University of Southern California Dissertations and Theses

A longitudinal study of job resources, burnout, and turnover among social workers: implications for evidence-based organizational practices

(USC Thesis Other)

A longitudinal study of job resources, burnout, and turnover among social workers: implications for evidence-based organizational practices

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content A LONGITUDINAL STUDY OF JOB RESOURCES, BURNOUT, AND TURNOVER AMONG SOCIAL WORKERS: IMPLICATIONS FOR EVIDENCE-BASED ORGANIZATIONAL PRACTICES by Hansung Kim A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (SOCIAL WORK) May 2008 Copyright 2008 Hansung Kim ii Acknowledgements I sincerely thank my parents for their faith in me and providing me with unending encouragement and support. I would like to gratefully thank Dr. Madeleine R. Stoner for her guidance, and most importantly, her patience during my doctoral program at USC. She inspired me to not only grow as a social work scholar but also as an independent thinker. It would not be possible for me to conduct my three-wave longitudinal survey research without her understanding and support. I would like to also thank Dr. Robert H. Nishimoto for his guidance and friendship during my program. He wrestled with my research and gave guidance regarding the structuring of my chaotic thinking. His encouragement and help made me feel confident to overcome every difficulty I encountered. Very special thanks are given to Dr. Devon Brooks. His unselfish help enabled me to develop the necessary research skills to undertake and complete the work. Additionally, I am also grateful for the friendship of all of members of his research group, especially Alice Kim and Dr. Jan Nissly. I would like to thank Dr. Peter Robertson for his valuable advice and comments, which enabled me to develop the theoretical framework of my dissertation. I also thank my doctoral colleagues for their generous help and support: particularly Marissa Hansen, Juye Ji, Dennis Kao, Dr. Jane Yoo and so many other colleagues. I would like to thank Dr. Kyu-Taik Sung. His wisdom, knowledge and commitment to the scholarly work inspired and motivated me. Finally, and most importantly, I would like to thank my wife Sun-Young Lee for her tremendous support, encouragement, patience and unwavering love iii Table of Contents ACKNOWLEDGEMENTS ii LIST OF TABLES vi LIST OF FIGURES vii ABSTRACT viii CHAPTER I: INTRODUCTION 1 1.1 PROBLEM STATEMENT 1 1.2 PREVIOUS STUDIES AND KNOWLEDGE GAPS 2 1.3 PURPOSE OF THE RESEARCH 8 1.4 FORMAT OF THE DISSERTATION 9 CHAPTER II: LITERATURE REVIEW 10 2.1 BURNOUT AND TURNOVER: DEFINITION AND SIGNIFICANCE 10 2.1.1 Definition of Burnout 10 2.1.2 The Significance of Social Worker Burnout 12 2.1.3 The Significance of Social Worker Turnover 14 2.1.4 The Relationships among Job Demands, Burnout, and Turnover Intention 16 2.2. JOB RESOURCES: THEORETICAL BACKGROUNDS 19 2.2.1 Stress and Coping Theory 19 2.2.2 Resource Theory and Job Resources 21 2.3 JOB RESOURCES FOR SOCIAL WORKERS 24 2.3.1 Job Autonomy 25 2.3.2 Effective Supervision 26 2.3.3 Social Support at Work 27 2.4 FUNCTIONS OF JOB RESOURCES 29 2.4.1 Edifying Professional Roles 30 2.4.2 Empowering Social Workers 31 2.4.3 Embedding Social Workers 34 2.5 FUNCTIONS OF JOB RESOURCES AND PROFESSIONAL DEVELOPMENT 37 2.5.1 Professional Development and Social Work Licensure 38 2.5.2 Job Resources for Social Workers in Different Stages of Professional Development 41 2.6 HYPOTHESES, CONCEPTUAL MODEL, AND RESEARCH QUESTIONS 42 2.6.1 Study Hypotheses and Conceptual Model 42 2.6.2 Research Questions 47 CHAPTER III: METHODS 48 iv 3.1 RESEARCH DESIGN AND PROCEDURE 48 3.2 SAMPLE 50 3.3 MEASURES 53 3.4 ANALYSIS 58 3.4.1 Overview of Analytical Methods 58 3.4.2 Preliminary Analysis 59 3.4.3 Descriptive Statistics and Correlations 61 3.4.4 Examining Study Hypotheses 62 3.4.5 Group Comparison 65 CHAPTER IV: RESULTS 66 4.1 RESULTS OF TOTAL SAMPLE 66 4.1.1 Preliminary Analysis 66 4.1.2 Descriptive Statistics and Correlations 69 4.1.3 Hypothesis Testing 70 4.2 RESULTS OF THE GROUP COMPARISON OF ASW AND LCSW 76 4.2.1 Group Characteristics 76 4.2.2 Levels of Burnout and Turnover Intention 78 4.2.3 Functions of Job Resources 79 CHAPTER V: DISCUSSION 86 5.1 MAJOR FINDINGS 86 5.1.1 Concurrent and Longitudinal Relationships among Job Demands, Burnout, and Turnover Intention 86 5.1.2 Functions of Job Autonomy and Supervisory Support 87 5.1.3 Importance of Effective Supervision 91 5.2 LIMITATIONS AND STRENGTHS 92 5.2.1 Limitations 92 5.2.2 Strengths 94 5.3 IMPLICATIONS FOR PRACTICE, EDUCATION, AND RESEARCH 95 5.3.1 Implications for Social Work Education 95 5.3.2 Implications for Organizational Practices 96 5.3.3 Implications for Social Work Research toward Evidence-Based Organizational Practices 99 5.4 CONCLUSION 102 REFERENCES 103 APPENDICES 121 Appendix A. COVER LETTER FOR THE FIRST SURVEY 121 Appendix B. INFORMATION SHEET FOR THE FIRST SURVEY 123 Appendix C. SURVEY QUESTIONNAIRE FOR THE FIRST SURVEY 126 v Appendix D. COVER LETTER FOR THE SECOND SURVEY 132 Appendix E. INFORMATION SHEET FOR THE SECOND SURVEY 133 Appendix F. SURVEY QUESTIONNAIRE FOR THE SECOND SURVEY 136 vi List of Tables Table 1: Conceptual Definition of Study Variables 46 Table 2: Respondents’ Characteristics 52 Table 3: Binary Logistic Regression Analysis 67 Table 4: Independent Sample T-test for Top Management Support 68 Table 5: Independent Sample T-test for Turnover Intention at Time 1 68 Table 6: Correlation Coefficient among Study Variables 69 Table 7: Fit Statistics for Comparative Models of Job Demands, Burnout, and 71 Turnover Intention Table 8: Chi-Square Difference between Freeing Paths to Differ Between 75 Groups and Constraining Paths to be Equal Across Groups Table 9: Demographic Characteristics of ASWs and LCSWs 77 Table 10: Between-Group Comparisons on Levels Burnout and 79 Turnover Intention Table 11:Chi-Square Difference between Freeing Paths to Differ Between 82 Groups and Constraining Paths to Be Equal Across Groups for Associate Social Workers Table 12: Chi-Square Difference between Freeing Paths to Differ Between 85 Groups and Constraining Paths to Be Equal Across Groups for Licensed Social Workers vii List of Figures Figure 1: Proposed Conceptual Model of Functions of Job Resources 45 Figure 2: The Two-wave, Two-variable Cross-lagged Panel Model 58 Figure 3: A Cross-lagged Model for Testing Longitudinal Relations between 64 Job Demands, Burnout, and Turnover Intention (Model1) Figure 4: Revised Models (Mode 2 and Model 3) 71 Figure 5: Standardized Path Coefficients for Model 3 with Total Sample 73 Figure 6: Regression Lines between Job Demands at Time 1 and Turnover 74 Intention at Time 2 for Two Groups of Job Autonomy Figure 7: Standardized Path Coefficient for Model 3 with Associate 80 Social Workers Figure 8: Standardized Path Coefficient for Model 3 with Licensed Clinical 81 Social Workers Graph 9: Regression Lines between Job Demands at Time 1 and Turnover 84 Intention at Time 2 for Two Groups of Supervisory Support with a data of ASWs Graph 10: Regression Lines between Burnout at Time 1 and Turnover 84 Intention at Time 2 for Two Groups of Supervisory Support with a data of ASWs viii Abstract The purpose of present study is to understand which job conditions is critical for social workers not only to prevent burnout, but also to cope with burnout experience. My central thesis is that different job resources can have distinctive functions such as 1) edifying roles and objectives of social workers, 2) empowering social workers to cope with job demands, and 3) embedding social workers in their organizations, depending on worker characteristics. In order to provide empirical evidence for these proposed functions of job resources, the current study tests longitudinal relationships among job demands, burnout, and turnover intention; 2) to test moderating effects of job resources in the framework of job demands-burnout- turnover intention; and 3) to explore how the moderating effects of job resources are influenced by social workers’ professional development. Study hypotheses and research questions are formulated and examined using a cross-lagged panel design and structural equation modeling (SEM) techniques. This study includes a sample of 365 social workers identified from a longitudinal survey of California registered social workers. Results showed that effective supervision was negatively associated with job demands, burnout, and turnover intention among social workers. Regarding the longitudinal relationships, job demands had positive causal effects on turnover intention only if job autonomy was low. Significant group differences were found across subgroups of Licensed Clinical Social Workers (LCSWs) and Associated Social Workers (ASWs). Supervisory support buffered the effects of job demands and burnout on turnover intention among Associate Social Workers; however, in the ix case of LCSWs, job resources did not have significant moderating effects in the framework of job demands-burnout-turnover intention. The current findings suggest that organizational practices should focus on developing effective supervision and enhancing job autonomy in frontline workers. Especially, organizational practices should focus on supporting social workers at the beginning of their professional careers. A supervisor plays a vital role helping inexperienced social workers cope with job demands and emotional distress at work. Implications for organizational practices, social work educations, and future research on social worker burnout are discussed. 1 CHAPTER I: INTROCUTION 1.1 Problem Statement Social work is an important profession that delivers vital social services to a vast array of populations in need of services. According to a report from the Bureau of Labor Statistics in the U.S. Department of Labor, social workers held about 562,000 jobs in 2004. Areas of practice within social work are quite broad and include mental health, child welfare/family, health, aging, and substance abuse. Job settings also vary, ranging from employment in hospitals, schools, private social service agencies, state and government agencies, and private practices. A recent report on the difficulties of the social work profession indicated that social workers are currently experiencing high levels of job demands such as increasing paperwork, unmanageable caseloads, and problems with difficult clients, as well as staff shortages and reduced availability of adequate supervision (Center for Workforce Studies, NASW, 2006). Furthermore, confusing legislation and concomitant guidelines have increased the conflicting and incompatible job demands of social workers (Bransford, 2005). The extant literature on burnout has suggested that these job demands are significant antecedents of social worker burnout (Söderfeldt, Söderfeldt, & Warg, 1995). Moreover, workers who feel burned out and 2 frustrated with their jobs are more likely to have higher turnover and be absent from work (De Croon, Sluiter, Blonk, Broersen & Frings-Dresen, 2004). Social worker turnover is a serious problem for social work administrators because social worker turnover negatively affects the quality, consistency, and stability of client services (Mor Barak, Nissly, & Levin, 2001). Worker turnover not only causes psychological distress in remaining staff members or new and inexperienced workers who fill the vacated positions (Powell & York, 1992), but it also leads to client mistrust of the system (Geurts, Schaufeli, & De Jonge, 1998) and financial problems for the organization (Kompier & Cooper, 1999). In light of the implications of burnout and high turnover rates in the social work profession, two crucial questions for human service managers and social work researchers arise: How do we prevent burnout among social workers and how do we retain these workers? 1.2 Previous Studies and Knowledge Gaps In previous literature, it is consistently suggested that job demands such as role overload, role conflict, and role ambiguity (see Table 1 for the definition of each construct) cause worker burnout (Cordes & Dougherty, 1993; Lee & Ashforth, 1996), and burned out workers are more likely to have higher levels of turnover intention (Harrington, Bean, Pintello, & Mathews, 2001; Huang, Chuang, & Lin, 2003). Knowing that certain job demands result in burnout and turnover, many 3 previous studies have explored ways to help workers cope with demanding jobs. Pearlin (1991) found that stress arising in the context of a bureaucratic workplace is an example of stress that cannot be resolved by the efforts of individuals. He suggested that, in this situation, coping strategies at the group and organizational levels may be the desired way to reduce the strain from stress. Previous researchers (e.g., Karasek & Theorell, 1990; Demerouti, Bakker, Nachreiner, & Schaufeli, 2001) have suggested that two critical job conditions influence job strain outcomes (e.g., burnout and turnover intention): job demands (e.g., role conflict, role ambiguity, role overload) and job resources (e.g., effective supervision, social support, job autonomy). The extant research studies have shown that both job demands and resources have unique main effects on workers’ burnout and turnover intention (Houkes, Janssen, de Jonge, & Bakker, 2003; Lee & Ashforth, 1996; Mor Barak et al., 2001; Söderfeldt, Söderfeldt, & Warg, 1995; Um & Harrison, 1998). Another group of studies have focused on understanding how job demands and job resources interact in explaining burnout or turnover intention (Bakker, Demerouti, & Euwema, 2005; Dollard, Winefield, Winefield, & de Jonge, 2000; Nissly, Mor Barak, & Levin, 2005; Posig & Kickul, 2003). Studies that focus on the interaction effects of job conditions are based on the hypothesis which postulates that the relationship between job demands and strain outcomes will be weaker for those in work places with resources to better cope with stress (Cordes & Doughety, 1993). This idea has been empirically supported by researchers showing that job resources moderate the relationship between job demands and burnout (Bakker, Demerouti, & Euwema, 2005; Dollard, Winefield, Winefield, & de Jonge, 4 2000; Posig & Kickul, 2003) as well as the relationship between job demands and turnover intention (Kim & Stoner, in press; Nissly, Mor Barak, & Levin, 2005). Previous studies on job resources suggest that resourceful job conditions help social workers better cope with their job demands. Based on this view, preliminary studies were conducted prior to the one reported here. In the first study, role-specific communication between social workers and their supervisors was found to directly reduce levels of job demands perceived by social workers (Kim & Lee, in press). In addition, when they experience high levels of job demands, upward communication was essential for social workers to cope with their job demands. Upward communication allows social workers to question organizational procedures or supervisors’ instructions and to exchange their own opinions with immediate supervisors (Kim & Lee, in press). The second study (Kim & Stoner, in press) examined the moderating effects of social support and job autonomy on the relationship 1) between job demands and burnout and 2) between job demands and turnover intention among social workers. The study found that social workers with higher job demands combined with lower job autonomy over demanding situations resulted in higher levels of burnout. In addition, although social workers perceive high levels of job demands, a supportive working environment helps maintain the social worker’s attachment to her/his organization. Unexpectedly, social support was not a significant moderator of the positive relationship between job demands and burnout. This seemed to indicate that social workers may be burned out from their demanding helping role even though they have adequate support from various organizational members. However, it is not 5 clear how social support helps social workers to maintain their attachment to their organizations even though they experience burnout, since how social support affects the relationship between burnout and turnover intention among social workers has not yet been examined. Therefore, there is a need to develop and empirically examine a conceptual model explaining how helpful job conditions such as job autonomy and social support help social workers not only to cope with job demands, but also to overcome emotional distress at work. In this respect, using the same cross-sectional data from the second study, the third study (Kim & Stoner, 2007) examined three different functions of resourceful job conditions in the framework of job demands-burnout-turnover relationships: 1) reducing perceived job demands; 2) buffering effects of job demands on burnout; and 3) buffering effects of job demands and burnout on turnover intention. Role- specific supervisory communication was significant in reducing perceived levels of job demands, and job autonomy and upward supervisory communication moderated the relationship between job demands and burnout. Finally, top managerial support significantly moderated the relationship between burnout and turnover intention among social workers. This suggests that supervision and job autonomy are critical for preventing burnout, while top managerial leadership is important to social workers’ attachment to their organizations. Although previous research studies have contributed substantially, there are knowledge gaps in the literature and these limitations make it difficult to comprehensively understand burnout and turnover among social workers. First, Few studies on social worker burnout and turnover have employed longitudinal research 6 design (e.g., Drake & Yadama, 1996; Nissly, 2004, Savicki & Cooley, 1994), and the lack of longitudinal studies on the issue makes it difficult to draw inferences about how job demands, burnout and turnover intention are casually associated with each other. Schaufeli and Enzmann (1998) reviewed longitudinal studies on burnout and found that many longitudinal studies failed to reproduce the results found in cross- sectional studies. Therefore, to fill the knowledge gaps, the use of a longitudinal research design is needed in order to examine the relationship between job demands, burnout, and turnover intention among social workers. In addition, few research studies have examined the different functions of job resources in a single conceptual model of burnout and turnover. Communication theory (e.g., social information processing theory, Salancik & Pfeffer, 1978; uncertainty reduction theory, Berger & Calabrese, 1975), cognitive empowerment theory (Thomas & Velthouse, 1990), and job embeddedness theory (Mitchell, Holtom, Lee, Sablynski, & Erez, 2001) suggest different functions of job resources in the framework of burnout and turnover. Communication theory explains how effective supervision reduces the role-related demands perceived by social workers. Empowerment theory focuses on why specific job resources are helpful for social workers in coping with their job demands and burnout experience, while the theory of job embeddedness focuses on how job resources influence workers’ decisions to remain in their organizations. According to these theories, job resources can: 1) edify the roles and objectives of social workers and minimize ambiguity and conflict associated with their helping-oriented roles, 2) empower social workers to cope with job demands and emotional stress, and 3) embed social workers in their 7 organizations. Developing a conceptual model elaborating various functions of job resources is important for developing a better understanding of burnout and turnover among social workers. In addition, most studies concerning social worker burnout have considered social workers as a single occupational group. It is generally known that social workers are more likely to experience burnout during the first two years (Cherniss, 1980). Lee and Ashforth (1993) found that inexperienced welfare workers are more likely to develop burnout compared with more experienced workers. A progression that social workers cumulate professional experience and knowledge at work can be conceptualized as a process of professional development. Tomas (1982) suggests that workers at lower levels of professional development are primarily concerned with survival through knowledge of agency rules and procedures required for maintaining service quality. Therefore, it is assumed that social workers at lower levels of professional development are more like to experience burnout compared with social workers at more advanced levels of professional development. However, a few studies have examined how professional development impacts social workers’ perceptions of job conditions or burnout experiences. Empirical studies identifying which job resources are critical for social workers at the early stages of professional development are still limited. 8 1.3 Purpose of the Research The purpose of this study was threefold: 1) to test the longitudinal relationships among job demands, burnout, and turnover intention among social workers; 2) to test the functions of job resources (e.g., supervision, job autonomy, and social support) in the framework of job demands-burnout-turnover intention; and 3) to understand how these functions of job resources are influenced by social workers’ professional development. Specifically, a model is proposed and empirically examined in order to explain the direct effects of supervision on perceived job demands as well as the moderating effects of job autonomy and social support on the relationships between job demands, burnout, and turnover intention among social workers. In addition, using the same model, the effects of job resources on social workers are compared at different levels of professional development, where licensing status is used as a proxy of professional development. Newly graduated MSWs (Associated Social Workers in California, ASW) are compared to those with supervised training (Licensed Clinical Social Workers in California, LCSW) in order to understand the relationship between functions of job resources and professional development. The conceptual model of hypothesized relationships is empirically examined using structural equation modeling (SEM) techniques. For the group comparison between ASWs and LCSWs, multi-group analysis techniques based on SEM are employed. This study included a sub-sample of 365 social workers (244 social workers at Time 2) identified from a longitudinal survey of state-registered social workers in California. 9 1.4 Formation of the Dissertation This dissertation is divided into six chapters. The first chapter is the introductory part. The second, the literature review, discusses the significance of burnout and turnover in the field of social work and the principles underlying the theoretical framework of job demands-burnout-turnover intention. The concept of job resources is also discussed. Resource theories as well as stress/coping theories provide theoretical backgrounds for understanding what job resources are. Next, three major job resources for social workers are discussed: job autonomy, effective supervision, and social support at work. The functions of these job resources in the framework of job demands-burnout-turnover intention are described based on communication theories, empowerment theory, and job embeddedness theory. In addition, it is discussed how the functions of job resources might be different depending on the professional development stages of social workers. Based on the literature review, the study hypotheses, a conceptual model describing the hypotheses, and the research questions are presented. The next chapter, the research methods section, describes sampling procedures, measures, and analysis plans. Results of the analysis (chapter 4) consist of two parts: an analysis of the total sample and a group comparison analysis. Finally, the last chapter (Chapter 5) summarizes the major findings and includes a discussion of implications for research and practice. 10 CHAPTER II: LITERATURE REVIEW 2.1 Burnout and Turnover: Definition and Significance 2.1.1 Definition of Burnout Burnout has been an important research topic in the field of occupational stress. The concept of burnout was fist discussed by Bradley (1969) and was further elaborated upon by Freudenberger (1980) based on his observations of extreme psychological strain often experienced by workers in the human service profession, including social workers (Cooper, Dewe, & O’Driscoll, 2001). Freudenberger (1980) defined burnout as a “state of fatigue or frustration brought about by devotion to a cause, way of life, or relationship that failed to produce the expected reward.” Pines, Aronson, and Kafry (1981) stated that burnout is “characterized by physical depletion, by feelings of helplessness and hopelessness, by emotional drain, and by the development of negative self-concept and negative attitudes toward work, life, and other people…” Maslach and Jackson (1981) defined burnout as “a syndrome of emotional exhaustion and cynicism that occurs frequently among individuals who do ‘people-work’ of some kind.” In addition, Maslach and Jackson (1981) addressed that burnout has three components: emotional exhaustion, depersonalization or cynicism, and diminished personal accomplishment. According to Maslach and Jackson’s (1981) conceptualization, emotional exhaustion represents feelings of being overextended and depleted of emotional and physical resources. Emotional 11 exhaustion has been considered as the essential aspect of burnout symptoms (Koeske & Koeske, 1989; Cordes, Dougherty, & Blum, 1997). When people describe themselves or others as experiencing burnout, they are usually referring to the experience of emotional exhaustion (Maslach, Schaufeli, & Leiter, 2001). The next component of burnout is depersonalization. Depersonalization or cynicism means a negative or excessively detached response to various aspects of the job, including clients, co-workers, and the organization in general. Depersonalization is often conceptualized as a coping response to emotional exhaustion (Cordes & Dougherty, 1993). Several empirical studies showed that depersonalization results from persistent emotional exhaustion (e.g., Koesk & Koeske, 1989; Cordes, Dougherty, & Blum, 1997). A final aspect of burnout syndrome, diminished personal accomplishment, is the tendency to evaluate oneself negatively, particularly with regard to one’s work with clients. Previous researchers have suggested that diminished personal accomplishment results, in part, from high levels of emotional exhaustion and depersonalization (Bakker, Schaufeli, Sixma, Bosveld, & Van Dierendonck, 2000; Cordes, Dougherty, & Blum, 1997; Toppinen-Tanner, Kalimo, & Mutanen, 2002). The majority of burnout studies have been based on Maslach and Jackson’s (1981) conceptualization of burnout. As suggested by Maslach and Jackson (1981), the present study defines burnout as a multi-dimensional construct. Examining burnout in terms of all three components would help us to look at the overall burnout phenomenon (Maslach, Schaufeli, & Leiter, 2001). 12 2.1.2 The Significance of Social Worker Burnout Historically, a social worker was considered as a profession at high risk of burnout. During the 1980s and 1990s, the field of social services experienced a continuous decline of the welfare state (Austin, 2003). As a result, the social work profession faced challenges such as political vulnerability and financial constraints (Hasenfeld, 1984). Facing such economic pressures, a major challenge for human service managers was to demonstrate their overall effectiveness and efficacy in programs and organizations. In such a working environment, social workers had to often follow not only program-specific standards defined by agencies, but also those outlined by funding sources. Dual standards are commonly found in managed care organizations and broader policies, such as those prescribed by the Government Performance and Results Act of 1993 (Patti, 2000). When organizations become so responsive to the standards of accountability imposed by various funding and policy bodies, frontline workers become confused and weary about meeting concomitant guidelines, and consequently are more likely to experience role-relevant stress (Arches, 1991). Over the past several decades, there has been considerable research on the phenomenon of burnout in the social work profession (see a review of burnout studies in the field of social work by Söderfeldt, Söderfeldt, & Warg, 1995). The majority of previous studies have consistently addressed stressful job conditions and emotional distress experienced in most social work positions. For example, Light (2003) used a survey of 1,213 human service workers in the U.S. and reported that 75 percent of these workers considered their work to be frustrating and 81 percent 13 agreeing that they often experienced burnout at work (Light, 2003). In addition, Priebe, Fakhoury, Hoffmann, and Powell (2005) described the extent of burnout among community mental health professionals in Berlin and London and found that social workers had a much higher burnout score than psychiatrists and community psychiatric nurses. Using the Maslach Burnout Inventory (MBI; Maslach & Jackson, 1986), a recent study of registered social workers in California (Kim & Stoner, in press) suggests that social workers are experiencing higher levels of emotional exhaustion compared to the normative information provided by Maslach and Jackson (1986). In the same study, the mean score of emotional exhaustion for the sample of social workers was 25.5 (SD=11.41), which is significantly higher than those of the subjects on whom the MBI (Maslach & Jackson, 1986) was normed (M=20.99, SD=10.75) (t=7.67, df=345, p<.05) and even the subgroup of social service workers in Maslach and Jackson (1986)’s normative sample (M=21.35, SD=10.51) (t=7.08, df=345, p<.05). However, even with high levels of emotional exhaustion, social workers were found to experience moderate depersonalization and high personal accomplishment (Kim & Stoner, in press). These results are consistent with Lloyd and King’s (2004) study reporting that Australian social workers in mental health care settings experienced high emotional exhaustion but moderate depersonalization and high personal accomplishment. It seems that, although social workers experience higher levels of emotional exhaustion, they do not necessarily have more negative attitudes toward their clients or negative perceptions regarding their professional accomplishments. 14 From this review, it simply cannot be concluded that social workers suffer excessive burnout, because burnout is a multi-dimensional concept and burnout experience can be differently understood based on which aspect of burnout is examined (i.e., emotional exhaustion, depersonalization, personal accomplishment). However, it is clear that social workers experience higher levels of emotional exhaustion compared to other occupational groups. 2.1.3 The Significance of Social Worker Turnover Social worker turnover has been recognized as a major problem in the social service delivery system and consistently documented in both practice and research, as well as policy reports. For example, the average turnover rate for child welfare case workers was 19.6 percent during 2005 (Rabe, 2006). Another previous study by the Alliance for Children and Family, the American Public Human Services Association, and the Child Welfare League of America (2001) reported that the social worker turnover rates in child protective services agencies were about 20 percent. High turnover rates contribute to staff shortages. Pasztor, Saint-Germain, and DeCrescenzo (2002) reported that, in California, there were an estimated 12,221 social worker positions at the county level in all 58 counties in California and the vacancy rate was estimated at 9.5%. These findings suggest that social service agencies were seeking to employ approximately 1,171 social workers at the time of the study. As a consequence of such a staff shortage, more social workers are forced to take on heavier caseloads. Reports from the Annie E. Casey Foundation (2006) 15 estimated the increase in occupational demands for social workers to be 36 percent between 1998 and 2008. At the same time, the number of students pursuing a bachelor’s or master’s degree in social work has recently decreased. Furthermore, a recent report by the National Association of Social Workers (Center for Workforce Studies, NASW, 2006) suggested that the supply of licensed social workers does not meet the needs of organizations serving children and families. In addition, the same report from NASW (2006) stated that the number of new social workers providing services to older adults is decreasing. Similarly, a recent report from the U.S. Department of Health and Human Services (2006) stated that there will be a huge demand for professional social workers providing long-term care services for the aging population. According to the report, approximately 36,100 to 44,200 professional social workers are currently employed in long-term care settings, but approximately 110,000 professional social workers will be needed by 2050 if the ratio of professional social workers to the population age 65 and older remains constant. In addition to the staff shortage, the effect of social worker turnover on outcomes for social services has been well documented. Worker turnover not only causes psychological distress for clients, but also impacts remaining staff members and new inexperienced workers who fill vacated positions (Powell & York, 1992). In addition, worker turnover leads to client mistrust of the system (Geurts, Schaufeli, & De Jonge, 1998) and financial problems for the organization (Kompier & Cooper, 1999). As a result, social worker turnover negatively affects the quality, consistency, and stability of client services (Mor Barak, Nissly, & Levin, 2001). 16 2.1.4. The Relationships between Job Demands, Burnout, and Turnover Intention Burnout is a function of job conditions, working environments, and interpersonal relationships (e.g., Schaueli, Van Dierendonck, & Van Gorp, 1996, Maslach & Leiter, 1997). Reviews of the burnout literature (Lee & Ashforth, 1996; Schaufeli & Enzmann, 1998) support the idea that burnout results from job demands such as role overload (Koeske & Koeske, 1989), role ambiguity, and role conflicts (Um & Harrison, 1998). Role overload is the perception of too much work to accomplish in the time available and with limited resources (Cordes & Dougherty, 2003, p. 640). The perceived workload is an important aspect of job demand in social service fields because social service workers often complain about their caseloads, documentation obligations, and heavy amounts of paperwork. More importantly, social workers who think they do too much work may feel that they cannot provide quality service to clients in need, and eventually this experience will negatively influence feelings of personal achievement at work. In addition to work overloads representing the quantitative aspects of job demands, qualitative aspects of job demands, such as role conflict or ambiguity, are also important in comprehensively understanding job demands for social workers. Role conflict is the perception of incongruity between the workers’ role expectations and actual tasks (Um & Harrison, 1998, p. 103). Role ambiguity is defined as the lack of predictability of the responses to one’s behavior and the existence or clarity of behavioral requirements that would serve to guide behavior and provide knowledge that the behavior is appropriate (Rizzo, House, & Lirtzman, 1970). 17 Workers may experience role ambiguity when proper procedures for performing job tasks or criteria for performance evaluations are not clear. Cordes and Dougherty (1993) explained that workers who perceive higher levels of role ambiguity are more likely to experience a sense of diminished personal accomplishment because it is difficult to develop strong feelings of efficacy when one feels unsure of what is expected of one’s performance or when there is ambiguity concerning performance feedback. Several studies have empirically shown that role ambiguity is associated with feelings of diminished personal accomplishment among human service workers in general (Brookings, Bolton, Brown, & McEvoy, 1985) and social workers in particular (Himle, Jayaratne, & Thyness, 1989). Second, burnout results in dysfunctional outcomes at the personal as well as at organizational levels. Previous research studies have found that burnout is associated with increased levels of depression (Glass & McKnight, 1996), absenteeism (Parker & Kulik, 1995), sick-leave absences (Toppinen-Tanner et al., 2005), actual turnover (Drake & Yadama, 1996), turnover intention (Harrington, Bean, Pintello, & Mathews, 2001; Huang, Chuang, & Lin, 2003; Lee & Ashforth, 1996), impaired performance (Parker & Kulik, 1995; Wright & Bonett, 1997), and self-reported health problems (Schawfeli & Enzmann, 1998). Longitudinal studies suggest that burnout among workers increases turnover rates. For example, Jackson, Schuler, and Schwab (1986) found that emotional exhaustion was related to actual job-leaving after one year. Firth and Britton (1989) studied a sample of nursing staff in long-term care settings and found that perceived emotional exhaustion predicted the frequency of absences, and depersonalization was correlated with turnover in the 18 future. Similarly, Drake and Yadama (1996) reported that burnout predicted turnover among child protective service workers. Wright and Cropanzano (1998) reported that emotional exhaustion was a significant predictor of voluntary turnover based on the results of a one-year longitudinal study of 52 social welfare workers. This study focuses on social workers’ intention to leave their job rather than actual turnover. Although it is known that intentions may change over time and are not necessarily predictive of future behavior, workers’ intention to leave is an important area to study in order to understand factors associated with turnover, because the turnover intention of workers has been widely agreed as a strong predictor for actual turnover (Alexahnder et al., 1998; Hendrix, Robbins, Miller, & Summers, 1999). The significant relationship between turnover intention and actual turnover has been identified by the previous reviews and meta-analyses of turnover studies. For example, a meta-analysis of longitudinal studies by Steel and Ovalle (1984) showed that an individual’s intention to leave was the strongest predictor of an actual decision to leave. More importantly, Mor Barrak, Nissly, and Levin’s (2001) extensive meta-analytic review of turnover studies in the fields of child welfare, social work, and human services also indicated that the strongest single predictor of actual turnover was turnover intention. Although previous researchers examined the long-term effects of burnout on actual turnover, only a small number of studies examined the longitudinal effects of burnout on workers’ intention to leave the organization, and the findings have been inconsistent. Lee and Ashforth’s (1993) longitudinal study found that emotional exhaustion resulted in turnover intention among supervisors and managers in the 19 public welfare setting. However, Jackson, Schuler, and Schwab (1989) reported teachers’ burnout at the initial survey to have only a marginal effect on turnover intention one year after the initial survey. Krausz, Koslowsky, Shalom, and Elyakim (1995) conducted a longitudinal study of 146 nurses working in a general hospital and found that burnout at Time 1 had only an indirect effect on turnover intention at Time 2 through turnover intention at Time 1. In summary, the previous literature on burnout suggests that burnout is a key mediator in the relationship between job demands and various strain outcomes, including actual turnover and turnover intention. However, the empirical findings of longitudinal studies are still limited. 2.2 Job Resources: Theoretical Background The following sections include a discussion of theories of stress and coping and key resource theories focusing on how these theories provide theoretical backgrounds for studying the job resources of social workers. 2.2.1 Stress and Coping Theory The term stress appeared initially in the index of Psychological Abstract in 1944 (Lazarus & Folkman, 1984), after which an enormous number of studies began to discuss the concept. The growth of the study of stress has led to various definitions of ‘stress’ and approaches to studying stress. The early literature on stress 20 focuses on the relationship between stressors (e.g., major life events or features of work design) and outcomes (e.g., turnover or cancer). For example, Selye (1993) viewed stress in physiological terms as the common response of the body to any demand made upon it. However, the perspective simply focusing on the relationship between stressors and outcomes has been criticized by inconclusive results regarding the relationships. For example, some people overcome stressful life events (e.g., serious illness, death of family members, major financial crisis) well, but others often fail to cope with major life events. As a result, researchers have become interested in examining the individual or environmental factors as the specific conditions explaining the relationship between stressors and strain outcomes. Based on this approach, many researchers tested social supports or personality traits as buffering factors which moderate the relationship between stressors and strain outcomes. Another important concept in stress research is coping. There have been two major approaches: the dispositional approach and the situational approach (Jones, Bright, & Clow, 2001). First, the dispositional approach focuses on individual differences in coping. For example, Byrne (1961) distinguished between repressors (i.e., those who react to anxiety-arousing stimuli and their consequences by avoidant behavior) and sensitizers (i.e., those who attempt to reduce anxiety by controlling threats) as two major coping styles (cited by Jones, Bright, & Clow, 2001). On the other hand, the situational approach considers coping as a process that is specific to situations of psychological stress (Lazarus & Folkman, 1984). Lazarus and Folkman (1984) defined stress as ‘a particular relationship between the person and the environment that is appraised by the person as taxing or exceeding his or her 21 resources and endangering his or her well-being’ (p. 19). According to this approach, the environment is not separate from the individual characteristics of the person exposed to the environment. Stress changes over time depending on how individuals appraise the stressors and the coping strategies they use (Lazarus & Fokman, 1984). Therefore, coping strategies can be good or bad depending on the situation. For example, Carver, Scheier, and Weintraub (1989) found that people are more likely to use problem-focused coping such as active coping (i.e., taking active steps to deal with stressors) and planning (i.e., thinking about how to deal with a stressor) in controllable situations. On the other hand, people are more likely to use emotionally- focused coping strategies such as denial (i.e., refusing to believe the stressor exists) or acceptance (i.e., accepting the reality of a stressful situation) in less controllable situations. The present study follows the situational approach of stress and coping. Therefore, it is assumed that the relationship between job demands and strain outcomes (i.e., burnout and turnover intention) are different depending on the environmental factors. In addition, certain types of coping strategies are effective depending on the situational factors. 2.2.2 Resource Theories and Job Resources Resource theories provide theoretical frameworks to explain why certain situational factors are critical in explaining the relationship between stressors and strain outcomes. There have been a large number of research studies focusing on the impact of peoples’ resources on their stress resistance and well-being. These early 22 studies on resources and well-being identified key resources of individual personalities such as mastery (Pearlin & Schooler, 1978), self-efficacy (Bandura, 1977), and high-self esteem (Hobfoll & Leiber, 1987). In addition, social support has been discussed as a key resource from the social environment (Cohen, Doyle, Skoner, Rabin, & Gwaltney, 1997; Cohen & Wills, 1985; Diener & Fujita, 1995). According to Hobfoll’s (2002) conservation of resources (COR) theory, when resources are lost or threatened, an individual experiences stress. COR theory also suggests that a resource gain becomes of increasing importance when a resource loss has occurred. Lazarus and Folkman (1984) addressed that the resources individuals have available to meet the threat and challenge it influence the degree to which they appraise something as threatening as well as the coping choices they make. The previous literature on resources suggests that peoples’ resources influence the degree of stress they experience and their coping responses to the stress. These approaches to resources have been adopted and received empirical support in the context of the workplace. Job stress is the harmful physical and emotional response that occurs when the requirements of the job do not match the capabilities, resources, or needs of the workers (NIOSH, 1999). In order to enable workers to meet this increasing pressure and respond to stress, much of the research has focused on identifying resourceful factors to reduce job stress. Demands-Control (DC) model (Karasek & Theorell, 1990) is an example of a theoretical framework of job stress, and focuses on identifying the resourceful factors. The Job Demands- Control (DC) model of job stress suggests that high demands combined with low control result in long-lasting psychological strain because the actions normally 23 needed to cope with job demands cannot be taken in low control situations. This long-lasting strain will be directly internalized with harmful consequences, including emotional exhaustion. The DC model of job stress suggests that social workers with higher levels of job control are more likely to overcome their perceived job demands. Furthermore, in the expanded job demands-control-support (DCS) model, Johnson and Hall (1988) hypothesized that workers with jobs combining high demands, low control, and low support are at the highest risk for psychological strain (Johnson & Hall, 1988). Based on the DCS model, a number of empirical studies (e.g., Chapman et al, 1990; Karasek and Theorell, 1990; Landsbergis et al., 1992; Munro et al., 1998) have examined the moderating effects of social support in the relations between stress and strain outcomes. These researchers reported that job control and social support in a working condition buffer the negative effects of job-related stress. The job demands-resources (JD-R) model elaborates the DC tradition in the direction of further specificity of relations (van Veldhoven, Taris, de Jonge, & Broersen, 2005). The central theme of the job demands-resource model (JD-R) of burnout is that “burnout develops when job demands are high and when job resources are limited because such negative working conditions lead to energy depletion and undermine employee’s motivation, respectively” (Demerouti, et al., 2001, p. 499). In this model, job demands are defined as the physical, social, or organizational aspects of the job that require sustained physical or mental efforts and are, therefore, associated with certain physiological and psychological costs (Demerouti et al., 2001, p. 501). Job resources are the physiological, psychological, social, or organizational aspects of the job that: (a) are functional in achieving work 24 goals, (b) reduce job demands at the associated physiological and psychological costs, or (c) stimulate personal growth and development (Demerouti et al., 2001, p. 501). According to the JD-R model, when the environment lacks resources, individuals cannot cope with the negative influences of environmental demands. Recently, Bakker, Demerouti, and Euwema (2005) examined whether job resources buffer the impacts of job demands on burnout. The results of the study showed that job demands (work overload, emotional demands, unfavorable working conditions, and work-home interference) had a weaker or no relationship with burnout in cases where job resources (autonomy, social support, high-quality relationship with the supervisor, and performance feedback) were available in the working environment. 2.3 JOB RESOURCES FOR SOCIAL WORKERS Based on the JD-R model, job resources for social workers are defined in the current study as positive job conditions that reduce job demands and are helpful for coping with job demands. Much of the research conducted on job resources has examined job autonomy (Bakker et al., 2003), social support (Bakker et al., 2005), and effective supervision (London, Larsen, & Thisted, 1999) as important protecting factors for workers. The following section summarizes the literature on these concepts and includes a discussion of how these concepts apply to the field of social work. 25 2.3.1 Job Autonomy The concept of job autonomy can be defined as the degree of control over the worker’s own immediate scheduling and tasks (Liu, Spector, & Jex, 2005). Previous studies supported the idea that burnout is triggered by individual perceptions of control on the job (Glass & Mcknight, 1996). For example, Maslach, Schfeli, and Leiter (2001) addressed that lack of job autonomy reduces feelings of accomplishment at work. Coders and Dougherty (1993) argued that lack of job autonomy engenders a depersonalized attitude among workers. Autonomy is especially important for social workers. Social workers who eventually left the field often mention ‘bureaucratic hassles’ as the main reason. Arches (1991) and Lecroy and Rank (1986) found that job autonomy was an important job condition of burnout among social service workers. In addition to burnout, job autonomy has also been found to be associated with turnover intention among workers. Spector’s (1986) meta-analysis on the effect of perceived autonomy showed a decrease in the likelihood of a worker quitting his or her job if they felt autonomy. According to Cherniss’ (1995) study on human service professionals, sufficient autonomy makes one’s work meaningful, and meaningful work helps professionals to remain dedicated. Mor Barrak, Nissly and Levin’s (2001) extensive meta-analytic review of turnover studies in the fields of child welfare, social work, and human services also indicated that the level of job autonomy perceived by social workers was a significant predictor of turnover intention. In the previous study (Kim and Stoner, in press), job autonomy was negatively associated with burnout among 346 registered social workers working in organizational settings. Furthermore, job 26 autonomy moderated the relationship between job demands and burnout, which suggests that job autonomy is an essential resource for social workers. 2.3.2 Effective Supervision Considering the fact that most social workers work and practice in the context of supervisory relations, effective supervision is important for social workers’ job performance and occupational well-being in general. Particularly, the supervisor needs to be a good teacher and trainer of professional skill and knowledge for frontline social workers. Therefore, the supervisor’s ability to effectively engage in role-specific communication (or job-relevant communication) is a key component of effective supervision. Supervisory communication regarding job-related matters embraces performance feedback, information about rules and policies, work schedules and assignments, task-specific instructions, and goals (Miles, Patrick, & King, 1996). Previous researchers have suggested that effective communication between supervisees and their supervisors can increase professional knowledge or skill levels (Albrecht & Adelman, 1987), improve role definition, and raise feelings of accomplishment at work (Ellis & Miller, 1994) among supervisees. Kadushin (1992) also found, in his national study of social work supervision, that the most useful function of social work supervision is educational. Therefore, role-specific communication with supervisors has been found to be the key predictor of supervisory performance (York & Denton, 1990). Kluger and DeNisi (1996) performed a meta-analysis of the effects of performance feedback and found that 27 performance feedback from the supervisor had a moderately positive effect on workers’ job performance. For social workers, receiving meaningful feedback specifically regarding their jobs can be a great help in maintaining attachment to their work. A study of 211 social workers in health care settings also found that role- specific communication was negatively associated with perceived role stress, and role specific communication had a direct effect on burnout and turnover intention through perceived role stress (Kim & Lee, in press). 2.3.3 Social Support at Work The concept of social support is generally defined as the supportive interactions in the context of both formal and informal relationships (House, 1981; Wilcox & Vernberg, 1985). In specific, House (1981) defined social support as interpersonal transactions that include affect, affirmation, and/or aid. When social support is investigated in work settings, it has been treated as important in the working environment, influencing job stress and consequently job-related strain (Karasek & Theorell, 1990). Researchers have found that perceived social support in the workplace decreases the likelihood of worker burnout (Houkes et al., 2003) and turnover intention (Mor Barak et al., 2001; Nissly et al., 2005). Karasek and Theorell (1990) addressed that social support at work refers to helpful social interactions available from organizational members including a supervisor, a top manager, and co-workers in the workplace. First, an immediate supervisor plays multiple roles not only as an instructor and manager, but also as a consultant and supporter for frontline social workers (Kadushin, 2002). Recent 28 developments in social work supervision have emphasized that the relationship between a supervisor and a social worker should be characterized as less hierarchical and more as one of mutual influence (Williams, 1997). Supervisors and frontline workers should exchange their opinions, feelings, and inputs related to decision- making with mutual trust, respect, and a certain rapport among them. Having a supportive supervisor has been identified as an important social condition for lower levels of perceived burnout. Moreover, evidence suggests that having a supportive supervisor is significantly related to organizational commitment and turnover intention (Jinnett & Alexander, 1999). Second, social support from co-workers (Janssen, 1999) has been identified as an important social condition for lower levels of perceived burnout. Baker et al. (1996) confirmed the significant and unique effects of co-worker support on negative job feelings and depression among workers in automotive factories. Support from coworkers has been shown to be significantly related to how individuals attach to their organization (Alexander et al., 1998; Jinnett & Alexander, 1999). Similarly, Landstrom, Biordi, and Gillies (1989) found that a lack of co-worker support motivates turnover intention among nurses. Third, top-managerial personnel are an important source of social support for social workers. However, top-managerial support has been a major subject in leadership studies rather than social support research. Based on behavioral approaches of leadership, consideration and initiating structure have been developed based on the two major clusters of leadership behavior. Consideration includes all forms of supportive interaction and concern for the subordinates’ needs, such as 29 allowing subordinates more participation in decision making and encouraging more communication (Oaklander & Fleishman, 1964). Consideration is based on mutual trust, respect, and a certain rapport between the top manager and the worker (Oaklander & Fleishman, 1964). Similarly, according to Burns’ (1978) transformational leadership theory, showing individualized support for organizational members is a key leadership behavior. Previous empirical studies have shown that top-managers’ support is related to positive job attitudes like job satisfaction (Butler & Cantrell, 1997; House & Baetz, 1979) and organizational commitment (Avolio, Zhu, Koh, & Bhatia, 2004). Findings showed that the greater the perceived quality of leaders’ supportive behavior, the greater the satisfaction with the job and the greater commitment to the organization. These findings were confirmed by previous studies on top-managerial leadership in human service organizations. For example, supportive behaviors from top managers in human service organizations were related to workers’ job satisfaction (Bordieri et al., 1988; Holdnak et al., 1993), organizational commitment (Bycio et al., 1995), and retention (Taunton et al., 1997). 2.4 FUNCTIONS OF JOB RESOURCES The previous literature suggests that job autonomy, effective supervision, and social support at work are important job resources for social workers. The following 30 sections discuss the theoretical background explaining how job resources function in preventing burnout and turnover among social workers. 2.4.1 Edifying Professional Roles Supervision serves a critical role in burnout models of social workers in terms of reducing the experienced job stressors. Social information processing theory (Salancik & Pfeffer, 1978) and uncertainty reduction theory (Berger & Calabrese, 1975) are particularly useful in considering the role of supervision in the process of burnout. Salancik and Pfeffer (1978) proposed the social information processing theory. This theory considers perceptions of the workplace as a function of the communication environment in which workers are embedded rather than as a function of the “objective” characteristics of jobs and needs of workers. Based on this perspective, Miller and his colleagues (1996) and Miller and Monge (1985) examined effects of supervisory communication on burnout. Their findings suggest that supervisory communication has a substantial impact on perceived workplace stress, burnout, satisfaction, commitment, (Miller et al., 1996) and anxiety at work (Miller & Monge, 1985). Another related theory is uncertainty reduction theory (Berger & Calabrese, 1975). This approach proposes that the need to reduce uncertainty provides an explanation for social behavior in developing relationships. Uncertainty is the cognitive response that occurs when one does not know how and why events are occurring. Although this approach focuses on the role of uncertainty about relational partners in interpersonal interaction, scholars have broadened the focus of 31 uncertainty theory to consider uncertainty about situations and events. For example, Albrecht and Adelman (1987) considered the role of social support in reducing uncertainty in organizational as well as interpersonal contexts. Based on this approach, Miller and his colleagues (1990) found that supervisory support reduced perceived role stress and workloads among permanent staff in a hospital. Social information processing theory and uncertainty reduction theory suggest that effective supervision can aid workers in interpreting and dealing with stressful situations. This idea has been postulated by previous scholars. Leiter and Maslch (1988) argued that the extent to which workers perceive supervisors to be making irreconcilable demands on their time and resources is a major part of role stress. Beehr (1995) suggested that supervisor contact is more likely to include instrumental support, positively affecting the conditions which lead to job stress. Based on this view, the previous study (Kim & Lee, in press; Kim & Stoner, 2007) found that role-specific communication between social workers and supervisors reduced levels of job demands perceived by social workers in health care settings. Therefore, it is hypothesized that effective supervision, especially role-specific supervision, will be negatively associated with perceived levels of job demands among social workers. 2.4.2 Empowering Social Workers Although the social information and uncertainty reduction theories explain the direct effects of supervision on perceived job demands, they are limited to explaining how job resources buffer the effects of perceived job demands on burnout 32 or turnover intention. Empowerment theory provides theoretical frameworks explaining the buffering effects of job resources. Kanter (1977) conceptualized empowerment as the process by which organizational members or structures allow workers to access power and opportunity. Opportunity refers to growth, mobility, and the chance to increase professional knowledge and skills, while power refers to the ability to access and mobilize resources, information, and support from one’s position in the organization in order to get the job done successfully (Kanter, 1977 Cited by Conger & Kanungo, 1988). Conger and Kanungo (1988) conceptualized empowerment as a motivational construct by defining empowerment as “a process of enhancing feelings of self-efficacy among organizational members through the identification of conditions that foster powerlessness and through their removal by both formal organizational practices and informal techniques of providing efficacy information” (p. 474). In this motivational view of empowerment, power refers to an intrinsic need for self-determination or a belief in personal self-efficacy (Conger & Kanungo, 1988). Conger and Kanungo (1988) suggested that fostering opportunities for subordinates to participate in decision making and providing autonomy from bureaucratic constraint are key leadership practices. Further improving upon Conger and Kanungo’s (1988) model, Thomas and Velthouse (1990) proposed a cognitive model of empowerment by defining empowerment as increased intrinsic task motivation. Key aspects of intrinsic motivation include competence (the degree to which a person can perform task activities skillfully when he or she tries) and choice (the issue of whether a person’s behavior is perceived as self-determined) at work. 33 Empowerment theory suggests that enabling workers to feel as if they can perform their work competently will empower them to better cope with their job demands. Hackman and Oldham (1976) proposed that intrinsic motivation is generated by characteristics of the job and work environments, suggesting that individuals might develop a sense of competency by working in a job with task significance, autonomy, and feedback. Similarly, Eby, Freeman, Rush, and Lance (1999) identified job autonomy, feedback, and supervisory satisfaction as key intrinsic motivators using structural equation modeling based on meta-analytic correlations. Following the previous literature, this study considers intrinsically rewarding job characteristics such as job autonomy, role-specific supervision, top- managerial support, and collegial support as key elements of empowerment strategies and practices for social workers. In other words, enhancing the motivating potential of jobs by increasing these supportive elements will empower social workers and will intensify affective reactions toward the job. Some studies have shown that resourceful job conditions such as social support, performance feedback, and autonomy can promote a motivational process leading to work engagement (e.g., Bakker, Hakanen, Demerouti, & Xanthopoulou, 2007; Salanova, Agut, & Peiró, 2005). Work engagement is often considered as a opposite concept of burnout. Bscmscphd (2007) empirically showed that improvement of job conditions is associated with a higher sense of empowerment, which, in turn, is related to lower levels of burnout. Therefore, it is hypothesized that intrinsic motivators such as job autonomy, supervisory support, top-managerial support, and collegial support will moderate the relationship between job demands and burnout among social workers. 34 2.4.3 Embedding Social Workers Social work is a value-driven profession. Social workers generally endorse altruistic work values, as they are in a helping-oriented profession. The evaluative principle by which workers can determine the meaning of their work and interpret their work experience is considering their work value (George & Jones, 1997), it is expected that social workers are more likely to become attached to their jobs when they consider their jobs to be intrinsically rewarding and important to their professional goals. The theory of job embeddedness provides a theoretical framework for understanding why people remain at their job (Mitchell et al., 2001). According to Mitchell et al. (2001), job embeddedness is a concept reflecting the totality of forces that constrain people from leaving their current employment. In their conceptual study, Mitchell et al. (2001) found job embeddedness to have three different aspects: link (the extent to which people have links to other people or activities), fit (the extent to which their jobs and communities are similar to or fit with the other aspects in their life spaces), and sacrifice (what they would give up if they were to leave their job). Several empirical researchers (Holtom & O’Neill, 2004; Mitchell et al., 2001) have supported the idea that people who are embedded in their jobs have less intent to leave and do not leave their jobs as readily as those who are not embedded. Lee and colleagues (2004) found that job embeddedness moderated the effects of absences and performance behaviors on actual turnover. The previous empirical findings demonstrate that the theoretical framework of job embeddedness provides a 35 conceptual foundation to explain why job resources prevent social workers from leaving their jobs. Early qualitative studies in the field of social work have offered important insight on which job conditions are essential to keep social workers at their jobs. First, Samantrai (1992) interviewed two groups of social workers: social workers who chose to leave their child welfare jobs and those who chose to stay in public child welfare agencies. The study found that burned-out child welfare social workers left their jobs when they found themselves unable to transfer to other positions within the organization and when they had unsupportive supervisors. In another study, Rycraft (1994) conducted in-depth interviews with a sample of 23 caseworkers in six public child welfare agencies. The qualitative information from the interviews revealed four critical factors that affect workers’ decisions to stay in their positions: (1) a sense of mission or commitment to the work, (2) a good fit or perceived suitability of their particular job assignments, (3) good supervision, which was generally perceived as consultative rather than instructive or mentoring in nature, and (4) a sense of personal and professional investment in child welfare. Similarly, Reagh (1994) conducted a study of 18 child welfare workers who portrayed their work altruistically and found the following to be important characteristics of devoted child welfare workers from the interview data: altruism, a connection between work and the individual’s personal identity, a sense of spiritual calling, meaning through work, personal accomplishment, the support of colleagues, the ability to accept incremental change, maturity, and values of protecting human rights or serving others. 36 Recently, two studies based on structural equation model (SEM) analyses have provided quantitative evidence related to the factors of child welfare workers’ retention. Smith (2005) found that the facilitation of life-work balance, supervisor support, and agency average turnover rate were important factors in retaining child welfare workers. Mor Barak, Levin, Nissly, and Lane (2006) used mixed methods with structural equation modeling and comparative content analyses. Quantitative findings from the SEM analyses indicate that the key predictors of turnover intention are lack of job satisfaction, low organizational commitment, younger age, high stress, and exclusion from the organizational decision-making process. In addition, good supervision, flexibility at work, and perception of achievement in making a difference in the clients’ lives were addressed by the interviewers as important factors for workers to stay in the organization. Finally, two extensive reviews of the literature on child welfare workers’ decisions to stay at their jobs have substantiated earlier findings. According to a literature review by Bednar (2003), workers who are most likely to remain in their positions despite burnout tend to be those who come to the work with a sense of personal and professional mission, who have been well-matched to their position, and who enjoy supportive relationships with supervisors. Similarly, another extensive literature review by Zlotnik, DePanfilis, Daining, and Lane (2005) shows that professional commitment, adequate supervision, and a supportive working environment attract workers and keep them at their work settings. Similarly, Nedd (2006) found that nurses who perceive higher levels of peer networking and positive supervisory relations were more likely to stay in their jobs. 37 Previous research consistently suggests that personal factors such as a sense of mission and commitment to the work, as well as positive working environmental factors such as a supportive supervisor and participatory decision-making, can retain workers despite demanding job conditions. According to the job embeddedness theory, social workers with higher levels of job resources (i.e., job autonomy, top- managerial support, and effective supervisory communication) are more likely to link to other organizational members, perceive high levels of fit with their jobs and supervisors, and believe they would sacrifice a great deal if they left their current organizations. Therefore, it is hypothesized that job resources such as job autonomy, supervisory support, top-managerial support, and collegial support will moderate the relationship between job demands and turnover intention and the relationship between burnout and turnover intention among social workers. 2.5 Functions of Job Resources and Professional Development To this point, the functions of job resources in preventing burnout and turnover among social workers have been discussed. However, it is questionable that the proposed functions of job resources can be explicitly applied to any social worker. Contextual factors such as the organizational setting or field of practice may influence the functions of job resources. Furthermore, individual factors such as a worker’s coping style or personality may be related to the effectiveness of job resources. A central question, then, is what significantly impacts functions of job 38 resources among social workers. Although there may be numerous factors to consider, this study focuses on social workers’ professional development. 2.5.1. Professional Development and Social Work Licensure The emergence of accountability mechanisms for social services has placed increasing emphasis on the professional development of social workers. When a social worker is involved with clients, the social worker empowers clients to contribute to an informed decision about treatment or service planning based on each client’s goals and values. At the same time, a practitioner’s professional skills, knowledge, and competence are also important for facilitating the effectiveness of services. Therefore, the professional development of social workers has been a major issue in social work education. Several models of professional development from cognitive psychology view the process of the accumulation of knowledge and skills as passing through several developmental stages. For example, Feiman-Nemser and Remillard (1996) explained teacher development as “an initial stage of survival and discovery, a second stage of experimentation and consolidation, and a third stage of mastery and stabilization” (p. 66). Similarly, Thomas (1982) presented the stages of professional development for child care workers, in which the new worker is primarily concern with survival through gaining knowledge of agency rules and service policies in the first stage. After this stage, the worker becomes confident about the values of practices and services, and moves beyond a formalized body of knowledge, developing a style of work from his or her own practice experience. 39 Dreyfus and Dreyfus (1986) proposed one of the most influential models of skill development, and this model has been widely used to study the professional development of various human service workers including nurses (Benner, Tanner, & Chesla, 1996), teachers (Berliner, 1994), and social workers (Ryan, Fook, & Hawkins, 1995). Dreyfus and Dreyfus (1986) criticized the view of human skill development as the accumulation of explicit rules to follow in order to perform a task, and characterized advanced skill levels as experience-based know-how that cannot be articulated entirely in the form of rules. They suggested that skill development proceeds through various skill levels: novice, advanced beginner, competent proficient, and expert. According to this model, more advanced professionals at the proficient and expert levels identify problems, goals, plans, and actions holistically and apply their professional skills to each situation intuitively (Dall’Alba & Sandberg, 2006). In contrast, those at lower levels of professional development, such as novices and advanced beginners, routinely approach each case in a deliberative way (Dall’Alba & Sandberg, 2006). Ryan, Fook, and Hawkins (1995) observed 37 social workers in order to delineate the process of how social work students become competent practitioners. They found that many social work students appeared to have reached the advanced beginner stage by the end of the two year course. Therefore, they argued that, in order to improve social work education, social work courses need to concentrate on the processes by which formal theory becomes practical knowledge through concrete experience in a practice situation. Social work licensure is a formal process of learning how to apply professional skills and knowledge, in which formal education has prepared workers 40 for practice related to various client problems. Fourteen states, including California, have two-tier licensing (Dyeson, 2004). In California, individuals with BSWs and MSWs can be employed as social workers without holding a license. However, becoming a registered social worker requires an MSW from an Accredited School of Social Work. First, a person with an MSW degree can register as an Associated Social Worker (ASW) with the California Board of Behavioral Sciences (BBS), after which he or she must meet the following requirements to become a Licensed Clinical Social Worker (LCSW): educational requirements (substance abuse: 15 hours; human sexuality: 10 hours; child abuse assessment and reporting: 7 hours; spouse/partner abuse assessment, detection, and intervention, 15 hours; aging and long-term care, 10 hours), 3,200 supervised clinical hours, the standard written exam, and the written clinical vignette exam (NASW California Chapter, 2007). During this process, social workers can learn how to make clinical judgments and use theory according to the particular situation. Dreyfus and Dreyfus (1986) explained this learning process as the development of ‘situational rules’. Ryan, Fook, and Hawkins (1995) used more traditional social work terminology and considered this learning process to be the development of ‘practice wisdom’. According to the literature review on professional development, it is clear that LCSWs are in a more advanced stage of professional development than ASWs. 41 2.5.2 Job Resources for Social Workers in Different Stages of Professional Development There are several reasons to expect that certain job resources are more salient for either ASWs or LCSWs. First, professional licenses, including the clinical social work license, are often related with an individual’s social position or organizational status. According to role theory (Biddle, 1986), social position, organizational status, or job levels often define organizational roles. Organizational roles influence an individual worker’s organizational behavior since the individual is evaluated positively in light of behaviors consistent with organizational role expectations (Tsui & Ohlott, 1988). In addition, organizational roles often influence how people are treated in organizations (Jackson, Schuler, & Rivero, 1989). Job level has been found to be associated with a difference in expectations about the superior-subordinate communication relationship (Dansereau & Markham, 1987). Miles, Patrick, and King (1996) also found that the effects of supervisory communication are different depending on job levels, and explained that the process of assimilation into an organization is dependent on the development of knowledge about acceptable behavior and role expectations. Therefore, Miles, Patrick, and King (1996) proposed that the causal influence of communication on role ambiguity might be especially acute for organizational newcomers. Therefore, it is expected that the influence of effective supervision will be more significant for ASWs than for LCSWs. Second, supervisees differ in their expectations of supervision according to the professional stage of development in which they are in (Shechtman & Wirzberger, 1999). The supervisee in the early stage of professional development is 42 dependent on the supervisor as the major source of knowledge. On the other hand, social workers at higher professional stage levels are essentially independent practitioners with secure professional identities, and supervision is consultative if it continues (Wiley & Ray, 1986, cited by Shechtman & Wirzberger, 1999). In this respect, it is assumed that job autonomy may be a more important resource for LCSWs than for ASWs. However, few researchers have explored these issues. Therefore, the current study explores group differences between LCSWs and ASWs in terms of the functions of job resources in preventing burnout and turnover. 2.6 Hypotheses, Conceptual Model, and Research Questions 2.6.1 Study Hypotheses and Conceptual Model Based on the literature review, the following relationships between job conditions and burnout and turnover intention are hypothesized: Hypothesis 1: Role-specific supervision in the initial phase will be negatively associated with job demands, which are characterized as role conflict, role ambiguity, and role overload at the initial phase. Hypothesis 2: Higher levels of job demands in the initial phase will result in significantly higher levels of burnout in the second phase after adjusting for the levels of burnout in the initial phase. 43 Hypothesis 3: Higher levels of job demands in the initial phase will result in significantly higher levels of turnover intention in the second phase after adjusting for the levels of turnover intention in the initial phase. Hypothesis 4: Higher levels of burnout in the initial phase will result in significantly higher levels of turnover intention in the second phase after adjusting for the levels of turnover intention in the initial phase. Hypothesis 5: The longitudinal effect of job demands on burnout will be moderated by job resources (i.e., job autonomy: H5-a, supervisory support: H5-b, top managerial support: H5-c, co-worker support: H5-d). That is, as social workers experience higher levels of job demands in the initial phase, burnout in the second phase will significantly increase when they have lower levels of job resources in the first phase. Hypothesis 6: The longitudinal effect of job demands on turnover intention will be moderated by job resources (i.e., job autonomy: H6-a, supervisory support: H6-b, top managerial support: H6-c, co-worker support: H6-d). That is, as social workers experience higher levels of job demands in the initial phase, turnover intention in the second phase will significantly increase when they have lower levels of job resources in the first phase. Hypothesis 7: The longitudinal effect of burnout on turnover intention will be moderated by job resources (i.e., job autonomy: H7-a, supervisory support: H7-b, top managerial support: H7-c, co-worker support: H7-d). That is, as social workers experience higher levels of burnout in the initial phase, 44 turnover intention in the second phase will significantly increase when they have lower levels of job resources in the first phase. To test the above hypotheses, a hypothetical model was developed based on the previous models of burnout including Cordes and Dougherty’s (1993) integrative model of burnout, Moore’s (2000) partial mediating model of burnout, the JDCS model of job stress (Johnson & Hall, 1988), and the JD-R model of burnout (Demerouti et al., 2001). First, the testing of Hypotheses 2 and 4 is based on a traditional mediating model of burnout (e.g., Cordes & Dougherty, 1993), which hypothesizes that the effects of job demands on turnover intention are fully mediated by burnout. This means that job demands have direct effects on burnout, and burnout affects turnover intention, but job conditions do not have direct effects on turnover intention. Second, the direct effects of job demands on turnover intention (Hypothesis 3) are included in the hypothesized model based on Moore’s (2000) partial mediating model, which suggests that burnout does not fully mediate the effect of job conditions on turnover intention. This means that job demands have unique effects on burnout as well as turnover intention. Third, the moderating effects of job resources on the relationship between job demands and burnout (Hypothesis 5), the relationship between job demands and turnover intention (Hypothesis 6), and the relationship between burnout and turnover intention (Hypothesis 7) were included. As discussed above, these hypothesized moderating effects of job resources are based on the JDCS model of job stress (Johnson & Hall, 1988) and the JD-R model of burnout (Demerouti et al., 2001). 45 Finally, the extant literature on burnout and turnover suggest that the relationships among job demands-burnout-turnover cannot be understood without accounting for the effects of individual characteristics. For example, Brewer and Shapard’s (2004) meta-analytic review of burnout research reported age as a negative correlate of worker burnout. Jackson (1993) also found that gender is a significant demographic factor of burnout (Brewer & Shapard, 2004). Other studies suggest that workers having a shorter organizational tenure (Somers, 1996) and higher dissatisfaction with salary (Curpall, Tooler, Judge, & Kohn, 2005) are more likely to quit their jobs. Therefore, the proposed conceptual model includes important sample characteristics as control variables. Figure 1 illustrates the hypothesized relationships between key constructs and control variables together. Figure 1. Proposed Conceptual Model of Functions of Job Resources + Job Resources • Job Autonomy • Supervisory Support • Collegial Support • Managerial Support Demographic variables • Gender • Age • Org. Tenure • Annual Salary H1 H2 H4 H3 H6 + - - H7 - H5 - Job Demands Job Resources • Role- specific Supervision + + Burnout Turnover Intention 46 Table 1 presents conceptual definitions of variables included in the conceptual model. Table 1. Conceptual Definitions of Study Variables Domains Variables Conceptual Definition Job Demands Role Conflict The perception of incongruity of the role expectations associated with a job (Um & Harrison, 1998) Role Ambiguity The perception of lack of clarity on proper procedure for performing job tasks or criteria for performance evaluations (Rizzo, House, & Lirtzman, 1970) Role Overload The perception of too much work to accomplishment in the time available (Cordes & Dougherty, 2003) Job Resources Role-specific Supervision Perceived quality of communication with one’s supervisor in role-specific matters (Miles, Patrick, & King, 1996) Job Autonomy The perceived degree of control over the worker’s own immediate scheduling and tasks (Liu, Spector, & Jex, 2005) Supervisory Support Perceived interpersonal transactions that include affect, affirmation, and /or aid from a immediate supervisor (House, 1981) Top-managerial Support Perceived interpersonal transactions that include affect, affirmation, and /or aid from top-managerial personnel (House, 1981) Co-worker Support Perceived interpersonal transactions that include affect, affirmation, and /or aid from co-workers (House, 1981) Burnout Emotional Exhaustion Feelings of being overextended and depleted of emotional resources (Maslach & Jackson, 1986) Depersonalization Negative or excessively detached responses to various aspects of the job (Maslach & Jackson, 1986) Personal Accomplishment Feelings of incompetence and a lack of achievement at work (Maslach & Jackson, 1986) Turnover Turnover Intention Intention to quit a current job 47 2.6.2 Research Questions In addition to examining the study hypotheses, the present study explores group difference between ASWs and LCSWs by attempting to answer the following three research questions: Research question 1: Are ASWs experiencing different levels of job demands, burnout, and turnover intention than LCSWs? Research question 2: Are the impacts of role-specific supervision on perceived job demands different between ASWs and LCSWs? Research question 3: Are the moderating impacts of job autonomy and social support on the relationships among job demands, burnout, and turnover intention different between ASWs and LCSWs? 48 CHAPTER III: METHODS 3.1 Research Design and Procedure This study employs a longitudinal panel design with two phases of mail- surveys. The initial survey (Time 1) was conducted between October and December 2005. A total of 1,500 registered social workers were randomly selected from the total population of state-registered social workers employed in California (N = 21,518). The registry of social workers was provided by the Department of Consumer Affairs (DCA) Public Information Unit in accordance with the Information Practices Act, Civil Code Section 1798.61. Each social worker was mailed a survey instrument, an introductory letter, an information sheet, and a return postage-paid envelope. The survey included questions regarding social workers’ thoughts and attitudes concerning their working environment and the extent to which they experience burnout at work. The survey took approximately 15-20 minutes to complete, and participation in this study was completely voluntary. The introductory letter in the survey mail package informed potential participants that they could withdraw at any time without any consequence, and they could also refuse to answer any questions they did not want to answer and sill remain in the study. Of the 529 questionnaires returned, 51 were ineligible because respondents were currently retired or working in other fields (N=43), had registered after the data 49 collection period (N=5), or had failed to answer a significant number of questions (N=3). From the original sample of 1,500, 478 questionnaires were eligible, yielding a survey response rate of 32 percent. Considering that 20 percent of licensed social workers are not active in the field, according to a recent report by the Center for Workforce Studies, NASW (2006), the 32 percent survey completion rate seems to be reasonable. This study only used a sub-sample of 405 social workers working in organizational settings by excluding 73 full-time private practitioners. The 73 full- time private practitioners were excluded since this study mainly focused on social workers’ turnover in organizational settings. A total of 405 study participants received a second-phase questionnaire (Time 2) approximately 1 year after the first survey. To reduce the attrition rate of the study, the same second-phase questionnaire was sent to the study sample of 405 social workers at 6 weeks and 10 weeks after the initial second-phase survey. Both questionnaire forms (i.e., Time 1 & Time 2) contained a unique study ID code for the purpose of matching information between the two surveys. Only the researchers knew how the study ID codes were assigned. Out of 405 participants in the first phase, 315 social workers returned their questionnaire in one of the second phases. Of the 315 returned, 284 were eligible, yielding a returning rate of about 70 percent. However, 40 social workers were excluded from the study sample since they changed their job settings or left the field during the study period. Therefore, the actual study sample for the analysis consists of 365 social workers, with 244 social workers having completed both the initial and follow-up surveys. 50 3.2 Sample Table 2 describes the study participants (n = 365) who participated in the initial survey and were working in organizational settings. In addition, Table 2 also describes the respondents (n = 244) who participated in both the first and second surveys and who did not move to other jobs or organizations during the study period. Regarding the characteristics of total study sample, Licensed Clinical Social Workers (LCSWs) represented 64.7% (n=236) of selected registered social workers working in organizational settings, while 35.3% (n=129) were Associate Social Workers (ASWs). Mental health was the largest practice area for active registered social workers in California, representing about 46.8% of respondents. Medical health (21.9%), child welfare/family (16.7%), and school social work (7.1%) were the next on the list of the largest practice areas represented. The mean age of all respondents was 46 years old and the average number of years in the field of social work was 17 years. Regarding ethnic groups, the survey respondents were distributed as follows: 69.2% (249) Caucasian, 11.7% (42) Latino, 7.5% (27) Asian American, and 5.3% (19) African American. Participants in the second survey were similar to the total study sample in terms of demographic characteristics. Licensed Clinical Social Workers (LCSWs) represented 66.7% (n=162) of selected registered social workers employed in organizational settings, and 33.3% (n=81) were Associate Social Workers (ASWs). Mental health was the largest practice area for active registered social workers in California, representing about 50.6% of respondents. Medical health (21.8%), child 51 welfare/family (15.2%), and school social work (6.2%) were the next largest practice areas represented. The mean age of all respondents was 46 years old and the average number of years in the field of social work was 17 years. Regarding ethnic groups, the survey percentages were as follows: 69.7% (166) Caucasian, 12.6% (30) Latino, 8.8% (21) Asian American, and 5.0% (12) African American. To check for sample representativeness, the participants were compared with the sample of a recent report of licensed social workers from the NASW Center for Workforce Studies (2006). First, a binomial test was performed to test if the proportion of participants in the current study falling in each category of gender differed from the observed probabilities of social workers in the NASW study. The study sample was not significantly different from the sample of the NASW study in terms of gender. Regarding age, the median age of the national sample was 49 years, while the median for the current study was 47 years. A statistical test was not available for the age difference since the NASW report did not present a mean age for the national sample. However, a sample of California-registered social workers was different from the national sample in terms of ethnic background. In the national sample, 84.5% of licensed social workers identified themselves as non-Hispanic White, but 69.2% of participants of the present study were non-Hispanic White, which suggests that social workers in California are more ethnically diverse. Regarding the educational background of the social workers, all participants in the study had at least a master’s degree in social work, since the sample for the current study consists of California-registered social workers. As mentioned before, the educational requirement for the social work licensure varies from state to state. Since 52 some states do not require a master’s degree be registered or licensed to perform tasks as a social worker, the findings of the current study cannot be generalized to registered social workers nationwide. Table 2. Respondents’ Characteristics Characteristics Total (N = 365) Participants at Time 2 (N= 244) Gender Male Female 70 (19.6%) 288 (80.4%) 52 (21.9%) 185 (78.1%) Social Work Licensure ASW LCSW 129 (35.3%) 236 (64.7%) 81 (33.2%) 163 (66.8%) Respondent’s Mean Age (Years) 45.8 (SD = 12.0) 46.4 (SD = 11.8) Respondent’s Licensing Status Associate Social Worker (ASW) Licensing Clinical Social Workers (LCSW) 129 (35.3%) 236 (64.7%) 81 (33.3%) 162 (66.7%) Organizational tenure (Years) 8.2 (SD = 7.8) 8.2 (SD = 7.5) Years in the Field of Social Work (Years) 17.1 (SD = 10.4) 17.5 (SD = 10.1) Respondent’s Race/Ethnicity Other Black or African American Asian or Pacific Islander Hispanic or Latino White or Caucasian 23 (6.3%) 19 (5.3%) 27 (7.5%) 42 (11.7%) 249 (69.2%) 9 (3.8%) 12 (5.0%) 21 (8.8%) 30 (12.6%) 166 (69.7%) Approximate Annual Salary Less than $20,000 $20,000 – $39,999 $40,000 – $59,999 $60,000 – $79,999 $80,000 or more 5 (1.4%) 46 (12.6%) 131 (35.9%) 137 (37.5%) 46 (12.6%) 4 (1.6%) 31 (12.8%) 85 (35.0%) 94 (38.7%) 29 (11.9%) Service Fields 1 School Social Work Child Welfare/Family Medical Health Mental Health 26 (7.1%) 61 (16.7%) 80 (21.9%) 171 (46.8%) 15 (6.2%) 37 (15.2%) 53 (21.8%) 123 (50.6%) 53 Note. 1. Participants were asked to check all that applied among seven service fields. Seven service fields include mental health, adolescents, child welfare/family, additions, aging, health, and school social work. In this table, statistics were presented for four major service fields. 3.3 Measures Job demands. Job demands were assessed in terms of role conflict, role ambiguity, and role overload. Role conflict (RC) and role ambiguity (RA) were measured using a shortened form of the Rizzo, House, and Lirtzman (1970) role conflict/role ambiguity questionnaire. The role conflict scale consisted of eight items (e.g., “I receive an assignment without the manpower to complete it”) including questions regarding: 1) conflict between the focal person’s internal standards and the defined role behavior; 2) conflict between the time, resources, or capabilities of the focal person and defined role behavior; 3) conflict between several roles for the same person that require different behaviors; and 4) conflict expectations and organizational demands in the form of conflicting requests from incompatible standards of evaluation. Role ambiguity consists of six items (e.g., “I know what my responsibilities are”). Rizzo et al. (1970) reported internal consistency reliabilities for role conflict and ambiguity averaging .82 and .80, respectively. Participants were asked to respond to the six-item RA and eight-item RC scales by indicating the degree to which—on a 7-point scale ranging from “very false” (1) to “very true” (7)—the condition applied to them. Cronbach’s alpha for the internal consistency 54 and reliability for this sample was 0.88 for the RC scale and 0.84 for the RA scale. Role overload was measured by a five-item scale of work overload, specifically designed to measure workload among human service workers (Lait & Wallace, 2002). This scale also employed a 7-point Likert-type scale ranging from “strongly disagree” (1) to “strongly agree” (7). The role overload scale includes the following items: 1) “I have to attend too many meetings in this job”; 2) “My job involves a lot of paperwork”; 3) I have to work very fast to get everything done in my job”; 4) “My workload is too heavy in my job”; and 5) “I do not have enough time to get everything done in my job.” The Cronbach’s alpha for the role overload measure’s internal consistency and reliability was 0.82. Job autonomy. Job autonomy was assessed with a three-item subscale of decision authority from the Job Content Questionnaire (JCQ), a self-administered instrument designed to measure social and psychological job characteristics using items such as “My job allows me to make a lot of decisions on my own” (Karasek, Brisson, Kawakami, Houtman, & Bongers, 1998). These items were rated on a 7- point Likert-type scale ranging from “strongly disagree” (1) to “strongly agree” (7). Substantial theoretical and empirical work has supported the reliability and validity of the JCQ measure (Karasek et al., 1998), and the Cronbach’s alpha for this study was 0.73. Role-specific supervision. Role-specific supervision was measured by a sub- scale of job-relevant supervisory communication from a short form of the Huseman, Hatfield, Boulton, and Gatewood (1980) 56-item instrument. Huseman et al.’s (1980) original measure represents eight message types: feedback, rationale, 55 information, direction, negative expression, positive expression, participation, and upward openness. Miles, Patrick, and King (1996) performed exploratory factor analysis for a 24-item short version of the measure and identified four types of supervisory communication: job relevant communication (ten items), upward openness communication (three items), positive relationship communication (seven items), and negative relationship communication (three items). In this study, three items of negative relationship communication were excluded because the social workers who participated in a pilot study (n=48) suggested that these items were not appropriate to social work research. These three items are: “My supervisor ridicules or makes fun of me”, “My supervisor criticizes my work in front of others”, and “My supervisor is critical of me as a person”. In my previous study on supervisory communication among social workers (Kim & Lee, in press), a principal component factoring analysis with varimax rotation was performed to examine the factor structure of a 21-item version of Huseman et al.’s (1980) measure for a sample of social workers in health care settings. The results showed three sub-factors (i.e., job-relevant communication, upward openness communication, positive relationship communication), which was consistent with Miles, Patrick, and King’s (1996) study. The current study used the ten-item job relevant communication scale to measure the quality of supervision perceived by social workers. Each question was measured with a five-point Likert-type scale. Cronbach’s alpha for the seven-item scale of job relevant communication was 0.91. Supervisory, top-managerial, and collegial support. Social support was measured by House and Wells’ (1981) social support measure, which has been 56 widely used in work settings because it is brief and specifies a variety of sources (i.e., coworkers, top manager) and types of social support (i.e., emotional, instrumental, and informational). Emotional support is the availability of a person or persons who can listen sympathetically when an individual is having problems and can provide indications of caring and acceptance. Instrumental support involves practical help when necessary, such as assisting with transportation, helping with childcare, and providing tangible aid. Informational support is defined as providing knowledge that is useful for solving problems, such as information or advice and guidance about alternative courses of action (Cohen, Underwood, & Gottlieb, 2000). This study focuses on supervisors, top managerial personnel, and co-workers as key sources of support. Participants were asked to rate the extent to which their supervisors, top managerial personnel, and co-workers in their organizations provided each type of support in the workplace. All items were rated on a 5-point Likert-type scale ranging from “not at all” (0) to “very much” (4). A social support score for each source (i.e., supervisors, top managerial personnel, and co-workers) was computed by averaging six questions. The Cronbach’s alpha was 0.96 for supervisory support, 0.92 for collegial support, and 0.96 for top-material support. Burnout. Burnout among social workers was assessed using Maslach’s Burnout Inventory–Human Service Survey (MBI–HSS; Maslach & Jackson, 1986), which measures burnout components by asking about the frequency with which workers experience feelings related to each aspect of the burnout syndrome. It includes nine questions on perceived emotional exhaustion, five questions on depersonalization, and eight questions on personal accomplishment (Maslach & 57 Jackson, 1986). Each statement is rated on a 7-point continuum from “never” experienced (0) to experienced “every day” (6). Acceptable levels of reliability and validity have been reported for the MBI (e.g., Cordes et al., 1997; Hallberg & Sverke, 2004). For a sample of social service workers, the MBI showed internal reliability (Cronbach’s alpha ranging from 0.71 to 0.90) and test-retest reliability (2- to 4-week intervals for all scales, ranging from 0.60 to 0.82; Maslach & Jackson, 1986). Cronbach’s alpha for the current study was 0.91 for emotional exhaustion, 0.75 for depersonalization, and 0.79 for personal accomplishment. However, the current study calculated burnout scores (i.e., burnout at Time 1 & burnout at Time 2) for each participant by averaging 14 items of emotional exhaustion and depersonalization for the following two reasons: 1) investigation of second-order factor models revealed the presence of a common burnout factor among emotional exhaustion, depersonalization, and personal accomplishment but suggested that depersonalization and emotional exhaustion are core components of burnout; and 2) longitudinal factorial invariance was achieved for emotional exhaustion and depersonalization, but not for personal accomplishment (Kim & Ji, under review). When measurement invariance does not hold over time, researchers would not be convinced with results from longitudinal studies. Since this study employs a two-wave longitudinal research design, personal accomplishment was excluded from the analysis. Turnover Intention. Organizational turnover intention was measured by the following three items from the four-item scale of intention to leave (Nissly, Mor Barak, & Levin, 2005): “In the next few months I intend to leave this organization”, 58 “In the next few years I intend to leave this organization”, and “I occasionally think about leaving this organization”. These items were rated on a 7-point Likert-type scale ranging from “strongly disagree” (1) to “strongly agree” (7), for which the Cronbach’s alpha was 0.76. 3.4 Analysis 3.4.1 Overview of Analytical Methods The current study employs cross-lagged panel model to examine casual relationships among the study variables (Campbell & Stanley, 1963). Figure 2. The two-wave, Two-variable Cross-lagged Panel Model (Taris, 2000, p. 68) Figure 2 and the discussion about the cross-lagged panel model were based on the Taris’ book (2000). As shown in Figure 2, six different correlations can be computed for the four variables (two variables at Times 1 and 2): two synchronous d a c b f e X 1 Y 2 X 2 Y 1 M 59 correlations (each variable with the other at the same point in time; c & d in Figure 2), two autocorrelations (each variable with itself at two time points; a & b in Figure 2), and two cross-lagged correlations (each variable with the other at a different point in time; e & f in Figure 2). The basic idea of the cross-lagged panel methodology is to identify the causal priority of the variables by comparing two cross-lagged correlations. Kessler and Greenberg (1981) suggested regressing X 2 on X 1 and Y 1 as well as regressing Y 2 on X 1 and Y 1 . The standardized regression estimates of the effect of X 1 on Y 2 and of Y 1 on X 2 can be compared. Therefore, if it is hypothesized that X causes Y, the analysis focuses on testing whether the cross-lagged correlation e is significantly larger than the correlation f after accounting for two autocorrelations (i.e., a & b). In addition, a hypothesized moderating effect of a variable M can be examined by using multi-group analysis techniques. Another way of expressing the question in terms of an interaction effect is this: does group membership based on the variable M influence the relations? For this analysis, the sample needs to be divided into two groups based on a median score for the variable M. The path coefficients of a focal relationship can be compared between two groups. A significant difference in the path coefficients suggests a significant moderating effect of the variable M on the path. 3.4.2 Preliminary data analyses Preliminary analyses consist of attrition analysis and normality tests. First, several methods of dealing with missing data were used. Traditional methods for 60 fixing problems of missing data are “listwise” deletion (i.e., deleting cases which are not complete and using only complete in analysis) or “pairwise” deletion (i.e., deleting cases in forming statistics to be used in analysis). These methods are simple and easy to understand, but deleting some cases in the data set has critical limitations such as loss of subjects, increased standard errors, and bias if the data is not MCR (Missing Completely at Random) (Rubin, 1987). There are other methods such as Multiple Imputation methods and Likelihood (FIML) methods. Multiple imputation fills in missing values by a method that randomly selects values from the scores that other cases like this case have (Loehlin, 2004). In the other hand, likelihood methods estimates missing information by using the raw data rather than to the covariance matrix, using maximum likelihood function (Loehlin, 2004). Although mathematical backgrounds of these methods are difficult to understand, several SEM software such as AMOS (Arbucke, 1997), Mplus (Muthén & Muthén, 1998), or Mx (Neale, Boker, Xie, & Maes, 1999) allow researchers to use these advanced methods dealing with missing information. Multiple imputation has advantage when only some data are missing (less than 25 %), while likelihood methods are better when a great deal of data are missing. In this study, both methods were used. Since each data set (i.e., Time 1 data set and Time 2 data set) has a few missing values (less than 10 %), missing values at the item level were replaced by plausible values based on the multiple imputation method using the Windows freeware, NORM (Schafer, 1999). Multiple imputations involve a regression approach and a data-augmentation algorithm to impute missing values. Following the method described by Olsen and Schafer (1998), I created three imputed data sets and 61 obtained a single-point estimate by averaging across the estimates from the Time 1 data set. The same procedure was performed for the Time 2 data set. However, the Time 2 data set still had missing values for 161 social workers who did not respond to the follow-up survey or who were not eligible for the follow- up survey. In addition, there were missing values for the control variables: the percentage missing for gender was 1.9%, 0.8% for age, 1.6% for organizational tenure, and 2.7% for annual wage. In order to handle these missing values with no loss of information contained in the available dataset, Full Information Maximum Likelihood Method (FIML; Arbuckle, 1996) was used. In order to use the FIML method, the data must be (1) missing at random and (2) multivariate normal (Raykov, 2005). To test whether attrition at Time 2 was random, a logistic regression analysis was performed. A dependent variable for the attrition analysis was a dichotomous variable indicating attrition at Time 2. Logistic regression analysis was used to test the unique contribution of the study variables (i.e., age, gender, organizational tenure, wage, job demands at Time 1, job resources at Time 1, three dimensions of burnout at Time 1, and turnover intention at Time 1) in predicting dropout of the participants during the longitudinal study. Regarding the normality, the skewness and kurtosis of each variable were examined. 3.4.3 Descriptive statistics and correlations The means and standard deviations of the study variables as well as bivariate correlations between them were calculated. These descriptive statistics allow 62 researchers not only to better understand phenomena of interest, but also to replicate the SEM analysis. 3.4.4 Examining study hypotheses The study hypotheses were examined using path analysis. Parameters in the hypothesized path models were estimated with a Full Information Maximum Likelihood (FIML) method by using the Mplus 4.0 software (Muthén & Muthén, 2006). To evaluate model fit, the χ 2 /df, Comparative Fit Index (CFI) and Root-Mean- Square Error of Approximation (RMSEA) were examined. A χ 2 /df value below 3 indicates a reasonable fit (Kline, 1998). CFI ranges from 0 to 1, where 0 represents the goodness of fit associated with the null model and 1 represents the goodness of fit associated with a saturated model. For the CFI, a value above 0.90 suggests an acceptable fit between the model and data (Kline, 1998). Finally, the RMSEA is the most sensitive index to models with misspecified factor pattern coefficients and less sensitive to the sample size than other model fit indexes. An RMSEA of 0.08 or less indicates a reasonable error of approximation (Browne & Cudeck, 1993). Following the idea of cross-lagged panel analyses (e.g., De Jonge, Dormann, Janssen, Dollard, Landeweerd, & Nijhuis, 2001; Feldt, Kivimäki, Rantala, & Tolvanen, 2004), concurrent associations (cross-sectional associations) between variables (i.e., job demands, burnout, and turnover intention) were included in the tested cross-lagged model. The concurrent associations seemed to be reasonable since cross-sectional relationships between job demands, burnout, and turnover intention have been widely identified in previous studies (see Cordes & Dougherty, 63 1993). A model also includes three autocorrelations and following hypothesized relationships: a structural path from role-specific supervision to job demands at Time 1 (Hypothesis 1), a cross-lagged structural path from job demands at Time 1 to burnout at Time 2 (Hypothesis 2), a cross-lagged structural path from job demands at Time 1 to turnover intention at Time 2 (Hypothesis 3), and a cross-lagged structural path from burnout at Time 1 to turnover intention at Time 2 (Hypothesis 4). Finally, cross-lagged structural paths in the opposite direction were included to assess the causal priority among pairs of variables (Campbell & Stanley, 1963). In order to test the study hypotheses, standardized regression estimates of hypothesized paths were examined. If, for a certain pair of variables (e.g., job demands and burnout), cross- lagged effects in both directions are significant, the standardized regression estimates of both cross-lagged effects are compared to identify the causal priority of the variables as suggested by Kessler and Greenberg (1981). To examine the hypothesized moderating effects of job resources (i.e., job autonomy, supervisory support, top managerial support, collegial support) in the framework of job demands-burnout-turnover intention, multi-group analyses were used. For example, the main question for Hypothesis 5 is whether job resources moderate the longitudinal relationship between job demands and burnout. For this analysis, the sample was first divided into two groups based on the median score for each job resource (i.e., job autonomy, top managerial support, co-worker support). Two estimated models were used to test the moderating effect on the path. The first model freely estimated structural paths between study variables across the groups. The second model constrained the path coefficient of the relation of interest (i.e., the 64 path between job demands at Time 1 and burnout at Time 2) to equality across the group. A significant chi-square difference suggests that the equality constraints were not consistent with the data and an interaction effect exists (Rigdon, Schumacker, & Wothke, 1998). The same procedure was used to examine the moderating effect of job resources on the relationship between job demands at Time 1 and turnover intention at Time 2 (Hypothesis 6), as well as the relationship between burnout at Time 1 and turnover intention at Time 2 (Hypothesis 7). Figure 3. A Cross-lagged Model for Testing Longitudinal Relations between Job demands, Burnout, and Turnover Intention (Model 1) Supervision T1 Burnout T1 Job Demands T1 Turnover Intention T1 Burnout T2 Job Demands T2 Turnover Intention T2 Supervision T1 Burnout T1 Job Demands T1 Turnover Intention T1 Burnout T2 Job Demands T2 Turnover Intention T2 Note. In case of revising this model, this model was named as Model 1. For the simple presentation, control variables (i.e., age, gender, organizational tenure, and annual salary) were omitted. 65 3.4.5 Group Comparison In addition to examining the study hypotheses, this study explored group differences between ASWs and LCSWs in terms of 1) demographic characteristics; 2) levels of job demands, burnout, and turnover intention; and 3) functions of job resources. First, the demographic characteristics of ASWs and LCSWs were explored. Descriptive statistics included frequencies, percentages, means, and standard deviations depending on the variable type. Two-sample Student’s T tests were performed for between-group comparisons of continuous variables. Pearson’s χ 2 or Kolmogorov-Smirnov two-sample Z tests were performed for between-group comparisons of nominal and ordinal variables. The Kolmogorov-Smirnov Z test is a non-parametric statistical significance test for assessing the difference between cumulative distributions of two samples in order to determine whether the two independent samples are from the same population (Siegel & Castellan, 1988). A P value below .05 was used to indicate statistical significance. Second, levels of job demands, burnout, and turnover intention were examined across groups. Two-sample Student’s T tests were performed for between-group comparisons for both Time 1 and Time 2, respectively. Third, the cross-lagged path model was fitted with the data of each sub-sample (i.e., ASWs and LCSWs) in order to explore the longitudinal relationships between role-specific supervision, job demands, burnout, and turnover intention. Multi-group analyses were then performed to explore the functions of job autonomy and social support for each group. 66 CHAPTER IV: RESULTS 4.1 Results of Total Sample 4.1.1 Preliminary Analysis In this study, 121 social workers out of 365 study participants did not respond to the second survey or were not eligible because they moved to other organizations. A binary logistic regression was performed to test whether attrition at Time 2 was at random (see Table 3 for detail). A dependent variable for the binary logistic regression was a dichotomous variable indicating attrition at Time 2. The results of the logistic regression analysis showed that a set of study variables was not significant in predicting attrition ( χ 2 = 14.43, df = 11, p > .05). However, top managerial support ( β = -.30, p < .05) and turnover intention at Time 1 ( β = .16, p < .05) were significantly related to the attrition after accounting for the effects of other variables on attrition. Subsequently, independent sample t-tests were performed to examine whether social workers who did not participate in the second survey were different from those who participated in the second survey in terms of levels of top managerial support and turnover intention in the first phase. Table 4 and Table 5 present the results of the t-tests, which show that the levels of top managerial support (t = 1.27, df = 363, p > .05) and turnover intention (t = 1.65, df = 363, p > .05) were not significantly different between the two groups. Although the t-test results indicate that attrition at Time 2 was random, the results of the present study should 67 be interpreted with special care due to the potential impacts of top managerial support and turnover intention on participant dropout. Table 3. Binary Logistic Regression Analysis Independent Variables Beta S.E. df Sig. Job Demands at Time 1 -.10 .16 1 .55 Burnout (EE+DP) at Time 1 .07 .14 1 .65 Turnover Intention at Time 1 .16 .08 1 .04* Role-specific Supervision at Time 1 -.10 .16 1 .54 Job Autonomy at Time 1 .14 .11 1 .20 Supervisory Support at Time 1 -.11 .17 1 .53 Collegial Support at Time 1 .10 .17 1 .55 Top Managerial Support at Time 1 .30 .14 1 .03* Gender 1 .44 .35 1 .21 Age -.02 .01 1 .20 Organizational Tenure .01 .02 1 .68 Annual Salary .01 .01 1 .48 Note. A dependent variable of the analysis was a dichotomous variable indicating attrition at Time 2 (Attrition = 1 and Non-attrition = 0). A sample size of this analysis is 330 because 35 cases have missing values in the independent variables. 1 Female =1 and Male =0 * p < 0.05 68 Table 4. Independent Sample T-test for Top Managerial Support Groups N Mean Std. Deviation t df p 1 Participants at both Time 1 & Time 2 244 1.78 1.02 1.27 363 .27 Participants who dropped out after Time 1 121 1.91 .95 Note. 1. 2-tailed. A dependent variable for the t-test is a variable of top managerial support. Table 5. Independent Sample T-test for Turnover Intention at Time 1 Groups N Mean Std. Deviation t df p 1 Participants at both Time 1 & Time 2 244 3.59 1.69 1.65 363 .10 Participants who dropped out after Time 1 121 3.91 1.78 Note. 1. 2-tailed. A dependent variable for the t-test is a variable of top managerial support. As mentioned before (Section 3.4.2), the normality of the study variables is an important assumption for a Full Information Maximum Likelihood (FIML) method. Therefore, skewness and kurtosis were examined for the study variables. The skewness ranged between −1.54 and 1.40, while the Kurtosis ranged between −1.03 and 1.56, so normal distributions were assumed for all variables. 69 4.1.2. Descriptive Statistics and Correlations Table 6 presents the means, standard deviations, and correlation coefficients for the 14 observed variables. Job demands, burnout, and turnover intention were positively associated with each other. In addition, these variables were negatively associated with the job resource variables. Table 6. Correlation Coefficients among Study Variables Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1. JD-T1 1.00 2. BO-T1 .61* 1.00 3. TI-T1 .36* .34* 1.00 4. JD-T2 .76* .51* .32* 1.00 5. BO-T2 .50* .75* .25* .60* 1.00 6. TI-T2 .31* .28* .67* .43* .36* 1.00 7. RS -.34* -.27* -.31* -.22* -.14* -.20* 1.00 8. JA -.41* -.30* -.32* -.36* -.30* -.25* .36* 1.00 9. SPS -.28* -.18* -.21* -.16* -.05 -.13* .66* .14* 1.00 10. TMS -.17* -.17* -.17* -.12 -.21* -.08 .28* .16* .39* 1.00 11. CWS -.06 .05 -.09 -.02 .10 -.10 .22*-.04 .31* .17* 1.00 12. Age .02 -.16* .05 -.04 -.23*-.07 -.10 -.01 -.12* .02 -.19* 1.00 13. Gender -.07 -.05 -.02 .04 .03 .00 -.05 .05 -.04 -.06 .06 -.25* 1.00 14. Tenure .02 -.02 .09 -.01 -.08 -.05 -.06 .00 -.11*-.05 -.04 .56*-.23* 1.00 15. Salary .02 .01 -.08 .00 -.02 -.15* .01 .04 .01 .05 .07 .22*-.23* .35* 1.00 Mean SD 4.0 1.0 2.2 1.1 3.70 1.72 3.9 1.0 2.1 1.0 3.7 1.8 3.5 1.1 5.2 1.3 2.8 1.0 1.8 1.0 3.1 .8 45.8 12.0 .8 .4 8.2 7.8 57.3 17.3 Note. JD-T1: Job demands; BO-T1: Burnout at Time 1; TI-T1: Turnover intention at Time 1; JD-T2: Job demands at Time 1; BO-T1: Burnout at Time 2; TI-T2: Turnover intention at Time 2; RS: Role-specific Supervision; JA: Job autonomy; SPS: Supervisory Support; TMS: Top managerial support; CWS: Co-worker Support; Tenure: Organizational tenure; Salary: Annual salary. *p < .05 (2-tailed). 70 4.1.3. Hypotheses Testing Table 7 presents results of a series of path analyses. Model 1 yielded an overall χ 2 (5) value of 56.13, with χ 2 /df = 11.22, CFI = .95, and RMSEA = 0.17 (PCLOSE = 0.00). The model-fit statistics indicated that Model 1 did not fit the data well. Therefore, a residual covariance of the matrix and modification index were further examined. The small and near-zero residuals suggested that the model fit the data well. A standardized residual for covariance between supervision and turnover intention at Time 1 was -.56, while a standardized residual for covariance between supervision and burnout at Time 1 was -.33. These two standardized residuals were larger than .20, which indicate significant misfit between the model and the data. The modification index suggested the same problems. Therefore, Model 1 was modified by adding a structural path from supervision to turnover intention at Time 1 (Model 2) (see Figure 4). Model 2 yielded an overall χ 2 (4) value of 36.68, with χ 2 /df = 9.17, CFI = 0.97, and RMSEA = 0.15 (PCLOSE = 0.00). The chi-square difference between Model 1 and Model 2 was significant ( Δχ 2 = 19.45, Δdf = 1, p < .05). However, the model fit for Model 2 was not yet acceptable since the χ 2 /df value was larger than 3 and the RMSEA value was larger than .08. Subsequently, in Model 3, a path from supervision to burnout at Time 1 was added to Model 2. Model 3 yielded an overall χ 2 (3) value of 1.99, with χ 2 /df = 0.66, CFI = 1.00, and RMSEA = .00 (PCLOSE = 0.83). Model 3 provided satisfactory fit to the data. A chi-square difference between Models 2 and 3 was significant ( Δχ 2 = 34.69, Δdf = 1, p < .05). The results of the model tests to this point suggest that role-specific supervision has significant effects on job demands, burnout, and turnover intention. 71 Figure 4. Revised Models (Model 2 and Model 3) Model 2 Model 3 Table 7. Fit Statistics for Comparative Models of Job Demands, Burnout, and Turnover Intention Models χ 2 /df CFI RMSEA (PCLOSE) MODEL COMPARION ∆χ 2 / ∆df Model 1: Hypothesized paths + cross-lagged paths in the opposite direction 56.13/5 .95 0.17 (0.00) Model 2: M 1 + a path from supervision to turnover intention (T1) 36.68/4 .97 0.15 (0.00) M1 & M1a 19.45/1* Model 3: M 2 + a path from supervision to burnout (T1) 1.99/3 1.00 0.00 (0.83) M1a & M1c 34.69/1* Note. Age, gender, organizational tenure and annual salary were control variables in the analysis. * p < 0.05 72 Since the model fit of Model 3 was satisfactory, regression coefficients of Model 2 were examined to test the study hypotheses 1, 2, 3, and 4. Researchers often revise structural equation models by omitting non-significant paths to obtain a more parsimonious model. However, this study used Model 3 as a final model to examine the hypothesized relationships because the purpose of the study was not building a best model of burnout and turnover intention, rather examining the hypothesized relationships. Figure 5 reports all standardized path coefficients and presents the relationships among variables in Model 3. Specifically, role-specific supervision was negatively associated with perceived job demands ( β = -.35) (Hypothesis 1). Unexpectedly, role-specific supervision was also associated with burnout ( β = -.30) and turnover intention ( β = -.31). The hypothesized longitudinal relationships among job demands, burnout, and turnover intention (Hypotheses 2, 3, and 4) were not significant. Since three auto-correlations (i.e., a path from job demands at Time 1 to job demands at Time 2, a path from burnout at Time 1 to burnout at Time 2, and a path from turnover intention at Time 1 to turnover intention at Time 2) were significant, role-specific supervision had indirect effects on job demands, burnout, and turnover intention at Time 2 through job demands, burnout, and turnover intention at Time 1. 73 Figure 5. Standardized Path Coefficients for Model 3 with Total Sample Supervision Job Demands T1 Job Demands T2 Turnover Intention T1 Burnout T1 Turnover Intention T2 Burnout T2 .72* .07 .04 .07 .71* .03 .02 -.04 .64* -.35* -.30* -.31* .22* .48* .24* .20* .16* .18* Supervision Job Demands T1 Job Demands T2 Turnover Intention T1 Burnout T1 Turnover Intention T2 Burnout T2 .72* .07 .04 .07 .71* .03 .02 -.04 .64* -.35* -.30* -.31* .22* .48* .24* .20* .16* .18* Note. N = 365. Effects of age, gender, organizational tenure, and annual salary were accounted. To test the hypothesized moderating effects of job autonomy, the model was first estimated for high and low job autonomy with all of the relevant paths freely estimated and then re-estimated with each path in interest constrained to equality across groups. Therefore, the two models were nested, and the difference in their chi- square statistics was a chi-square difference with 1 degree of freedom. The difference in the chi-square statistics was statistically significant only for the path between job demands at Time 1 and turnover intention at Time 2 (see Table 8). It suggested that job autonomy significantly moderated the longitudinal relationship between job demands and turnover intention (Hypothesis 6a). In order to better understand how job autonomy influenced the relationship between job demands and 74 turnover intention, standardized coefficients for the path were examined across the low and high job autonomy groups. The standardized coefficients for the paths between job demands at Time 1 and turnover intention at Time 2 were .15 (p < .05) for the low job autonomy group and -.06 (p > .05) for the high job autonomy group. Graph 1 presents the regression lines of the path between job demands at Time 1 and turnover intention at Time 2 for the two groups, which clearly demonstrates that job demands at Time 1 are positively associated with turnover intention at Time 2 when job autonomy is low. Figure 6. Regression Lines between Job Demands at Time 1 and Turnover Intention at Time 2 for Two Groups of Job Autonomy Job Demands at Time 1 7 6 5 4 3 2 1 Turnover Intention at Time 2 4 3 2 1 0 -1 -2 -3 High Job Autonomy Low Job Autonomy Note. Variances of turnover intention at Time 2 explained by other variables were accounted by using unstandardized residuals. The same procedure was applied to test the hypothesized moderating effects of supervisory support, top-managerial support, and collegial support. However, the 75 results of the chi-square difference tests showed that there was no significant moderating effect of supervisory support, top-managerial support, and collegial support. TABLE 8— Chi-Square Difference between Freeing Paths to Differ between Groups and Constraining Paths to be Equal across Groups Path Δχ 2 High (n = 206) and Low (n = 159) on Job Autonomy Job Demands → Burnout (Hypothesis 5a) .47 Job Demands → Turnover Intention (Hypothesis 6a) 4.34* Burnout → Turnover Intention (Hypothesis 6a) .74 High (n = 187) and Low (n = 178) on Supervisory Support Job Demands → Burnout (Hypothesis 5b) .44 Job Demands → Turnover Intention (Hypothesis 6b) 1.44 Burnout → Turnover Intention (Hypothesis 7b) .38 High (n = 185) and Low (n = 180) on Top Managerial Support Job Demands → Burnout (Hypothesis 5c) 1.39 Job Demands → Turnover Intention (Hypothesis 6c) .75 Burnout → Turnover Intention (Hypothesis 7c) .38 High (n = 191) and Low (n = 174) on Co-worker Support Job Demands → Burnout (Hypothesis 5d) 2.26 Job Demands → Turnover Intention (Hypothesis 6d) .06 Burnout → Turnover Intention (Hypothesis 7d) 1.32 * p < .05 76 4.2 Results of the Group Comparison of ASW and LCSW 4.2.1 Group Characteristics ASWs were significantly different from LCSWs in terms of various demographic characteristics. As is shown in Table 9, the age of the ASWs ranged from 24 to 70 years, with an average age of 37.94 (SD=10.16) years, while the age of the LCSWs ranged from 28 to 77 years, with an average age of 50.11 (SD=10.61). As expected, when years in the current organization and years in the field of social work were compared, ASWs were less experienced than LCSWs. It is not surprising that ASWs are significantly different from LCSWs in terms of age, organizational tenure, and field tenure since obtaining a clinical social worker license requires at least three years of registration as an ASW. Regarding annual salary, 66 percent of LCSWs had an annual salary greater than $60,000, but only 21 percent of ASWs had an annual salary greater than $60,000. About 59 percent of ASWs had an annual salary between $40,000 and $59,999. This findings are consistent with the Dall’Alba and Andberg’s (2006) argument that a higher salary or other benefits often accompany with professional development. Regarding demographic characteristics, 27 percent of LCSWs were male, while only 6 percent of ASWs were male. In terms of ethnic background, the distribution for ASWs was as follows: 58 percent Caucasian, 18 percent Hispanic or Latino, 9 percent Asian or Pacific American, and 6 percent African American. However, 75 percent of LCSWs identified themselves as Caucasian, followed by 77 Hispanic or Latino (8.2%), Asian American (6.5%), and African American (4.7%). When asked to indicate what types of services they provide, 23.3% of ASWs and 13.1% of LCSWs indicated that they were in the field of child and family services, while 40.3% of ASWs and 50.4% of LCSWs were involved with mental health services. The significant group differences in terms of gender and ethnicity were unexpected. Based on the current findings, it is hard to find out why these group differences exist. In addition, it seems that the social work workforce has become diverse in terms of their ethnic background. Future researchers should explore these issues. Table 9: Demographic Characteristics of ASWs and LCSWs Characteristics ASW (n = 129) LCSW (n = 236) Sig. Gender Male Female 8 (6.3%) 120 (93.8%) 62 (27.0%) 168 (73.0%) p < .05 2 Respondent’s Mean Age (Years) 37.94 (SD = 10.16) 50.11 (SD = 10.61) p < .05 3 Organizational tenure (Years) 4.28 (SD = 4.53) 10.31 (SD = 8.45) p < .05 4 Years in the Field of Social Work (Years) 9.97 (SD = 7.69) 21.07 (SD = 9.48) p < .05 5 Respondent’s Race/Ethnicity Other Black or African American Asian or Pacific Islander Hispanic or Latino White or Caucasian 11 (8.6%) 8 (6.3%) 12 (9.4%) 23 (18.0%) 74 (57.8%) 12 (5.2%) 11 (4.7%) 15 (6.5%) 19 (8.2%) 175 (75.4%) p < .05 6 Approximate Annual Salary Less than $20,000 $20,000 – $39,999 $40,000 – $59,999 $60,000 – $79,999 $80,000 or more 1 (0.8%) 25 (19.4%) 76 (58.9%) 21 (16.3%) 6 (4.7%) 4 (1.7%) 21 (8.9%) 55 (23.3%) 116 (49.2%) 40 (16.9%) p < .05 7 78 Table 9. Continued Service Fields 1 School Social Work Child Welfare/Family Medical Health Mental Health 12 (9.3%) 30 (23.3%) 23 (17.8%) 52 (40.3%) 14 (5.9%) 31 (13.1%) 57 (24.2%) 119 (50.4%) - Note. 1. Participants were allowed to select multiple responses. 2 & 6 Chi-Square tests. 3, 4, & 5 Student’s t tests. 7 Kolmogorov-Smirnov Z test. 3 t = 10.56, df=360, p < .05. 4 t = 8.80, df=357, p < .05. 5 t = 12.07, df=309, p < .05. 7 Kolmogorov-Smirnov Z = 4.13, n=365, p < .05. 4.2.2 Levels of Burnout and Turnover Intention Two-sample t-tests were performed to test whether the subgroup means differed significantly from one other. The means, standard deviations, and results of the multi-group comparisons for job demands, burnout, and turnover intention are provided in Table 10. There were no significant differences in job demands. However, the levels of burnout and turnover intention were different between ASWs and LCSWs at Time 2. In specific, ASWs experienced significantly higher levels of burnout and turnover intention at Time 2 compared to LCSWs. 79 Table 10. Between-Group Comparisons on Levels of Burnout and Turnover Intention Variables Samples N Mean SD t-value Job Demands at Time 1 ASW LCSW 129 236 4.03 3.99 0.97 1.02 -.31 Job Demands at Time 2 ASW LCSW 81 163 3.99 3.79 1.04 1.01 -1.47 Burnout at Time 1 ASW LCSW 129 236 2.33 2.16 1.08 1.07 -1.48 Burnout at Time 2 ASW LCSW 81 163 2.36 2.02 1.04 0.96 -2.55* Turnover Intention at Time 1 ASW LCSW 129 236 3.73 3.68 1.67 1.75 -.32 Turnover Intention at Time 2 ASW LCSW 81 163 4.19 3.44 1.85 1.73 -3.11** * <.05; ** < .01 (2-tailed) 4.2.3 Functions of Job Resources By using multi-group structural equation modeling techniques, Model 3 was fitted across ASWs and LCSWs to identify group differences in terms of: 1) longitudinal relationships among job demands, burnout, and turnover intention; and 2) the effects of role-specific supervision on job demands, burnout, and turnover intention. The model fit was good. The χ 2 was 6.52 (df = 6), the CFI was 1.00, and the RMSEA was .02. The results of the standardized path coefficients of Model 3 for ASWs and LCSWs are given in Figures 7 and 8. All relationships between study variables were similar in the two groups. Only one noticeable difference was the relationship between role-specific supervision and turnover intention. A standardized path coefficient between role-specific supervision and turnover intention was -.24 for 80 ASWs and -.35 for LCSWs. In order to test whether this difference was statistically significant, the path between role-specific supervision and turnover intention was constrained to be equal across groups. When the model was estimated without the equality constraint, the chi-square statistic was 6.5 (df = 7). And then, the path coefficient between role-specific supervision and turnover intention was fixed to be equal across groups and re-estimated. The chi-square statistic for the constrained model was 6.9 (df = 6). The chi-square difference between the two nested models was 0.4, which was not statistically significant at one degree of freedom. The standardized path coefficients for both ASWs and LCSWs were consistent with those of the total sample (see Figure 5). The results suggest that the relationships among role-specific supervision, job demands, burnout, and turnover intentions did not differ between ASWs and LCSWs. Figure 7. Standardized Path Coefficient for Model 3 with Associate Social Workers Supervision Job Demands T1 Job Demands T2 Turnover Intention T1 Burnout T1 Turnover Intention T2 Burnout T2 .67* .07 .01 .05 .75* .09 .12 -.01 .62* -.34* -.29* -.24* .15* .39* .28* .17* .15* .15* Supervision Job Demands T1 Job Demands T2 Turnover Intention T1 Burnout T1 Turnover Intention T2 Burnout T2 .67* .07 .01 .05 .75* .09 .12 -.01 .62* -.34* -.29* -.24* .15* .39* .28* .17* .15* .15* Note. N = 129. Effects of age, gender, organizational tenure, and annual salary were accounted. 81 Figure 8. Standardized Path Coefficient for Model 3 with Licensed Clinical Social Workers Supervision Job Demands T1 Job Demands T2 Turnover Intention T1 Burnout T1 Turnover Intention T2 Burnout T2 .75* .11 .07 .05 .68* -.03 -.03 -.04 .65* -.35* -.31* -.35* .24* .56* .23* .20* .19* .16* Supervision Job Demands T1 Job Demands T2 Turnover Intention T1 Burnout T1 Turnover Intention T2 Burnout T2 .75* .11 .07 .05 .68* -.03 -.03 -.04 .65* -.35* -.31* -.35* .24* .56* .23* .20* .19* .16* Note. N = 236. Effects of age, gender, organizational tenure, and annual salary were accounted. Next, the moderating effects of job autonomy and social support within the longitudinal relationships among job demands, burnout, and turnover intention were examined for both ASWs and LCSWs. Table 11 presents the results of multi-group analyses with the data for the ASWs. Results show that supervisory support had significant moderating effects on the relationship between job demands and turnover intention as well as the relationship between burnout and turnover intention. 82 Table 11— Chi-Square Difference between Freeing Paths to Differ Between Groups and Constraining Paths to Be Equal Across Groups for Associate Social Workers Path Δχ 2 High (n = 73) and Low (n = 56) on Job Autonomy Job Demands → Burnout (Hypothesis 5a) .06 Job Demands → Turnover Intention (Hypothesis 6a) .40 Burnout → Turnover Intention (Hypothesis 7a) 2.20 High (n = 75) and Low (n = 54) on Supervisory Support Job Demands → Burnout (Hypothesis 5b) .02 Job Demands → Turnover Intention (Hypothesis 6b) 5.42* Burnout → Turnover Intention (Hypothesis 7b) 4.00* High (n = 69) and Low (n = 60) on Top Managerial Support Job Demands → Burnout (Hypothesis 5c) .06 Job Demands → Turnover Intention (Hypothesis 6c) .31 Burnout → Turnover Intention (Hypothesis 7c) .65 High (n = 75) and Low (n = 54) on Co-worker Support Job Demands → Burnout (Hypothesis 5d) .14 Job Demands → Turnover Intention (Hypothesis 6d) .03 Burnout → Turnover Intention (Hypothesis 7d) .00 * p < .05 Figure 9 presents the regression lines of association between job demands at Time 1 and turnover intention at Time 2 for the two sub-groups of ASWs (i.e., ASWs with low social support and ASWs with high social support). This clearly 83 demonstrates that job demands at Time 1 were positively associated with turnover intention at Time 2 among ASWs when job autonomy was low ( β = .24, p < .05). However, job demands at Time 1 were not associated with turnover intention at Time 2 among ASWs when they had higher levels of job autonomy ( β = -.18, Not Sig.). Figure 10 presents the regression lines of association between burnout at Time 1 and turnover intention at Time 2 for the two sub-samples of ASWs (i.e., ASWs with low social support and ASWs with high social support). The results of the chi-square different test showed significant moderating effects of supervisory support on the relationship between burnout at Time 1 and turnover intention at Time 2 for ASWs. The path coefficient between burnout at Time 1 and turnover intention at Time 1 was significant for ASWs with low supervisory support ( β = 0.31, p < .05) and not significant for ASWs with high supervisory support ( β = -0.04, Not sig.) (See Figure 9). 84 Figure 9. Regression Lines between Job Demands at Time 1 and Turnover Intention at Time 2 for Two Groups of Supervisory Support with data from ASWs Note. Variances of turnover intention at Time 2 explained by other variables were accounted for by using unstandardized residuals. Figure 10. Regression Lines between Burnout at Time 1 and Turnover Intention at Time 2 for Two Groups of Supervisory Support with data from ASWs Burnout at Time 1 3 2 1 0 -1 -2 Turnover Intention at Time 2 4 3 2 1 0 -1 -2 -3 High Supervisory Support Low Supervisory Support Note. Variances of turnover intention at Time 2 explained by other variables were accounted for by using unstandardized residuals. 85 Table 12 presents the results of the multi-group analyses with data on the LCSWs. The results show that the hypothesized moderating effects of job autonomy and social support were not significant among LCSWs. TABLE 12— Chi-Square Difference between Freeing Paths to Differ Between Groups and Constraining Paths to Be Equal Across Groups for Licensed Social Workers Path Δχ 2 High (n = 133) and Low (n = 103) on Job Autonomy Job Demands → Burnout (Hypothesis 5a) .34 Job Demands → Turnover Intention (Hypothesis 6a) 2.44 Burnout → Turnover Intention (Hypothesis 7a) .28 High (n = 112) and Low (n = 124) on Supervisory Support Job Demands → Burnout (Hypothesis 5b) 2.21 Job Demands → Turnover Intention (Hypothesis 6b) .12 Burnout → Turnover Intention (Hypothesis 7b) .01 High (n = 116) and Low (n = 120) on Top Managerial Support Job Demands → Burnout (Hypothesis 5c) 3.03 Job Demands → Turnover Intention (Hypothesis 6c) .06 Burnout → Turnover Intention (Hypothesis 7c) .48 High (n = 116) and Low (n = 120) on Co-worker Support Job Demands → Burnout (Hypothesis 5d) 3.19 Job Demands → Turnover Intention (Hypothesis 6d) 1.69 Burnout → Turnover Intention (Hypothesis 7d) 2.22 * p < .05 86 CHAPTER 5: DISCUSSION 5.1 Major Findings 5.1.1 Concurrent and Longitudinal Relationships among Job Demands, Burnout, and Turnover Intention This study investigates the longitudinal relationships among job demands, burnout, and turnover intention for social workers. The results showed that a social worker with higher job demands at Time 1 did not increase the likelihood of turnover intention at Time 2, after adjusting for levels of burnout and turnover intention at Time 1. In addition, higher burnout at Time 1 did not increase the likelihood of turnover intention at Time 2, after accounting for levels of turnover intention at Time 1. In other words, the level of job demand did not have a causal effect on burnout or turnover intention, and burnout did not have a causal effect on turnover intention one year later. However, there were some significant concurrent correlations among job demands, burnout, and turnover intention (relationships among variables at Time 1 or Time 2). Social workers with higher levels of job demands were more likely to experience burnout and to think about quitting their jobs at the same time. In addition, three autocorrelations (i.e., relationships between job demands at Time 1 and burnout at Time 2, between job demands at Time 1 and turnover intention at Time 2, and between burnout at Time 1 and turnover intention at Time 2) were significant, suggesting that these phenomena were stable over time. In other words, a 87 social worker who experiences higher levels of burnout was more likely to experience higher levels of burnout later. In summary, the results of a cross-lagged path analysis with a sample of registered social workers showed that job demands had only indirect effects on burnout and turnover intention at Time 2. According to these findings, burnout may or may not be causally related with job demands and turnover intention in social workers but may be a momentary and concurrent phenomenon that happens when a social workers experience high levels of job demands. Thus, even though a social workers experience high levels of job demands now, current job demands may not influence on his/her burnout experience or turnover intention in the future. Explanations for the findings of non-significant casual relationships among job demands, burnout, and turnover intention include the suggestion that job demands may be causally related to burnout and turnover intention only under certain conditions. This study found that low levels of job autonomy and lack of social support were key conditions for the causal effects of job demands on burnout and turnover intention and the causal effects of burnout on turnover intention. The next section discusses how job autonomy and social support impact on the causal relationships among job demands, burnout, and turnover intention. 5.1.2 Functions of Job Autonomy and Social Support at work This study found a significant moderating effect of job autonomy on the long-term relationship between job demands and turnover intention. In specific, social workers with higher levels of job demands had increasing levels of turnover 88 intention over time when they did not have enough job autonomy. Previous literature addressed the importance of job autonomy in preventing turnover among workers. For example, a meta-analysis study by Spector’s (1998) showed that job autonomy perceived by workers was negatively associated with role stress, absenteeism, turnover intention, and actual turnover. Another meta-analysis by Mor Barak and colleagues (2001) also found that job autonomy was a significant and negative correlate of turnover intention among child welfare workers, social workers, and other human service workers. Although the direct effect of job autonomy on turnover intention has been supported, there has been less confirmation of the buffering effects of job autonomy. For example, Fox, Dwyer, and Ganster (1993) found significant interactions between job demands and job autonomy in explaining job satisfaction among registered nurses. However, O’Driscoll and Beehr (2000) reported that control at work did not buffer the effect of role stress on job satisfaction in a sample of U.S. and New Zealand workers. In addition to the inconsistent findings, few studies have examined the moderating effects of job autonomy with a sample of social workers. Therefore, the current study provides important empirical evidence for the salience of job autonomy as a buffer of the adverse consequence of job demands on turnover intention among social workers. An unexpected finding was that social support did not moderate the longitudinal relationships between job demands, burnout, and turnover intention. Social support has consistently been found to buffer the negative effects of job demands on burnout (Bakker, Demerouti, & Euwema, 2005) and turnover intention (Kim & Stoner, in press; Nissly, Mor Barak, & Levin, 2005). Considering that the 89 previous studies exhibit the limitations of cross-sectional research design, the findings of the current longitudinal study are vital in understanding the impacts of social support on social worker burnout and turnover. However, caution should be taken in concluding that social support does not buffer the long-term effects of job demands on burnout and turnover intention. The major issue in the longitudinal study of burnout and turnover intention is dealing with individual differences in the development of burnout or turnover intention over time. Previous research studies have addressed impacts of personality types (Zellars, Perrewe, & Hochwarter, 2000), working experience (Lee & Ashforth, 1993), and coping resource (Corde & Dougherty, 1993) on the development of burnout. A research by Trevor (2001) showed that individual attributes such as levels of education and occupation-specific training moderated the relationship between job satisfaction and voluntary turnover, after accounting for the effects of job market. Therefore, it is important to understand how job resources and individual characteristics interact with each other in explaining the development of burnout and turnover intention. This study proposed that the function of coping resources in the development of burnout and turnover intention would depend on the professional development of social workers. Professional development is somewhat different from working experience, which is often measured by years in the field of social work. High levels of professional development require not only professional experience, but also the professional training and education that endorse professional skills and competency. The current study used the social work licensing status as an approximate indicator of level of professional development. When the moderating effects of job resources 90 were examined across ASWs and LCSWs, job autonomy was not found to be significant in either group, and levels of job autonomy were not different between LCSWs and ASWs (t = .60, p > .04). Considering the significant moderating effect of job autonomy in the analysis of the total sample, the non-significant effects of job autonomy in the sub-group analysis may suggest that job autonomy is associated with the social workers’ licensing status. Based on these results, it is not clear why the effects of job autonomy are not consistent across the total sample and sub- sample. To help clarify this issue, future research should focus on the effects of job autonomy in social workers in specific situations. For example, the importance of job autonomy can be explored for child protective service workers, who utilize a structured decision-making process. In addition, the current study conceptualized job autonomy as a one-dimensional concept, so using multidimensional approaches of job autonomy (Breaugh, 1999) may reveal more comprehensive information regarding which facet of job autonomy is more important for a certain group of social workers. When the buffering effects of social support were examined with a sub- sample of ASWs, supervisory support was found to significantly moderate. The relationship between job demands at Time 1 and turnover intention at Time 2 and the relationship between burnout at Time 1 and turnover intention at Time 2 suggest that, although ASWs perceive high levels of job demands or burnout, support from the supervisor can help them to maintain attachment to their organization. However, the moderating effects of supervisory support were not significant in LCSWs, suggesting that a social worker in the early stages of professional development is 91 more dependent on the supervisor as a major source of support at work. This is consistent with the Shechtman and Wirzberger’s (1999) argument that workers differ in their expectations of supervision according to their stage of professional development. 5.1.3 Importance of Effective Supervision The results of the path analyses showed significant direct impacts of supervision on not only perceived job demands, but also burnout and turnover intention among social workers. The present study defines effective supervision as high-quality supervisory communication regarding social workers’ roles and tasks. Effective social work supervision requires not only professional knowledge of practice, but also skills in coordinating work and workload, setting limits and manageable goals, and monitoring the work process for frontline social workers (Kadushin, 2002). Therefore, supervisory communication regarding job-related matters is especially important. The findings regarding the significant impact of supervision on job demands support the notion that effective supervisory communication will reduce social workers’ uncertainty and conflict at work. According to Jain’s (1973) study on the relationship between the effectiveness of supervisory communication and a supervisor’s performance in hospitals, supervisees perceive as satisfactory the information they receive regarding hospital policies and tasks if they perceive the communication from the supervisor as generally positive and helpful. Jain (1973) also argued that effective supervisory communication encourages the supervisees to exchange a great deal of information about task-related 92 matters and helps supervisors to understand social workers’ opinions about their current working experience, including aspects such as workload or role conflict, based on direct input from the frontline social workers. If social workers express their role as unmanageable and stressful, supervisors could facilitate information sharing up and down the hierarchy to help administrators understand how management processes affect frontline workers’ ability to commit to their jobs. Although significant direct effects of effective supervision on burnout and turnover intention were not hypothesized, effective supervision had unique negative effects on burnout and turnover intention. Cohen and Gagin (2005) found that staff development actually reduces burnout among hospital social workers. A previous study on social workers in health care settings (Kim & Lee, in press) found that job- relevant supervisory communication had significant direct effects on turnover intention. The current findings confirm that effective supervision is associated not only with perceived job demands, but also with social workers’ burnout and their attachment to organizations. 5.2 Limitations and Strengths 5.2.1 Limitations Several caveats should be taken into consideration when interpreting the findings of this study. First, it is unclear how well these findings can be generalized to the greater social worker population. Although selected randomly, the sample was limited to registered social workers in California. According to the Current 93 Population Survey (a monthly survey of the U.S. labor market administered by the U.S. Bureau of the Census), nearly 30 percent of self-reported social workers have less than a Bachelor’s degree, while 10 percent of them have no college education (Barth, 2003). Thus, new research could to reexamine the conceptual model using a sample of entry-level social workers without MSW degrees. Overall, results based on a more nationally representative sample would increase the understanding of the burnout phenomenon in social work. The low return rate with the initial survey was a second limitation. Higher return rates might be achieved by a smaller sample size and more sophisticated research designs with multiple follow-ups. However, multiple follow-ups were not possible at Time 1 because of limited financial resources. About 25% of participants from the initial survey dropped out failed to respond to the follow-up survey. According to the study’s results of attrition analysis, it is possible that attrition from the study was related to top managerial support and turnover intention among social workers at the beginning of the research. Third, the study used self-reported data exclusively, so the possibility exists that the results of the current study were biased because of the lack of objective measures. Finally, the current study indicated that there are no significant longitudinal relationships among job demands, burnout, and turnover intention when the relationships were examined with a sample of registered social workers. The two- wave longitudinal panel design with 1-year intervals might be too long for research on social worker stress and burnout, and future research should attempt a replication of this study using shorter time intervals, such as 3 or 6 months apart. 94 5.2.2 Strengths Despite its limitations, the study makes contributions to the literature on social worker burnout and turnover issues. Strengths of this work include not only the random sample of social workers and the use of standardized instruments, but also the study’s longitudinal nature. It is generally agreed that longitudinal research is more powerful than cross-sectional design. The current study proposed a model of job resources, which allows us to understand the direct effects of job resources on job demands as well as the moderating effects of job resources on the relationships among job demands, burnout, and turnover intention among social workers. According to the literature review, the various effects of job resources in the framework of job demands-burnout-turnover intention have not, until now, been integrated as a single conceptual model. The proposed model will provide an important conceptual framework with which to understand the different effects of job resources. 5.3 Implications for Education, Practices, and Research The results of the current study suggest that 1) effective supervision reduces perceived job demands among social workers, 2) job autonomy helps social workers better cope with job demands, and 3) supervisory support is critical in preventing turnover among social workers in the early stage of professional development. These empirical findings have significant implications for social work education, practices and research in the field of social work. 95 5.3.1 Implications for Social Work Education Interventions for social worker burnout and turnover should start from the degree programs in social work. The primary focus of burnout intervention studies has been educational intervention to enhance the ability of workers to cope with the stress (Maslach, Schaufeli, & Leiter, 2001). Educational intervention to prevent burnout, based on the stress literature, demonstrated that people can learn new ways of coping (Maslach, Schaufeli, & Leiter, 2001). Therefore, the first objective of social work education is to provide a framework with which social workers can effectively confront the many problems at work. Social work education should include ample content to increase students’ awareness of symptoms and to teach ways to detect, prevent, and ameliorate the effects of burnout (Jones, 2007). Having realistic expectations for work is important for students who are about to start their careers (Edelwich & Brodsky, 1980). A field practicum can allow social work students to learn first hand what they can expect from work. Instructors should focus on not only developing students’ intervention skills and knowledge but also acknowledging students’ difficulties at their field practices. Instructors can lead students to share their experiences with other students and to discuss about ways to deal with the difficulties. In most social work programs, students may learn about theories and symptoms of burnout. During this course, social workers should learn that they are professionally responsible for taking care of themselves and avoiding burnout. Coping skills and communication techniques, which are practically useful in the field, can be included in the curriculum. 96 The second objective of social worker education is to develop social workers to become effective supervisors and managers in the future. To be an effective social work supervisor, social work students should learn the important practical and leadership skills of a social work supervisor. Kadushin (2002) suggested that effective social work supervision requires not only professional knowledge of acceptable practices, but also skills in coordinating work and workload, setting limits and manageable goals, and monitoring work process for frontline workers. Tropman, Faleer, and Feldt (2007) suggested that essential supervisory skills for child welfare supervisors include 1) planning work, 2) training/coaching/teaching/mentoring, 3) supportive communication, 4) working with workers with strong opinions, 5) running effective meetings, 6) decision-making/problem-solving, 7) monitoring/reviewing/evaluating and 8) management of self. Many social work students learn little about how to supervise others and how to perform managerial tasks. For social workers at the beginning of their professional careers, a poorly prepared supervisor could be a major reason to leave their jobs. Social work educators need to develop and to improve the curriculum of supervisory skills for social work students. 5.3.2 Implications for Organizational Practices Organizational practices can focus on organizational contexts to enhance job autonomy and support for frontline social workers. Enhancing job conditions such as job autonomy and support at work can be conceptualized as changing the organizational climate. The organizational climate is the shared perceptions of 97 practices and procedures in the organization (Schneider, 1975). Scholars of organizational climates have contended that top managerial leadership (Grojean et al., 2004) and top managerial practices (Rogg, et al., 2001) determine the organizational climates. In specific, social work administration can improve the organizational climate by focusing on developing effective supervision (Kim & Lee, in press; Hopkins & Austine, 2004a), allowing means and time flexibility (Tropman, Faller, & Feldt, 2007) and promoting participatory/collective decision-making (Hopkins & Austin, 2004b). The importance of effective supervision in preventing burnout and turnover among social workers has been widely accepted (Kim & Lee, In press; Mor Barak, Nissly, & Levin, 2001; Um & Harrison, 1998). The current study also shows that supervision is significantly associated with job demands perceived by social workers, as well as burnout and turnover intention in social workers. Considering the importance of effective supervision, developing organizational support for supervisors is essential. However, certain types of supervisor training or support may not be available at the organizational level, since most organizations do not have enough resources for staff development. In public child welfare, there have been statewide efforts to develop, implement, and evaluate supervisor-training programs (e.g., Landsman, 2007; Tropman, Faller, & Feldt, 2007). Brannon and Berry (2004) conducted a training evaluation program for supervisors in health care settings. In this study, Brannon and Berry showed that the training program improved supervisors’ behavioral supervision skills and, consequently, enhanced performance among direct care workers in long-term-care organizations. Based on their findings, 98 they argued that good supervision should become a higher priority in improving the quality of direct care work. More attention should be paid to supervisor training, since supervisors’ skills and relationships with front-line workers are critical assets and resources for the organization. One of the important findings in the current study is that supervisory support is more important for social workers at the lower skill levels. Tropman and his colleagues (2007) suggested that supervisors need to provide different time frames for workers at different levels of skills and experience to perform tasks. In addition, they suggested that supervisors need to set realistic work objectives for supervisees depending on the workers’ skill levels and experience. Inexperienced workers tend to pay close attention to rules and procedures. Considering that social workers in the beginning stage of their careers are more likely to experience burnout (Cherniss, 1980), supervisory support and coaching are essential for inexperienced social workers in these demanding job settings (Hopkins & Austin, 2004a). Another issue is to improve front-line social workers’ autonomy at work. Job autonomy has multiple facets. Breaugh (1985) proposed three types of autonomy as work: work method autonomy (i.e., the degree of discretion/choice individuals have regarding the procedures/methods they utilize in going about their work), work scheduling autonomy (i.e., the extent to which workers feel they can control the scheduling/sequencing/timing of their work activities), and work criteria autonomy (i.e., the degree to which workers have the ability to modify or choose the criteria used for evaluating their performance). One or another of these aspects of job autonomy may be more important for social workers depending on type of tasks they 99 work on or the goals they are supposed to accomplish. Human service managers should understand which aspects of job autonomy are most needed for workers in their organizations. The final issue is to build a positive culture in the organization through participatory/collective decision-making processes (Hopkins & Austin, 2004b). Social workers are often not encouraged to participate in making decisions regarding the social workers’ own practices and the organization (Packard, 1993). Therefore, supervisors have an important role in helping frontline workers engage with organizational decision-making. Effective supervisors should be able to communicate with frontline workers regarding areas in need of change and adjustment and then help mangers to recognize the issues addressed by frontline workers. When social workers perceive their supervisory relationships as trustworthy, they will be more likely to perceive the organizational climate as more participatory. Of course, having formal channels through which to promote frontline social workers’ voices, such as regular team meetings, suggestion boxes, and anonymous internal survey for organizational members, is also important. 5.3.3 Implication for Research toward Evidence-Based Organizational Practices Many administrators and scholars in social work may perceive that the findings presented here are not new because there has already been a substantial amount of literature on social worker burnout and turnover. However, most of this material has dealt with the questions of what causes burnout and turnover and what 100 can help social workers to cope with them. Perhaps the most daunting challenge for enhancing job resources in social work is to draw attention from schools, social welfare organizations, community, and governments on the issue of social workers’ occupational well being. Recently, Soydan (2007) discussed the importance of extra- institutional sources, such as governments and professional organizations like the Council on Social Work Education (CSWE) and the National Association of Social Workers (NASW), in order to change social work agencies’ approach to managing social workers. His discussion was intended to explain how to change social work education to embrace Evidence-Based Practices (EBP). However, this idea can also be applied to changing the service organizations to employ various interventions for social worker burnout and turnover. In fact, NASW has undertaken research on social workers’ job demands, which is critical to supporting organizational interventions for social worker burnout and turnover. (See a report of workforce study from NASW, 2006.) According to Thyer’s (2006) discussion, the EBP process consists of five steps: 1. Converting one’s need for information into an answerable question. 2. Tracking down the best available evidence to answer that question. 3. Critically appraising that evidence for its validity, impact and applicability. 4. Integrating this critical appraisal with our practice skills and our client’s unique values, preferences, and circumstances. 5. Evaluating our effectiveness and efficiency in executing Steps 1-4 and seeking ways to improve them both for next time. 101 A detailed discussion of each step of the EBP process is beyond the scope of this study. However, it is clear that EBP can be applied to macro social work practices only if there is valid empirical evidence to support the practices. Development of organizational practices to prevent burnout and turnover in social workers should be based on clear-cut empirical evidence showing significant impacts of social workers’ occupational well being on client outcomes. A study by Yoo and Brooks (2005) is one of a few studies that has presented empirical evidence of the impacts of organizational variables in child welfare outcomes (i.e., out-of-home placement outcomes). Studies by Glisson and Hemmelgarn (1998) and by Rogg et al. (2001) demonstrated that the organizational context has significant effects on client outcomes. However, empirical evidence linking organizational factors, service providers’ burnout, and client outcomes is still limited. In order to improve the empirical knowledge on the issue, several topics can be examined in the future. First, there is a need to conduct more longitudinal research to understand burnout and turnover intention among service providers, including social workers, child welfare workers, and other human service workers. In particular, understanding how burnout among inexperienced workers develops over time is of practical as well as theoretical interest. The future study needs to examine the developmental process of burnout with a cohort of social workers at the beginning of their professional careers. At the same time, the study should examine the effects of social work education and training on the trajectories of burnout and turnover intention among social workers. It is also important to develop an organized screening and care program for social workers with high levels of job demands (e.g., 102 child protection service workers and mental health service workers for clients with severe mental illness). Well designed program evaluation research can examine the impacts of the program on the trajectories of burnout and turnover among social workers, and ultimately client outcome. The current research can be expanded by examining the impacts of job stress and coping resources on various strain outcomes, such as mental and physical health problems (e.g., burnout, depression and somatic symptoms) and work performance outcomes (e.g., service fidelity and job-seeking behaviors). This research will provide important implications for social work education and training programs. 5.4. Conclusions This study provides substantial empirical evidence that job conditions such as effective supervision and job autonomy are important in preventing burnout and turnover among social workers and that supervisory support is critical for social workers in the early stage of professional development. Clients are served better when the staff is more optimistic and more satisfied. Given the prevalence of the burnout and turnover problem, social work administrators and researchers must turn their attention to how best to develop these resources in organizational settings and how to support social workers more effectively and efficiently. The ultimate goal of social work practice is to help clients in needs. In order to achieve this goal, more efforts and resources should be given to occupational well being of service providers. 103 REFERENCES Albrecht, T. L., & Adelman, M. B. (1987). Communicating social support. Newbury Park, CA: Sage. Alexander, J. A., Lichtenstein R, Oh, H. J., and Ullman, E., (1998). A Causal Model of Voluntary Turnover among Nursing Personnel in Long-Term Psychiatric Settings. Research in Nursing and Health, 21, 415-427. Alliance for Children and Families, American Public Human Services Association, and Child Welfare League of America. (May 2001). The child welfare workforce challenge: results from a preliminary study. Washington, DC: Child Welfare League of America. Arbuckle, J. L. (1996). Full information estimation in the presence of incomplete data. In G. A. Marcoulides & R. E. Schumacker (Eds.), Advanced structured equation modeling (pp. 243-247). Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Arbuckle, J. L. (1997). Amos User’s guide version 4.0. Chicago: Small Waters Corporation. Arches, J. (1991). Social structure, burnout, and job satisfaction. Social Work, 36, 202-206. Austin, M. (2003). The changing relationship between nonprofit organizations and public social service agencies in the era of welfare reform. Nonprofit and Voluntary Sector Quarterly, 29, 69-97. Avolio, B. J., Zhu, W., Koh, W., & Bhatia, P. (2004). Transformational leadership and organizational commitment: mediating role of psychological empowerment and moderating role of structural distance. Journal of Organizational Behavior, 25, 951-968. Baker, E., Israel, B., & Schurman, S. (1996). Role of control and support in occupational stress: an integrated model. Social Science and Medicine, 43, 1145-1159. Bakker, A. B., Hakanen, J. J., Demerouti, E., & Xanthopoulou, D. (2007). Job resources boost work engagement, particularly when job demands are high. Journal of educational psychology, 99, 274-284. Bakker, A. B., Demerouti, E., Taris, T. W., Schaufeli, W. B., & Schreurs, P. J. G. (2003). A multigroup analysis of job demands-resources model in four home care organizations. International Journal of Stress Management, 10, 16-38. 104 Bakker, A. B., Demerouti, E., & Euwema, M. C. (2005). Job resources buffer the impact of job demands on burnout. Journal of Occupational Health Psychology, 10(2), 170-180. Bakker, A. B., Schaufeli, W. B., Sixma, H., Bosveld, W., & Van Dierendonck, D. (2000). Patient demands, lack of reciprocity, and burnout: A five-year longitudinal study among general practitioners. Journal of Organizational Behavior, 21, 425-441. Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice-Hall. Barth, M. C. (2003). Social work labor market: A first look. Social Work, 48(1), 9- 19. Bednar, S. G. (2003). Elements of satisfying organizational climates in child welfare agencies. Families in Society, 84(1), 7-12. Beehr, T. A. (1995). Psychological stress in the workplace (2 nd ed.). NY: Routledge, Co. Benner, P., Tanner, C. A., & Chesla, C. A. (1996). Expertise in nursing practice: Caring, clinical judgment, and ethics. New York: Springer. Berger, C. R., & Calabrese, R. J. (1975). Some explorations in initial interaction and beyond: Toward a developmental theory of interpersonal communication. Human Communication Research, 1(2), 99-112. Berliner, D. (1994). Expertise: The wonder of exemplary performance. In J. Mangieri & C. Block (Eds.), Creating powerful thinking in teachers and students: Diverse perspectives (pp. 161-186). Fort Worth, TX: Harcourt Brace College. Biddle, B. J. (1986). Recent developments in role theory. Annual Review of Sociology, 12, 67-92. Bordieri, J.E., Reagle, D.Y., & Coker, C.C., (1988). Job satisfaction and leadership style of rehabilitation facility personnel. Rehabilitation Counseling Bulleting, 32, 149-160. Brannon, S., & Barry, T. (2004). Supervision and the quality of direct care work. The Gerontologist, 44(1), 154. Bransford, C. L. (2005). Conceptions of authority within contemporary social work practice in managed mental health care organizations. American Journal of Orthopsychiatry, 75(3), 409–420. 105 Breaugh, J. A. (1985). The measurement of work autonomy. Human Relations, 38, 551-570. Breaugh, J. A. (1999). Further investigation of the work autonomy: two studies. Journal of Business and Psychology, 13(3), 357- 373. Brewer, E. W., & Shapard, L. (2004). Employee burnout: A meta-analysis of the relationship between age or years of experience. Human Resource Development Review, 3(2), 102-123. Brookings, J. B., Bolton, B., Brown, C., & McEvoy, A. (1985). Self-reported job burnout among female human service professionals. Journal of Occupational Behavior, 6, 143-150. Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen, & J. S. Long (Eds.), Testing structural equation models (pp. 136- 162). Newbury Park, CA: Sage. Bscmscphd, J. H. (2007). Research Article: The psychosocial work environment and burnout among Swedish registered and assistant nurses: The main, mediating, and moderating role of empowerment. Nursing and Health Sciences, 9(3), 205-211. Bureau of Labor Statistics, U.S. Department of Labor, Occupational Outlook Handbook, 2006-07 Edition, Social Workers, on the Internet at http://www.bls.gov/oco/ocos060.htm (visited August 31, 2006). Burns, J. M. (1978). Leadership. New York: Harper & Row, Co. Butler, J. K., & Cantrell, R. S., (1997). Effects of perceived leadership behaviors on job satisfaction and productivity. Psychological Report, 80, 976-978. Bycio, P., Hackett, R.D., & Allen, J.S., (1995). Further assessment of Bass’s (1985) conceptualization of transactional and transformational leadership. Journal of Applied Psychology, 80(4), 468-478. Campbell, D. T., & Stanley, J. C. (1963). Experimental and quasi-experimental designs for research on teaching. In Handbook of Research on Teaching, ed. N. L. Gage, 5, 171-246. New York: Rand-McNally. Carver, C., Scheier, M. F., & Weintraub, J. K. (1989). Assessing coping strategies: A theoretically based approach. Journal of Personality and Social Psychology, 56(2), 267-283. 106 Center for Workforce Studies, NASW (2005). Assuring the sufficiency of frontline workforce: A national study of licensed social workers preliminary report. <http://www.socialworkers.org/resources/workforce/files/NASW_SWCassur ing_3.pdf> (Accessed: 22-July-07). Center for Workforce Studies, NASW. (2006). Licensed Social Workers in the U.S., 2004. http://workfroce.socialworkers.org/studies/intro.pdf. (Accessed: 22- June-06). Cherniss, C. (1980). Professional burnout in the human service organizations. New York: Praeger. Cherniss, C. (1995). Beyond burnout: helping teachers, nurses, therapists, and lawyers recover from stress and disillusionment. New York, NY: Routledge. Cohen, S., Doyle, W. J., Skoner, D. P., Rabin, B. S., & Gwaltney, J. M. (1997). Social ties and susceptibility to the common cold. Journal of the American Medical Association, 277, 1940-1944. Cohen, M., & Gagin, R. (2005). Can skill-development training alleviate burnout in hospital social workers? Social Work in Health Care, 40(4), 83-97. Cohen, S., Underwood, L. G., & Gottlieb, B. H. (2000). Social support measurement and intervention: A guide for health and social scientists. Oxford, NY: Oxford University Press. Cohen, S., & Wills, T. A. (1985). Stress, social support and buffering hypothesis. Psychological Bulletin, 98, 310-357. Conger, J. A., & Kanungo, R. N. (1988). The empowerment process: Integrating theory and practices. The Academy of Management Review, 13(3), 471-482. Cooper, C. L., Dewe, P. J., & O’Driscoll, M. P. (2001). Organizational stress: a review and critique of theory, research, and applications. Thousand Oaks, CA: Sage. Cordes, C. L., & Dougherty, T. W. (1993). A review and an integration of research on job burnout. Academy of Management Review, 18(4), 621–656. Cordes, C. L., Dougherty, T. W., & Blum, M. (1997). Patterns of burnout among mangers and professionals: A comparison of models. Journal of Organizational Behavior, 18(6), 685-701. 107 Curpall, S. C., Towler, A. J., Judge, T. A., & Kohn, L. (2005). Pay satisfaction and organizational outcomes. Personnel Psychology, 58(3), 613-640. Dall’Alba, G., & Sandberg, J. (2006). Unveiling professional development: A critical review of stage models. Review of Education Review, 76(3), 383-412. Dansereau, F., & Markham, S. E. (1987). Superior-subordiante communication: Multiple levels of analysis. In F. M. Jablin, L. L. Putnam, K. H. Bobert & L. W. Porter (Eds), Handbook of Organizational Communication, pp. 343-388. Newbury Park, CA: Sage. De Croon, E. M., Sluiter, J. K., Blonk, R. W., Broersen, J. P., & Frings-Dresen, M. H. (2004). Stressful work, psychological job strain, and turnover: A 2-year prospective cohort study of truck drivers. Journal of Applied Psychology, 89(3), 442-454. De Jonge, J., Dormann, C., Janssen, P. P. M., Dollard, M. F., Landeweerd, J. A., & Nijhuis, F. J. N. (2001). Testing reciprocal relationships between job characteristics and psychosocial well-being: A cross-lagged structural equation model. Journal of Occupational and Organizational Psychology, 74, 29-46. Demerouti, E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001). The job demands-resources model of burnout. Journal of Applied Psychology, 86(3), 499–512. Diener, E., & Fujita, F. (1995). Resources, personal strivings, and subjective well- being: A nomothetic and idiographic approach. Journal of Personality and Social Psychology, 68, 926-935. Dollard, M., Winefield, H. R., Winefield, A. H., & de Jonge, J. (2000). Psychological job strain and productivity in human service workers: A test of the demand-control-support model. Journal of Occupational and Organizational Psychology, 73(4), 501-510. Drake, B., & Yamada, G. N. (1996). A structural equation model of burnout and job exit among child protective service workers. Social Work Research, 20, 179– 189. Dreyfus, H. L., & Dreyfus, S. E. (1986). Mind over machine: The power of human intuition and expertise in the era of the computer. New York: Free Press. Dyeson, T. B. (2004). Social work licensure: A brief history and description. Home Health Care Management Practice, 16(5), 408-411. 108 Eby, L. T., Freeman, D. M., Rush, M. C., & Lance, C. E. (1999). Motivational bases of effective organizational commitment: A partial test of an integrative theoretical model. Journal of Occupational and Organizational Psychology, 72(4), 463-482. Edelwich, J., & Brodsky, A. (1980). Burn-out: Stages of disillusionment in the helping professions. New York: Human Sciences Press. Ellis, B. H., & Miller, K. I. (1994). Supportive communication among nurses: effects on commitment, burnout, and retention. Health Communication, 6(2), 77-96. Feiman-Nemser, S., & Remillard, J. (1996). Perspectives on learning to teach. In F. B. Murray (Ed.), The teacher educator’s handbook (pp. 63-91). San Francisco: Jossey-Bass. Feldt, T. Kivimäki, M., Rantala, A., & Tolvanen, A. (2004). Sense of coherence and work characteristics: A cross-lagged structural equation model among managers. Journal of Occupational and Organizational Psychology, 77, 323- 342. Firth, H., & Britton, P., (1989). “Burnout,” absence and turnover among British nursing staff. Journal of Occupational Psychology, 62(1), 55-59. Fox, M. L., Dwyer, D. J., & Ganster, D. C. (1993). Effects of stressful job demands and control on physiological and attitudinal outcomes in a hospital setting. Academy of Management Journal, 36(2), 289. Freudenberger, H. (1980). Burnout. New York, NY: Doubleday. Geurts, S., Schaufeli, W., & De Jonge, J. (1998). Burnout and intention to leave among mental health-care professionals: A social psychological approach. Journal of Social and Clinical Psychology, 17(3), 341-362. George, J. M., & Jones, G. R. (1997). Experiencing work: Values, attitudes and moods. Human Relations, 50, 393–416. Glass, D. C., & McKnight, J. D. (1996). Perceived control, depressive symptomatology, and professional burnout: A review of the evidence. Psychology and Health, 11, 23-48. Glisson, C., & Hemmelgarn, A. (1998). The effects of organizational climate and interorganizational coordination on the quality and outcomes of children’s service systems. Child Abuse & Neglect, 5, 401-421. 109 Grojean, M. W., Resick, C. J., Dickson, M. W., & Smith, D. B. (2004). Leaders, values, and organizational climate: Examining leadership strategies for establishing an organizational climate regarding ethics. Journal of Business Ethnic, 55, 223-241. Hackman, J. R., & Oldham, G. R. (1976). Motivation through the design of work: Tests of a theory. Organizational Behavior and Human Performance, 16, 250-279. Hallberg, U. E., & Sverke, M. (2004). Construct validity of the Maslach Burnout Inventory: Two Swedish health care sample. European Journal of Psychological Assessment, 20(4), 320-338. Harrington, D., Bean, N., Pintello, D., & Mathews, D. (2001). Job satisfaction and burnout: Predictors of intentions to leave a job in a military setting. Administration in Social Work, 25(3), 1–16. Hasenfeld, Y., (1984). The changing context of human services administration. Social Work, 29, 522-529. Hendrix, W. H., Robbins, T., Miller, J., & Summers, T. P., (1999). Effects of procedural and distributive justice on factors predictive of turnover. Journal of Social Behavior and Personality, 13(4), 611-632. Himle, D. P., Jayaratne, S., & Thyness, P. (1989). The effects of emotional support on burnout, work stress and mental health among Norwegian and American social workers. Journal of Social Service Research, 13, 27–45. Hobfoll, S. E. (2002). Social psychological resources and adaptation. Review of General Psychology, 6(4), 307-324. Hobfoll, S. E., & Leiberman, Y. (1987). Personality and social resources in immediate and continued stress resistance amog women. Journal of Personality and Social Psychology, 52, 18-26. Holdnak, B.J., Harsh, J., & Bushardt, S.C., (1993). An examination of leadership style and its relevance to shift work in an organizational setting. Health Care Management Review, 18(3), 21-30. Holtom, B. C., & O’Neill, B. S. (2004). Job embeddedness: A theoretical foundation for developing a comprehensive nurse retention plan. Journal of Nursing Administration, 34(5), 216-227. 110 Hopkins, K. M., & Austin, M. J. (2004a). Coaching employees with performance problems. In K. M. Hopkinds & M. J. Austin (Eds.), Supervision as collaboration in the human services (pp. 215-226). Thousand Oaks, CA: Sage Publication, Inc. Hopkins, K. M., & Austin, M. J. (2004b). The changing nature of human services and supervision. In K. M. Hopkinds & M. J. Austin (Eds.), Supervision as collaboration in the human services (pp. 215-226). Thousand Oaks, CA: Sage Publication, Inc. Houkes, I., Janssen, P. P. M., de Jonge, J., & Bakker, A. B. (2003). Specific determinants of intrinsic work motivation, emotional exhaustion and turnover intention: A multissample longitudinal study. Journal of Occupational and Organizational Psychology, 76, 427-450. House, J. S. (1981). Work stress and social support. MA: Addison-Wesley Publishing Company. Hourse, R. J., & Baetz, M. L. (1979). Leadership: Some empirical generalizations and new research directions. Research in Organizational Behavior, 1, 341- 423. Huang, I., Chuang, C. J., & Lin, H. (2003). The role of burnout in the relationship between perceptions of organizational politics and turnover intentions. Public Personnel Management, 32(4), 519–531. Huseman, R. C., Hatfield, J. D., Boulton, W., & Gatewood, R. (1980). Development of a conceptual framework analyzing the communication-performance relationship. In K. H. Chung (Ed.), Academy of Management Proceedings, 178-182. Wichita, KS: Academy of Management. Jackson, S. E., Schuler, R. S., & Rivero, J. C. (1989). Organizational characteristics as predictors of personnel practices. Personnel Psychology, 42, 727-786. Jackson, S. E., Schwab, R. L., & Schuler, R. S. (1986). Toward an understanding of the burnout phenomenon. Journal of Applied Psychology, 71, 630-640. Jain, H. C. (1973). Supervisory communication and performance in urban hospital. The Journal of Communication, 23(1), 103-117. Janssen, P. M. (1999). Work-related and individual determinants of the three burnout dimensions. Work & Stress, 13(1), 74-86. 111 Jinnett, K., & Alexander, J. A. (1999). The influence of organizational context on quitting intention. Research on Aging, 21(2), 176-204. Jones, F., & Kinman, G., (2001). Approaches to studying stress. In Stress: Myth, theory, and research ed. F. Jones & J. Bright, pp. 17-45. UK: Person Education Limited. Johnson, J. V., & Hall, E. M. (1988). Job strain, work place social support, and cardiovascular disease: A cross-sectional study of a random sample of the Swedish working population. American Journal of Public Health, 78, 1336- 1342. Jones, F., Bright, J., & Clow, A. (2001). Stress: myth, theory and research. New York: Prentice Hall. Jones, S. H. (2007). Secondary trauma and burnout in child protective workers: Implications for preparation of social workers. Paper presented at the 11 th Annual Conference of the Society for Social Work and Research, San Francisco, CA. Kadushin, A. (1992). Social work supervision: An updated survey. The Clinical Supervisor, 10(2), 9-27. Kadushin, A. (2002). Supervision in social work (3 rd ed.). New York: Columbia University Press. Karasek, R. A., Brisson, C., Kawakami, N., Houtman, I. L. D., Bonders, P., & Amick, B. (1998). The Job Content Questionnaire (JCQ): An instrument for internationally comparative assessments of psychological job characteristics. Journal of Occupational Health Psychology, 3(4), 322-355. Karasek, R., & Theorell, T. (1990). Healthy work. New York: Basic Books. Kessler, R. C., & Greenberg, D. F. (1981). Linear panel analysis: Models of quantitative change. New York: Academic Press. Kim, H., & Ji, J. (in press). Factor structure and longitudinal invariance of Maslach Burnout Inventory (MBI-HSS) for social work research. Research on Social Work Practice. Kim, H., & Lee, S. Y. (in press). Supervisory communication, burnout, and turnover intention among social workers in health care settings. Social Work in Health Care. 112 Kim, H., & Stoner, M. (in press). Burnout and turnover intention among social workers: Effects of role stress, job autonomy and social support. Journal of Administration in Social Work. Kim, H., & Stoner, M. (2007). An Empirical test of conceptual model of turnover intention among registered social workers. Paper presented at the Annual Conference of the Society for Social Work and Research, San Francisco, CA. Kline, R. B. (1998). Principles and practice of structural equation modeling. New York: Guilford Press. Kluger, A. N., & DeNisi, A. (1996). The effects of feedback interventions on performance: A historical review, a meta-analysis, and preliminary feedback intervention. Psychological Bulletin, 119, 254-284. Koeske, G. F., & Koeske, R. D. (1989). Work load and burnout: Can social support and perceived accomplishment help? Social Work, 34, 243–248. Kompier, M., & Cooper, C. (1999). Preventing stress, improving productivity: European case studies in the workplace. London: Routledge. Krausz, M., Koslowsky, M., Shalom, N., & Elyakim, N. (1995). Predictors of intentions to leave the ward, the hospital, and the nursing profession: A longitudinal study. Journal of Organizational Behavior, 16, 277-288. Landsman, M. (2007). Supporting child welfare supervisors to improve worker retention. Child welfare, 86(2), 105. Landstrom, G. L., Biordi, D. L., & Gillies, D. A. (1989). The emotional and behavioral process of staff nurse turnover. Journal of Nursing Administration, 19(9), 23-28. Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer Pub. Co. Lecroy, C. W., & Rank, M. R. (1987). Factors associated with burnout in the social services: An exploratory study. Journal of Social Service Research, 10, 23- 56. Lee, R. T. & Ashforth, B. E. (1993). A longitudinal study of burnout among supervisors and mangers: Comparisons between the Leiter and Maslach (1988) and Golembiewski et al. (1986) Models. Organizational Behavior and Human Decision Process, 54, 369-398. 113 Lee, R. T., & Ashforth, B. E. (1996). A meta-analytic examination of the correlates of the three dimensions of job burnout. Journal of Applied Psychology, 81(2), 123–133. Lee, T. W., Mitchell, T. R., Sablynski, C. J., Burton, J. P., Holtom, B. C. (2004). The effects of job embeddedness on organizational citizenship, job performance, volitional absences, and voluntary turnover. Academy of Management Journal, 47(5), 711. Leiter, M. P., & Maslach, C. (1988). The impact of interpersonal environment on burnout and organizational commitment. Journal of Organizational Behavior, 9, 297-308. Light, P. D. (2003). The Health of the Human Service Workforce. The Brookings Institution. Liu, C., Spector, P. E., & Jex, S. M. (2005). The relation of job control with job strains: A comparison of multiple data sources. Journal of Occupational and Organizational Psychology, 78(3), 325-336. Lloyd, C., & King, R. (2004). A survey of burnout among Australian mental health occupational therapists and social workers. Social Psychiatry and Psychiatric Epidemiology, 39, 752-757. Loehlin, J. C. (2004). Latent variable models: an introduction to factor, path, and structural equation analysis (4 th ed.). New Jersey: Lawrence Erlbaum Associates, Publishers. London, M., Larsen, H. H., & Thisted, L. N. (1999). Relationships between feedback and self-development. Group & Organization Management, 24(1), 5-27. Maslach, C., & Jackson, S. E. (1981). The measurement of experienced burnout. Journal of Occupational Behavior, 2, 99-113. Maslach, C., & Jackson, S. E. (1986). The Maslach Burnout Inventory. Palo Alto, CA: Consulting Psychologists Press. Maslach, C., & Leiter, M. P. (1997). The truth about burnout: How organization cause personal stress and what to do about it. San Francisco, CA: Jossey- Bass. Maslach, C., Schaufeli, W. B., & Leiter, M. P. (2001). Job burnout. Annual Review of Psychology, 52, 397-422. 114 Miller, K. I., Ellis, B. H., Zook, E. G., & Lyles, J. S. (1990). An integrated model of communication, stress, and burnout in the workplace. Communication Research, 17(3), 300-326. Miller, K. I., & Monge, P. R. (1985). Social information and employee ansiety about organizational change. Miles, E. W., Patrick, S. L., & King, W. C. (1996). Job level as a systemic variable in predicting the relationship between supervisory communication and job satisfaction. Journal of Occupational and Organizational Psychology, 69, 277-292. Mitchell, T. R., Holtom, B. C., Lee, T. W., Sablynski, C. J., & Erez, M. (2001). Why people stay: Using job embeddedness to predict voluntary turnover over. Academy of Management Journal, 44(6), 1102-1121. Moore, J. E. (2000). One road to turnover: An examination of work exhaustion in technology professionals, MIS Quarterly, 24(1), 141-168. Mor Barak, M. E., Nissly, J. A., & Levin, A. (2001). Antecedents to retention and turnover among child welfare, social work, and other human service employees: What can we learn from past research? A review and metaanalysis. Social Service Review, 75, 625-661. Muthén, L. K., & Muthén, B. (1998). Mplus user’s guide. Los Angeles: Muthén & Muthén. National Association of Social Workers California Chapter (2007). Ten basic steps to a California LCSW. <http://www.naswca.org/licensing/10steps.pdf>. (Accessed: 1-July-07). Neale, M. C., Boker, S. C., Xie, G., & Maes, H. H. (1999). Mx: Statistical modeling (5 th ed.). Richmond: Medical College of Virginia. Nedd, N. (2006). Perceptions of empowerment and intent to stay. Nursing Economics, 24(1), 13-18. NIOSH, (1999). Stress at work. DHHS (NIOSH) publication No. 99-101. Nissly, J. A. (2004). Prospective and retrospective examinations of factors related to intention to leave and turnover among public child welfare workers. Unpublished doctoral dissertation, University of Southern California, California. 115 Nissly, J. A., Mor Barak, M. E., & Levin, A. (2005). Stress, social support, and workers’ intentions to leave their jobs in public child welfare. Administration in Social Work, 29(1), 79–100. Oaklander, H., & Fleishman, E. A. (1964). Patterns of leadership related to organizational stress in hospital settings. Administrative Science Quarterly, 8, 520-532. O’Driscoll, M. P., & Beehr, T. A. (2000). Moderating effects of perceived control and need for clarity on the relationship between role stressros and employee affective reactions. The Journal of Social Psychology, 140(2), 151-159. Olsen, M. K., & Schafer, J. L. (1998). Parameter estimates for semicontinuous longitudinal data using an EM algorithm. Technical Report #98–31, Pennsylvania State University, Methodology Center. Packard, T. (1993). Managers’ and workers’ views of the dimensions of participation in organizational decision making. Administration in Social Work, 17, 53-65. Parker, P. A., & Kulik, J. A. (1995). Burnout, sef-and supervisor-rated job performance, and absenteeism among nurses. Journal of Behavioral Medicine, 18(6), 581-599. Pasztor, E. M., Saint-Germain, M., DeCrescenzo, T. (2002). Demands for social workers in California. Retrieved October 22, 2006, from http://www.csus.edu/calst/government_affairs/reports/Demand_for_Social_ Workers.pdf Patti, R. J. (2000). The handbook of social welfare management. Thousand Oaks, CA: Sage Publications, Inc. Pearlin, L. I. (1991). The study of coping: An overview of problems and directions. In J. Echenrode (Ed.), The social context of coping (pp. 261-176). New York, NY: Plenum. Pearline, L. I., & Schooler, C. (1978). The structure of coping. Journal of Health and Social Behavior, 19, 2-21. Pine, A., Aronson, E., & Kafry, D. (1981). Burnout: From Tedium to Personal Growth, Free Press, New York. Posig, M., & Kickul, J. (2003). Extending our understanding of burnout: Test of an integrated model in nonservice occupations. Journal of Occupational Health Psychology, 8(1), 3–19. 116 Powell, M. J., & York, R. O. (1992). Turnover in county public welfare agencies. Journal of Applied Social Science, 16(2), 111-127. Priebe, S., Fakhoury, W. K. H., Hoffmann, K., & Powell, R. A. (2005). Morale and job perception of community mental health professionals in Berlin and London. Social Psychiatry and Psychiatric Epidemiology, 40(3), 223-232. Rabe, J. (2006, May 23). Report shows juvenile agency can’t keep social workers. Daily Oklahoman. Retrieved September 5, 2006, from http://web.lexis- nexis.com. Raykov, T. (2005). Analysis of longitudinal studies with missing data using covariance structure modeling with full-information maximum likelihood. Structural Equation Modeling, 12(3), 493-505. Reagh, R. (1994). Public child welfare professionals: Those who stay. Journal of Sociology & Social Welfare, 21(3), 69-78. Rigdon, E. E., Schumacher, R. E., & Wothke, W. (1998). A comparative review of interaction and nonlinear modeling. In R. E. Schumacher & G. A. Marcoulides (Eds.), Interaction and nonlinear effects in structural equation modeling. Mahwah, NJ: Erlbaum. Rizzo, J. R., House, R. J., & Lirtzman, S. I. (1970). Role conflict and ambiguity in complex organizations. Administrative Science Quarterly, 15, 150-164. Rogg, K. L., Schmidt, D. B., Shull, C., & Schmitt, N. (2001). Human resource practices, organizational climate, and customer satisfaction. Journal of Management, 27, 431-449. Rubin, D. B. (1987). Multiple imputation for nonresponse in surveys. John Wiley & Sons, New York. Rycraft, J. (1994). The party isn’t over: The agency role in the retention of public child welfare caseworkers, social work, 39, 75-80. Salancik, G. R., & Pfeffer, J. (1978). A social information processing approach to job attitudes and task design. Administration Science Quarterly, 23, 224-253. Salanova, M., Agut, S., & Peiró, J. M. (2005). Linking organizational resources and work engagement to employee performance and customer loyalty: the mediation of service climate. Journal of Applied Psychology, 90, 1217-1227. Samantrai, K. (1992). Factors in the decision to leave: Retaining social workers with MSWs in public child welfare. Social Work, 37, 454-458. 117 Savicki, V., & Cooley, E. J. (1994). Burnout in child protective service workers: A longitudinal study. Journal of Organizational Behavior, 15, 655-666. Schafer, J. L. (1999). NORM: Multiple imputation of incomplete multivariate data under a normal model, version 2. Software for Windows 95/98/NT. Available from http://www.stat.psu.edu/~jls/misoftwa.html. Schaufeli, W. B., & Enzmann, D. (1998). The burnout companion to study and practice: a critical analysis. London: Taylor & Francis. Schaufeli, W. B., Van Dierendonck, D., & Van Gorp, K. (1996). Burnout and reciprocity: Towards a dual-level social exchange model. Work & Stress, 3, 225-237. Shechtman, Z., & Wirzberger, A. (1999). Needs and preferred style of supervision among Israeli school counselors at different states of professional development. Journal of counseling and development, 77(4), 456-464. Schneider, B. (1975). Organizational climates: An essay. Personnel Psychology, 28, 447-479. Selye, H. (1993). History and present status of the stress concept. In L. Goldberger & S. Breznitz (Eds.), Handbook of stress: Theoretical and clinical aspects (pp. 7-17), New York: Free Press. Siegel, S., & Castellan, N.J., Jr. (1988). Nonparametric statistics for the behavioral sciences. New York: McGraw-Hill. Smith, B. D. (2005). Job retention in child welfare: Effects of perceived organizational support, supervisor support, and intrinsic job value. Children and Youth Service Review, 27(2), 153-169. Somers, M. J. (1996). Modeling employee withdrawal behavior over time: A study of turnover using survival analyses. Journal of Occupational and Organizational Psychology, 69, 315-326. Söderfeldt, M., Söderfeldt, B., & Warg, L. E. (1995). Burnout in social work. Social Work, 40, 638–646. Soydan, H. (2007). Improving the teaching of evidence-based practice: Challenges and priorities. Research on Social Work Practice, 17(5), 612-618. Spector, P. E. (1986). Perceived control by employees: A meta-analysis of studies concerning autonomy and participation at work. Human Relations, 39, 1005- 1016. 118 Spector, P. E. (1998). A control theory of the job stress process. In C. L. Cooper (Ed.), Theories of organizational stress (pp. 153-169). Oxford: Oxford University Press. Steel, R. P., & Ovalle, N. K. (1984). A review and meta-analysis of research on the relationship between behavioral intentions and employee turnover. Journal of Applied Psychology, 69, 673-686. Taris, W. T. (2000). Primer in Longitudinal Data Analysis. London, GBR: Sage Publications, Inc. Taunton, R.L, Boyle, D.K., Woods, C.Q., Hansen, H.E., & Bott, M.J., (1997). Manager leadership and retention of hospital staff nurses. Western Journal of Nursing Research, 19(2), 205-226. The Annie E. Casey Foundation 2006, The Advocasey Index. (n.d.). Trouble on the human services frontline. Retrieved September 25, 2006, from http://www.aecf.org/publications/advocasey/spring2004/advocasey_index.ht m. Thomas, D. C. (1982). Stages of professional development in child care work: A proposed model. Child Care Quarterly, 11(2), 147-149. Thomas, K. W., & Velthouse, B. A. (1990). Cognitive elements of empowerment: An “interpretive” model of intrinsic task motivation. Academy of Management Review, 15(4), 666-681. Thyer, B. (2006, October). Evidence-based macro practice: Addressing the challenges and opportunities for social work education. Paper presented at the symposium “ Improving the Teaching of Evidence-Based Practice,” University of Texas at Austin School of Social Work. Retreived October 18, 2007, from http://www.utexas.edu/ssw/ceu/practice/papers/EBPpaperThyer.pdf Toppinen-Tanner, S., Kalimo, R., & Mutanen, P. (2002). The process of burnout in white-collar and blue-collar jobs: Eight-year prospective study of exhaustion. Journal of Organizational Behavior, 23, 555–570. Toppinen-Tanner, S., Ojajärvi, A., Väänaanen, A., Kalimo, R., & Jäppine, P. (2005). Burnout as a predictor of medically certified sick-leave absences and their diagnosed causes. Behavioral Medicine, 31(1), 18-32. Trevor, C. O. (2001). Interactions among actual ease-of-movement determinants and job satisfaction in the prediction of voluntary turnover. Academy of Management Journal, 44(4), 521-638. 119 Tropman, J., Faller, K. C., & Feldt, S. (2007). Essential of Supervisory Skills for Child Welfare Managers. Retrieved October 1, 2007, from http://www.ssw.umich.edu/tpcws/articles/EntireEditedManual_9-14- 04CE.pdf. Tsui, A. S., & Ohlott, P. (1988). Multiple assessment of managerial effectiveness: Interrater agreement and consensus in effectiveness models. Personnel Psychology, 41, 779-803. Um, M. Y., & Harrison, D. F. (1998). Role stressors, burnout, mediators, and job satisfaction: A stress-strain- outcome model and an empirical test. Social Work Research, 22(2), 100–115. U.S. Department of Health and Human Services. 2006. The supply and demand of professional social workers providing long-term care services: Report to congress. Retrieved September 25, 2006, from http://aspe.hhs.gov/daltcp/reports/2006/SWsupply.htm. van Veldhoven, M., Taris, T. W., de Jonge, J., & Broersen., S. (2005). The relationship beween work characteristics and employee health and well- being: How much complexity do we really need? International Journal of Stress Management, 12(1), 3-28. Wilcox, B. L., & Vernberg, E. M. (1985). Conceptual and theoretical dilemmas facing social support research. In I. G. Sarason & B. R. Sarason (Eds.), NATO ASI Series, series D: behavioral and social science: No. 24. Social support: Theory, research, and applications (pp. 3-20). Dordrecht: Martinus Nijhoff. Williams, A. B. (1997). On parallel process in social work supervision. Clinical Social Work Journal, 25(4), 425-435. Wright, T. A., & Bonett, D. G. (1997). The contribution of burnout to work performance. Journal of Organizational Behavior, 18, 491-499. Wright, T. A., & Cropanzano, R. (1998). Emotional exhaustion as a predictor of job performance and voluntary turnover. Journal of Applied Psychology, 83(3), 486-493. Yoo, J., & Brooks, D. (2005). The role of organizational variables in predicting service effectiveness: An analysis of a multilevel model. Research on Social Work Practice, 15(4), 267-277. York, R. O., & Denton, R. T. (1990). Leadership behavior and supervisory performance: The view from below. The Clinical Supervisor, 8(1), 93-108. 120 Zellars, K. L., Perrewe, P. L., & Hochwarter, W. A. (2000). Burnout in health care: The role of the five factors of personality. Journal of Applied Social Psychology, 30(8), 1570-1598. Zlotnik, J. L., DePanfilis, D., Daining, C., & Lane, M. (2005). Factors influencing retention of child welfare staff: A systematic review of research. Washington, DC: Institute for the Advancement of Social Work Research. 121 APPENDICES APPENDIX A. Cover Letter for the First Survey UNIVERSITY OF SOUTHERN CALIFORNIA School of Social Work Dear social workers: You may be aware that the National Association of Social Workers (NASW) (2005) reported that social workers are experiencing increases in paperwork, the severity of client problems, and caseload size as well as reduced levels of reimbursement and the availability of social work staff. However, there has been little done with the objective of discovering how these changes contribute to work stress among social workers. The purpose of the enclosed questionnaire is to identify the main sources of work stress among social workers and to try to suggest ways of reducing it. All questions are about your experiences while working for the organization. Your response is important, to help us get meaningful results on which to base recommendations for change. The survey will take 15-20 minutes to complete. The questionnaire form itself carries no identification and all the information from you will be entirely confidential. In addition, no one except our research staff will have access to the information. All the collected information will be stored for five years in a locked file cabinet where only the investigator will have access the information. Data from the questionnaire will be coded and entered into computerized data files. All the computerized files will be stored in password-protected computers that are accessible only to investigators. Information will be analyzed for overall group patterns and not for individual responses. Your participation is completely voluntary. You do not have to answer any questions you do not wish to answer, nor do you have to complete the questionnaire if it causes you discomfort. We would greatly appreciate your willingness to respond to this survey honestly and thoroughly. If you have any questions or concerns about the research, please feel free to contact Hansung Kim, MSW, Co Principal Investigator, at hansungk@usc.edu, or Professor Madeleine R. Stoner, Principal Investigator/Faculty Sponsor, at mstoner@usc.edu. Please return it to us in the enclosed self-addressed envelope as soon as possible. 122 The final report and recommendations will be presented in the scholarly literature of social work. The more responses we get, the more convincing the recommendations will be, so please respond. Yours sincerely, Madeleine R. Stoner, PhD Hansung Kim, MSW Date of Preparation: 9/20/2005 USC UPIRB# UP-05-00040 123 APPENDIX B. Information Sheet for the First Survey University of Southern California School of Social Work INFORMATION SHEET FOR NON-MEDICAL RESEARCH A Study of Burnout among Social Workers Completion and return of the questionnaire will constitute consent to participate in this research project. You are asked to participate in a research study conducted by Professor Madeleine R. Stoner and Hansung Kim, MSW, from the School of Social Work at the University of Southern California. You were selected as a possible participant in this study because you are listed in the registry of social workers, which is provided by the Department of Consumer Affairs (DCA) Public Information Unit in accordance with the Information Practices Act, Civil Code Section 1798.61. A total of 1,000 social workers will be selected from the list to participate. Your participation is voluntary. PURPOSE OF THE STUDY We are trying to learn more about job stress and burnout among social workers. Findings from this survey will help us understand how we can provide clinical social workers with appropriate support to prevent job stress and burnout. PROCEDURES You will be asked to complete a survey questionnaire asking about work-related stress and burnout among social workers. The survey will ask how you think about your working environment and how you experience burnout at work. The survey will take 15-20 minutes to complete. When you complete the questionnaire, you will be asked to mail your completed questionnaire with a provided return-envelope to the investigator. POTENTIAL RISKS AND DISCOMFORTS There are no foreseeable risks to your physical or psychological well-being from being involved in this study. However, you may feel mild discomfort or inconvenience with questions that seek an opinion of your working experience. All information will be kept absolutely confidential. Further, you can refuse to answer any questions that cause you any discomfort. POTENTIAL BENEFITS TO SUBJECTS AND/OR TO SOCIETY 124 You may not directly benefit from participating in this research study. However, for those of you who agree to participate in this research, the information we collect will provide an outlet to share concerns about the job stress and its impact on your experiences of burnout and overall workplace well-being in the field of social work. Also, findings from this study may make valuable contributions to the scholarly literature regarding burnout among social workers. Furthermore, the results will be useful information for policy makers and organizational administrators to develop appropriate interventions and support to prevent job stress and burnout among social workers. The results will be publicly available through the scholarly literature in the field of social work. PAYMENT/COMPENSATION FOR PARTICIPATION You will not be paid for participating this research study. However, we enclosed a USC pen as appreciation. Please, keep this pen even if you refuse to participate in this study. CONFIDENTIALITY The survey is anonymous with no identifier. Other organizational members will not know your responses to the questions because you have to mail your survey directly to the co-principal investigator as soon as you complete it, and co-principal investigator will handle all completed survey questionnaires. All the collected information will be stored in a locked file cabinet. The information will be available only to the investigators. Data from the questionnaire are coded and entered into computerized data files. All the computerized data files are located in password- protected computers that are accessible to study personnel only. Five years from the end of the study all coded and un-coded data will be destroyed and computerized data files will be erased. PARTICIPATION AND WITHDRAWAL You can choose whether to be in this study or not. If you volunteer to be in this study, you may withdraw at any time without consequences of any kind. You may also refuse to answer any questions you don’t want to answer and still remain in the study. The investigator may withdraw you from this research if circumstances arise which warrant doing so. IDENTIFICATION OF INVESTIGATORS If you have any questions or concerns about the research, please feel free to contact Hansung Kim, MSW, Co Principal Investigator at hansungk@usc.edu, or Professor Madeleine R. Stoner, Principal Investigator/Faculty Sponsor, at mstoner@usc.edu. RIGHTS OF RESEARCH SUBJECTS You may withdraw your consent at any time and discontinue participation without penalty. You are not waiving any legal claims, rights or remedies because of your participation in this research study. If you have questions regarding your rights as a research subject, contact the University Park IRB, Office of the Vice Provost for 125 Research, Grace Ford Salvatori Building, Room 306, Los Angeles, CA 90089-1695, (213) 821-5272 or upirb@usc.edu. Date of Preparation: 9/20/2005 USC UPIRB # UP-05-00040 126 APPENDIX C. Survey Questionnaire for the First Survey Study Questionnaire Please clearly mark the box that represents the best answer to each question. Brief Demographic Profile 1. Educational Background: 2. Sex: 3. Ethnic/Racial Background: □ 1 Bachelor’s degree □ 1 Female □ 1 African-American □ 2 Master’s degree □ 2 Male □ 2 Asian-American □ 3 Doctoral degree □ 3 Caucasian □ 4 Other □ 4 Latino □ 5 Other ________ 4. Age: ________ 5. Years in the field of social services: ________ 6. Since you started your first job in the field of social work, how many times have yo u changed your job or agency?: ________ 7. Type of work setting: □ 1 Organizational setting - Years at current organization: ________ □ 2 Private practice - Years at Private practice: ________ 8. If you are working at the organization, indicate the types of the organization (Check all that apply): □ 1 Public organization □ 2 Private organization □ 3 Non-profit organization □ 4 Profit organization 9. Type of work you are currently doing: □ 1 Mental health □ 6 Health □ 2 Adolescents □ 7 School social work □ 3 Child welfare/family □ 8 Other _________ □ 4 Addictions □ 5 Aging 10. How many hours per day do you spend to finish any paperwork on the average? _________ 11. How many cases per week do you have on the average? _________ 12. What is your total annual wage or salary from this job? (To nearest thousand dollars) 127 Listed below are a series of statements that represent possible feelings that individuals might have about their jobs. With respect to your own feelings about your job, please indicate the degree of your agreement or disagreement with each statement by checking one of the following alternatives: 1 2 3 4 5 6 7 Strongly Disagree Disagree quite a lot Disagree just a little I am not sure about this Agree just a little Agree quite a lot Strongly Agree WO1. I have to attend too many meetings in this job. ---------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 WO2. My job involves a lot of paperwork. ----------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 WO3. I have to work very fast to get everything done in my job. ------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 WO4. My workload is too heavy in my job. ---------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 WO5. I do not have enough time to get everything done in my job. ------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 ED1. On my job, I experience emotional burden from working with clients. ----------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 ED2. On my job, I get to make difficult decisions for clients. □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 ED3. My job involves conflicting demands that clients make. □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 UE1. All in all, I am disappointed in this job. ------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 UE2. My experience in this job have been better than I originally expected. ------------------------------------------ □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 VC1. I often experience that my professional values conflict with the job. --------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 UE4. This job has lived up to the expectations I had when I first started. ---------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 UE3 Generally, this job is not what I thought it would be. --- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 OI. I feel responsible for my client’s well being. ------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RC1. I have to do things that should be done differently. ----- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RC2. I receive an assignment without the manpower to complete it.----------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 128 RC3. I have to buck a rule or policy in order to carry out an assignment. ---------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RC4. I work with two or more groups who operate quite differently.----------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RC5. I receive incompatible requests from two or more people. --------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RC6. I do things that are apt to be accepted by one person and not accepted by another. -------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RC7. I receive an assignment without adequate resources and materials to execute it. --------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 JI1. The major satisfaction in my life comes from my job. -- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 JI2. The most important things that happen to me involve my work. ------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 JI3. I’m really a perfectionist about my work. ----------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 JI4. I live, eat, and breathe my job. ----------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 JI5. I am very much involved personally in my work. ------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 JI6. Most things in life are more important than work. ------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 OT1. In the next few months I intend to leave this organization. -------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RA1. I feel certain about how much authority I have. ---------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RA2. I have clear and planned goals and objectives for my job. ------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RA3. I know that I have divided my time properly. ------------ □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RA4. I know what my responsibilities are. ---------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RA5. I know exactly what is expected of me. ------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RA6. Explanation is clear of what has to be done. -------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 OT2. In the next few years I intend to leave this organization. □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 OT3 I occasionally think about leaving this organization. ---- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 OT4 I’d like to work in this organization until I reach retirement age. ------------------------------------------------ □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 129 Think of support you receive from various people while you work for the current agency and please indicate your agreement or disagreement with each statement by checking one of the following four alternatives: 1 2 3 4 Not at all A little Somewhat Very much SS1. How much does each of the following people listen to your work-related problems? a. Your immediate supervisor (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A b. Co-workers (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A c. Top manager/Administrator (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A d. Family/Significant others (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A SS2. How much does each of the following people show you concern towards your job-related problems? a. Your immediate supervisor (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A b. Co-workers (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A c. Top manager/Administrator (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A d. Family/Significant others (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A SS3. How much does each of the following people give you aid in dealing with your work-related problems? a. Your immediate supervisor (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A b. Co-workers (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A c. Top manager/Administrator (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A d. Family/Significant others (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A SS4. How much does each of the following people give you tangible assistance to deal with your work related stress? a. Your immediate supervisor (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A b. Co-workers (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A c. Top manager/Administrator (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A d. Family/Significant others (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A SS5. How much does each of the following people give you sound advice about problems encountered on the job? a. Your immediate supervisor (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A b. Co-workers (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A c. Top manager/Administrator (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A d. Family/Significant others (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A SS6. How much does each of the following people give you useful suggestions in order to get though difficult times? 130 a. Your immediate supervisor (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A b. Co-workers (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A c. Top manager/Administrator (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A d. Family/Significant others (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A The following set of questions concern about your perception of YOUR SUPERVISOR. 1 2 3 4 5 Strongly Disagree Disagree I am not sure about this Agree Strongly Agree CS1. My supervisor gives recognition for good work. ------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS2. My supervisor lets me know why changes are made in work assignments. ------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS3. My supervisor keeps me informed about rules and policies. ------ □ 1 □ 2 □ 3 □ 4 □ 5 CS4. My supervisor gives clear instructions to me. ----------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS5. I question my supervisor’s instructions when I don’t understand them. ---------------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS7. My supervisor jokes good-naturedly with me. ---------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS8. My supervisor asks for my suggestions about how work should be done. ------------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS9. My supervisor lets me know when I’ve done a good job. --------- □ 1 □ 2 □ 3 □ 4 □ 5 CS10. My supervisor tells me the reasons for work schedules. ----------- □ 1 □ 2 □ 3 □ 4 □ 5 CS11. My supervisor informs me about future plans for my work group. --------------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS12. My supervisor sets useful goals for me to meet. -------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS13. I tell my supervisor when I think things are being done wrong. -- □ 1 □ 2 □ 3 □ 4 □ 5 CS15. My supervisor asks me about my interests outside of work. ------ □ 1 □ 2 □ 3 □ 4 □ 5 CS16. My supervisor seeks my input on important decisions. ------------ □ 1 □ 2 □ 3 □ 4 □ 5 CS17. My supervisor praises good work. ------------------------------------ □ 1 □ 2 □ 3 □ 4 □ 5 CS18. My supervisor tells me the reasons for rules and policies. -------- □ 1 □ 2 □ 3 □ 4 □ 5 CS19. My supervisor keeps me informed about what’s happening in the company.–------------------------------------------------------------ □ 1 □ 2 □ 3 □ 4 □ 5 CS20. My supervisor asks versus tells me to do things. ------------------- □ 1 □ 2 □ 3 □ 4 □ 5 131 CS21. I question my supervisor’s instructions when I think they are wrong.--------------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS23. My supervisor strikes up casual conversations with me. ---------- □ 1 □ 2 □ 3 □ 4 □ 5 CS24. My supervisor asks me for suggestions for improvements in my department.--------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 You’ve reached the end of the questionnaire. Please return it to us in the enclosed self-addressed envelop as soon as possible. The more responses we get, the more convincing the recommendations will be, so please respond. We thank you very much for your time and your response! All the best to you in your future career endeavors. Madeleine Stoner and Hansung Kim School of Social Work University of Southern California Los Angeles, California 90089-0411 Tel: 213-740-2711 Date of Preparation: 9/20/2005 USC UPIRB # UP-05-00040 Note. The original survey questionnaire was modified in order to fit to the format of the dissertation. The modified version of survey questionnaire which presented here does not include Maslach’s Burnout Inventory-Human Service Survey (MBI-HSS; Maslach & Jackson, 1986) and scales from the Job Content Questionnaire (JCQ) (Karasek, Brisson, Kawakami, Houtman, Bongers, 1998), as authors of the measures request. 132 APPENDIX D. Cover Letter for the Second Survey UNIVERSITY OF SOUTHERN CALIFORNIA School of Social Work Dear social workers: About 12 months ago, you responded to our first survey, which is a part of a longitudinal study that seeks to identify the long-term effects of work stress and burn-out on professional social workers. As a follow-up, we have enclosed the second (and last) questionnaire. The survey will take approximately 30-40 minutes to complete and we would greatly appreciate your willingness to respond to this survey honestly and thoroughly. Again, your participation is completely voluntary. In accordance to federal law and school policies, all of the information you provide in this survey will remain entirely confidential. The questionnaire form contains study ID codes so that the information from each survey can be matched. No one – except for ourselves and our research staff – will have access to the information. The collected information will be stored for five years in a locked file cabinet where only the investigators will have access. Data from the questionnaire will be coded and entered into computerized data files. All the computerized files will be stored in password-protected computers that are accessible only to investigators. Finally, information will be analyzed for overall group patterns and not for individual responses. Thank you again for your participation. Once completed, please return it to us in the enclosed self-addressed envelope as soon as possible. If you have any questions or concerns about the research, please feel free to contact Hansung Kim, MSW, Co Principal Investigator, at hansungk@usc.edu, or Professor Madeleine R. Stoner, Principal Investigator/Faculty Sponsor, at mstoner@usc.edu. Sincerely, Madeleine R. Stoner, PhD Hansung Kim, MSW Date of Preparation: 6/21/2006 USC UPIRB# UP-05-00040-AMD001 133 APPENDIX E. Information Sheet for the Second Survey University of Southern California School of Social Work INFORMATION SHEET FOR NON-MEDICAL RESEARCH A Longitudinal Study of Burnout among Social Workers: A Follow-Up Survey Completion and return of the questionnaire will constitute consent to participate in this research project. You are received this follow-up survey because you participated in a survey approximately 12 months ago. Your participation is completely voluntary. You do not have to participate in this follow-up survey. PURPOSE OF THE STUDY We are trying to learn more about long-term effects of job stress and burnout on professional social workers. From the results, we hope to make recommendations to address what appears to be serious issues for many social workers. PROCEDURES You will be asked to complete a survey questionnaire asking about work-related stress and burnout among social workers. The survey will take about 30-40 minutes to complete. When you complete the questionnaire, you will be asked to mail your completed questionnaire with a provided return-envelope to the investigator. The questionnaire contains a study ID code for the purpose of matching information between the first survey and this follow-up survey. In addition, all the information you provide on both surveys will remain confidential. POTENTIAL RISKS AND DISCOMFORTS There are no foreseeable risks to your physical or psychological well-being with involvement in this study. However, you may feel mild discomfort or inconvenience with questions that request information regarding your opinion of your working experience. Further, you can refuse to answer any questions that may cause you any discomfort. POTENTIAL BENEFITS TO SUBJECTS AND/OR TO SOCIETY You may not directly benefit from participating in this research study. However, for those of you who agree to participate in this research, the information we collect may provide an outlet to share concerns about the job stress and its impact on your 134 experiences of burnout and overall workplace well-being in the field of social work. Also, findings from this study may make valuable contributions to the scholarly literature regarding burnout and turnover among social workers. Furthermore, the results may be useful information for policy makers and organizational administrators and may help to inform the development of appropriate interventions and support to prevent job stress and burnout among social workers. The results will be publicly available through the scholarly literature and conferences in the field of social work. PAYMENT/COMPENSATION FOR PARTICIPATION You will not be paid for participating in this research study. However, we enclosed a key holder as appreciation. Please, keep this key holder even if you refuse to participate in this study. CONFIDENTIALITY The questionnaire contains study ID codes, which were assigned to each participant in order to match information between the initial and follow-up survey. Only the researchers will know how study ID codes are assigned. All of the collected information will be stored in a locked file cabinet and will be available only to the investigators. Data from the questionnaire are coded and entered into computerized data files in password-protected computers that are accessible to study personnel only. After five years from the end of the study, all the survey will be destroyed and computerized data files will be erased. PARTICIPATION AND WITHDRAWAL Your participation is this study is completely voluntary. If you choose to be in this study, you may withdraw at any time without any consequences. You may also refuse to answer any questions you do not want to answer and still remain in the study. The investigator may withdraw you from this research if circumstances arise which warrant doing so. IDENTIFICATION OF INVESTIGATORS If you have any questions or concerns about the research, please feel free to contact Hansung Kim, MSW, Co Principal Investigator at hansungk@usc.edu, or Professor Madeleine R. Stoner, Principal Investigator/Faculty Sponsor, at mstoner@usc.edu. RIGHTS OF RESEARCH SUBJECTS You may withdraw your consent at any time and discontinue participation without penalty. You are not waiving any legal claims, rights or remedies because of your participation in this research study. If you have questions regarding your rights as a research subject, contact the University Park IRB, Office of the Vice Provost for Research, Grace Ford Salvatori Hall, Room 306, Los Angeles, CA 90089-1695, (213) 821-5272 or upirb@usc.edu. Date of Preparation: 6/21/2006 135 USC UPIRB # UP-05-00040-AMD 001 136 APPENDIX F. Survey Questionnaire for the Second Survey Study Questionnaire Listed below are a series of statements that represent possible feelings that individuals might have about their jobs. With respect to your own feelings about your job, please indicate the degree of your agreement or disagreement with each statement by checking one of the following alternatives: 1 2 3 4 5 6 7 Strongly Disagree Disagree quite a lot Disagree just a little I am not sure about this Agree just a little Agree quite a lot Strongly Agree WO1. I have to attend too many meetings in this job. ---------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 WO2. My job involves a lot of paperwork. ----------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 WO3. I have to work very fast to get everything done in my job. ------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 WO4. My workload is too heavy in my job. ---------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 WO5. I do not have enough time to get everything done in my job. ------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 ED1. On my job, I experience emotional burden from working with clients. ----------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 ED2. On my job, I get to make difficult decisions for clients. □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 ED3. My job involves conflicting demands that clients make. □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 UE1. All in all, I am disappointed in this job. ------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 UE2. My experience in this job have been better than I originally expected. ------------------------------------------ □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 VC1. I often experience that my professional values conflict with the job. --------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 UE4. This job has lived up to the expectations I had when I first started. ---------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 UE3 Generally, this job is not what I thought it would be. --- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 OI. I feel responsible for my client’s well being. ------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 137 RC1. I have to do things that should be done differently. ----- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RC2. I receive an assignment without the manpower to complete it.----------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RC3. I have to buck a rule or policy in order to carry out an assignment. ---------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RC4. I work with two or more groups who operate quite differently.----------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RC5. I receive incompatible requests from two or more people. --------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RC6. I do things that are apt to be accepted by one person and not accepted by another. -------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RC7. I receive an assignment without adequate resources and materials to execute it. --------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 JI1. The major satisfaction in my life comes from my job. -- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 JI2. The most important things that happen to me involve my work. ------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 JI3. I’m really a perfectionist about my work. ----------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 JI4. I live, eat, and breathe my job. ----------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 JI5. I am very much involved personally in my work. ------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 JI6. Most things in life are more important than work. ------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 OT1. In the next few months I intend to leave this organization. -------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RA1. I feel certain about how much authority I have. ---------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RA2. I have clear and planned goals and objectives for my job. ------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RA3. I know that I have divided my time properly. ------------ □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RA4. I know what my responsibilities are. ---------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RA5. I know exactly what is expected of me. ------------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 RA6. Explanation is clear of what has to be done. -------------- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 OT2. In the next few years I intend to leave this organization. □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 OT3 I occasionally think about leaving this organization. ---- □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 138 OT4 I’d like to work in this organization until I reach retirement age. ------------------------------------------------ □ 1 □ 2 □ 3 □ 4 □ 5 □ 6 □ 7 Think of support you receive from various people while you work for the current agency and please indicate your agreement or disagreement with each statement by checking one of the following four alternatives: 1 2 3 4 Not at all A little Somewhat Very much SS1. How much does each of the following people listen to your work-related problems? a. Your immediate supervisor (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A b. Co-workers (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A c. Top manager/Administrator (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A d. Family/Significant others (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A SS2. How much does each of the following people show you concern towards your job-related problems? a. Your immediate supervisor (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A b. Co-workers (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A c. Top manager/Administrator (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A d. Family/Significant others (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A SS3. How much does each of the following people give you aid in dealing with your work-related problems? a. Your immediate supervisor (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A b. Co-workers (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A c. Top manager/Administrator (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A d. Family/Significant others (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A SS4. How much does each of the following people give you tangible assistance to deal with your work related stress? a. Your immediate supervisor (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A b. Co-workers (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A c. Top manager/Administrator (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A d. Family/Significant others (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A SS5. How much does each of the following people give you sound advice about problems encountered on the job? a. Your immediate supervisor (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A b. Co-workers (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A c. Top manager/Administrator (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A 139 d. Family/Significant others (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A SS6. How much does each of the following people give you useful suggestions in order to get though difficult times? a. Your immediate supervisor (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A b. Co-workers (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A c. Top manager/Administrator (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A d. Family/Significant others (Not at all) □ 1 □ 2 □ 3 □ 4 (Very much) □ N/A The following set of questions concern about your perception of YOUR SUPERVISOR. 1 2 3 4 5 Strongly Disagree Disagree I am not sure about this Agree Strongly Agree CS1. My supervisor gives recognition for good work. ------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS2. My supervisor lets me know why changes are made in work assignments. ------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS3. My supervisor keeps me informed about rules and policies. ------ □ 1 □ 2 □ 3 □ 4 □ 5 CS4. My supervisor gives clear instructions to me. ----------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS5. I question my supervisor’s instructions when I don’t understand them. ---------------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS7. My supervisor jokes good-naturedly with me. ---------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS8. My supervisor asks for my suggestions about how work should be done. ------------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS9. My supervisor lets me know when I’ve done a good job. --------- □ 1 □ 2 □ 3 □ 4 □ 5 CS10. My supervisor tells me the reasons for work schedules. ----------- □ 1 □ 2 □ 3 □ 4 □ 5 CS11. My supervisor informs me about future plans for my work group. --------------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS12. My supervisor sets useful goals for me to meet. -------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS13. I tell my supervisor when I think things are being done wrong. -- □ 1 □ 2 □ 3 □ 4 □ 5 CS15. My supervisor asks me about my interests outside of work. ------ □ 1 □ 2 □ 3 □ 4 □ 5 CS16. My supervisor seeks my input on important decisions. ------------ □ 1 □ 2 □ 3 □ 4 □ 5 CS17. My supervisor praises good work. ------------------------------------ □ 1 □ 2 □ 3 □ 4 □ 5 CS18. My supervisor tells me the reasons for rules and policies. -------- □ 1 □ 2 □ 3 □ 4 □ 5 140 CS19. My supervisor keeps me informed about what’s happening in the company.–------------------------------------------------------------ □ 1 □ 2 □ 3 □ 4 □ 5 CS20. My supervisor asks versus tells me to do things. ------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS21. I question my supervisor’s instructions when I think they are wrong.--------------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 CS23. My supervisor strikes up casual conversations with me. ---------- □ 1 □ 2 □ 3 □ 4 □ 5 CS24. My supervisor asks me for suggestions for improvements in my department.--------------------------------------------------------------- □ 1 □ 2 □ 3 □ 4 □ 5 1. In general, would you say your health is □ 1 Excellent □ 2 Good □ 3 Fair □ 4 Poor □ 5 Very Poor 2. Since you received the first questionnaire for this study, have you exp erienced any changes in your job : (Please check all that apply) □ 1 I have become a supervisor. □ 2 My supervisor has been changed. □ 3 I have become a licensed social worker. □ 4 Organizational policy or service procedure has been significantly changed. □ 5 I have moved to a different organization. 3. You are: □ 1 Top Management/administration □ 2 Program director □ 3 Field supervisor □ 4 Direct-service worker You’ve reached the end of the questionnaire. Please return it to us in the enclosed self-addressed envelop as soon as possible. The more responses we get, the more convincing the recommendations will be, so please respond. We thank you very much for your time and your response! All the best to you in your future career endeavors. 141 Madeleine Stoner and Hansung Kim School of Social Work University of Southern California Los Angeles, California 90089-0411 Tel: 213-740-2711 Date of Preparation: 6/21/06 USC UPIRB # UP-05-00040-AMD001 Note. The original survey questionnaire was modified in order to fit to the format of the dissertation. The modified version of survey questionnaire which presented here does not include Maslach’s Burnout Inventory-Human Service Survey (MBI-HSS; Maslach & Jackson, 1986), scales from the Job Content Questionnaire (JCQ) (Karasek, Brisson, Kawakami, Houtman, Bongers, 1998), and some standardized measures, as authors of the measures request.

Abstract (if available)

Abstract The purpose of present study is to understand which job conditions is critical for social workers not only to prevent burnout, but also to cope with burnout experience. My central thesis is that different job resources can have distinctive functions such as 1) edifying roles and objectives of social workers, 2) empowering social workers to cope with job demands, and 3) embedding social workers in their organizations, depending on worker characteristics. In order to provide empirical evidence for these proposed functions of job resources, the current study tests longitudinal relationships among job demands, burnout, and turnover intention

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Examining the impact of job burnout on the well-being of human service workers

PDF

Transformational leadership theory aligned-practices and social workers' well-being: an exploratory study of leadership practices in the context of stress and job burnout

PDF

Parenting practices among non-offending mothers of sexually abused girls and its impact on the abused girls' behavioral adjustment: perspectives from a multigenerational, longitudinal study

PDF

Understanding employees' efforts to seek change or quit: a longitudinal study of human and social service workers' cognitive and behavioral responses to their work experiences

PDF

SafeGuard: enhancing psychological safety for child protection supervisors and workers in New Jersey

PDF

The influence of organizational ethics on job attitudes and government performance

PDF

SafeGuard: enhancing psychological safety for child protection supervisors and workers in New Jersey

PDF

Child welfare turnover: an evaluation study of work-life balance practices

PDF

The impact of high turnover and burnout among behavioral health's clinical workforce (clinicians)

PDF

The impact of child maltreatment on the mental health of adolescents: a longitudinal study of social anxiety and self-perception

PDF

The role of professional development and certification in technology worker turnover: An evaluation study

PDF

Longitudinal relationships of cognitive deficits, symptoms, and social functioning outcomes in community-based psychosocial rehabilitation programs: mechanisms of longitudinal change

PDF

Racial/ethnic diversity and disparities in the child welfare system: a multilevel perspective

PDF

Sociocultural influences on mental health functioning: implications for the design of community-based services

PDF

Adjunct life in a full time world: evaluation of worklife experiences and risk for burnout in social work field faculty

PDF

Physician burnout in the COVID-19 pandemic: Healthcare organization and leadership implications for patient care and clinical team leadership with nurses

PDF

Exploring the development of psychological empowerment among survivors of intimate partner violence: does the Personal Empowerment Program live up to its name?

PDF

Stop the revolving door: the influence of emotionally intelligent leadership practices on employee retention in non‐profit human service organizations

PDF

Compassion fatigue among social work students: a gap analysis

PDF

A case study of an evidence-based approach to resource allocation at a school for at-risk and incarcerated students in Hawaii

Asset Metadata

Creator Kim, Hansung (author)

Core Title A longitudinal study of job resources, burnout, and turnover among social workers: implications for evidence-based organizational practices

School School of Social Work

Degree Doctor of Philosophy

Degree Program Social Work

Publication Date 04/15/2008

Defense Date 03/03/2008

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Burnout,job resources,OAI-PMH Harvest,organizational practices,social worker,turnover

Place Name California (states), USA (countries)

Language English

Advisor Stoner, Madeleine (committee chair), Nishimoto, Robert (committee member), Robertson, Peter John (committee member)

Creator Email hansungk@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m1111

Unique identifier UC1221244

Identifier etd-Kim-20080415 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-54573 (legacy record id),usctheses-m1111 (legacy record id)

Legacy Identifier etd-Kim-20080415.pdf

Dmrecord 54573

Document Type Dissertation

Rights Kim, Hansung

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu