., 1993; Jansson & Dodd, 2002), and others that likely included hospital social workers (Csikai & Bass, 2000; Grady, et al., 2008; Kaplan, 2006; Manetta & Wells, 2001). Kaplan (2006) found that the two most predictive factors of moral reasoning are having an undergraduate degree in liberal arts, rather than in social work, which indicated higher levels of moral reasoning, and having had a discrete ethics course in graduate school, which actually correlated with lower levels of moral reasoning. Even though undergraduate education and an ethics course are the most predictive here, they only accounted for five percent of the variance in level of moral reasoning, so there are still other unaccounted 52 for factors. There are conflicting findings about the impact of education and training, in part because each study uses a different dependent variable, such as the ethical reasoning, ethical activism, ethical confidence, ethical participation, or the ethical influence of social workers (Table 3.). As to ethical reasoning, Manetta and Wells (2001) found that social workers with ethical training (in school and after graduation) are slightly more thoughtful in their ethical reasoning and tend to come to ethical conclusions more in line with social work values and ethics. The small improvement in ethical reasoning from training and education is confirmed by other research as well (Boland, 2006; Csikai, 1999; DiFranks, 2008). As for ethical activism, involvement, and participation, there is a difference between discrete ethics courses as part of social work education, and training received after graduation, most often in the form of continuing education seminars. In earlier research, continuing education was not at all helpful (Joseph & Conrad, 1989), but more recent studies find that continuing education has a slightly greater effect than university–based instruction in improving actions, activism, involvement, and participation (Doyle, et al., 2009; Grady, et al., 2008; Jansson & Dodd, 2002; O'Donnell, et al., 2008). The research nearly always concludes that more ethics education and training, and more effective ethics education and training are needed for social workers. Without adequate ethics instruction, ethical decision-making tends to be influenced by potentially biased personal values (Csikai, 1999; Doyle, et al., 2009; Joseph & Conrad, 1989). Social workers seem to have more training in ethics than other professions, such as nurses (Grady, et al., 2008), but often request even more (Bronstein, Kovacs, & Vega, 2007; Csikai & Bass, 2000; Foster, et al., 1993) which researchers state is necessary (Csikai, 1997, 1999; Csikai & Bass, 2000; Doyle, et al., 2009; 53 Foster, et al., 1993; Grady, et al., 2008; Landau, 2000a). In particular, further and better education on ethical theory, decision-making models and processes, and the separation of personal values from professional practice is suggested (Csikai, 1999; Csikai & Bass, 2000; Cummings & Cockerham, 1997; Doyle, et al., 2009; Grady, et al., 2008; Landau, 2000a). Of special note in the perceptions research is the work by O'Donnell, et al. (2008). This is the same social worker sample of study as Grady, et al. (2008), and as Ulrich, et al. (2007) from the Ethical Environment literature (Table 1). Whereas Ulrich, et al, (2007) focused primarily on nurses, Grady, et al. (2008) focused on social workers and education, while O'Donnell, et al (2008) focused on ethical stress, moral action, and job satisfaction among social workers. O'Donnell, et al., found that the more supportive the ethical climate, with more ethics resources (such as an ethics committee), the lower the ethical stress of social workers, the more likely they are to engage in moral action, and the higher their job satisfaction. Research Gaps in Hospital Social Work Ethics Education. As noted in the literature on ethical reasoning, social workers are not explicitly utilizing any of the formal ethical decision-making models (Doyle, et al, 2009; Holland & Kilpatrick, 1991; Joseph & Conrad, 1989; Kugelman, 1992; Mattison, 2000; Walden, et al., 1990). This needs to be explored in detail as to why, and what education or training might improve ethical decision-making (or at least move decision-making away from personal beliefs or biases). Clearly, there needs to be more research on ethical practices and deliberations in the hospital context, and more needs to be done to determine the amount, type, and most effective mode of social work training and education in ethics. Training and education also needs to include more on aspects of interprofessional communication in order to increase social work skills in the information exchange necessary for participation in ethical issues, and for 54 demonstrating the value of the contributions that social work makes (Davidson, 1990; Joseph & Conrad, 1989; Landau, 2000b). The current requirement that ethics content be infused throughout the social work educational curriculum attests to the importance of ethics education, but does not specify exactly what should be taught or how (Goldstein, 1998; Sanders & Hoffman, 2011). Social work education may need to consider the needs of hospital social workers explicitly given the unique nature and setting of the practice (Keigher, 1997). The research to date, although calling for additional and/or new approaches to social work ethics education, and having delineated something as to the content needed, has not clarified how this is to be achieved. The weaknesses in the research undermine the call for greater social work ethics education, and render the suggestions as to how that education should be implemented effectively as speculative. Personal, professional, and organizational factors. There is some implicit use of the profession's values, so long as the individual social worker's values are congruent with the profession's and are understood, accepted, and respected by the hospital and the medical staff. This raises questions about the issue of personal, professional, and organizational values congruence, and suggests that organizational variables, including the Ethical Environment of the hospital, may be barriers and sources of conflict for social workers in hospital ethics (Csikai & Sales, 1998; Jansson & Dodd, 1998, 2002; Landau, 2000b; O'Donnell, et al., 2008; WalshBowers, et al., 1996). The difference between the socialization of the other professions in the hospital, and the disparate status and authority levels they hold are other aspects of the environment that need to be considered (Abramson & Mizrahi, 1996; Cowles & Lefcowitz, 1992; Csikai & Sales, 1998; Kadushin & Kulys, 1995; Landau, 2000b). These contextual, organizational environment factors are an element that is not well accounted for in any ethics 55 training or education (Jansson & Dodd, 1998), although there is the suggestion that the knowledge and skills of social workers in communication, mediation, and advocacy could be quite useful in dealing with the environment and status issues (Auerbach, et al., 2007; Csikai & Sales, 1998; Landau, 2000b; Nelson & Merighi, 2003). It is interesting to note that the literature seems to suggest that it is always the responsibility of social work to address these organizational environment issues, which seems to belie the concomitant status issues. Summary and Theoretical Model of Research on Ethics in the Hospital Setting In each of the identified themes (dilemmas, decision-making, perceptions, and education), there are elements that suggest the Ethical Environment at work. Personal, professional, and organizational variables are commonly mentioned throughout the hospital social work ethics literature, but they seem to be poorly understood with conflicting findings. The organizational variables in particular have not often been systematically investigated. None of the research combines exclusivity to hospital social workers with representative samples. The inconsistent dependent variables (Table 3) and conflicting findings about independent personal and professional variables leave doubt as to what elements are important. The findings about independent variables include many of the same variables of interest noted in the Ethical Environment literature. The hospital social work ethics literature suggests that race and religiosity appear to be related to outcomes (Csikai, 1999; Doyle, et al., 2009), although O'Donnell, et al. (2008) found no differences by race. Results on gender are inconclusive (Doyle, et al., 2009; O'Donnell, et al., 2008). Formal social work education seems helpful (Boland, 2006; Csikai, 1999; Doyle, et al., 2009; Foster, et al., 1993; Joseph & Conrad, 1989), although O'Donnell, et al. (2008) didn't find this to be so for those with ethics classes or higher degrees (BSW vs. MSW). Continuing education training both internal and external to 56 hospitals has been positively related to outcomes (Csikai, 1999; Foster, et al., 1993; Grady, et al., 2008; Jansson & Dodd, 2002; O'Donnell, et al., 2008), but not by all (Joseph & Conrad, 1989). Tenure in the form of years of service in the social work profession may be related (Boland, 2006; Csikai, 1999; Jansson & Dodd, 2002) or it may not (Foster, et al., 1993; O'Donnell, et al., 2008). Years of service within the hospital may also correlate (Jansson & Dodd, 2002) or it may not (Boland, 2006; O'Donnell, et al., 2008), while others find that age and tenure seem unrelated to ethical outcome variables when you consider education (Grady, et al., 2008). On the level of the frequently mentioned but rarely measured organizational variables, it seems the size of the social work department is not related to ethical outcomes (Jansson & Dodd, 2002), but service on an ethics committee is related (Foster, et al., 1993). There also seems to be theoretical support for considering organizational constraints arising from the nature of the practice (discharge planning and psychosocial), and from concerns about communication and collaboration. Kadushin and Kulys (1995) looked at these types of organizational constraints as elements predictive of job satisfaction. However, O'Donnell, et al. (2008) found no differences in ethical stress, actions, or job satisfaction by the status of the hospital as for-profit or nonprofit. As for other dependent or outcome variables, they include a host of what can be summarized as the ethical "actions" of social workers (Table 3), such as ethical reasoning or decision making (Clemens, 1995; Kugelman, 1992; Walden, et al., 1990; Proctor, et al, 1993; Landau 2000a, 2000b), confidence and moral action (Grady, et al., 2008; O'Donnell, et al., 2008), ethical competence (Boland, 2006), ethical activism (Jansson & Dodd, 2002), ethical preparedness and participation (Foster, et al., 1993), and influence in resolving ethical dilemmas (Joseph & Conrad, 1989). Some of the same actions can be applied to ethics committees as well, 57 such as committee competence, involvement, and participation (Csikai, 1997; Csikai & Sales, 1998; Hoffman, et al., 2000). Based on the literature a theoretical model representing the potential relationships is presented (Figure 3). Note that in this model, due to a lack of definitive supporting information, job satisfaction is considered an independent personal variable contributing to the ethical actions of social workers. In addition, Ethical Environment is listed under organizational variables to 1. Personal Variables: Race, Age, Gender, Religiosity, Years of Practice, Job Satisfaction 2. Professional Variables: Ethics Education, Training, and Committee Service Social Workers' Ethical Actions 3. Organizational Variables: Nature of Practice, Communication and Collaboration, Department Size, Ethical Environment Figure 3. A Theoretical Model of Social Work Hospital Ethics Literature denote the vague suggestion in the social work literature that organizational variables need to be explored, and to allow from some comparison with the work of O'Donnell et al. (2008). The literature on hospital social work ethics and the literature on hospital Ethical Environment suggest that the two have similar or overlapping concepts. In each body of literature is a series of common or related personal, professional, and organizational explanatory variables, all of which have shown some degree of relationship with a number of outcome variables. The two primary outcome variables are ethical actions and Ethical Environment. Combining the Ethical Environment and social work ethics theoretical constructs (Figure 2 and 58 Figure 3) suggests a fourth theoretical model (Figure 4). In Figure 4 the list of variables have been omitted under the category headings for simplicity of presentation. 1. Personal Variables 2. Professional Variables 3. Organizational Variables B A Social Workers' Ethical Actions Perceptions of Ethical Environment C Figure 4. A Combined Theoretical Model: Social Work Hospital Ethical Environment The model in Figure 4 is the basis for the research questions at issue in this dissertation. Given the existing social work literature on the relationships between personal, professional, and organizational variables and ethical actions (line A in Figure 4), and the difficulties in measuring ethical actions, this area is not the focus of the current study. With the apparent importance of Ethical Environment for hospital social work practice, and the lack of social work literature on Ethical Environment, the decision is to focus on the relationships suggested by line B in Figure 4, with the three variable categories as predictors, and Ethical Environment as the response variable of interest. Exploring the relationships suggested by line B for social workers can be compared to the Ethical Environment research conducted with nurses, with a few indications available from O'Donnell, et al. (2008) as well. Explanations of line C relationships are present in the nursing and Ethical Environment literature as well, and there is the single study by O'Donnell, et al. (2008) that indicates some of this relationship for social workers, so commentary on the relationships of line C for social workers will be possible. Detailed 59 investigations of line C (and to some extent definitive findings for line A) will be left for future research. The investigation of the line B relationships will also attempt to address the lack of a representative sample of hospital social workers, which in turn, might settle some of the contradictory findings noted above in the Ethical Environment literature (and supported by the line A social work research). Although no hypotheses were offered as part of this dissertation research, this review of the literature suggests that there are a number of issues, barriers, and complications to the Ethical Environment of hospital social workers. A tentative hypothesis might then be that hospital social workers would be more critical in their perceptions of the Ethical Environment, although O'Donnell et al. (2008) found that social workers rated the ethical climate of hospitals favorably. Of particular importance are the independent professional variables related to ethical reasoning, education, and training, as well their interaction with personal variables, since the interference of the personal with the professional is a noted concern in the literature. The role of the organizational variables may indicate what policies hospital social workers and management should promote and implement to improve Ethical Environment, and theoretically, subsequent ethical actions by organizational members. 60 III. Methodology In order to explore perceptions of Ethical Environment, answer the research questions, and to capture a representative sample exclusive to hospital social workers, a number of methodological challenges needed to be addressed. Reaching the social workers and surveying a large, geographically dispersed sample by traditional methods (a mass mailing campaign), was considered too time consuming and expensive. Thus, a web-based survey was conducted, which came with its own methodological quirks. Nevertheless, informational resources were available to design and implement the study within a restricted period of time and with minimal resources. The details of the sampling, design, measures, process, and data analysis are as follows. Sampling Frame There are currently 3,558 non-federal hospitals in the U.S. (American Hospital Directory, n.d.). Federal (Military and Veterans’ Administration) facilities operate differently with regard to social work services, so are excluded from the sample. According to the U.S. Department of Labor, Bureau of Labor and Statistics (2006), there are 38,194 social workers employed in nonfederal hospitals nation-wide. A random sample of approximately 1000 social workers would provide a 95% confidence interval, with a margin of error of ± 3% (Sue & Ritter, 2007). However, the number of social workers in any given hospital is not readily available, so the sampling frame needs to be approached by an alternate method. Starting with the number of hospitals, and considering the size of those hospitals allows for some calculations. Small rural hospitals often do not have social workers, tending to provide ad hoc social services via other hospital staff. Although excluding them might bias the sample toward social workers in large, urban hospitals, these facilities would also tend to have a standard social work service model. By selecting a cluster of hospitals with more than 200 beds, the number of 61 hospitals remaining in the sample is 1404 (American Hospital Directory, n.d.). Assuming at least two social workers at each hospital completed the survey, the requirement for adequate sample size would easily be met. In fact, a random sample of 500 of those 1404 hospitals would likely be sufficient. This represents a hybrid approach to sampling, with both random and convenience elements. First is the random closed population cluster sample of hospitals (Sue & Ritter, 2007), which can also be examined for regional (Census Region and Division) differences, provided there is a sufficient nation-wide response rate. As to the population of hospital social workers, given that they will then self-select into the survey, this would be considered a convenience sample (Schonlau, Fricker, & Elliott, 2002), although the representational weaknesses associated with a convenience sample are attenuated by other elements of the survey design and process (Dillman, 2007). Unfortunately, there is not one comprehensive and easily accessible list of all the information needed about the hospitals within this sampling frame. Thus, a list was constructed by combining data from a number of sources. First, the American Hospital Association (AHA) Guidebook (American Hospital Association, 2010) was consulted for an initial list of hospitals with 200 beds or more. This list was then collated with information from the American Hospital Directory (AHD) web site (www.ahd.com), the Hospital-Data hospital and nursing home profiles web site (www.hospital-data.com), and from the Baby Boomer Caretaker web site (www.babyboomercaretaker.com/hospitals/). Each source provided different details about the hospitals, such as ownership, profit status, number of beds, combination hospital systems, and occasional reports as to the number of social workers employed at the hospital. At times, the information was conflicting or absent from one or more sources. In the end, a comprehensive list of the 1400 hospitals within the sampling frame was constructed. This list was randomized and 62 the first 500 hospitals were selected for inclusion. The web sites of each of the selected hospitals were reviewed for contact information for the social work department, or, if available, for the director, manager, or supervisor of the department. Phone numbers were recorded, and if no specific information was available, the general contact number for the hospital was noted. During data collection, each hospital was telephoned directly to request participation in the study. Design and Procedures Given the size and geographical dispersion of the desired sample, the study was implemented via the World Wide Web. This is still a somewhat experimental survey medium with a limited amount of research supporting its effectiveness (Dillman, 2007; Schonlau, et al., 2002; Sue & Ritter, 2007). In general, it seems that there are differential responses between email, web, and mail surveys, but the differences are too inconsistent to draw any conclusions (Schonlau, et al., 2002). However, there is enough survey design information to extrapolate techniques to the web environment, and some research to suggest an expected response rate of approximately 30% (Sue & Ritter, 2007). The design and process is based on the Dillman (2007) Tailored Design Method, with some modifications to facilitate a web-based survey methodology based on web survey design resources (Schonlau, et al., 2002; Sue & Ritter, 2007). The most important modification for web-based surveys is brevity. Multiple studies have shown that long, complex surveys in the web environment have very high dropout rates (Schonlau, et al., 2002; Sue & Ritter, 2007). This caveat will inform the design of the study throughout, and was a major point in the decision to focus on the relationships between the independent variables and Ethical Environment (Figure 4, line B). Although the need exists for a representative sample in the ethical actions relationships (Figure 4, line A), the vignettes 63 methodology does not translate well to a web survey. Since the Ethical Environment to ethical actions relationship (Figure 4, line C) has never been explored in the social work research, to do so here would require a very lengthy survey. In addition to overall research questions, the texts of all communications (initial invitation, welcome and consent letter, and reminder messages) were informed by the social exchange theory supporting survey design and web survey design (Dillman, 2007; Schonlau, et al., 2002; Sue & Ritter, 2007). This design approach attends to the specific wording and content of communications such that they include key terms that encourage participation and response, such as thanking participants and communicating to them the value and necessity of their participation. The Dillman (2007) method was also used to inform and revise all survey questions for clarity, specificity, and brevity. The study design and processes were reviewed and approved by the Michigan State University Human Research Protection Program Institutional Review Board (IRB) as exempt (IRB# X10-628). The social work department manager, director, supervisor, lead social worker, or in the case of decentralized departments (where social workers are not part of a singular social work unit, but assigned to medical units and report to management for the medical units themselves), a "point person" social worker was the point of contact. Occasionally, a department administrative assistant functioned as a contact between the investigator and the management. If an individual was reached, that person was provided a brief introduction to the investigator and a brief description of the study. The method for participation was also described as on online survey for which an email invitation, or "evite" would be sent (see Evite Text, Appendix A). Utilizing an email invitation to a web survey can increase responses (Sue & Ritter, 2007), in part due to the 64 individual invitation and direct communication. If the person contacted was willing to participate, that person provided a single email address for the evite. The evite contained a link to the survey site, the password to access the web survey site, and could simply be forwarded to all the social workers at the hospital. Due to IRB concerns about anonymity, a single password was used for all participants in the study. Neither the participants nor the hospital at which they worked could be identified from participation or the data collected. Maintaining this level of anonymity is preferable in a web survey, especially when potentially sensitive questions are asked (Sue & Ritter, 2007). Since this approach also precluded tracking response rates, a single reminder email was sent to the contact person at every hospital one week (seven days) after the initial evite (see Reminder Text, Appendix A). Near the end of the study, the reminder email was modified to warn participants of the time limit (Modified Reminder Text, Appendix A). The single reminder strategy was supported by the survey design literature, which shows that additional reminders can increase responses only by very small increments (Sue & Ritter, 2007) If necessary, the brief description of the study and how to participate was left as a voicemail message. The message included phone and email contact information for the investigator and participants were encouraged to email. If there was no response to the message, the hospital was called again one week (seven days) later. The study description, invitation, and method of participation were repeated to the contact or again left as a voicemail message. There were no further contact attempts after the second message. Although multiple contacts above the two attempted here are recommended for some survey methods (Dillman, 2007), there is a diminishing return for these efforts after two contacts, so two was deemed sufficient in light of time and resource limits. This approach was frequently successful in obtaining email addresses 65 and sending evites. Occasionally a contact person or managerial staff wanted to asked questions, or preview the survey before agreeing to participate, which was permitted. The web survey was hosted on the commercial service “Surveygizmo” (surveygizmo.com). A number of free and commercial web survey software services and packages were reviewed and Surveygizmo was selected based on the combination of features needed to complete this study, and cost considerations. All completely free services lacked necessary features, such as ease of use, data security, and compatible download formats. Software to purchase for ownership and server install by the department or college would likely be in excess of $10,000. Surveygizmo presented a balance of cost (at $50.00 per month), with necessary features. The Surveygizmo system hosted and sent the initial evite to the respondents, and automatically included the direct URL link to the study survey embedded in the evite, and the password to enter the site. The survey site link connects respondents to a welcome screen that includes a preemptive thank you for participating, asks them to enter the password provided, and instructs them on how to proceed from there (Dillman, 2007). Requiring a password prevents the uninvited from completing the survey, and helps increase the validity of the sample. After entering the password, the following screen is a combination welcome and consent letter (Online Consent Form Letter, Appendix B). After ticking the “I agree” box, participants can then click on the “Start Survey” button and advance to the first page of the survey. If participants click on “I do not agree to take this survey,” they will be directed to a confirmation page asked them to confirm that they do not want to participate, or allowing them to state that they meant to agree and to continue with the survey. Those who confirm that they do not want to participate are forwarded to the end screen thank you page. 66 In keeping with the web survey design specifications for improving participation and response, the designs of the web pages are simple and direct, with high contrast text and background, and with a minimal but attractive color scheme (Dillman, 2007; Schonlau, et al., 2002; Sue & Ritter, 2007). There are no graphics on any of the pages. Each page also includes a “report problems” button, in case of any malfunctions with passwords or any part of the questionnaire, as well as a “progress bar” indicating how much of the survey is completed. There is a single error message if a question is not answered when the continue button is clicked. The error message states, “A question was not answered. Please fill in the answer, or click the next button to continue.” If the unanswered question is on the scale instrument, the message reads, “A question was not answered. Please make sure you intended to leave this question blank before clicking the next button.” Aside from the entry of the password and consent agreement, no questions required an answer in order to proceed after the error message, as this is contrary to both participant rights and effective web survey design (Schonlau, et al., 2002). Incomplete survey reports were reviewed for patterns or problematic areas of the questionnaire (see process section, below). Questions are grouped together by theme, and in small numbers that will likely fit on a single screen to avoid the need for ‘scrolling’, which can increase respondent fatigue (Dillman, 2007; Sue & Ritter, 2007). Categorical questions have radio buttons with some text boxes for write-in answers. Single rating scale questions have radio buttons along the spectrum of the scale with the scale categories directly above the buttons. Scale questions that form the instrument are presented as a matrix question with the rating scale and question stem visible in the header row on each page. After the consent form page, the set of questions about education 67 and ethics education are asked (Survey Questionnaire, Appendix C), followed by the dependent variable instrument (explanation below). The next question set is about job tenure, and then department and organizational context, followed by job satisfaction and then sociodemographics. The next page asks if participants are willing to be contacted via email for follow-up questions as an option to include anecdotal or qualitative information in the study (Follow Up Contact, Appendix C). The follow-up contact page includes text explaining that the email address will not be attached or connected to survey responses. The following page asks if participants wish to enter an email address in order to be entered into the lottery for the electronic gift certificates (Gift Cards, Appendix C). A drawing for one of twenty $10 Amazon.com electronic gift cards is offered as an incentive to increase response rates. The use of incentives is known to be effective in increasing response rates; however, the best strategy is to provide the incentive first, before completing the survey (Dillman, 2007). This is not possible in a web survey setting and there is little definitive research on incentives in web surveys, other than to note that they most likely do increase participation (Schonlau, et al., 2002; Sue & Ritter, 2007). A generic, easily accessible, and broad use incentive such as Amazon.com gift certificates are recommended in geographically dispersed samples (Sue & Ritter, 2007). As with the follow-up contact, text here also advises that email addresses are not connected to responses. The final page thanks respondents for their participation. Measures Dependent Variable and Instrument: Ethical Environment After a careful review of available Ethical Environment instruments (see Table 1), the Ethical Environment Questionnaire (EEQ) by McDaniel (1997) was selected. The EEQ has 68 previously been used to explore the Ethical Environment of hospitals, although its development was tested only on nurses. Other previous research with the EEQ has focused on the hospital Ethical Environment as rated by nurses (Corley, et al., 2005; McDaniel, 1998), nurses and physicians (Hamric & Blackhall, 2007), and on the corporate Ethical Environment as rated by employees of a large firm (McDaniel, et al., 2001). The EEQ was also used in a sample of nurse practitioners about the Ethical Environment of their primary care setting (Ulrich & Soeken, 2005). The EEQ has not previously been utilized with social workers. An alternate instrument (the Hospital Ethical Climate Scale or HECS) was used in a large sample of nurses (Ulrich, et al., 2007), which included a small sample of hospital social workers. This is the same instrument and hospital social worker sample utilized by O'Donnell, et al. (2008), and the sample was drawn from only a few states with available social work licensing lists. The EEQ measure required the least modification for use with social workers (simple editing will suffice), as well as for use in a web-hosted survey environment. It is also a measure of the shared perceptions of Ethical Environment of the hospital as opposed to specific dimensions of ethical climate, while at the same time offering comparable findings, as explained in the literature review. In addition, the author and copyright holder of the EEQ granted permission for its use for this dissertation, provided steps are taken to protect the copyright, such as not producing the questionnaire here. Attempts to contact the copyright holder of the HECS (Olson, 1998) were unsuccessful. The copyright and use agreement for the EEQ requires that the instrument is not reproduced in this dissertation. The instructions on the EEQ ask respondents to click on the button that represents their agreement or disagreement with the statements, and that there are no right or wrong answers. The EEQ consists of 20 questions, all rated on a five-point Likert scale from 1 = Strongly Agree, 69 to 5 = Strongly Disagree. Several items are reverse scored. The question stem is "Regarding this hospital..." and example questions include "The administration is concerned with ethical practice" and "Ethics accountability is not rewarded." There are no subscales such that the scale produces a single average value per respondent to measure Ethical Environment. The previous studies report good psychometric properties, with a test-retest Pearson's r of 0.88 and Cronbach’s alphas of 0.90 to 0.94 (Corley, et al., 2005; McDaniel, et al., 2001; McDaniel, 1997; Ulrich & Soeken, 2005). McDaniel (1997) addressed content, construct, and criterion validity and reported that the EEQ has a single factor structure on exploratory and confirmatory factor analysis. The single factor structure was confirmed with the current sample, along with good internal consistency reliability, both reported in the results section. Independent Variables Independent variables were derived from the literature review and include those variables that might have a relationship with Ethical Environment. Under the category of professional variables, information about participants' ethics education included highest degree earned, having taken an ethics course or a course with ethics content during their professional education, and having participated in ethics training within the past two years (Survey Questionnaire, Appendix C). Ethics training was specified as continuing education seminars, both within and outside of the employing hospital. Participants were also asked if they had ever served on a hospital ethics committee, and if that committee was at their current hospital. A single question Likert-type measure anchored by "strongly agree" and "strongly disagree" of how well participants felt their professional education prepared them to address ethical issues in practice was also included (adapted from Deshpande, et al., 2006). 70 After completing the EEQ, the next set of questions pertain to personal work and organizational variables, and include years of service in the hospital social work profession and in the current hospital of employment, respondent age, and full time or part time status. Primary job function asks about whether they are providing discharge planning services or psychosocial assessment and support. Notably, this is a forced choice question between only these two options, with an optional "other, please specify" answer (Survey Questionnaire, Appendix C). This was a conscious strategy designed to have the social workers choose one as "primary," although in many hospitals they may be expected to attend to both functions. Research suggests that this type of question design can increase the validity of participant responses by forcing them to carefully contemplate their answers (Dillman, 2007; Sue & Ritter, 2007). They were also asked if their primary assignment was in an inpatient or outpatient service, and again, this was a forced choice of only one as primary, with the "other" option. Participants were asked to define the social work department as either centralized (all social workers in one department reporting to one supervisor), or decentralized (social workers reporting to separate departments or units), and to specify if their immediate supervisor was a social worker, nurse, or other. They were asked to identify their hospital as either for-profit or non-profit, and to indicate the size of their social work staff, both inpatient and outpatient, at the hospital within specific ranges. The range of options for size of social work staff started at one to five, with mutually exclusive categories in increments of five up to 16 to 20, with a final category of more than 20. It may have been better to ask for a set number of social workers, however due to IRB concerns about the ability to identify the hospitals in the sample, the ordinal question was required. Respondents also identified in which State their hospital was located, and provided a self-reported number for how many inpatient beds the hospital has. 71 As part of the personal variable set, a single-question Likert-type measure anchored by "strongly agree" and "strongly disagree" about general job satisfaction was asked, adapted from existing literature (Kadushin & Kulys, 1995; Siefert, Jayaratne, & Chess, 1991; Trevino, et al., 1998). This single-item measure has been used extensively in the job satisfaction literature, and has been determined by meta-analysis to be an acceptable, valid, and reliable measure of the job satisfaction construct (Cunningham & Sagas, 2004). Another similarly constructed singlequestion measure was asked about the social workers' intent to leave their position or hospital within the next six months (this question was subsequently dropped from the analysis due to confounding correlations with job satisfaction). These questions are followed by the sociodemographic questions of gender, sexual orientation, and race. The final question asks participants to indicate approximately how often they attend religious services of any kind, with the ordinal categories of never, annually, monthly, and weekly as response options. This question is used as a proxy measure of how religious the participant is (religiosity). There are a number of multiple-item instruments designed to measure various dimensions of religiosity (Hall, Meador, & Koenig, 2008), and no clear consensus on the best measure. The single-item question of religious attendance has been used extensively, and has been found to demonstrate strong associations with numerous outcomes (Mahoney, Pargament, Tarakeshwar, & Swank, 2008). It is believed to be a measure of something unique, real, and significant, although a thorough interpretation of meaning is difficult with a single question. In the cases of job satisfaction and religiosity, the single-question measures have their weaknesses; especially with such multidimensional constructs (e.g., for religiosity see Hill & Pargament, 2003; King & Crowther, 2004). The use of any of the more complex, multi-question 72 instruments was detrimental to the required brevity of the web survey method in this study. Some research has attempted to look at the validity and reliability of single-item self-report measures, given their importance in studies where brevity is a necessity. The conclusion drawn is that such items are legitimate proxy measures (Dollinger & Malmquist, 2009). Process and Data Collection As recommended (Dillman, 2007; Schonlau, et al., 2002; Sue & Ritter, 2007), the web survey questions, text, design and process were pre-tested by colleagues and associates representing current social workers (both hospital and not), and research experts for clarity, understanding, ease of use, completion time, and any other problems or concerns. Based on feedback from the pre-test, a number of changes were made, and a second pre-test by some of the same participants as the first was conducted. Due to the volume of telephone calls required for the recruitment process, a prepaid phone card was utilized, and the calls and evites were staggered in sets of 20 to 70 per week, depending on time available to the investigator. Staggered evites are also recommended to prevent potentially overwhelming the web service (Schonlau, et al., 2002), and to allow for observation of the data as it is incoming and submitted. Invitations and data collection began Monday, August 30th, 2010. The final email invitations were sent on November 8th and the final reminder emails were sent on November 16th. Access to the web survey site was closed on November 19th, 2010. After launch of the web survey site and observations of incoming data for the first two weeks, a pattern of respondents abandoning the survey before completion was noted. Abandonments are common in web surveys, but can indicate a need to address content and navigation steps (Sue & Ritter, 2007). A review of web survey design steps revealed that the 73 question order of the survey was incorrect, with demographics following the educational questions and before the EEQ. Demographics are considered boring questions and are best placed near the end of the survey (Sue & Ritter, 2007). The question order was therefore changed on September 15th and an immediate decrease in abandonments was observed. At the same time as the review of question order, an additional error was noted in that the question asking for the number of inpatient beds had a required minimum of 250, when the sampling frame specified hospitals of 200 beds or more. To fix this and avoid any future problems, the set floor for the question was reduced to 100, with the intent of screening out ineligible hospitals during data cleaning. However, the ongoing observation of incoming data indicated that the social workers completing the survey were providing a wide range of reported bed numbers for the same hospital, so these numbers will be of dubious accuracy, much the same as the bed numbers sources identified in the sampling frame. Another interesting pattern that emerged during data collection was the response to participants being asked to identify their sexual orientation. Of course, the question could be skipped and unanswered, as 12 respondents chose to do. However, seven individuals wrote comments about this question in the "other" box, such as "none of your business," "not an appropriate question to be asking," and "very invalid question for a doctoral thesis." Two individuals also took the additional step to email the investigator to complain and/or seek an explanation about this question. They were informed that the question was included as a basic descriptive variable for two reasons: as an ethical best practice for its inclusion of populations typically unrecognized or acknowledged, and because of indications that ethical perceptions may vary by sexual orientation. Neither respondent initiated any further contact. 74 There is little research that clearly supports the use of incentives in the web environment, and incentive effects on response rates vary in such a way that the lottery conducted here may not have been effective. Still, there was evidence to suggest that the chance to win an Amazon.com electronic gift certificate could increase participation. The entry into the lottery was the second to last question on the survey (preceding the option to provide an email for potential follow-up contacts), and 649 respondents (66.7%) who completed the survey entered the lottery. It is speculative, but the high entry numbers may indicate that the lottery had some effect. Email addresses of those who entered the lottery were entered into a separate spreadsheet file, randomized, and the top 20 selected. The 20 winners were emailed their electronic gift certificates on December 6th, 2010. Although not utilized in this study, the responses to the option to be contacted for follow-up questions or clarifications after the study also resulted in a large number of willing respondents. There were 449 participants (46%) who entered their email addresses for possible follow-up contact. Data Analysis Data collected as part of the sampling frame were maintained in an Excel spreadsheet, separate from the survey data. The sampling frame data were cleaned to retain only basic information about bed numbers, profit status, and location by state. This information was uploaded into SPSS statistical software for analysis of any potential differences between participating hospitals and those that declined or failed to respond to contact attempts. This included t-tests to compare numbers of beds per hospital, and Chi-square analysis of profit status and Census Region and Census Division. Data from the survey were downloaded from the web hosting service into SPSS for analysis. As the bed numbers, profit status, and location from the sampling frame were not connected to the survey data, these two files were kept separate. The 75 participating hospital data were examined for coherence between the data from the sampling frame as compared with the self-reported survey information for bed numbers (one sample ttest). The collected survey data were examined for cleaning, coding, and other potential errors. Missing values were minimal across the majority of variables, accounting for less than one percent of all values for any given variable with the exception of the sexual orientation question (20 or 2.1% missing), and the religiosity question (14 or 1.4% missing). The SPSS missing values analysis was conducted to examine the missing data for any patterns, none of which were found among the variables collected. Little's MCAR test was not significant (χ2 (107) = 111, p = .372), so the missing values can be inferred to be missing completely at random (Tabachnick & Fidell, 2007). Given the sample size, minimal missing values, and the apparently completely random pattern, no imputations or other replacement strategies were utilized for missing values. The data were also examined for outliers that may influence inferential statistical analyses. Extreme values were also minimal, and in fact, only one total score on the EEQ was unusually low. The EEQ was examined for factor structure and reliability (Cronbach's alpha). Some analyses were run with and without detected outliers and no differences were observed in the results, so all cases were retained in testing except where the deletion of outliers improved the strength of relationships. In all analyses, post-hoc outlier influence statistics and residual plots were examined (Tabachnick & Fidell, 2007). There were no notable ceiling or floor effects on any variable. Some of the collected variables had multiple categories resulting from write-in or "other" answers. Other multiple category response variables had insufficient responses in some categories. These were examined for potential recoding where necessary to facilitate analysis 76 and to attend to the research questions. Multiple degree responses, such as high school diploma, associates, or PhD, were recoded as missing values (n = 16) in order to focus on the Bachelors/Masters comparison. Sexual orientation was recoded into either heterosexual or sexual minority due to low numbers in the more specific sexual minority identities (lesbian, gay, or bisexual), with no loss of data. Respondent race was also recoded into minority or Caucasian due to low counts, again with no loss of data. Job function was recoded into the two primary functions of discharge planning (DCP) and psychosocial assessment and support (PAS). Due to a number of managers, ER social workers, "other," and already missing values, this resulted in a working sample for this variable of 833, with 140 missing values. Recoding of the assignment of the social workers to either inpatient or outpatient hospital units also resulted in the loss of some data, again due to managers, ER social workers, and so on, resulting in a sample for this variable of 864 (109 missing). To facilitate analysis, the profession of the supervisor of the social worker was recoded into either a social worker, or not a social worker, with no loss of data. Examination of the data during submission confirmed the inconsistent and inaccurate reports of bed numbers per hospital. Given both under and overestimations, and utilizing the AHA and AHD numbers as a guide, the numbers reported were windsorized with those less than 200 rounded up to 200, and those over 1000 rounded down to 1000. Finally, the self-reported state in which the hospital is located was coded into additional variables representing Census Region and Census Division. There are four Census Regions, the Northeast, Midwest, South, and West, each of which subsumes two or three of the nine Census Divisions. The Northeast Region contains the New England (CT, ME, MA, NH, RI, and VT) and Middle Atlantic (NJ, NY, and PA) Divisions. The Midwest Region contains the East North Central (IN, IL, MI, OH, and WI) and West North Central (IA, KS, MN, MO, NE, ND, and SD) 77 Divisions. The South Region contains the South Atlantic (DE, DC, FL, GA, MD, NC, SC, VA, WV), East South Central (AL, KY, MS, TN), and West South Central (AR, LA, OK, TX) Divisions. The West Region contains the Mountain (AZ, CO, ID, NM, MT, UT, NV, WY) and Pacific (AK, CA, HI, OR, and WA) Divisions. Regions and Divisions are simply groupings of States for presenting statistical data (US Census Bureau, n.d.). They have no cultural, historical, or social significance. Bivariate relationships in the survey data were explored with Pearson correlation coefficients, paired and independent samples t-tests to compare groups, and ANOVA and Chisquare for multiple response measures. As suggested by the research model, the personal, professional, and organizational variables were entered into a linear regression model in a hierarchical or sequential block fashion (Tabachnick & Fidell, 2007), with significant predictors retained by p-values until a parsimonious solution was achieved. Attention was also given to potential interaction effects, mediating and moderating variables, and to effect sizes (Phi, Cramer's V, eta-squared, and partial eta-squared). Notably, the organizational variables suggest a different unit of measurement from the personal and professional variables, which in turn suggests analysis by hierarchical linear modeling (HLM). However, as explored in the literature review the dependent measure is one of individual shared perceptions of an organizational construct, which may serve to bridge the units of measurement. This conceptualization and the randomness of the sample suggest that the differences between HLM and the disaggregated linear regression approach utilized here are minimal. 78 IV. Results Hospital Sample and Response Rates From the original 500 randomly selected hospitals, 10 were determined to have been ineligible and were dropped from the sample and replaced by the next ten facilities on the randomized list. Seven of those hospitals reported that they employed no social workers, two of them had recently ceased operating, and one had been reduced in beds to less than 200. After the survey site was closed to further participation, 293 of the 500 hospitals approached had agreed to participate, an initial hospital response rate of 58.6%. One hospital initiated contact with the investigator over four weeks after the second message was left for that hospital, and more than two weeks after the closing date, and was not included. Three additional hospitals agreed to participate, but submitted no completed surveys (identifiable as the only facilities from their respective states), thus resulting in 290 participating, a final hospital response rate of 58%. Of the 210 hospitals not participating, the vast majority (174) simply did not respond to messages left about the study. A total of 24 hospitals declined to participate, 11 of which did not provide a reason, eight of which stated the social workers did not have the time to complete a survey, and three that cited hospital policies that forbid participation in surveys, the use of company email, or work time. Of the final two hospitals that declined, one manager felt the questions pertaining to religiosity and sexual orientation would make the staff "uncomfortable," and one manager reported that the hospital administration would require the study to be approved by their own internal IRB before the staff could participate. In addition to those that failed to respond and those that declined, eight hospitals were unreachable. Two were due to a lack of voicemail or full voicemail boxes despite multiple contact attempts, and six provided email addresses that 79 were not accessible (all email contacts "bounced" as undeliverable). In the cases of email bounces, calls to clarify email addresses were completed and the email addresses confirmed. It may be that the protective software (firewalls) of the hospital email system prevented the evites. The sampling frame data were utilized to compare those hospitals that participated with those that did not participate. A necessary caveat is that the data from the sampling frame sources was at times contradictory or missing. For hospital size as indicated by the number of beds, the available information from the AHA and AHD was examined for discrepancies. The mean number of beds of those hospitals participating is 416 by the AHA, and 421 by the AHD, whereas the mean beds of nonparticipating hospitals are 378 by both sources. For the analysis, with differences evident, the beds reported by the AHA and the AHD were averaged together (the results are the same). If only one source reported bed numbers, or if the two sources agreed, the number was retained. The mean numbers of beds for participating hospitals by this method are 419. This is a significant difference from the non-participating hospitals 378 bed average by independent samples t-test (t = 2.36(496), p = .019, two-tailed). The bed numbers as reported by social workers completing the questionnaire were also different from the numbers indicated for participating hospitals (419), with an even higher mean of 458 beds per hospital, which is statistically significant (one sample t-test; t = 5.50 (926), p = .000, two-tailed). All of the bed numbers are of dubious accuracy, but given the consistency of findings, it does seem that smaller hospitals did not participate in the study as often as larger hospitals. There was also a significant difference by profit status between participating and nonparticipating hospitals, although there is again some question as to the accuracy of the selfreport data. Profit status was only available from the AHA data source, which indicates that there are 26 for-profit and 264 non-profit hospitals participating, whereas there are 43 for-profit 80 and 167 non-profit that are not participating. A Chi-square test of association confirms these differences as significant (χ2 (1) = 13.57, p < .001), but the effect size is negligible with a Phi coefficient of -.165 (Field, 2009). An examination of the contingency table (Table 4) shows that the main significant Table 4. Hospital participation by profit status contingency table. difference is between the participation or not of the for-profit hospitals, but given the minimal effect For-Profit Non-Profit Total Participation n % n % n Yes 26* 5.2 264 52.8 290 No 43** 8.6 167 33.4 210 Total 69 13.8 431 86.2 500 * Standardized Residual = -2.2 (p < .05) ** Standardized Residual = 2.6 (p < .01) % 58 42 100 size, this may be an artifact of the large sample size. These numbers are not directly comparable to the questionnaire data, since responses are not tied to hospitals. However, examining the percentages shows that the 264 participating nonprofit hospitals would be 91% of the sample. The self-reported profit status from the social workers shows that 81.4% are non-profit (n = 966). Given the apparent lack of accuracy in the AHA data, and the lack of accuracy in the self-report data on profit status, it is difficult to draw any conclusions, other than to state that the overall results suggest that for-profit hospitals did not participate in the study as often as non-profit hospitals. In the final comparison of participating and non-participating hospitals, Census Regions are used to compare the locations of the hospitals. A Chi-square test of association is significant (χ2 (3) = 13.23, p = .004), but with a negligible effect size (Cramer's V = .163) (Field, 2009). There are slightly more participating than non-participating hospitals in the Northeast and Midwest Regions, and nearly equal participation and not in the South and West Regions. However, an examination of the contingency table and standardized residuals (Table 5) shows 81 that the only significant difference is the non-participating Table 5. Hospital participation by Census Region contingency table. South West Total Northeast Midwest Participation n % n % n % n % n % Yes 58 11.6 80 16 98 19.6 54 10.8 290 58 No 39 7.8 33* 6.6 99 19.8 39 7.8 210 42 Total 97 19.4 113 22.6 197 39.4 93 18.6 500 100 * Standardized Residual = -2.1 (p < .05) number in the Midwest Region, which is lower than expected. In all, 43 states are represented in the sample. By self-selection, there are no participating hospitals in Alaska, Arkansas, Delaware, Hawaii, Idaho, Vermont, Wyoming, or the District of Columbia. A small number of states had only one (three states) or two (three states) respondents, while some states had more than 70 (four states). The mean number of respondents per State is 22.6. Social Worker Sample and Response Rates From the 290 participating hospitals, there were 981 completed questionnaires. Eight of these respondents were nurses, not social workers, and were excluded. There were no other exclusions necessary, bringing the final number of completed questionnaires to 973. The number of missing values are minimal, as addressed in the Methodology section, and met criteria for missing completely at random (MCAR), so no imputations or other efforts were made to replace missing values. Calculating a response rate for the social workers presents some challenges. Approaching the task from national numbers reported in the sampling frame, there are 38,194 hospital social workers at 3558 hospitals in the US (not including federal or specialty hospitals). Based on these numbers, there is an average of 10.7 social workers per hospital, which makes the social worker response rate for this study 31.4% (973 / (290 x 10.7)). However, in the sampling frame, only hospitals with 200 inpatient beds or larger were selected, which may indicate a higher average number of social workers per hospital. The survey did ask 82 the participants to specify how many social workers are employed at their respective hospitals, but the question was composed of ordinal ranges, not true values. Taking the median of those ranges (and 21 for the "more than 20" category; n = 968) results in a sample average of 13.8 social workers per hospital, which is a response rate of 24.3% (973 / (290 x 13.8)). Sample Characteristics The sample of hospital social workers is primarily female at 91% or 889 (Table 6). The sample is 87% (848) Caucasian, and 92% heterosexual (891). The categories for identifying sexual orientation included gay, lesbian, and bisexual, but due to low numbers are collapsed into a single sexual minority category. Despite the apparent discomfort that this question Table 6. Sample demographic and descriptive elements (n = 973). Variable Gender Category n (a) % (b) Female 889 91 81 8.3 Male Race Caucasian 87 848 African American 48 4.9 Hispanic 38 3.9 22 2.3 Asian/Pacific Other/Multiracial 7 0.7 Sexual Orientation Heterosexual 891 92 Sexual Minority 62 6.4 Degree MSW 79 770 BSW 112 11.5 82 8.4 Other (MA, BA, PhD, JD, Other) Job Status Full Time 796 82 Part Time 172 18 Primary Inpatient 78 759 Assignment Outpatient 145 15 Emergency Department 51 5.2 11 1.1 Management Hospital Profit Non-Profit 792 81 Status For-Profit 174 18 Department Centralized 816 84 Structure Decentralized 153 16 Primary Function Discharge Planning 47 479 Psychosocial Support 446 44 53 5.4 Management 19 2.0 Other 11 1.1 Emergency Department Supervisor Social Worker 59 569 348 36 Nurse 52 5.3 Other (Admin., MD, PhD, Other) (a) Sample size may not equal 973 due to missing values. (b) Percentages may be more or less than 100 due to rounding and/or missing values 83 caused some respondents, only 20 (less than 2%) chose not to respond to the question. Table 6 also shows that the majority of the respondents, 770, or 79%, have an MSW degree, 796 (82%) work full time, and 759 (78%) in the inpatient setting. The majority, at 792, or 81%, work in a nonprofit hospital, and 816, or 84%, work as part of a centralized social work department. These social workers identify their primary job function as nearly equally split between either discharge planning (479, or 47%), and psychosocial assessment and support (446, or 44%). Most of them report that their immediate supervisor is a social worker (569, or 59%), while 348, or 36%, report their supervisor as a nurse. One note of clarification for Table 6 is that those participants in the sample that identified their primary function as management (n = 53) are inconsistent in defining their primary assignment as management as well (n = 11). The two questions are on the same page of the web survey, with primary assignment first, immediately followed by primary function. Further descriptive variables include the age or respondents and their number of years of hospital social work (social work tenure) and number of years in their current hospital (hospital tenure). The mean age of the sample is 43.5 (SD = 11.6, n = 963), and they have worked in hospital social work a mean of 11 years (SD = 9.1, n = 968), with a mean of 8.8 years (SD = 8.1, n = 969) at their current hospital. The self-reported distribution of the sample by Census Region and Census Division is presented in Table 7. There is strong representation from the Midwest and South regions at 345, or 36%, and 285, or 29% respectively, with moderate representation in the West (164 or 17%) and the Northeast (179 or 18%). Once again, the data here do not necessarily match the data from the sample frame source (AHA Guide) and in this case, there is no doubt that the errors lie in the self-report data. The distribution by Census Region and Division from the AHA guide is presented in Table 7 as well. Although the numbers are not 84 comparable given that they are based on the respondent level in the Table 7. Sample self-report and AHA guide distributions by Census Region and Census Division. Census Region West Census Division Pacific Mountain self-report column and Midwest on the hospital level in the AHA Guide column, the East North Central West North Central Northeast Middle Atlantic New England South South Atlantic West South Central East South Central percentages do show some differences. The Total (n) Self-Report n % 164 17 129 13 35 3.6 345 36 224 23 121 12 179 18 111 11 68 7.0 285 29 173 18 80 8.2 32 3.3 973 AHA Guide n % 54 18.6 41 14.1 13 4.5 80 27.6 58 20 22 7.6 58 20 42 14.5 16 5.5 98 33.8 55 19 25 8.6 18 6.2 290 overall pattern of distribution is relatively the same with a slightly higher drift in the percentage for the South and more so the Midwest. Due to small cell sizes at the Census Division level (Table 7), analyses will be conducted with the Region-level data. Dependent Variable: Ethical Environment The measures of central tendency for the EEQ are reported in Table 8. The distribution of scores is approximately normal without floor or ceiling effects. Normality tests such as Kolmogorov-Smirnov are contraindicated as Table 8. Distribution of EEQ scores (n = 972). they would all be significant given the large Mean Median Mode SD Range sample size (Field, 2009), however visual inspection of the histogram and normal Q-Q 100 point scale 73.2 74 78 12.13 22 - 100 5 point scale 3.66 3.70 3.90 0.607 1.10 - 5.00 plot (not shown) show approximately normal distributions, with a slight left-skew and leptokurtosis. The effect this may have on subsequent tests will be examined by checking the testing of each tests' assumptions. The EEQ performed well in this study with a Cronbach alpha coefficient of .93 (n = 967), and principal component analysis (PCA) demonstrating a single 85 factor structure (not shown), with an acceptable KMO value of sampling adequacy at .96, and a significant Bartlett's test of sphericity (χ2 (190) = 9417.3, p < .001), indicating sufficient correlations for PCA (Field, 2009). Research question one is answered by the mean score of 3.66 for social workers, which suggests a very positive perception of the Ethical Environment of the hospitals in the sample (McDaniel, 1997, 1998). It is a significantly higher score (one sample t-test, t = 23.6 (971), p < .001, two-tailed) than previous studies with nurses (McDaniel, 1997; Corley, et al., 2005), and with employees in a large business firm (McDaniel, et al., 2001), all of which consistently scored a mean of 3.2. Independent Variables The means, standard deviations, and Pearson's correlations for all variables by category are presented in Table 9. The majority of the significant correlations among the personal variables are logical and expected relationships. The same is true for the professional and organizational variables, as well as Census Region. For example, the correlations indicate that ethics education and training is related to age, social work tenure, and hospital tenure (reported earlier in sample characteristics section, above), and that internal ethics trainings are more likely to occur at larger hospitals with larger social work departments. Many of the correlations in Table 9, although significant, are quite small (less than .10). Additional independent variables not reported previously or in Table 9 include beliefs about professional education in preparing to deal with ethical issues. Figure 5 shows the distribution of the ordinal responses to this survey question, where the majority felt that their ethical preparation was adequate. Notably, there is no correlation between degree (MSW/BSW) and the ethical preparation question, and only moderate correlations with ethics class (r = .35) and content (r = .44) (both p < .01; Table 9). Participants were also asked about job satisfaction 86 Table 9. Means, Standard Deviations, and Pearson's correlations for all variables by category (listwise n = 937). Mean A. Personal (1) Race (2) Age (3) Gender (4) Sexual Orient (5) SW Tenure (6) Hospital Tenure (7) Religiosity (8) Job Satisfaction B. Professional Ethics Education (9) BSW/MSW (10) Ethics Class (11) Ethics Content (12) Ethics Prep. Ethics Training (13) Internal CE (14) External CE (15) Committee C. Organizational Job Functions (16) Full/Part Time (17) In/Outpatient (18) DCP/PAS Organization (19) Beds (20) Profit Status (21) SW Dept. Size (22) SW Dept. Org. (23) Supervisor D. (24) Census Region E. (25) EEQ SD 0.87 43.5 0.92 0.93 10.9 8.79 1.59 3.01 0.34 11.6 0.28 0.25 9.14 8.06 1.18 0.84 0.85 0.49 0.70 2.61 1 2 3 4 5 6 7 8 9 10 11 0.05 0.07* 0.01 0.04 0.06 -0.06 0.02 -0.08* -0.05 0.66** 0.60** 0.07* 0.09** 0.16** -0.06 -0.05 0.01 -0.01 -0.01 -0.01 0.15** 0.01 0.85** 0.07* 0.08* 0.08* 0.06 0.07* 0.36 0.50 0.46 0.99 -0.08* -0.08* -0.01 -0.06 -0.06 -0.17** -0.21** -0.19** -0.05 -0.01 -0.01 -0.04 -0.01 -0.01 0.02 0.05 -0.05 -0.16** -0.22** -0.24** -0.12** -0.15** -0.21** -0.25** 0.70 0.67 0.26 0.46 0.47 0.44 0.02 0.05 0.05 0.13** 0.14** 0.23** 0.00 0.01 -0.06 0.02 -0.01 0.02 0.17** 0.16** 0.32** 0.82 0.86 0.52 0.38 0.35 0.50 -0.04 0.02 0.07* -0.00 -0.06 -0.07 -0.05 0.02 0.02 -0.02 0.08* 0.04 458 0.18 3.30 0.84 0.59 2.45 73.2 218 0.38 1.47 0.37 0.49 0.98 12.1 -0.01 -0.09** -0.01 0.03 -0.12** -0.11** 0.04 -0.06 -0.09** -0.04 0.04 -0.07* -0.02 0.12** -0.01 0.06 0.01 0.05 -0.02 0.05 -0.09** 0.06 0.05 -0.03 0.03 -0.01 0.03 0.02 12 -0.18** 0.07* 0.07* 0.05 -0.01 0.05 0.11** 0.07* -0.07* -0.10** 0.00 0.49** 0.35** 0.44** 0.17** 0.12** 0.24** 0.05 0.08* 0.01 -0.01 0.08* 0.01 -0.01 0.07* 0.05 -0.02 -0.06* -0.07* -0.04 -0.03 -0.07* -0.05 -0.01 -0.15** 0.04 -0.05 -0.03 0.01 -0.08* -0.03 -0.04 0.01 0.09** -0.01 -0.01 -0.09* 0.01 -0.08* -0.25** -0.04 0.03 0.06 -0.02 -0.01 0.00 -0.05 -0.02 0.01 -0.01 -0.13** 0.00 0.09** -0.07* -0.07* 0.13** 0.00 -0.14** 0.02 0.08* -0.07* -0.09* 0.09** 0.04 0.02 -0.00 0.01 -0.09** -0.02 0.07* -0.04 -0.06 0.00 0.09** 0.02 0.09** 0.49** 0.10** 0.04 0.14** -0.07* 0.14** -0.00 -0.01 -0.03 -0.00 -0.00 -0.04 -0.01 0.04 0.00 -0.05 0.04 -0.03 -0.04 -0.03 0.06 0.01 0.03 0.01 0.01 -0.08* 0.02 -0.00 0.12** Table continues on the next page 87 Table 9, Continued. Means, Standard Deviations, and Pearson's correlations for all variables by category (listwise n = 937). Mean SD 13 14 15 16 17 18 19 Ethics Training 0.70 0.46 (13) Internal CE 0.67 0.47 0.11** (14) External CE 0.26 0.44 (15) Committee 0.05 0.19** C. Organizational Job Functions 0.82 0.38 0.08** (16) Full/Part Time 0.00 0.11** 0.86 0.35 (17) In/Outpatient -0.03 -0.03 0.11** -0.02 0.52 0.50 (18) DCP/PAS 0.02 -0.03 0.02 0.01 0.36** Organization 458 218 0.12** (19) Beds 0.04 -0.09** 0.02 -0.04 -0.01 0.18 0.38 (20) Profit Status -0.06 -0.09** -0.07* -0.00 0.07 -0.05 -0.08* 3.30 1.47 0.18** (21) SW Dept. Size 0.04 -0.16** -0.04 -0.15** -0.05 0.58* 0.84 0.37 0.09** (22) SW Dept. Org. -0.00 0.11** 0.02 0.15** 0.07* -0.04 0.59 0.49 (23) Supervisor 0.05 -0.06 -0.16* -0.09** -0.05 -0.16** 0.16** D. (24) Census Region 2.45 0.98 -0.11** -0.07* 0.06 -0.04 0.08* 0.06 -0.06 E. (25) EEQ 73.2 12.1 0.10** 0.09** 0.10** 0.02 0.05 0.03 0.04 * p < 0.05, ** p < 0.01 (1) Coding: 0 = Person of Color, 1 = Caucasian (3) Coding: 0 = Male, 1 = Female (4) Coding: 0 = Sexual Minority, 1 = Heterosexual (7) Coding: 0 = Low, 3 = High (8) (18) (20) Coding: 1 = Low, 5 = High (9) Coding: 0 = BSW, 1 = MSW (10) (11) (13) (14) (15) Coding: 0 = No, 1 = Yes (16) Coding: 0 = Part Time, 1 = Full Time (17) Coding: 0 = Outpatient, 1 = Inpatient (18) Coding: 0 = Psychosocial Assessment and Support (PAS), 1 = Discharge Planning (DCP) (20) Coding: 0 = Non Profit, 1 = For Profit (21) Coding: 1 = 1 – 5, 2 = 6 – 10, 3 = 11 – 15, 4 = 16 – 20, 5 = More than 20 (22) Coding: 0 = Decentralized, 1 = Centralized (23) Coding: 0 = Not a Social Worker, 1 = Social Worker (24) Coding: 1 = Northeast, 2 = Midwest, 3 = South, 4 = West (25) EEQ = Ethical Environment Questionnaire 88 20 21 22 23 24 -0.02 -0.00 -0.02 0.02 -0.11** -0.08* 0.30** -0.10** 0.05 0.07* 0.03 0.12** -0.05 -0.01 0.03 "I feel my professional education adequately prepared me to address ethical issues in my practice." (Figure 6), which is also generally positive, and correlates most strongly with having had ethics 451 content infused in professional social work 143 197 161 19 education (r = .11, p <.01; Table 9). Finally, respondents were asked about religiosity as Strongly Disagree Neutral Disagree Agree Strongly Agree Figure 5. Ethics preparation frequency distribution (n = 971). measured by the frequency of attendance at religious services (Figure 7), which was nearly evenly distributed across response categories and most strongly negatively correlated with having an MSW degree (r = -.18, p < .01, Table 9). "In general, I am satisfied with my job." 554 "Approximately how often do you attend religious services (of any kind)?" 316 236 238 251 14 54 169 96 Strongly Disagree Neutral Disagree Agree Strongly Agree Never Figure 6. Job satisfaction frequency distribution (n = 969). Annually Monthly Weekly Figure 7. Religiosity as measured by religious service attendance (n = 959). Relationships with Ethical Environment: Personal Variables The correlations of the EEQ scores with the independent variables are also presented in Table 9. The second research question set regarding relationships between personal variables and Ethical Environment is partly addressed here. Significant differences in perceptions of Ethical Environment by age, social work tenure and hospital tenure are notable in Table 9, where 89 increasing age and tenure results in higher EEQ scores, although the correlations are very small (r = .09 - .13). Other categorical relationships suggested by Table 9 were further explored in a series of independent sample t-tests. Table 10 presents the results of the t-tests, and demonstrates that there is a significant difference in Ethical Environment perceptions by gender among hospital social workers, where men rate the environment higher than women (or, that women are more critical of Ethical Environment), although the effect size is small, suggesting this is not a very strong relationship. There are no differences detected by race or sexual orientation (Table 10). The final two parts of the research question about personal variables ask if there are differences by job satisfaction (Figure 6) or by religiosity (Figure 7). These relationships were examined with one-way, between groups ANOVA tests. No differences in EEQ scores are found for religiosity (F (3, 954) = 2.45, p = .062), but there is a significant difference where EEQ scores increase as job satisfaction increases (Welch F (4, 76.3) = 73.7, p < .001), and with a strong effect size by partial eta-squared (ŋp2 = .25). The ANOVA for job satisfaction was subject to unequal variances (Levene's F (4, 963) = 11.4, p < .001) due to unequal groups, small cell sizes at the lower end of the satisfaction scale (few respondents reporting very poor job satisfaction), and a large sample size, therefore the more robust Welch test was used (Field, 2009). In addition, a post-hoc Games-Howell test for unequal variances and unequal groups (Field, 2009) confirmed significant differences (at p ≤ .05) in EEQ scores at all levels of satisfaction. Relationships with Ethical Environment: Professional Variables For the category of professional variables, the relationships between perceptions of Ethical Environment and ethics education and training were investigated. Neither having had a 90 Table 10. Group differences in Ethical Environment Questionnaire scores (two-tailed independent samples t-tests). n Group B Variable Group A Mean (SD) Gender Male 76.7 (11.8) 81 Female Race Minority 72.0 (12.5) 125 Caucasian Sexual Orientation Minority 72.2 (13.6) 62 Heterosexual Ethics Class Yes 73.2 (12.5) 494 No Ethics Content Yes 73.3 (12.2) 676 No Degree BSW 73.5 (11.3) 145 MSW Internal Ethics Training Yes 74.0 (12.2) 684 No External Ethics Training Yes 74.0 (12.2) 650 No Ethics Committee Service Yes 75.3 (13.2) 250 No Profit Status Non Profit 73.8 (12.2) 791 For Profit Primary Function DCP 73.1 (11.3) 433 PAS Supervisor Social Work 73.1 (11.7) 569 Not Social Work Social Work Dept. Org. Centralized 73.9 (12.0) 815 Decentralized * p ≤ .006 ŋ2 = eta squared (effect size) (a). Levene's significant (F = 8.37, p = .004) for unequal variance. ns = Not significant DCP = Discharge Planning PAS = Psychosocial Assessment and Support 91 Mean (SD) 72.8 (12.1) 73.4 (12.1) 73.2 (12.1) 73.2 (11.8) 73.0 (12.1) 73.2 (12.2) 71.3 (11.7) 71.6 (11.9) 72.5 (11.7) 70.4 (11.4) 72.4 (12.4) 73.3 (12.7) 69.7 (12.1) n 888 847 890 475 292 811 287 320 720 174 399 399 153 t (df) Difference ŋ2 2.78 (967) 3.90* .01 -1.21 (970) ns --.629 (950) ns --.043 (976) ns --.380 (966) ns -.259 (954) ns --3.09 (969) -2.63* .01 -2.92 (968) -2.41* .01 -3.00 (392)(a) -2.83* .01 3.34 (963) 3.38* .01 -.880 (830) ns -.191 (966) ns --3.89 (966) -4.13* .02 separate ethics course nor substantive ethics content during one's social work education had an effect on the EEQ scores (Table 10). There was also no significant difference in perceptions of Ethical Environment based on having a MSW or BSW degree. In contrast, Table 10 shows that having had ethics training internal to one's hospital of employment, having had continuing education training outside of the hospital, and having served on an ethics committee were related to higher perceptions of Ethical Environment, although the effect sizes are very small (all ŋ2= .01), suggesting weak relationships. The final professional variable item is the single-question measure of how well respondents believed their professional education prepared them to deal with ethical issues (Figure 5). A one-way, between groups ANOVA test finds that those who more strongly believe their education did prepare them also report higher perceptions of Ethical Environment (Welch F (4, 117) = 6.16, p < .001), with a small effect size (ŋp2 = .03). This test was also subject to unequal variances (Levene's F = 2.74 (4, 965), p = .027) due to unequal groups, small cell sizes at the lower end of the ethical preparation scale, and a large sample size, therefore the more robust Welch test was used (Field, 2009). The post-hoc Games-Howell test for unequal variances and unequal groups (Field, 2009) showed that there is no significant difference in EEQ scores at the lowest "strongly disagree" level of ethical preparation, but significant differences (at p ≤ .05) are present at all other levels in relation to the highest "strongly agree" level. Relationships with Ethical Environment: Organizational Variables Hospital size as measured by number of beds showed no significant relationship with EEQ scores (Pearson's r = .038, p = .248, n = 926). The size of the social work department as measured by the number of social workers at ordinal levels is significant by one-way between groups ANOVA (F (4, 962) = 2.82, p = .024), with a post-hoc Hochberg's GT2 test for unequal 92 groups (Field, 2009) showing that the differences are between the 16 - 20 social workers in the department group, who rate the Ethical Environment higher than the 1 - 5 and 6 - 10 groups (at p ≤ .05), but not with the 11 - 15 or more than 20 groups. However, the effect size for the strength of this relationship is so weak as to be negligible (ŋp2 = .002). The result may be a product of the large sample size, so the practical conclusion is that there is not a significant difference in EEQ scores by the size of the social work department, a finding consistent with the lack of differences by hospital size by bed numbers. The EEQ scores are significantly different between for-profit and non-profit hospitals as seen in Table 10, with respondents from the non-profit hospitals reporting higher ratings of the Ethical Environment with a small effect size (ŋ2 = .01). There are no differences noted in EEQ scores by whether the social workers are full time or part time (t = -.628 (966), p = .530), or whether or not they work in the inpatient or outpatient setting (t = -1.38 (861), p = .169). There is also no difference by whether or not the social workers in the sample are engaged in discharge planning (DCP) or psychosocial assessment and support (PAS), and no difference by whether they are supervised by a social worker or another profession than social work (both Table 10). The final test of differences in perceptions of ethical climate by an organizational variable is the assessment of the organization of the social work department as either centralized or decentralized. Table 10 shows that social workers in centralized departments report EEQ scores higher than those in decentralized departments, with a small effect size (ŋ2 = .02). The final unanswered research question asks if there are differences in ratings of Ethical Environment across Census Regions of the United States (see Table 4 for Census Region and Division representation). A one-way between groups ANOVA indicates that there are no differences (F (3, 968) = .368, p = .776). 93 Regression Model Testing Given the noted relationships between Ethical Environment as measured by the EEQ and a number of other variables, a sequential (hierarchical) linear regression model was constructed. Independent variables with relationships suggested by the above testing were entered into the regression model in sequential blocks representing each of the personal, professional, and organizational categories. Variables were retained in the model based on initial p-values of ≤ .05, and subject to removal in later analyses if the p-value dropped. The final model, standardized regression coefficients, adjusted R2 and changes in R2 (∆R2) with effect sizes (ŋp2) are presented in Table 11. Unstandardized Beta's are presented for the final model only. The regression models were examined for violations of assumptions of normality, independence, homoscedasticity, independence of errors, and for the influence of outliers and multicollinearity. The normal probability plot, residual histogram, standardized residuals plot, and partial plots indicated no deviation from normality, independence, or homoscedasticity (Field, 2009; Tabachnick & Fidell, 2007). The initial Durbin-Watson test for the independence of errors was 0.539, which is an unacceptable value indicating a positive correlation between adjacent residuals (Field, 2009). Examination of the Mahalanobis distances indicated four outliers exceeding the χ2 critical value of 27.88 (df = 9, α = .001) (Tabachnick & Fidell, 2007). The outlier cases were excluded and determined not to affect the overall conclusions when the regression model was repeated (all significant and non-significant loadings remained the same), but their exclusion did correct the Durbin-Watson test to 1.94, which is acceptable (Field, 2009). The four excluded cases were left out of further analyses. The tolerance is within suggested limits with all results > .82, and VIF's (variance inflation factors) are as well with values of 1.0 - 94 Table 11. Sequential regression model predictors of Ethical Environment Questionnaire scores (n = 942). Block 1 Variables β p A. Personal Social Work Tenure .117 .000 Female Gender -.101 .002 Religiosity .064 .047 Satisfaction B. Professional Internal Ethics Training Ethics Committee Service Ethics Preparation C. Organizational Centralized Department Non Profit Status * all p ≤ .005 β = Standardized Coefficients ∆R2 = Change in R2 ŋp2 = Partial eta squared (effect size) Beta = Unstandardized Coefficients Block 2 β p .079 -.075 .026 .474 .006 .008 .335 .000 Block 3 β p .071 -.065 .012 .463 .020 .019 .662 .000 .099 .000 .099 .001 .159 .000 Beta Block 4 β Adj. R2 ∆R2* .085 .064 .036 -2.96 -.066 .017 .133 .013 .641 6.66 .453 .000 2.41 2.47 2.00 .090 .001 .089 .002 .162 .000 2.49 .074 .008 -1.77 -.055 .048 95 ŋp2 p .027 .247 .004 .006 -.221 .229 .284 .010 .008 .038 .031 .291 .008 .008 .004 1.2, all of which indicates no multicollinearity (Field, 2009; Tabachnick & Fidell, 2007). All variables with significant β values represent the unique effect of that predictor when the effects of all other predictor variables are held constant (Field, 2009). Of the personal variables entered into the regression model, social work tenure is directly and positively related to perceptions of Ethical Environment, and being female is directly and negatively related, but both with very small effect sizes (Table 11). Race, age, sexual orientation, and hospital tenure were not significant and were removed from the model. Religiosity was initially positively related to EEQ scores; however, religiosity is mediated by the inclusion of job satisfaction in the model, which eliminates the significance of religiosity in the model. Table 11 shows that of the personal variables, job satisfaction has the strongest positive relationship with perceptions of Ethical Environment with an effect size of ŋp2 = .229. The adjusted R2 for the model of personal variables is .247, explaining over 24% of the variance in EEQ scores, primarily (22%) attributable to job satisfaction. The inclusion of the professional variables into the model results in the retention of internal ethics training and service on an ethics committee as significant but weak predictors (Table 11). Professional degree, having had an ethics class, and having had ethics content were not significant contributors and were removed from the model. External ethics training (continuing education) was a significant predictor when considered among only the professional variables, but not when entered into the model after the personal variables block. Table 11 also shows that beliefs about ethics preparation are also retained as a positive predictor of Ethical Environment perceptions, although with a small effect size. The inclusion of the professional variables do not cause any of the personal variables to drop out of the model, and results in a 96 small but significant increase in the overall R2 to .284, explaining 28% of the variance in EEQ scores. Of the organizational variables, Table 11 shows that working in a centralized social work department was positively predictive of EEQ scores, while working at a non-profit hospital was negatively predictive. Both the social work department organization and the hospital profit status' effects on Ethical Environment are very weak, with small effect sizes shown in Table 11. Working full time or part time, inpatient or outpatient, or engaging in discharge planning or psychosocial assessment and support as job functions were not significant predictors and are omitted from the model. Also omitted are number of beds, size of the social work department, and whether a social worker or another professional supervises the social work staff. The significant and non-significant loadings do not change when considered in a single-level regression model of just organizational variables. The inclusion of the organizational variables do not cause any of the personal or professional variables to drop out of the model, and results in a small but significant increase in the R2 to .291. The final complete and yet parsimonious model explains 29.1% of the variance in perceptions of Ethical Environment, the majority of this attributable to job satisfaction. Taking the results of the regression model and applying them to the initial proposed research model (Figure 1), as derived from literature (Figure 2, Figure 3, and Figure 4), results in Figure 8, a model representing the relationships detected in the sample. Summary and Results Model To summarize, the results as presented in the model (Figure 8) indicate a number of variables from each category that have relationships with perceptions of Ethical Environment. The model and results show that among the personal variables, having a longer tenure in medical 97 social work predicts higher Ethical Environment ratings, while female gender indicates lower ratings (or more critical appraisals of Ethical Environment). Religiosity shows that the higher levels of religious attendance predict higher ratings of Ethical Environment. However, the Personal Variables: Years in Social Work (+) Female Gender (-) Religiosity (+) Job Satisfaction (+) Professional Variables: Ethics Preparation Belief (+) Ethics Committee Service (+) Perceptions of Ethical Environment Internal Ethics Training (+) Organizational Variables: Centralized Department (+) Non-Profit Status (-) Figure 8. Research Results Model. religiosity relationship is mediated by the most important positive predictor of job satisfaction where the more satisfied, the higher the perception of Ethical Environment (or vice-versa). The professional variables of ethical preparation belief, serving on an ethics committee, and having had ethics trainings within the hospital are all positive predictors of higher Ethical Environment ratings. On the organizational level, working in a centralized social work department is predictive of higher environment ratings, while working in a non-profit hospital is predictive of lower ratings, or conversely, working in a for-profit hospital predicts more positive perceptions of Ethical Environment. 98 V. Discussion Perceptions of Ethical Environment are thought to mediate the relationship between personal, professional, and organizational factors and the ethical practices and behaviors of hospital employees, including social workers. The organizational psychology literature suggests that a poor Ethical Environment, with few ethics resources and little ethical accountability, allows for unethical behavior (Ambrose, et al., 2008; Deshpande & Joseph, 2009; Goldman & Tabak, 2010; Gray-Toft & Anderson, 1985; Joseph & Deshpande, 1997; McDaniel, 1998; Peterson, 2002; Rathert & Fleming, 2008; Schwepker, 2001; Trevino, et al., 1998; Ulrich & Soeken, 2005), although the actual effect on behavior is still being investigated, and far from conclusive (Deshpande, et al., 2006). The independent variables identified in the organizational research literature as contributing to perceptions of Ethical Environment are very similar to those believed to affect the ethical actions of social workers (Boland, 2006; Grady et al., 2008; Jansson & Dodd, 2002; Joseph & Conrad, 1989; Kugelman, 1992; Walden, et al., 1990), whether those actions were termed ethical reasoning, decision making, involvement, influence, activism, or participation. The organizational literature supports this same connection between Ethical Environment and the ethical actions of hospital nurses (Corley, et al., 2005; Deshpande & Joseph, 2009; Peterson, 2002; Ulrich, et al., 2007). In this dissertation study, a sample of hospital social workers rated the Ethical Environment higher than nurses, which is consistent with the findings of O'Donnell, et al. (2008), and contrary to the tentative hypothesis that social workers would be more critical. This may be due to a number of factors. Since social workers are professionally trained and socialized to a different system of ethics, this may play a role. Perhaps not having been familiar with the medical ethics model, the social workers find it more concrete and useful than the 99 sequential social work models, and thus perceive that ethics are more attended to in the hospital than their experiences with social work practice. It may be more likely that the combination of a foundation in social work ethics with subsequent experience and exposure to the medical model results in a synergistic expansion of ethical understanding. The two approaches can be complimentary, and a form of teaching comparative ethics may already exist in social work education (the content of social work education was not explored here). It is also possible that hospital social workers are simply predisposed to favor the medical ethics environment, given that they professionally self-select into hospital social work. Alternatively, they may align with the medical approach to ethics as a function of their socialization into the organizational culture after entry into hospital social work. Hospitals do represent a unique institutional environment, and given the hierarchical dynamics of this host setting, ascribing to the medical model of ethics may be required of hospital social workers in order to maintain their place within the organization. Personal Variables In the literature review many studies express concerns about social workers allowing personal variables to interfere with or cloud their ethical perceptions, reasoning, and actions (Csikai, 1999; Doyle, et al., 2009; Holland & Kilpatrick, 1999; Joseph & Conrad, 1989; Kugelman, 1992). In many respects, the results of this study show that this is not the case for hospital social workers and Ethical Environment. There were no differences detected by racial/ethnic identification, which aligns with some of the organizational findings (Deshpande, et al., 2006), but is contrary to the work done on hospital Ethical Environment with nurses (McDaniel, 1998; Ulrich et al., 2007; Corley, et al., 2005). Perhaps the social work educational 100 emphasis on examining personal biases is effective in this regard. Given the strength of the sample in this study, it may be that this is a true difference between social work and nursing. Although differences were detected in this study by gender and religiosity, the effect sizes were minimal (or completely mediated) and may well be an artifact of the large sample size. The relationship between ethics and religiosity was only hinted at in the existing literature, with no conclusive findings (Doyle, et al., 2009), and this dissertation can only add that it may not be influential, at least for hospital social workers. However, the role of religiosity is still very speculative, especially considering that this study did not identify the religious affiliations of the hospitals in the sample. It may be that the personal religiosity of the social workers aligned with the affiliation of their hospitals, and this may explain why higher religiosity was predictive of higher Ethical Environment ratings before the mediating effect of job satisfaction. The finding that female social workers rated the Ethical Environment lower than males is consistent with the findings of the organizational and Ethical Environment literature pertaining to nurses (Hart, 2005; McDaniel, et al., 2001). The social work literature on gender has been inconclusive (Doyle, et al., 2009; O'Donnell, et al., 2008), so the current study helps in suggesting that gender has an effect, although it is minimal. Although female social workers may be more critical of professional and organizational ethics, as Tyson (1990) suggested for business ethics, their perceptions may well be more astute and accurate. Still, the miniscule statistical support for this finding calls into question the practical usefulness or applicable meaning of any gender differences, especially when extrapolating beyond perceptions of Ethical Environment to actual ethical action and behavior. The difference in Ethical Environment by years of service in hospital social work is expected, given that the longer any given social worker remains within this area of practice, the 101 more comfortable and attached they are to the work environment. This supports previous findings about tenure increasing the ethical actions in nurses (Corley, et al., 2005; Goldman & Tabak, 2010; Rathert & Flemming, 2008), but is counter to the increase in critical appraisals detected with longer tenure in other nurse samples (McDaniel, 1998; Ulrich, et al., 2007). The previous social work research supports the positive relationship between tenure and ethical actions (Boland, 2006; Csikai, 1999; Jansson & Dodd, 2002), especially given that the other social work research findings were of no effect, rather than negative correlations (Foster, et al., 1993; Grady, et al., 2008; O'Donnell, et al., 2008). For hospital social workers, and perhaps most social workers and nurses, ethical wisdom does come with age and experience. The finding that job satisfaction is the single largest and only practically significant predictor of perceptions of Ethical Environment is very interesting. It is consistent with the organizational and Ethical Environment research (Ambrose, et al., 2008; Goldman & Tabak, 2010; Joseph & Deshpande, 1997; Schwepker, 2001; Ulrich, et al., 2007). It is also consistent with the hospital social work ethics literature that reports positive correlations between job satisfaction and ethical actions (Joseph & Conrad, 1989; Kadushin & Kulys, 1995). In one respect, the strong relationship of Ethical Environment with job satisfaction supports the suggestion that some of the other confounding personal factors are not at issue for hospital social workers. However, the literature on job satisfaction repeatedly shows that the foremost predictor of job satisfaction is financial compensation, or economic recognition and reward (Acker, 2004; Butler, 1990; Sze & Ivker, 1986; Vinokur-Kaplan, 1991; Gellis, 2001, 2002; Kadushin & Kulys, 1995; Siefert, et al., 1991). Thus, in another respect, the strength of job satisfaction seems to detract from the ethical foundations, training, and education of the social work profession. This seems consistent with the study of O'Donnell, et al. (2008), where they found both income and 102 ethical climate negatively predictive of ethical stress, and ethical climate positively predictive of career satisfaction. However, job satisfaction among social workers, much more so than nurses, has also been found to be strongly related to other factors beyond pay. Social workers derive job satisfaction from the challenge of their work, and feelings of autonomy and control over their own professional decision-making (Gellis, 2001, 2002). The organizational environment is one key element to these satisfaction items, especially in the form of organizational support and collaboration (Kadushin & Kulys, 1995). The positive perceptions of Ethical Environment found in this dissertation suggest that hospital social workers are pleased with the support and collaboration within the hospital, and this may well be a more important variable than pay in their job satisfaction. It may also be that there is an alignment between the ethics and values of people who self-select into the social work profession, and the mission, values, and ethics of the agencies where they work, including hospitals. This could certainly contribute to their job satisfaction. Professional Variables The contributions of the professional variables to the results of this study are also notable in a number of ways. Social work ethics courses, substantive ethics content, or whether one had an MSW degree or not, made no difference in perceptions of Ethical Environment. The literature was inconsistent in findings here related to ethical actions among social workers, and related to perceptions of Ethical Environment among nurses. In the organizational literature, the research sometimes supported education (Deshpande, et al., 2006; Malloy & Agarwal, 2003; Schluter, et al., 2008), while at other times it supported continuing education programs (Ulrich & Soeken, 2005), or that education made employees more critical (Ulrich, et al., 2007). In the 103 social work research, education was most often found to positively correlate with ethical actions (Boland, 2006; Csikai, 1999; Doyle, et al., 2009; Foster, et al., 1993; Joseph & Conrad, 1989), and that continuing education was often even more important (Csikai, 1999; Foster, et al., 1993; Grady, et al., 2008; Jansson & Dodd, 2002). The results of the current study support the findings that continuing education or other trainings outside of professional social work education are more influential when it comes to perceptions of Ethical Environment. In fact, the findings of this study indicate that continuing education trainings conducted by the hospitals themselves is the most effective format when it comes to contributions to perceptions of Ethical Environment. O'Donnell, et al. (2008) also found no relationships between course work or degree when it comes to ethical stress or moral action, but they did find that ethical resources provided by the employer were consistently helpful. Ethics resources were not well defined, but can be assumed to include internal ethics educational events. Given that the form and content of these ethics events are unknown, one can only speculate that a hospital that sponsors and invests in ethical trainings and discussions demonstrates to employees that ethics are important. This explanation does align with the existing literature and the findings of the current dissertation research. Another element to consider is that the effect of any continuing education program is, in part, dependent upon the ethics education foundation established during formal social work education. The findings of no relationship between formal education and Ethical Environment may be subject to time and distance, so the findings do not necessarily detract from the prior contributions of ethics education. Although the direct questions about social work ethics education and training were not significant in this study, the single question about the participants' beliefs about the adequacy of 104 their professional ethical preparation was contributory. It makes sense that perception is a stronger predictive variable than fact, but this does not necessarily connect with the ethical behavior that perceptions of Ethical Environment are meant to influence. Deshpande, et al. (2006) actually found that this type of belief question was not related to ethical practices and behaviors. The findings here that the ethical preparation question was not at all correlated with degree and only moderately correlated with having had an ethics class or substantive ethics content support the disconnect between these measures. Ethics education and training are also related to social work involvement, participation, and competence when it comes to ethics committees (Csikai, 1997; Csikai & Sales, 1998; Foster, et al., 1993; Hoffman, et al., 2000). In this dissertation research, it was found that service on an ethics committee predicts higher Ethical Environment ratings, and this speaks to the ethics resources the hospital provides, consistent with the findings of O'Donnell, et al. (2008). It is clear that the presence of an ethics committee is a valuable resource for hospital employees, including social workers, and that social workers should be ethics committee members. The apparent trend in declining ethics committee prominence and social work participation (Foster, et al., 1993; Guo & Schick, 2003) is concerning, and should be addressed. Social workers need to volunteer and advocate for serving on hospital ethics committees, both to increase their ethical actions, and to increase the voice and value of social work in this forum. This is also another opportunity for the multidisciplinary collaboration that may be linked to the job satisfaction of hospital social workers. Organizational Variables The ratings of Ethical Environment by departmental configuration support the continuation of centralized social work departments, and this suggests that a central, organized 105 social work presence in the hospital is supportive of social workers in a complex way. Since they perceive the Ethical Environment more positively than nurses and assuming that this is, in part, related to having aligned with the medical ethical model, it seems that if they were more integrated into the hospital in a decentralized way, that this ascription would be stronger. It may be that it is the ability to stand somewhat aside and compare the two professional ethical stances that brings social workers to appreciate the Ethical Environment. This is one of the organizational variables that has not been studied well in other research, with only Jansson and Dodd (2002) reporting that departmental size (not configuration) had no relationship with ethical activism in their sample of hospital social workers. Nevertheless, some have noted a trend toward dismantling centralized social work departments, which should be reconsidered (Berger, Robbins, Lewis, Mizrahi, & Fleit, 2003; Judd & Sheffield, 2010). Another complex pattern to emerge in this study relates to the finding that the Ethical Environment of for-profit hospitals is perceived more favorably than non-profit hospitals. This may tie into the larger for-profit hospitals having more internal ethics trainings, and potentially more opportunity for social work contact with, and participation in, ethics committee services. O'Donnell, et al. (2008) suggest that part of a positive Ethical Environment is the level of ethics resources provided by a hospital, such as continuing education seminars and ethics committees, and that these contribute to both job satisfaction and ethical action among social workers. However, they did not find profit status to correlate with ethical stress, moral action, or job satisfaction. The relationship between profit status and Ethical Environment may be unique, and the effect on ethical actions mediated by the environment. In another interpretation of the interaction between profit status and employees, some of the organizational research suggests that non-profit employees are predisposed to focus on their own personal growth and social 106 responsibility, while for-profit employees are more focused on, and identify with, the organization (Agarwal & Malloy, 1999; Brower & Shrader, 2000). Other organizational factors, such as the job functions of the social workers, were not predictive of Ethical Environment perceptions. Whether the social workers worked inpatient or outpatient, whether they did discharge planning or provided psychosocial support, and whether they were supervised by a professional social worker or not made no difference. It is curious that social workers perceive the Ethical Environment more positively from a centralized social work department, but without necessarily having another social worker as a supervisor. Perhaps the centralized department is yet another function of a hospital with the wherewithal and resources to dedicate to the work environment, Ethical Environment and competent supervision included. It is also possible that the centralized model allows for professional and peer interaction, support, and informal supervision, such that the social workers gain more from each other than from the immediate supervisor. Strengths and Limitations This study succeeded in obtaining a large, national, and representative sample of hospital social workers in the United States. The size and randomness of the sample support the strength of the findings, while at the same time large samples tend to detect very small differences as statistically significant, with dubious practical significance. This actually supports the representativeness of the sample, given the miniscule differences noted between hospitals that participated and those that did not, and in geographical representation. The generalizability of the study is high when it comes to hospital social workers in the United States given the size and representativeness of the sample. However, the sample selection criteria of hospitals over 200 beds in size does limit the generalizability to smaller hospitals, and perhaps a great many hospital 107 social workers in the thousands of hospitals not included in the sampling frame. In addition, as with any cross-sectional design, the results here indicate relationships and correlations between the variables, not causality. Those social workers who completed the study questionnaire may well be different from those that did not. The self-selection by hospital social workers into the study may introduce response biases. Their responses may also be subject to social desirability bias, in that they may have answered the questions, especially sensitive questions about ethics, in a way other than how they truly felt. The anonymous nature of the survey certainly helped with controlling for this concern, as well as the fact that the respondents were only asked about their perceptions of the organization, rather than their own ethical behavior. The lack of missing data and missing completely at random pattern are also strengths that suggest the minimization of biased responses. The variables chosen for this study are not exhaustive, and some items were necessarily brief to facilitate the web survey. Given the relatively low explanatory power of the regression model, there are clearly other variables unaccounted for in this research. Considering the role of job satisfaction and its connection to financial compensation, current salary or perceptions of salary adequacy would have been excellent additional questions. The question about the profit status of the hospitals may have benefited from refined categories, since the differences between private and public hospitals may be important. Another missing piece of information in this study is the religious affiliation of hospitals in the sample. The data about profit status and religious affiliations may have added a great deal to this study, especially regarding the relationships between ethics, religiosity, and job satisfaction. In addition, this study did not 108 explore the actual ethical behaviors of the social workers, nor the relationship between Ethical Environment and ethical actions. Finally, there is the consideration of the linear regression modeling in light of organizational level variables, as mentioned in the methodology section. Although hierarchical linear modeling may have been generally indicated, there are a number of facts that help to reduce any statistical error related to the choice of statistical testing. The randomness of the sample contributes to the strength of the linear regression. The grouped individual perceptions that make up the organizational outcome variable suggest a natural connection between the levels of the variables. Finally, the fact that the regression modeling steps involving controlling for the organizational variables and included running that block independently without a change in results supports the strength of the relationships detected. Implications Overall, this study has explained a great deal about the correlations between numerous independent variables and perceptions of Ethical Environment among hospital social workers. It has also raised many interesting questions, and the findings are subject to a lot of speculation. Although the sample is large and representative, and this lends strength to the findings, the contributions of the variables selected to perceptions of Ethical Environment are plagued by small effect sizes. Additionally, this study focused on perceptions of Ethical Environment, which although a factor in the ethical actions of hospital social workers, does not necessarily explain ethical actions, nor the role of the variables studied in those ethical actions. Nonetheless, the findings do suggest some ideas around social work education, policy, practice, and directions for future research. 109 Education. The lack of relationships between perceptions of Ethical Environment and professional social work education raises a number of questions. Of course, it would be preferable to see professional education variables having a larger impact on social workers in the field, whether in perceptions of Ethical Environment, utilization of ethical reasoning models, or in the as yet undetermined effect on ethical actions. Social work ethics education strategies and content are supported as effective in diminishing the effect of personal biases, but the literature suggests the relationship to ethical reasoning and decision-making is somewhat tenuous. More research is needed to explain, strengthen, and develop the links between ethics education and ethical actions, including new educational approaches. This study can only speak to the link between past educational practices (given the age and years of service of the sample) and perceptions of Ethical Environment, and the current results do not investigate the eventual ethical actions of social workers. The fact that the social workers in this sample rated the Ethical Environment highly may actually suggest that social work ethics education, when combined with hospital ethics and the principlist medical model, has a more profound impact on understanding ethics. Social work ethics education is different from the medical model, and there is a significant difference in how well these are utilized in the field. But, recalling the literature review on professional ethics, the differences between them is not as great as it may seem given that they are both founded in the same ethical theory, and that they each contain similar ethical principles. Given the contributions of the two approaches and their possibly having a positive synergistic effect in the hospital setting, perhaps some form of education that speaks to comparative ethics would be beneficial, both for social workers, and for medical professionals. 110 In congruence with some of the literature, this dissertation research supports continuing ethics education efforts. Although internal hospital trainings were found to be particularly effective when it came to perceptions of Ethical Environment, this does not necessarily diminish the value of other professional continuing education. In fact, the literature suggests that professional social work continuing education on ethics may well have an impact on ethical actions and behavior, which this dissertation did not explore. The professional social work licensure and code of ethics requirements for continuing education, and in some cases mandated ethics hours, may be the best form of ethics education for those in practice. This allows the profession to address current and pressing issues encountered in all areas of social work practice, including hospitals. It is important to note that continuing education is certainly building upon the foundation established during formal social work education, so ethics education must continue and has definite value. This is particularly evident in reducing the influence of personal biases in social workers. Policy. The role of public financing and the provision of medical care in the United States is an ongoing and controversial issue. The findings here related to job satisfaction, hospital-based ethics trainings, and for-profit status all suggest economic forces. The level of resources a given hospital might direct at ethical training and just compensation is an apparently powerful contributor to a positive Ethical Environment. On a smaller scale than the overall financing of medical care nationwide, the information from this dissertation indicates that hospital administrations should consider carefully attending to organizational resources that can improve the Ethical Environment. This is especially important given that the organizational psychology research with hospital nurses does point to a direct relationship between Ethical Environment and ethical behavior. 111 For social work, the policy implications include the profession supporting health care reform, particularly in the area of adequate public financing. For hospital social workers, social work departments, and social work management, greater involvement in hospital ethical policy is suggested. Hospital social work departments should encourage, support, and even sponsor ethics education events, as well as ensuring social work involvement in ethics committees. Given the possibilities of comparative ethics education and the importance of multidisciplinary collaboration, all hospital employees could benefit from expanding their understanding of ethics and the contributions social work can make in this area. The ability of social work ethics approaches to minimize the affect of personal bias is something that this dissertation research supports, and something that the literature suggests would be beneficial for nurses and physicians. This is a unique contribution that social work could make to the hospital, and one that besides being necessary, could be quite helpful in clarifying ethical reasoning. More research needs to be done on what hospital policies contribute to Ethical Environment, and how Ethical Environment influences hospital policies. There are a number of current and controversial issues that may play a role in this policy dynamic, including family visitation rights, the treatment of patients with and without insurance, the distribution of resources along profitable service lines, and the strength and nature of relationships between hospitals and their employees. For example, there may be Ethical Environment implications related to staffing levels and union representation, related to the mix of private, public, and uninsured patients, or related to the status and prestige of certain work areas. None of these variables were explored here, but the correlation between Ethical Environment and resources suggests that these are important policy areas that this dissertation supports investigating. 112 Practice. The practice of hospital social work is still highly concentrated around discharge planning activities, although it is notable that the sample here was almost evenly split between identifying with discharge planning or psychosocial support. It may be that some considered the strong presence of psychosocial support issues in modern, complex discharge planning. In any event, the literature review on these topics indicated that the ongoing challenges of hospital social work practice are intricately interconnected with the theoretical content of Ethical Environment perceptions. And yet, the social workers in this sample rated the Ethical Environment positively, so there is something about hospital practice that, despite its challenges, is positively experienced and appreciated by social workers. As mentioned above, this may be correlated with the job satisfaction social workers derive from the challenges of their work. Of course, this refers to their perceptions of the overall hospital Ethical Environment, not their perceptions of their smaller work climates, and there may be a disconnect between job challenges and Ethical Environment for social workers. Given that they are non-medical personnel in a host setting, they may sense or maintain a distance between the hospital Ethical Environment and their own work. For the ongoing practice of hospital social work, it appears that social workers should continue their roles in both discharge planning and psychosocial support. Their perceptions of a positive Ethical Environment need to be articulated and communicated to their colleagues, and to the patients and families with whom they work. It is speculative, but this dissertation research may indicate that social workers' perceptions are both accurate and valuable, and if understood by others could help in addressing ethical concerns as they arise in practice. It seems as if hospital social workers are negotiating a complex and nuanced Ethical Environment, and with some success given their longevity in the field. There may be something to social work practice 113 in the hospital setting regarding ethics that has yet to be appreciated or utilized effectively. This may be one way for hospital social work to improve and enhance their contribution and value to the hospital. Research. This dissertation has raised more questions than it has provided answers. Perceptions of Ethical Environment among hospital social workers are tied to job satisfaction more than anything else. This leads to the need for more research on job satisfaction. There is an existing body of research in this area, and some of it does pertain to hospital social workers, although it is limited and overwhelmingly based on small, nonrepresentative samples. Job satisfaction is apparently determined by financial compensation among nurses, but there are other factors to consider when it comes to social workers. Still, some research into the economics of hospital services, and the contributions (both monetary and otherwise) of the social work profession are recommended. More research is needed on the role of personal factors and beliefs and their potential influence on the ethical actions in social workers, as well as the effectiveness of social work ethics education. The literature suggests that social workers are not directly utilizing social work ethical models, or the NASW code of ethics. However, social workers are making daily ethical decisions, and more research is needed to determine the basis of their ethical reasoning. A reliable and valid measure of the effect of their formal education is needed. In the meantime, social work educators can continue to adjust and improve educational strategies and content, preferably with some research as to immediate and long-term outcomes. It may be a need for different ethics content, or it may be a question of new and novel educational methods. The findings of this dissertation research suggest a disconnect between ethics education and perceptions of Ethical Environment, but not necessarily between that education and ethical 114 actions or behaviors. Whether or not ethics education has a lasting effect on social workers in the field is still debatable, so many questions remain about on what social workers are basing their ethical reasoning and decision-making. This is especially true in the morass of complex and competing interests that characterize hospital social work practice. However, the social work research has yet to determine how foundational ethics education combines with continuing education programs. It seems that crediting continuing education over formal education discounts that foundation, which is contrary to the long process required to develop and refine one's ethical reasoning. Other future research should attend to the measurement of professional educational variables. The single-item measure of ethical preparation may be a poor substitute for actual education, and yet the direct measures of education so far seem weakly related to outcomes, both in this study and in the research literature. More research is needed to determine the best way to capture the multidimensional effects of professional social work education and training, as well as the complex interactions between formal education and continuing education. Further investigations into single-item measures are needed as well, as these are critical to efficient survey research, especially in the realm of web surveys, which may be the most economical and popular survey method of the future. Conclusion This exploration of hospital social workers and their perceptions of Ethical Environment has achieved a great deal, both in terms of understanding hospital social workers and in terms of the challenges of the hospital setting, and in web survey research. It has demonstrated the use of the web survey methodology to secure a large, representative sample, while at the same time highlighting some of the issues that need to be attended to in this process. This dissertation has 115 demonstrated that Ethical Environment is in some way tied to job satisfaction, at least for hospital social workers. The direction of this relationship is unknown, as it may be that Ethical Environment precedes job satisfaction, or vice-versa. In addition, there are many more unaccounted for variables that may be related to either job satisfaction, Ethical Environment, or both. This study has indicated that much more research needs to be done on the complex interactions of personal, professional, and organizational variables, but has provided some insight into the concept of Ethical Environment in the hospital setting, and suggests interesting differences between social workers and nurses. Perhaps more importantly, this study has shown that hospital social workers think quite highly of their employing hospitals, and has delineated some of the factors that are important to that perception. In this way, it is clear that centralized social work departments should be maintained and that hospitals need to dedicate resources to ethics education for their employees. At the same time, it is clear that the connection between social work ethics education and the ethical actions of social workers has not been determined. Ethics research has not yet found a way to adequately measure ethical actions, nor the impact of ethics education. Exploring this line of research is an exciting opportunity for social work researchers, and would be strengthened by working with social work educators to measure social work ethics education content and strategies. The hospital social workers who participated in this study have helped immensely in this exploration of Ethical Environment, and in raising these pressing questions, and their contributions are greatly appreciated. 116 APPENDICIES 117 Appendix A. Evite Text and Reminder Texts 1. Evite Text HelloThank you so much for providing an email contact for the survey information. Below is the link to the survey and the current password. I would greatly appreciate all the social workers at your hospital completing the survey. This is part of a nationwide study of hospital social workers’ perceptions of the Ethical Environment of hospitals. It should only take about 15 minutes to complete. Everyone who completes the survey may enter to win a $10.00 Amazon.com gift certificate! There is only one password for all of the social workers at your hospital. The questionnaires are anonymous. No individual or hospital can be linked to the responses. The password and the link below will allow access to the study, including an explanation of the study, an online consent to participate, and the questionnaire. The study site is open for access immediately. Please distribute the password and link to the social work staff of your hospital. It may be easiest to forward this email to the social workers. Thank you for your time and consideration. Only with the generous help of yourself and your colleagues can this research project be successful. If anyone has any questions or concerns, please contact me at any time. Survey Site Link: PASSWORD: Sincerely, -greg Greg L. Pugh, MSW, LICSW, ACSW Clinical Instructor, Doctoral Candidate Michigan State University School of Social Work 254 Baker Hall, East Lansing, MI 48824-1118 Email: gpugh@msu.edu Phone: 517-353-0770 P.S. If there is any problem with the password or study web site, please contact me immediately. 118 2. Reminder Email Text & Modified Reminder Email Text HelloOnce again, I would like to thank you for responding to my contacting you about the online ethics survey I am conducting for my dissertation research. It is crucial to hear the voice of hospital social workers on Ethical Environment, given that there are no previous studies exclusive to our perspective. Since the survey method is both anonymous as to the individual participants, and anonymous as to the hospital where they work, I am unable to track response rates. As such, I am sending this note as a reminder. This will be the only reminder. The survey site is open and can be accessed at any time. The link and password are below, and again, this email may simply be forwarded to the social workers at your facility. [modified text section] Since the survey method is both anonymous as to the individual participants, and anonymous as to the hospital where they work, I am unable to track response rates. As such, I am sending this note as a reminder. This will be the only reminder. The survey site is open and can be accessed at any time. However, the survey is nearly complete and the web site will close on Friday, November 19th, at 5:00pm EST. The link and password are below, and again, this email may simply be forwarded to the social workers at your facility. If there are any questions or concerns about the survey or having the social workers at your facility respond, please contact me at the email or phone below. Survey Site Link: Password: Thanks again, -greg Greg L. Pugh, MSW, LICSW, ACSW Clinical Instructor, Doctoral Candidate Michigan State University School of Social Work 254 Baker Hall, East Lansing, MI 48824-1118 Email: gpugh@msu.edu Phone: 517-353-0770 119 Appendix B: Online Consent Form Letter Hospital Social Worker's Perceptions of Ethical Environment Study Greetings fellow social worker, Purpose: I am writing to you to ask for your help with a nationwide research study of hospital social workers; my dissertation research at Michigan State University School of Social Work. This study is part of an effort to understand how social workers feel about their work in hospitals. The results of this study may improve not only the practice of hospital social work, but also inform how we educate future medical social workers. Procedures: You have been selected as part of a random sample of hospitals in the United States. As such, it is incredibly important to hear your voice in this survey. It should take less than 15 minutes to complete the questionnaire. As a hospital social worker myself for 12 years, I understand how valuable your time is, so I have kept the survey as brief as possible. You may notice that the questionnaires do not include any way to identify you personally or connect you with your responses. The password used to access this site is only to prevent uninvited guests from completing the survey, and is in no way linked to you or your responses. Your answers are anonymous, and results from the study will only be presented in aggregate form. Your participation is voluntary, you may choose not to participate at all, or you may refuse to answer certain questions or discontinue your participation at any point. If you have any questions, comments, or concerns, please feel free to contact me at any time at the email, phone, or address below. If you have any questions about your role and rights as a research participant, or would like to register a complaint about this study, you may contact, anonymously if you wish, the Michigan State University Human Research Protection Program at 517-355-2180, FAX 517-432-4503, or email irb@msu.edu, or regular mail at: 207 Olds Hall, MSU, East Lansing, MI 48824. You may indicate your consent to participate in this study simply by clicking the "I Agree" button below, and then clicking "Next" to begin the survey. Thank you so much for taking the time to complete the questionnaire. As there is no identifier linking you to the survey, the entire survey needs to be completed at one time. It is not possible to stop and resume your survey later. However, if for some reason you must leave the survey before it is complete, you may restart at the beginning at any time. Only complete responses will be used. A progress bar at the bottom of the screen will indicate how much of the survey is complete. As a thank you at the end of the survey, you will have the option to enter a drawing to win one of twenty $10 Amazon.com electronic gift cards! Sincerely, -greg Greg L. Pugh, MSW, LICSW, ACSW, Doctoral Student Michigan State University School of Social Work: www.socialwork.msu.edu 254 Baker Hall, East Lansing, MI 48824. email: gpugh@msu.edu phone: 517-353-0770 P.S. If there are any difficulties with the online survey, please contact me immediately. 120 Appendix C: Survey Questionnaire (Final version, in order as presented online) Hospital Social Workers’ Perceptions of Ethical Environment Part A: Please select one box for each of the following questions. 1. What is your job title? Social Worker Discharge Planner Care Coordinator Case Manager Other, please specify: _________________________ 2. What is your highest educational degree? BASW Other Bachelors degree: _________________________ MSW Other Masters degree: _________________________ Other, please specify: _________________________ 3. During your education, did you take a discrete course about ethics (a course where ethics was the primary subject of the entire course)? Yes No 4. During your education, did you take a course that was in large part about ethics (a significant section or theme throughout)? Yes No 5. I feel my professional education adequately prepared me to address ethical issues in my practice. Strongly Agree Agree Neutral Disagree Strongly Disagree 6. Have you attended any continuing education seminars (OUTSIDE of your hospital) on ethics within the past 2 years? Yes No 7. Have you attended any continuing education seminars offered WITHIN your hospital within the past two years? Yes No 121 8. Have you ever served on a hospital ethics committee? Yes No 8b. If yes, was this at your current hospital? Yes No Part B: The Ethical Environment Questionnaire [due to copyright laws, the EEQ is not reproduced here] 9. How many years overall have you worked as a social worker in the hospital setting? (If less than one year, enter “1”) _____ (number of years) 10. How many years have you been at the hospital where you currently work? (If less than one year, enter “1”) ______ (number of years) 11. What is your age? _____ (age in years) 12. Do you work: Full-Time Part-Time 13. What is your PRIMARY job function in the hospital? Discharge planning Psychosocial assessment and support Other, please specify _________________________ 14. What is your PRIMARY assignment? Inpatient Outpatient Other, pleased specify ______________________________ 15. Is your hospital social work department: Centralized (all social workers in one department reporting to one supervisor) Decentralized (social workers reporting to separate departments or units) 16. Is your immediate supervisor a: Social Worker Nurse Other, please specify _________________________ 17. Is your hospital… For Profit Not For Profit 122 18. How many social workers (both inpatient and outpatient) are employed at your hospital? 1–5 6 – 10 11 – 15 16 – 20 More than 20 19. In which State is your hospital located? ____________________ [drop down menu for selecting the state] 20. How many inpatient beds does your hospital have? ____________________ [text box for entering the number] 21. In general, I am satisfied with my job: Strongly Strongly Agree Agree Neutral Disagree Disagree 22. I intend to leave my position / hospital in the next six months: Strongly Strongly Agree Agree Neutral Disagree Disagree 23. What is your gender? Female Male Transgender 24. What is your sexual orientation? Gay / Lesbian Heterosexual Bisexual Other, please specify _________________________ 25. Are you of Spanish, Hispanic or Latino origin, including Mexican-American, Chicano, Mexican, Puerto Rican, Cuban, Central or South American, or other Hispanic? Yes No 26. Are you… (check all that apply) Asian (including Chinese, Korean, Japanese and Southeast Asian) Black or African American Native American or Alaskan Native Other, please specify _________________________ Pacific Islander White 123 27. Approximately how often do you attend religious services (of any kind)? Never Annually Monthly Weekly Part C: Enter to Win one of TWENTY $10 Amazon.com electronic gift cards! Thank you for completing this survey and for participating in this study! Your help is greatly appreciated. As a further thank you for participating, we would like to offer you a chance to win one of twenty $10 Amazon.com electronic gift cards! If you would like to be entered into the drawing to win one of the electronic gift cards, please enter your email address in the box below. Your email address will be kept on a secure server for the sole purposes of the drawing and then deleted. Your email address is in no way linked to your survey responses. Email: ___________________________________ Confirm Email: ___________________________________ At the completion of the study (on or about January 1st, 2011), an email address will be randomly selected and the electronic gift card codes emailed to the winners. Part D: Follow up contact: If you are willing to be contacted via email to answer some follow up questions about this study and our preliminary findings, please enter your email address below. Again, these email addresses will be kept secure, separate from the survey responses, and deleted at the end of the study. Email: ___________________________________ Confirm Email: ___________________________________ 124 REFERENCES 125 References Abbott, A. (1983). Professional ethics. American Journal of Sociology, 88(5), 855-885. Abramson, J.S. (1988). Participation of elderly patients in discharge planning: Is selfdetermination a reality? Social Work, 33(5), 443-448. Abramson, J.S., & Mizrahi, T. (1996). When social workers and physicians collaborate: Positive and negative interdisciplinary experiences. Social Work, 41(3), 270-281. Abramson, J.S., Donnelly, J., King, M.A., & Mailick, M.D. (1993). Disagreements in discharge planning: A normative phenomenon. Social Work, 18(1), 57-64. Abramson, M. (1981). Ethical dilemmas for social workers in discharge planning. Social Work in Health Care, 6(4), 33-42. Abramson, M. (1983). A model for organizing an ethical analysis of the discharge planning process. Social Work in Health Care, 9(1), 45-52. Abramson, M. (1996a). Reflections on knowing oneself ethically: Toward a working framework for social work practice. Families in Society, 77(4), 195-201. Abramson, M. (1996b). Toward a more holistic understanding of ethics in social work. Social Work in Health Care, 23(2), 1-14. Acker, G. (2004). The effect of organizational conditions (role conflict, role ambiguity, opportunities for professional development, and social support) on job satisfaction and intention to leave among social workers in mental health care. Community Mental Health Journal, 40(1), 65-73. Agarwal, J., & Malloy, D.C. (1999). Ethical work climate dimensions in a not-for-profit organization: An empirical study. Journal of Business Ethics, 20(1), 1-14. Alvesson, M. (2002). Understanding organizational culture. Thousand Oaks, CA: SAGE. Ambrose, M.L., Arnaud, A., & Schminke, M. (2008). Individual moral development and ethical climate: The influence of person-organization fit on job attitudes. Journal of Business Ethics, 77, 323-333. American Hospital Association (2010). AHA Guide. Chicago, IL: Health Forum. American Hospital Directory (n.d.). State and National Hospital Statistics. Retrieved June 20, 2008 from http://www.ahd.com/ 126 Anderson, N.R., & West, M.A. (1998). Measuring climate for work group innovation: Development and validation of the team climate inventory. Journal of Organizational Behavior, 19(3), 235-258. Antle, B.J., & Regehr, C. (2003). Beyond individual rights and freedoms: Metaethics in social work research. Social Work, 48(1), 135-144. Appelbaum, S.H., Deguire, K.J., & Lay, M. (2005). The relationship of ethical climate to deviant workplace behavior. Corporate Governance, 5(4), 43-55. Auerbach, C., Mason, S.E., & LaPorte, H.H. (2007). Evidence that supports the value of social work in hospitals. Social Work in Health Care, 44(4), 17-31. Backof, J.E., & Martin, C.L. (1991). Historical perspectives: Development of the codes of ethics in the legal, medical and accounting professions. Journal of Business Ethics, 10(2), 99110. Bartlett, H. M. (1975). Ida M. Cannon: Pioneer in medical social work. Social Service Review, 49(2), 208-229. Beauchamp, T.L., & Childress, J.F. (2009). Principles of biomedical ethics (6th ed.). New York, NY: Oxford University Press. Beckett, C., & Maynard, A. (2005). Values and ethics in social work: An introduction. Thousand Oaks, CA: SAGE. Beder, J. (2006). Hospital social work: The interface of medicine and caring. New York, NY: Routledge. Berger, C., Robbins, C., Lewis, M., Mizrahi, T., & Fleit, S. (2003). The impact of organizational change on social work staffing in a hospital setting: A national, longitudinal study of social work in hospitals. Social Work in Health Care, 37(1), 1-18. Berkman, B., Bonander, E., Kemler, B., Rubinger, M., Rutchick, I., & Silverman, P. (1996). Social work in the academic medical center: Advanced training – A necessity. Social Work in Health Care, 24(1/2), 115-135. Berlant, J.L. (1978). Medical ethics and professional monopoly. Annals of the American Academy of Political and Social Science, 437, 49-61. Blumenfield, S., & Lowe, J.I. (1987). A template for analyzing ethical dilemmas in discharge planning. Health & Social Work, 12(1), 47-56. Boland, K. (2006). Ethical decision-making among hospital social workers. Journal of Social Work Values and Ethics, 3(1). Retrieved from http://socialworker.com/jswve/ 127 Boutin-Foster, C., Euster, S., Rolon, Y., Motal, A., BeLue, R., Kline, R., & Charlson, M.E. (2005). Social work admission assessment tool for identifying patients in need of comprehensive social work evaluation. Health & Social Work, 30(2), 117-125. Bronstein, L., Kovacs, P., & Vega, A. (2007). Goodness of fit: Social work education and practice in health care. Social Work in Health Care, 45(2), 59-76. Brower, H.H., & Shrader, C.B. (2000). Moral reasoning and ethical climate: Not-for-profit vs. for-profit boards of directors. Journal of Business Ethics, 26(2), 147-167. Butcher, A.H. (1994). Supervisors matter more than you think: Components of a missioncentered organizational climate. Hospital & Health Services Administration, 39(4), 505520. Butler, B. (1990). Job satisfaction: Management's continuing concern. Social Work, 35, 112117. Callahan, D. (2003). Principlism and communitarianism. Journal of Medical Ethics, 29(5), 287291. Cannon, I. M. (1923). Social work in hospitals: A contribution to progressive medicine. New York, NY: Russell Sage Foundation. Cannon, I. M. (1952). On the social frontier of medicine: Pioneering in medical social service. Cambridge, MA: Harvard University Press. Caputi, M. (1978). Social work in health care: Past and future. Health & Social Work, 3(1), 929. Carlton, T.O. (1989). Education for health social work: Opportunities and constraints in schools and hospitals [Editorial]. Health & Social Work, 14(3), 147-152. Christ, G.H., Siegel, K., & Weinstein, L. (1995). Developing a research unit within a hospital social work department. Health & Social Work, 20(1), 60-69. Clark, P.G. (1997). Values in health care professional socialization: Implications for geriatric education in interdisciplinary teamwork. The Gerontologist, 37(4), 441-451. Clemens, E.L. (1995). Multiple perceptions of discharge planning in one urban hospital. Health & Social Work, 20(4), 254-261. Congress, E. (1999). Social work values and ethics: Identifying and resolving ethical dilemmas. Belmont, CA: Wadsworth Group/Thomson Learning. Corley, M.C., Minick, P., Elswick, R.K., & Jacobs, M. (2005). Nurse moral distress and ethical work environment. Nursing Ethics, 12(4), 381-390. 128 Cowles, L.A. (2003). Social work in the health field: A care perspective (2nd ed.). New York, NY: The Haworth Press. Cowles, L.A., & Lefcowitz, M.J. (1992). Interdisciplinary expectations of the medical social worker in the hospital setting. Health & Social Work, 17(1), 57-65. Cowles, L.A., & Lefcowitz, M.J. (1995). Interdisciplinary expectations of the medical social worker in the hospital setting: Part 2. Health & Social Work, 20(4), 279-286. Csikai, E.L. (1997). Social workers’ participation on hospital ethics committees: An assessment of involvement and satisfaction. Arete, 22(1), 1-13. Csikai, E.L. (1999). The role of values and experience in determining social workers’ attitudes toward euthanasia and assisted suicide. Social Work in Health Care, 30(1), 75-95. Csikai, E.L., & Bass, K. (2000). Health care social workers’ views of ethical issues, practice, and policy in end-of-life care. Social Work in Health Care, 32(2), 1-22. Csikai, E.L., & Sales, E. (1998). The emerging social work role of hospital ethics committees: A comparison of social work and chair perspectives. Social Work, 43(3), 233-242. Cullen, J.B., Parboteeah, K.P., & Victor, B. (2003). The effects of ethical climates on organizational commitment: A two-study analysis. Journal of Business Ethics, 46(2), 127-141. Cummings, S.M. (1999). Adequacy of discharge plans and rehospitalization among hospitalized dementia patients. Health & Social Work, 24(4), 249-259. Cummings, S.M., & Cockerham, C. (1997). Ethical dilemmas in discharge planning for patients with Alzheimer’s disease. Health & Social Work, 22(2), 101-108. Cunningham, G., & Sagas, M. (2004). Examining the main and interactive effects of deep- and surface-level diversity on job satisfaction and organizational turnover intentions. Organizational Analysis, 12(3), 319-332. Dabby, M., Faisal, H., Holliman, D., Karliner, S., & Silverman, B. (2008). Ethics as activity: Building collaborative, expansive, and just social work. Journal of Social Work Values and Ethics, 5(2). Retrieved from http://socialworker.com/jswve/ Davidson, K.W. (1990). Role blurring and the hospital social worker’s search for a clear domain. Health & Social Work, 15(3), 228-234. Davis, C. (2004). Hospital social work: Are we conducting the right type of research? Social Work in Health Care, 38(3), 67-79. 129 Davis, R.B. (1995). The principlism debate: A critical overview. The Journal of Medicine and Philosophy, 20(1), 85-105. Deshpande, S.P., & Joseph, J. (2009). Impact of emotional intelligence, ethical climate, and behavior of peers on ethical behavior of nurses. Journal of Business Ethics, 85, 403-410. Deshpande, S.P., Joseph, J., & Prasad, R. (2006). Factors impacting ethical behavior in hospitals. Journal of Business Ethics, 69(2), 207-216. DiFranks, N.N. (2008). Social workers and the NASW code of ethics: Belief, behavior, disjuncture. Social Work, 53(2), 167-176. Dillman, D. (2007). Mail and internet surveys: The tailored design method (2nd ed.). Hoboken, NJ: John Wiley & Sons. Dolgoff, R., Loewenberg, F.M., & Harrington, D. (2009). Ethical decisions for social work practice (8th ed.). Belmont, CA: Brooks/Cole. Dollinger, S., & Malmquist, D. (2009). Reliability and validity of single-item self-reports: With special relevance to college students' alcohol use, religiosity, study, and social life. The Journal of General Psychology, 136(3), 231-241. Doyle, O.Z., Miller, S.E., & Mizra, F.Y. (2009). Ethical decision-making in social work: Exploring personal and professional values. Journal of Social Work Values and Ethics, 6(1). Retrieved from http://socialworker.com/jswve/ Evans, J.H. (2000). A sociological account of the growth of principlism. Hastings Center Report, 30(5), 31-38. Feather, J. (1993). Factors in perceived hospital discharge planning effectiveness. Social Work in Health Care, 19(1), 1-14. Field, A. (2009). Discovering statistics using SPSS (3rd ed.). Thousand Oaks, CA: SAGE. Fleck-Henderson, A. (1991). Moral reasoning in social work practice. Social Service Review, 65(2), 185-201. Fletcher, C. (1973). Who should teach medical ethics? Hastings Center Report, 3(6), 4-6. Foster, L.W., Sharp, J., Scesny, A., McLellan, L., & Cotman, K. (1993). Bioethics: Social work’s response and training needs. Social Work in Health Care, 19(1), 15-38. Freud, S., & Krug, S. (2002a). Beyond the code of ethics, part I: Complexities of ethical decision making in social work practice. Families in Society, 83(5/6), 474-482. 130 Freud, S., & Krug, S. (2002b). Beyond the code of ethics, part II: Dual relationships revisited. Families in Society, 83(5/6), 482-492. Furlong, R.M. (1986). The social worker’s role on the institutional ethics committee. Social Work in Health Care, 11(4), 93-100. Furman, R. (2003). Frameworks for understanding value discrepancies and ethical dilemmas in managed mental health for social work in the United States. International Social Work Journal, 46(1), 37-52. Galambos, C. (1999). Resolving ethical conflicts in a managed health care environment. Health & Social Work, 24(3), 191-197. Gehlert, S. (2006). The conceptual underpinnings of social work in health care. In S. Gehlert & T.A. Browne (Eds.). Handbook of health and social work (pp. 3-22). Hoboken, NJ: John Wiley & Sons. Gellis, Z. (2001). Job stress among academic health center and community hospital social workers. Administration in Social Work, 25(3), 17-33. Gellis, Z. (2002). Coping with occupational stress in healthcare: A comparison of social workers and nurses. Administration in Social Work, 26(3), 37-52. Goldman, A., & Tabak, N. (2010). Perception of ethical climate and its relationship to nurses’ demographic characteristics and job satisfaction. Nursing Ethics, 17(2), 233-246. Goldstein, H. (1998). Education for ethical dilemmas in social work practice. Families in Society, 79(3), 241-253. Gracia, D. (2003). Ethical case deliberation and decision-making. Medicine, Health Care, and Philosophy, 6(3), 227-233. Grady, C., Danis, M., Soeken, K.L., O’Donnell, P., Taylor, C., Farrar, A., & Ulrich, C.M. (2008). Does ethics education influence the moral action of practicing nurses and social workers? The American Journal of Bioethics, 8(4), 4-11. Gray-Toft, P.A., & Anderson, J.G. (1985). Organizational stress in the hospital: Development of a model for diagnosis and prediction. Health Services Research, 19(6), 753-774. Greene, G.J., & Kulper, T. (1990). Autonomy and professional activities of social workers in hospital and primary health care settings. Health & Social Work, 15(1), 38-44. Gregorian, C. (2005). A career in hospital social work: Do you have what it takes? Social Work in Health Care, 40(3), 1-14. 131 Guo, L., & Schick, I.C. (2003). The impact of committee characteristics on the success of healthcare ethics committees. HEC Forum, 15(3), 287-299. Hadjistavropoulos, T. (1996). The systematic application of ethical codes in the counseling of persons who are considering euthanasia. Journal of Social Issues, 52(2), 169-188. Hall, D., Meador, K., & Koenig, H. (2008). Measuring religiousness in health research: Review and critique. Journal of Religion and Health, 47, 134-163. Hamric, A.B., & Blackhall, L.J. (2007). Nurse-physician perspectives on the care of dying patients in intensive care units: Collaboration, moral distress, and ethical climate. Critical Care Medicine, 35(2), 422-429. Hart, A.E. (2005). Hospital ethical climates and registered nurses’ turnover intentions. Journal of Nursing Scholarship, 37, 173-177. Hartsell, B.D. (2006). A model for ethical decision-making: The context of ethics. Journal of Social Work Values and Ethics, 3(1). Retrieved from http://www/socialworker.com/jswve/ Hellriegel, D., & Slocum, J.W. (1974). Organizational climate: Measures, research and contingencies. Academy of Management Journal, 17(2), 255-280. Hill, P., & Pargament, K. (2003). Advances in the conceptualization and measurement of religion and spirituality. American Psychologist, 58(1), 64-74. Hinman, L.M. (2008). Ethics: A pluralistic approach to moral theory (4th ed.). Belmont, CA: Thompson Wadsworth. Hoffmann, D., Tarzian, A., & O’Neil, J.A. (2000). Are ethics committee members competent to consult? Journal of Law, Medicine & Ethics, 28(1), 30-40. Holden, G.H., Cuzzi, L.F., Grob, G.G., & Bazer, C. (1995). Decisions regarding the order of opening multiple high-risk cases: A pilot study in an urban hospital. Social Work in Health Care, 22(1), 37-55. Holland, T.P., & Kilpatrick, A.C. (1991). Ethical issues in social work: Toward a grounded theory of professional ethics. Social Work, 36(2), 138-144. Holliman, D., Dziegielewski, S.F., & Teare, R. (2003). Differences and similarities between social work and nurse discharge planners. Health & Social Work, 28(3), 224-231. Holstein, M. (1995). Multidisciplinary ethical decision-making: uniting different professional perspectives. Journal of Elder Abuse & Neglect, 7(2/3), 169-182. 132 Hurlbutt, M.E. (1934). The rise of social work. Annals of the American Academy of Political and Social Science, 176, 1-13. James, L.R., James, L.A., & Ashe, D.K. (1990). The meaning of organizations: The role of cognitions and values (pp. 40-84). In B. Schneider (Ed.). Organizational climate and culture. San Francisco: Jossey-Bass. Jansen, E., & Chandler, G. (1994). Innovation and restrictive conformity among hospital employees: individual outcomes and organizational considerations. Hospital & Health Services Administration, 39(1), 63-80. Jansson, B.S., & Dodd, S.J. (1998). Developing a social work research agenda on ethics in health care. Health & Social Work, 23(1), 17-23. Jansson, B.S., & Dodd, S.J. (2002). Ethical activism: Strategies for empowering medical social workers. Social Work in Health Care, 36(1), 11-28. Jecker, N.S. (2008). Reasoning in ethics. In D.M. Hester (Ed.). Ethics by committee: A textbook on consultation, organization, and education for hospital ethics committees (pp. 27-47). Lanham, MD: Rowman & Littlefield. Jonsen, A.R. (1993). The birth of bioethics [Special Supplement]. Hastings Center Report, 23(6), S1-S4. Jonsen, A.R. (2000). A short history of medical ethics. New York, NY: Oxford University Press. Jonsen, A.R., Siegler, M., & Winslade, W.J. (2006). Clinical ethics: A practical approach to ethical decisions in clinical medicine (6th ed.). New York, NY: McGraw-Hill Joseph, J., & Deshpande, S.P. (1997). The impact of ethical climate on job satisfaction of nurses. Health Care Management Review, 22(1), 76-81. Joseph, M.V., & Conrad, A.P. (1989). Social work influence on interdisciplinary ethical decision making in health care settings. Health & Social Work, 14(1), 22-30. Judd, R., & Sheffield, S. (2010). Hospital social work: Contemporary roles and professional activities. Social Work in Health Care, 49, 856-871. Kadushin, G., & Kulys, R. (1993). Discharge planning revisited: What do social workers actually do in discharge planning? Social Work, 38(6), 713-726. Kadushin, G., & Kulys, R. (1995). Job satisfaction among social work discharge planners. Health & Social Work, 20(3), 174-186. 133 Kanoti, G., & Vinicky, J. (1987). The role and structure of hospital ethics committees. In G. Anderson & V. Glesnes-Anderson (Eds.). Health care ethics: A guide for decision makers (pp. 293-307). Rockville, MD: Aspen. Kaplan, L.A. (2006). Moral reasoning of MSW social workers and the influence of education. Journal of Social Work Education, 42(3), 507-522. Keigher, S.M. (1997). What role for social work in the new health care paradigm? [National Health Line]. Health & Social Work, 22(2), 149-155. King, J.E., & Crowther, M.R. (2004). The measurement of religiosity and spirituality: Examples and issues from psychology. Journal of Organizational Change, 17(1), 83-101. Kossman, H.D., Lamb, J.M., O’Brien, M.W., Predmore, S.M., & Prescher, M.J. (2005). Measuring productivity in medical social work. Social Work in Health Care, 42(1), 1-16. Kugelman, W. (1992). Social work ethics in the practice arena: A qualitative study. Social Work in Health Care, 17(4), 59-77. Kultgen, J. (1988). Ethics and professionalism. Philadelphia, PA: University of Pennsylvania Press. Landau, R. (2000a). Ethical dilemmas in general hospitals: Differential perceptions of direct practitioners and directors of social services. Social Work in Health Care, 30(4), 25-44. Landau, R. (2000b). Ethical dilemmas in general hospitals: Social workers’ contribution to ethical decision-making. Social Work in Health Care, 32(2), 75-92. Lane, M.S., & Schaupp, D. (1989). Ethics in education: A comparative study. Journal of Business Ethics, 8(12), 943-949. Levi, B.H. (1996). Four approaches to doing ethics. The Journal of Medicine and Philosophy, 21(1), 7-39. Levine, C. (1984). Questions and (some very tentative) answers about hospital ethics committees [A New and Largely Untested Approach]. The Hastings Center Report, 14(3), 9-12. Linzer, N. (1999). Resolving ethical dilemmas in social work practice. Boston, MA: Allyn & Bacon. Loewenberg, F.M., & Dolgoff, R. (1992). Ethical decisions for social work practice (4th ed.). Itasca, IL: F.E. Peacock. Loewy, E.H., & Loewy, R.S. (2004). Textbook of healthcare ethics (2nd ed.). Dordrecht, The Netherlands: Kluwer Academic. 134 Luptak, M. (2004). Social work and end-of-life care for older people: A historical perspective. Health & Social Work, 29(1), 7-15. Lyon, H.L., & Ivancevich, J.M. (1974). An exploratory investigation of organizational climate and job satisfaction in a hospital. Academy of Management Journal, 17(4), 635-648. Mahoney, A., Pargament, K., Tarakeshwar, N., & Swank, A. (2008). Religion in the home in the 1980s and 1990s: A meta-analytic review and conceptual analysis of links between religion, marriage, and parenting. Psychology of Religion and Spirituality, S(1), 63-101. Malloy, D.C., & Agarwal, J. (2003). Factors influencing ethical climate in a nonprofit organisation: An empirical investigation. International Journal of Nonprofit and Voluntary Sector Marketing, 8(3), 224-250. Manetta, A.A., & Wells, J.G. (2001). Ethical issues in the social worker’s role in physicianassisted suicide. Health & Social Work, 26(3), 160-166. Manning, S.S. (1997). The social worker as moral citizen: Ethics in action. Social Work, 42(3), 223-230. Mattison, M. (2000). Ethical decision-making: The person in the process. Social Work, 45(3), 201-212. McCarthy, J. (2003). Principlism or narrative ethics: Must we choose between them? Journal of Medical Ethics, 29(2), 65-71. McDaniel, C. (1997). Development and psychometric properties of the ethics environment questionnaire. Medical Care, 35(9), 901-914. McDaniel, C. (1998). Ethical environment: Reports of practicing nurses. Nursing Clinics of North America, 33(3), 363-372. McDaniel, C., Schoeps, N., & Lincourt, J. (2001). Organizational ethics: Perceptions of employees by gender. Journal of Business Ethics, 33(3), 245-256. McGee, G., Spanogle, J.P., Caplan, A.L., Penny, D., & Asch, D.A. (2002). Successes and failures of hospital ethics committees: A National survey of ethics committee chairs [Breaking Bioethics]. Cambridge Quarterly of Healthcare Ethics, 11, 87-93. Meacham, M.G. (2007). Ethics and decision making for social workers. Journal of Social Work Values and Ethics, 4(3). Retrieved from http://socialworker.com/jswve/ Mizrahi, T., & Berger, C.S. (2005). A longitudinal look at social work leadership in hospitals: The impact of a changing health care system. Health & Social Work, 30(2), 155-165. 135 Nacman, M. (1976). Social work in health settings: A historical review. Social Work in Health Care, 2(4), 407-418. National Association of Social Workers (2006). NASW code of ethics. Washington, DC: Author. Nelson, K.R., & Merighi, J.R. (2003). Emotional dissonance in medical social work practice. Social Work in Health Care, 36(3), 63-79. Neuman, K.M. (2003). Developing a comprehensive outcomes management program: A tenstep process. Administration in Social Work, 27(1), 5-23. O'Donnell, P., et al. (2008). Predictors of ethical stress, moral action, and job satisfaction in health care social workers. Social Work in Health Care, 46(3), 29-51. Olson, L.L. (1998). Hospital nurses’ perceptions of the ethical climate of their work setting. Journal of Nursing Scholarship, 30(4), 345-349. Osman, H., & Perlin, T.M. (1994). Patient self-determination and the artificial prolongation of life. Health & Social Work, 19(4), 245-252. Ott, J.S. (1989). The organizational culture perspective. Chicago: Dorsey Press. Parker, C.P., Baltes, B.B., Young, S.A., Huff, J.W., Altmann, R.A., Lacost, H.A., & Roberts, J.E. (2003). Relationships between psychological climate perceptions and work outcomes: A meta-analytic review. Journal of Organizational Behavior, 24, 389-416. Parsons, R. (2001). The ethics of professional practice. Boston, MA: Allyn & Bacon. Peterson, D.K. (2002). The relationship between unethical behavior and the dimensions of the ethical climate questionnaire. Journal of Business Ethics, 41(4), 313-326. Peterson, S. (1987). Professional codes and ethical decision-making. In G. Anderson & V. Glesnes-Anderson (Eds.). Health care ethics: A guide for decision makers (pp. 321-329). Rockville, MD: Aspen. Pettigrew, A.M. (1990). Conclusion: Organizational climate and culture: Two constructs in search of a role (pp. 413-433). In B. Schneider (Ed.). Organizational climate and culture. San Francisco: Jossey-Bass. Ponton, R.F., & Duba, J.D. (2009). The ACA code of ethics: Articulating counseling’s professional covenant. Journal of Counseling & Development, 87(1), 117-121. Post, L., Blustein, J., & Dubler, N. (2007). Handbook for health care ethics committees. Baltimore, MD: The Johns Hopkins University Press. 136 Proctor, E.K., Morrow-Howell, N., & Lott, C.L. (1993). Classification and correlates of ethical dilemmas in hospital social work. Social Work, 38(2), 166-177. Rachels, S. (2007). The elements of moral philosophy (5th ed.). NY: McGraw Hill. Rathert, C., & Fleming, D.A. (2008). Hospital ethical climate and teamwork in acute care: The moderating role of leaders. Health Care Management Review, 33(4), 323-331. Rawwas, M.Y.A., Strutton, D., & Pelton, L. (1994). Ethical attitudes of mental health practitioners: Balancing therapeutic practices and treatments. Journal of Business Ethics, 13(8), 597-608. Reamer, F.G. (1979). Fundamental ethical issues in social work: An essay review. Social Service Review, 53(2), 229-242. Reamer, F.G. (1985). The emergence of bioethics in social work. Social Work, 10(4), 271-281. Reamer, F.G. (1987). Ethics committees in social work. Social Work, 32(3), 188-192. Reamer, F.G. (1993). The philosophical foundations of social work. New York, NY: Columbia University Press. Reamer, F.G. (1998). The evolution of social work ethics. Social Work, 43(6), 488-500. Reamer, F.G. (2006a). Social work values and ethics (3rd ed.). New York, NY: Columbia University Press. Reamer, F.G. (2006b). Ethical standards in social work: A review of the NASW code of ethics (2nd ed.). Washington, DC: NASW Press. Reichers, A.E., & Schneider, B. (1990). Climate and culture: An evolution of constructs (pp. 539). In B. Schneider (Ed.). Organizational climate and culture. San Francisco: JosseyBass. Rhodes, M. (1991). Ethical dilemmas in social work practice. Milwaukee, WI: Family Service America. Sanders, S., & Hoffman, K. (2011). Ethics education in social work: Comparing outcomes of graduate social work students. Journal of Social Work Education, 46(1), 7-22. Schluter, J., Winch, S., Holzhauser, K., & Henderson, A. (2008). Nurses’ moral sensitivity and hospital ethical climate: A literature review. Nursing Ethics, 15(3), 304-321. Schneider, B. (Ed.). (1990). Organizational climate and culture. San Francisco: Jossey-Bass. 137 Schonlau, M., Fricker, R., & Elliott, M. (2002). Conducting research surveys via email and the web. Santa Monica, CA: RAND. Schwepker, C.H. (2001). Ethical climate’s relationship to job satisfaction, organizational commitment, and turnover intention in the sales force. Journal of Business Research, 54, 39-52. Siefert, K., Jayaratne, S., & Chess, W (1991). Job satisfaction, burnout, and turnover in health care social workers. Health & Social Work, 16, 193-202. Silverman, E. (1992). Hospital bioethics: A beginning knowledge base for the neonatal social worker. Social Work, 37(2), 150-154. Simon, E.P. (1991). Research for the research phobic: Developing research expertise in hospital social work. Health & Social Work, 16(2), 118-122. Sims, R.L., & Keon, T.L. (1999). Determinants of ethical decision-making: The relationship of the perceived organizational environment. Journal of Business Ethics, 19(4), 393-401. Sommers-Flanagan, R., & Sommers-Flanagan, J. (2007). Becoming an ethical helping professional: Cultural and philosophical foundations. Hoboken, NJ: John Wiley & Sons. Sparks, J. (2006). Ethics and social work in health care. In S. Gehlert & T.A. Browne (Eds.). Handbook of health and social work (pp. 43-69). Hoboken, NJ: John Wiley & Sons. Strom-Gottfried, K. (2007). Straight talk about professional ethics. Chicago, IL: Lyceum. Stuart, P. H. (2004). Individualization and prevention: Richard C. Cabot and early medical social work. Social Work in Mental Health, 2(2/4), 7-20. Sue, V.M., & Ritter, L.A. (2007). Conducting online surveys. Thousand Oaks, CA: SAGE. Sze, W., & Ivker, B. (1986). Stress in social workers: The impact of setting and role. Social Casework, 67, 141-148. Tabachnick, B.G., & Fidell, L.S. (2007). Using multivariate statistics (5th ed.). Boston, MA: Pearson Education/Allyn & Bacon. Todd, A.J. (1930). Common grounds in social case work. Social Forces, 8(4), 550-557. Trevino, L.K., Butterfield, K.D., & McCabe, D.L. (1998). The ethical context in organizations: Influences on employee attitudes and behaviors. Business Ethics Quarterly, 8(3), 447476. Tyson, T. (1990). Believing that everyone is less ethical: Implications for work behavior and ethics instruction. Journal of Business Ethics, 9(9), 715-721. 138 U.S. Census Bureau (n.d.). Census 2000 geographic definitions. Available at http://www.census.gov/geo/www/geo_defn.html U.S. Department of Labor, Bureau of Labor and Statistics (2006). National employment matrix, employment by industry, occupation, and percent distribution. Retrieved June 21, 2008 from ftp://ftp.bls.gov/pub/special.requests/ep/ind-occ.matrix/occ_pdf/occ_21-1022.pdf Ulrich, C., O’Donnell, P., Taylor, C., Farrar, A., Danis, M., & Grady, C. (2007). Ethical climate, ethics stress, and the job satisfaction of nurses and social workers in the United States. Social Science & Medicine, 65, 1708-1719. Ulrich, C.M., & Soeken, K.L. (2005). A path analytic model of ethical conflict in practice and autonomy in a sample of nurse practitioners. Nursing Ethics, 12(3), 305-316. Vandenberghe, C. (1999). Organizational culture, person-culture fit, and turnover: A replication in the health care industry. Journal of Organizational Behavior, 20, 175-184. Vardi, Y. (2001). The effects of organizational and ethical climates on misconduct at work. Journal of Business Ethics, 29(4), 325-337. Victor, B., & Cullen, J.B. (1988). The organizational bases of ethical work climates. Administrative Science Quarterly, 33(1), 101-125. Vidaver-Cohen, D. (1998). Moral climate in business firms: A conceptual framework for analysis and change. Journal of Business Ethics, 17, 1211-1226. Vinokur-Kaplan, D. (1991). Job satisfaction among social workers in public and voluntary child welfare agencies. Child Welfare, 155, 81-91. Wade, K., & Neuman, K. (2007). Practice-based research: Changing the professional culture and language of social work. Social Work in Health Care, 44(4), 49-64. Walden, T., Wolock, I., & Demone, H.W. (1990). Ethical decision making in human services: A comparative study. Families in Society, 71(2), 67-75. Walker, T. (2009). What principlism misses. Journal of Medical Ethics, 35, 229-231. Wallace, M.J., Ivancevich, J.M., & Lyon, H.L. (1975). Measurement modifications for assessing organizational climate in hospitals. Academy of Management Journal, 18(1), 82-97. Walsh-Bowers, R., Rossiter, A., & Prilleltensky, I. (1996). The personal is the organizational in the ethics of hospital social workers. Ethics & Behavior, 6(4), 321-335. Walz, T., & Ritchie, H. (2000). Gandhian principles in social work practice: Ethics revisited. Social Work, 45(3), 213-222. 139 Wenston, S.R. (1987). Applying philosophy to ethical dilemmas. In G. Anderson & V. GlesnesAnderson (Eds.). Health care ethics: A guide for decision makers (pp. 22-33). Rockville, MD: Aspen. Wimberley, E.T. (1988). Using productivity measures to avoid reductions in force [Briefly Stated]. Social Work, 33(1), 60-61. Zussman, R. (1997). Sociological perspectives on medical ethics and decision-making. Annual Review of Sociology, 23, 171-189. 140