THE WELL-BEING OF ADULTS WHO VOLUNTEER WITH CHILDREN AT RISK OF CHILD MALTREATMENT By Joshua Daniel Bishop A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Social Work—Doctor of Philosophy 2019 ABSTRACT THE WELL-BEING OF ADULTS WHO VOLUNTEER WITH CHILDREN AT RISK OF CHILDHOOD MALTREATMENT By Joshua Daniel Bishop The lack of sufficient foster care homes, their inconsistent quality, and their risk of increasing negative outcomes for children highlight the need for more people to be involved in roles that support children at risk of maltreatment and foster care. While volunteer opportunities exist for supporting children after foster care placements, few opportunities exist to care for children who are at risk of maltreatment and foster care. Innovative approaches are being developed to provide such opportunities. These approaches may find support from an emerging literature that has found a positive relationship between volunteerism and well-being. However, no studies have investigated the well-being of those who volunteer with children at risk of maltreatment and/or child welfare involvement. This dissertation, which is an exploratory, cross-sectional, quantitative study, will address this gap with a sample of volunteers (N = 302) from Safe Families for Children (SFFC), a faith-based organization that works to keep children safe during family crises, prevent child maltreatment, and reduce the number of children entering the child welfare system. The aim of the dissertation is to investigate whether volunteering and/or motivation are associated with seven dimensions of well-being: Happiness, Physical Health, Life Satisfaction, Self-Mastery, Self-Esteem, Anxiety, and Depression. Results demonstrate limited evidence of significant relationships between volunteering and well-being dimensions. There is also limited evidence of significant relationships between motivation and well-being. However, an important finding of this study is that despite the high time and emotional demands of doing this type of volunteer work, there is no apparent decrease or drop-off in the well-being of the volunteers. Rather, they are happy and physically healthy. They report very low levels of anxiety and depression, and they demonstrate a high degree of Self-Esteem, Self-Mastery, and Life Satisfaction. While some may believe that working with children at risk of maltreatment is stressful and may result in a decrease in well-being (Tyebjee, 2003), the results of this study suggest that it is not the case for Host Families from Safe Families for Children. The Confirmatory Factor Analysis used in this study is a unique contribution to the literature. It robustly demonstrates very reliable methods for operationalizing and measuring seven well-being dimensions as well as four dimensions of motivation. The analyses and results in this study go beyond typically used measurements of reliability and offer strong evidence for reliably measuring well-being in future studies. The most important limitation in this study is the lack of control or comparison group that would allow for investigating the difference in well-being among SFFC volunteers and those who are not SFFC volunteers. This study offers reliable options for future studies to operationalize well-being and motivation in a way that encourages accurate comparison between studies. Future studies should consider whether using measurement scales that can detect small changes in well-being among populations that may have a high level of well-being are important. Implications for practice include recommendations for volunteer managers to focus on volunteer efficiency, the importance of social support, and motivation. Copyright by JOSHUA DANIEL BISHOP 2019 ACKNOWLEDGEMENTS I would like to acknowledge and thank the Host Families who took the time to participate in this study. Host Families: Thank you for your dedication to children and families and for being willing to share your experience and voice in this study. I would also like to thank my dissertation committee chair, Dr. Gary Anderson, as well as the other members of my committee whose support and feedback were instrumental in completing this project and writing this dissertation. Dr. Anderson’s feedback during the final stages of the writing process were helpful, valued, and appreciated. Finally, I am thankful to my wife and children for their support, patience, and love while I worked on this dissertation. v TABLE OF CONTENTS LIST OF TABLES ....................................................................................................................... viii LIST OF FIGURES ....................................................................................................................... ix INTRODUCTION .......................................................................................................................... 1 HISTORICAL AND THEORETICAL FOUNDATIONS ......................................................... 4 Altruism. ................................................................................................................................. 4 Conceptual frameworks for volunteerism and well-being. ..................................................... 8 Pathological altruism. ............................................................................................................. 9 Social work values and ethics. .............................................................................................. 10 LITERATURE REVIEW ............................................................................................................. 12 SEARCH PROCESS ................................................................................................................ 12 LITERATURE REVIEW OVERVIEW ................................................................................... 13 MAIN LITERATURE REVIEW FINDINGS .......................................................................... 15 DATA SOURCE AND SAMPLES .......................................................................................... 20 METHODS AND DESIGNS IN THE LITERATURE ............................................................ 22 CONCEPTUALIZATION AND OPERATIONALIZATION OF VOLUNTEERISM IN THE LITERATURE .......................................................................................................................... 22 OPERATIONALIZATION OF WELL-BEING IN THE LITERATURE ............................... 24 CONTROLS AND COVARIATES IN THE LITERATURE .................................................. 26 CRITICAL ANALYSIS OF THE LITERATURE ................................................................... 27 Operationalization in the Literature. ..................................................................................... 27 External Validity. .................................................................................................................. 31 Causation. .............................................................................................................................. 32 Epistemology. ....................................................................................................................... 38 RESEARCH GAPS .................................................................................................................. 39 RESEARCH QUESTIONS ...................................................................................................... 42 METHOD ..................................................................................................................................... 43 PROGRAM DESCRIPTION .................................................................................................... 43 Assumptions. ......................................................................................................................... 47 Short-term Outcomes. ........................................................................................................... 47 Long-term Outcomes. ........................................................................................................... 48 SAMPLE ................................................................................................................................... 48 OPERATIONALIZATION AND MEASURES ...................................................................... 49 Demographics and Program Information .............................................................................. 49 Dependent Variable: Well-Being .......................................................................................... 49 Independent Variable: Volunteerism .................................................................................... 53 Covariate: Motivation ........................................................................................................... 54 Covariates ............................................................................................................................. 55 vi PROCEDURES ........................................................................................................................ 55 ANALYSIS PLAN ................................................................................................................... 56 RESULTS ..................................................................................................................................... 59 RELIABILITY ANALYSIS ..................................................................................................... 59 RESEARCH QUESTION 1 ...................................................................................................... 62 Demographics. ...................................................................................................................... 63 Other Generosity. .................................................................................................................. 69 Volunteerism. ........................................................................................................................ 70 Well-being. ............................................................................................................................ 73 RESEARCH QUESTION 2 ...................................................................................................... 80 RESEARCH QUESTION 3 ...................................................................................................... 90 RESEARCH QUESTION 4 ...................................................................................................... 96 DISCUSSION ............................................................................................................................... 98 WELL-BEING .......................................................................................................................... 98 VOLUNTEERISM ................................................................................................................... 99 VOLUNTEERING AND WELL-BEING .............................................................................. 101 MOTIVATION ....................................................................................................................... 103 MOTIVATION AND WELL-BEING .................................................................................... 104 LIMITATIONS ....................................................................................................................... 104 IMPLICATIONS AND CONCLUSION .................................................................................... 108 IMPLICATIONS FOR FUTURE RESEARCH ..................................................................... 108 IMPLICATIONS FOR PRACTICE ....................................................................................... 112 IMPLICATIONS FOR POLICY ............................................................................................ 114 IMPLICATIONS FOR EDUCATION ................................................................................... 115 CONCLUSION ....................................................................................................................... 116 APPENDICES ............................................................................................................................ 120 APPENDIX A: SURVEY PROTOCOL ................................................................................. 121 APPENDIX B: INVITATION EMAIL .................................................................................. 127 APPENDIX C: INFORMED CONSENT ............................................................................... 128 APPENDIX D: ADDITIONAL BIVARIATE ANALYSIS ................................................... 129 APPENDIX E: MULTIPLE REGRESSION ANALYSES WITH FACTOR SCORES ....... 130 APPENDIX F: CURVILINEAR ANALYSES ...................................................................... 138 REFERENCES ........................................................................................................................... 142 vii LIST OF TABLES Table 1. Definitions and Operationalization of Key Terms ............................................................ 4 Table 2. Summary of Literature Review ....................................................................................... 16 Table 3. Frequency of Control Variables in the Literature ........................................................... 27 Table 4. Standardized Measures ................................................................................................... 50 Table 5. Chronbach’s alpha. ......................................................................................................... 59 Table 6. Confirmatory Factor Analysis for Well-Being and Motivation Scales .......................... 61 Table 7. Characteristics and Background ..................................................................................... 63 Table 8. Volunteerism and Other Generosity ............................................................................... 70 Table 9. Well-Being Scores .......................................................................................................... 73 Table 10. Tests of Normality for Well-Being Variables ............................................................... 80 Table 11. Difference in Well-Being Means for Placement in Past Year ...................................... 81 Table 12. Bivariate Analysis – Spearman’s rho ............................................................................ 82 Table 13. Bivariate Analysis Covariates: Spearman's rho ............................................................ 83 Table 14. Regression Models with All Covariates Included ........................................................ 85 Table 15. Motivation ..................................................................................................................... 91 Table 16. Motivation Bivariate: Spearman's rho .......................................................................... 96 Table 17. Bivariate Analysis IV and DV - Pearson Correlation ................................................. 129 Table 18. Bivariate Analysis Control Variables and Well-being - Pearson Correlation ............ 129 Table 19. Regression Analyses with Factor Scores .................................................................... 130 viii LIST OF FIGURES Figure 1. Age Histograms ............................................................................................................. 64 Figure 2. Education Histogram ..................................................................................................... 65 Figure 3. Income Histogram ......................................................................................................... 66 Figure 4. Caregiving Burden Histogram ....................................................................................... 68 Figure 5. Social Support Histogram .............................................................................................. 69 Figure 6. Money Donated in Past 12 Months ............................................................................... 71 Figure 7. Hostings in the Past 12 Months Histogram ................................................................... 71 Figure 8. Hosting Days in the Past 12 Months Histogram ........................................................... 72 Figure 9. Years Served as a Host Family Histogram .................................................................... 73 Figure 10. Life Satisfaction Histogram ......................................................................................... 74 Figure 11. Anxiety Score Histogram ............................................................................................ 75 Figure 12. Depression Score Histogram ....................................................................................... 76 Figure 13. Physical Health Index Histogram ................................................................................ 77 Figure 14. Self-Mastery Histogram .............................................................................................. 78 Figure 15. Self-Esteem Histogram ................................................................................................ 78 Figure 16. General Happiness Histogram ..................................................................................... 79 Figure 17. Physical Motivation Histogram ................................................................................... 92 Figure 18. Personal Motivation Histogram ................................................................................... 93 Figure 19. Social Motivation Histogram ...................................................................................... 94 Figure 20. Cultural Motivation Histogram ................................................................................... 95 Figure 21. Happiness and Intensity ............................................................................................. 138 ix Figure 22. Self-Mastery and Intensity ........................................................................................ 138 Figure 23. Self-Esteem and Intensity .......................................................................................... 139 Figure 24. Health and Intensity ................................................................................................... 139 Figure 25. Life Satisfaction and Intensity ................................................................................... 140 Figure 26. Depression and Intensity ........................................................................................... 140 Figure 27. Anxiety and Intensity ................................................................................................ 141 x INTRODUCTION Although substantial work has been done to prevent child maltreatment and neglect, it continues to permeate cultures, socioeconomic statuses, genders, races, and ethnicities, leaving many children in or at risk of entering the child welfare system. A primary response to this problem has been foster care. In the United States (U.S.), child protective services receive referrals for more than seven million children each year and more than a quarter million children are placed in foster care (U.S. DHHS, 2017). Nearly a half million children in the U.S. are in a foster care placement at any given time, where the typical child spends more than a year of her/his life (M = 20.4 months; Mdn = 12.6 months) (Child Welfare Information Gateway, 2016b). Involvement with the child welfare system can be an extremely difficult experience for a child and can sometimes be an intervention that adds additional risk to a child’s life. Unfortunately, children placed in foster care face a lifetime of struggles, including outcomes such as decreased academic achievement and performance, lower earning potential, and a lack of social support (Perry, 2006; Salazar, 2013). The vast majority of children in foster care have been abused, neglected, experienced a traumatic loss or bereavement, or some other form of trauma (Burns et al., 2004; Greeson et al., 2011; Racusin, Maerlender, Sengupta, Isquith, & Straus, 2005). Additionally, foster children have the experience of being separated from their family or a caregiver, which can also be traumatic (Folman, 1998; Racusin et al., 2005). Experiencing a traumatic event has been linked with problems ranging from emotional dysregulation, increased risk of mental health diagnosis, behavioral problems, health problems, cognitive impairment, decreased quality of life, and premature death (Cohen, Deblinger, & Mannarino, 2006; Cook et al., 2005; Felitti et al., 1998; Greeson et al., 2011). 1 Challenges in the quantity and quality of foster homes make the situation more complicated. Supply of foster care homes does not meet the demand and many foster homes are minimally active. One large study found that 20% of foster homes in the U.S. are providing 60- 72% of all placements (Gibbs, 2005). Data from the same study also highlighted that the median length of service provided by foster homes is 8-14 months, suggesting that the careers of many foster parents are shorter than the amount of time many children spend in foster care. In addition, foster parenting is considered unfavorable or impossible for many Americans, likely because the children are seen as a burden (Tyebjee, 2003). The lack of sufficient foster care homes, the questionable quality of the placements, and the detrimental effects foster care can have on children highlight the need for more opportunities for people to support children who are at risk of maltreatment and entering foster care. While recent legislation, the Families First Preservation Services Act (Buchanan, 2017), has provided funding for evidence-based prevention services, there are still very few volunteer roles that a concerned citizen can play to care for children at risk of child maltreatment. There are some opportunities for concerned citizens to engage in supportive roles after a child is placed into foster care (e.g., mentoring programs, Court Appointed Special Advocates, faith-based programs), however, there are very few ways for individuals to support children who are at risk of entering the child welfare system. Innovative approaches are being developed that allow citizens to voluntarily show compassion and generosity toward these vulnerable children, however, these innovative approaches often struggle with funding and recruitment and retention of volunteers (Nolan, 2015). 2 The emerging Science of Generosity may offer support for these innovative approaches by calling attention to the benefits of being generous. A growing literature has shown that generosity is related to greater well-being and quality of life, including increased happiness and improved physical and mental health (Anderson et al., 2014; Herzog & Price, 2016; Morrow- Howell, Hinterlong, Rozario, & Tang, 2003; Musick, Herzog, & House, 1999; Nelson, Layous, Cole, & Lyubomirsky, 2016; Piliavin & Siegl, 2007; Post, 2005, 2007; Smith & Davidson, 2014; van Campen, de Boer, & Iedema, 2013). Generosity is “the virtue of giving good things to others freely and abundantly” (Smith & Davidson, 2014, Introduction, paragraph 12). For the purposes of this study, generosity is defined only as a behavior, with no consideration given to altruistic intention (See “Historical and Theoretical Foundations” for further discussion). The primary focus of this study is on volunteering, which is one of the most common forms of generosity. Herzog and Price (2016) identified nine categories of generosity from their nationally representative study on generosity. The most common forms are Financial Giving, Volunteerism, and Political Action. The remaining six are Blood Donation, Organ Donation, Estate Giving, Environmentally Sustainable Consumption, Possession Lending, and Relational Giving to friends and family. Volunteerism, defined as “donating time or services to charitable causes” (Herzog & Price, 2016, chapter 1, paragraph 2), is the primary independent variable of interest in this study. This dissertation will explore the concept of generosity and its relation to well-being from a social work perspective by examining theory, history, current research, and research gaps. The body of research that will be critically analyzed will be narrowed to volunteerism, as other forms of generosity (donating money, giving blood, political involvement) are not directly applicable to the concept of providing direct care to children at risk of maltreatment. Following the literature 3 review, research questions and methods for an exploratory study are described. Finally, results of the study are thoroughly reported and discussed. Table 1. Definitions and Operationalization of Key Terms (Centers for Disease Control, 2018) via time, talents, or services Term Concept Definition Volunteerism Contributing to a charitable cause Well-being People’s judgments and feelings regarding whether their health and lives are desirable, rewarding, and satisfying Children at risk of maltreatment Children for whom individual, familial, socioeconomic, or community characteristics increase the probability of child maltreatment and predict child welfare involvement Operationalization The volunteers in this study are people filling the roll of Host Families for Safe Families for Children. Volunteerism is measured as intensity and consistency. Well-being is operationalized in this study as self-reported Physical Health (Physical Health Index), Self- Esteem (Rosenberg Self-Esteem Scale), Self-Mastery, (Pearlin Mastery Scale), Depression (Patient Health Questionnaire-9), and Anxiety (General Anxiety Disorder- 7), Life Satisfaction (Personal Wellbeing Index), and Happiness. For the purpose of this study, “children at risk of maltreatment” is defined as children who have been placed by SFFC. These children are experiencing or are at risk of experiencing some form of neglect or maltreatment, and are therefore considered to be at risk of involvement with the child welfare system and of being removed from their home and placed in foster care. HISTORICAL AND THEORETICAL FOUNDATIONS Altruism. The concept of generosity is rooted in a spiritual, philosophical, and psychological discussion about altruism that is both ancient and modern. Auguste Comte, the nineteenth century French philosopher, is credited with coining the word altruism (Feigin, 4 Owens, & Goodyear-Smith, 2014). It is derived from the Latin “alteri” which means “the others,” and has a variety of definitions that can be generally summarized as behavior that benefits others. The concept, of course, predates Comte and is a primary moral teaching of many of history’s most revered religious and spiritual teachers. • Moses wrote, “Love your neighbor as yourself” (Leviticus 19:18) • Solomon wrote, “A generous man will prosper, he who refreshes others will himself be refreshed” (Proverbs 11:25) • Muhammad taught, “None of you [truly] believes until he loves for his brother that which he loves for himself” (Hadith 13). • Krishna said, “A gift is pure when it is given from the heart to the right person at the right time and at the right place, and when we expect nothing in return. But when it is given expecting something in return, or for the sake of a future reward or a specific type of sentiment in return, the gift is of Rajas (impurity). And a gift given to the wrong person, at the wrong time and the wrong place, or a gift which comes not from the heart, and is given with proud contempt, is a gift of darkness” (Bhagavad Gita 17:20-22) • Jesus taught, “It is more blessed to give than to receive” (Acts 20:35). • St. Paul wrote, “Be devoted to one another in love. Honor one another above yourselves” (Romans 12:10) • The Buddha taught, “One should seek for others the happiness one desires for himself." • Mahatma Gandhi is credited with saying, ‘‘The best way to find yourself is to lose yourself in the service of others.’’ 5 • A majority of ancient and modern religions, including Confucianism, Christianity, Buddhism, Islam, Judaism, Taoism, Native American religions, Hinduism, African tribal religions, Sikhism, and Zoroastrianism, invoke some version of the Golden Rule: treat others how you would want to be treated. Altruism has also been a frequent topic among philosophers and psychologists, with all falling into one of two categories: altruism and egoism (Bar-Tal, 1986; Feigin et al., 2014). Altruists describe humans with the ability to engage in self-sacrificial, other-oriented behavior without expectation of reward, while egoists describe humans as solely self-serving, so that even behavior that appears to be for the benefit of others is truly motivated by concern for the self (Feigin et al., 2014; Kitzrow, 1998). The egoists include Freud (and the major psychoanalytic psychologists) as well as Darwin and many evolutionary biologists (Bragg, n.d.; Kitzrow, 1998; Oord, 2008). While Freud and Darwin are very different, they both explain human altruism as a means to benefit the self. Freudian psychology explains all behavior as a form of instinctual gratification while Darwin theorizes that even seemingly altruistic behaviors exist to promote the survival of the self (Bragg, n.d.; Kitzrow, 1998). Evolutionary biologists have been challenged to explain the human capacity for altruism, however. It has most often been described as kin or group selection: the hypothesis that caring for others enhances the survival of the group or species because groups whose members are the most caring toward one another are more likely to survive and reproduce (Post, 2005). Another explanation has been the concept of generativity in which humans have adapted to care for their young and the next generation in order to promote the survival of the species (Erikson, 1964; Post, 2005; Wink & Dillon, 2007). The result of the egoist perspective has been the phrase homo economicus, which refers to the idea that if a human 6 cares only about her or his own pleasure, s/he will have a higher quality of life, and if being generous elicits positive emotions or pleasurable sensations, generous behavior will ultimately be reinforced (Konow & Earley, 2008; Meier & Stutzer, 2008). A subset of egoism is a neutral group that includes both the behaviorists and those who have theorized about developmental stages. This group essentially sees humans as “blank slates” that must learn all behaviors. Included in this group are Locke, Watson, Skinner, Piaget, Kohlberg, Gilligan, Bandura, and Erickson. None of these thinkers focus primarily on the concept of altruism, but in some fashion, they theorize that our altruistic behaviors are learned and developed over the course of our lifetime. They do not see humans as fundamentally altruistic or egotistic (Kitzrow, 1998). This neutral group is considered a subset of egoism because it does not meet the strict qualifications of being purely altruistic (self-sacrificial with no expectation of reward). Those who promote the theory that people are intrinsically good or altruistic are the humanists, such as Hume, Rogers, and Maslow (Bragg, n.d.; Kitzrow, 1998; Koltko-Rivera, 2006). Those in this category see altruism as the natural state of humans and believe that humans are capable of pure altruism. They emphasize the human capacity for empathy and believe that compassion and generosity are not just virtues to be practiced (Kitzrow, 1998). There are some who are unconcerned with this debate over pure altruism because they disregard motivation and focus solely on the generous behavior, believing that there is always some sort of reward or benefit to the giver (Bar-Tal, 1986; Smith & Davidson, 2014). This is the primary perspective that the studies in the literature review in the next section take, as most do not try to evaluate or measure motivation (except as a covariate), and all are interested in the benefits experienced by the giver. They define generosity or prosocial behavior as the “virtue of 7 giving good things to others freely and abundantly” (Smith & Davidson, 2014), or “any act with the goal of benefitting another person” (Nelson et al., 2016, p. 851) Conceptual frameworks for volunteerism and well-being. As will be discussed below, the majority of authors who have studied the relationship between volunteerism and well-being agree that the relationship is at least reciprocal and likely causal (volunteering causes increases in well-being). There is no consensus in the literature on a single framework for explaining this relationship, but the most prominent will be highlighted here. Role Identity theory is the most widely cited conceptual framework (Matz-Costa, Besen, Boone James, & Pitt-Catsouphes, 2014; Morrow-Howell, Hinterlong, Rozario, & Tang, 2003; Musick, Herzog, & House, 1999; Musick & Wilson, 2003; Piliavin & Siegl, 2007; Van Willigen, 2000). This framework posits that the human self-concept is a collection of identities based on the roles played (Matz-Costa et al., 2014). A significant body of literature has found that well- being increases as the number of roles increase (Thoits, 2012), which is likely due to the sense of purpose or meaning gained from the role. This theory is applied to the role of volunteer, which explains why, in some studies, only the status of volunteer is necessary in order to experience increases in well-being. Another framework is based in Positive Psychology. This framework hypothesizes that the well-being benefits of volunteering are based on the positive emotional response that is felt from engaging in activities that benefit others (Fredrickson, 2003; Post, 2005). The positive emotions (kindness, compassion, etc.) displace negative emotions, which can improve psychological health. Additionally, because consistent negative or stressful emotions have adverse impacts on physical health, it is suggested that the positive emotions felt during volunteer are protective factors from stress-related health problems (Post, 2005). 8 Finally, a conceptual framework provided by Social Network theory has also been employed to explain the association between volunteering and well-being (Borgonovi, 2008; Musick et al., 1999; Musick & Wilson, 2003; Piliavin & Siegl, 2007). Volunteer roles can increase the quantity and quality of relationships, reduce loneliness, and give people opportunities to fill roles that are socially valued. This increase in social integration can result in improved well-being. Pathological altruism. One important nuance in the relationship between volunteerism and well-being is the concept of pathological altruism. Pathological altruism is a recently coined term that describes altruistic activity that results in a decrease in well-being for the giver (and can potentially be detrimental to the recipient) (Oakley, 2011, 2013, 2014; Rubin, 2014; Smith, 2015; Smith & Davidson, 2014). In regard to volunteering, it is also referred to as the Burden Assumption (van Campen et al., 2013) and Role Strain (Son & Wilson, 2012), which describe the scenario in which a person overcommits, becomes emotionally overwhelmed by a volunteer role, or engages in too many roles (Morrow-Howell et al., 2003; Son & Wilson, 2012; Windsor, Anstey, & Rodgers, 2008). In the sources of this literature review, it is most often reported as a curvilinear relationship between volunteering and well-being in which higher levels of volunteering result in a tapering effect on well-being. Pathological altruism fits into Gilligan’s theory of moral development: The Pre- Conventional stage is self-survival, the Conventional Stage is self-sacrifice for others, and the Post-Conventional stage demonstrates a balance of caring for self and others (do no harm to self or others) (Gilligan, 1977; Rhodes, 1985). In her description of these stages, Gilligan is congruent with the hypotheses and findings of many of the studies (discussed later) that there is a 9 point at which too much of a good thing is literally too much, and that it is necessary to balance the well-being of the self with the well-being of others. A related question that is beyond the scope of this review, but nevertheless important, is whether the work- and stress-load of caring for children at risk of maltreatment is creating a scenario in which the caregivers are being taken advantage of or mistreated. According to van Campen and colleagues (2013), several European governments have implemented policies to protect a variety of social caregivers from maltreatment (for example, the 1995 Carers Recognition and Services Act in the UK and the 2007 Social Support Act in the Netherlands). This risk may more deeply threaten women, who more frequently fill the roles of caregiver (Harway & Nutt, 2006; MacDonald, Phipps, & Lethbridge, 2005; Windsor et al., 2008). Social work values and ethics. There are many perspectives from which Social Work values and ethics can be applied to this discussion. In this section, the value of Service will be highlighted in two ways. First, the social work profession is primarily a service profession (National Association of Social Workers, 2017). Social workers serve others and create opportunities for citizens to join in service activities. The primary function of this service is to help those in need, to solve social problems, and to advocate for justice. This study is focused on the important social problem of child welfare, and focuses on a recent innovation. In addition, the focus of this study is on volunteerism, which is a significant part of social work service. The second service-related perspective to highlight is in regard to well-being. As will be discussed at length below, there is a substantial body of evidence that suggests that volunteering improves physical and psychological health, as well as Life Satisfaction and happiness. Volunteering as an intervention has only been explored shallowly (Post, 2007, 2011; Schwartz & Sendor, 1999), but there is promise in further research. As more is learned about how 10 volunteering improves well-being, social workers will be able to consider how volunteering can be incorporated in treatment plans of those who struggle with illnesses and social support deficits. In this way, social workers can serve and empower clients, promote their worth as a person, and encourage the development of relationships. 11 SEARCH PROCESS LITERATURE REVIEW The search process for the literature review had two components: the use of multiple databases and snowballing. ProQuest was used to search 89 databases, including ERIC, International Bibliography of the Social Science, PsycARTICLES, Psychology Database, PsycINFO, Sociological Abstracts, and Sociology Database. The following search string was used: (volunteer* OR altruism OR generosity OR prosocial OR "social interest") AND (well- being). A final search was conducted on September 15, 2017 to ensure the most up-to-date results. After filtering for peer-reviewed material in English, 602 results were found. There were two search terms that were problematic. First, the word “health” was initially included in the search string, but was eventually removed to narrow results from over 6,300 hits. Second, the term, “volunteer” was problematic because so many studies use volunteers as study participants. The 602 results were sorted manually to exclude irrelevant uses of the term “volunteer”, resulting in 167. This list was further sorted to with the following criteria: 1. 2. 3. Must explore an empirical relationship between altruism and well-being Must include volunteering as an independent variable Sample is not from a narrowly defined population (e.g., inmates, migrants, victims of natural disasters, specific race/gender, sports volunteering, sample of immigrants, cancer patients, post-communist countries, etc.) Studies whose participants were not from the U.S. or Canada and/or were only a specific age group (i.e., older adults, adolescents) were reviewed, but most were excluded from the final reference list. However, some of these articles were included because of their relevance (i.e. methodology, seminal status, etc.). 12 A snowballing technique was also used to create the reference list. Most articles cited sources that had been found in the database search process, but six peer-reviewed articles were found in other bibliographies that had not been included in the search process because of their atypical terminology or dependent variables. Additionally, four scholarly books were found via the snowballing process. The final number of sources included is 37: four scholarly books, 32 peer-reviewed journal articles, and one non-scholarly article. Nineteen of these studies are from articles published in 2010 or later. Two of the peer-reviewed articles (Anderson et al., 2014; Post, 2005) and one book (Post, 2007) are reviews of empirical research, and the remainders are based on original empirical research or secondary data analysis. A final list of relevant sources is found on Table 2. Many more articles and sources are cited in this review to add to discussions on history, theory, and other important topics. LITERATURE REVIEW OVERVIEW The literature in this review is broad and fairly recent. There is a wide variety of authors from many disciplines and backgrounds who have studied the relationship between volunteering and well-being. Only five authors contributed more than one publication: Peggy Thoits, Carolyn Schwartz, Stephen Post, John Wilson, and Marc Musick. All of the peer-reviewed studies come from 1999-2017. During that time period, this mostly nonexistent area of study evolved into a growing and robust topic of research. Table 2 displays the included sources along with a summary of information about their design, data source, sample, and findings. As will be discussed below, there is nearly complete consensus among the sources of this literature review that there is a positive association between well-being and volunteering. The association was first noticed by Allan Luks (1988), who coined the term, “Helper’s High,” after a 13 project with Better Homes and Gardens. Sixty-eight percent of their readers and 88% of an additional sample reported experiencing a positive physical sensation when they were helping others. As mentioned above, it took another decade before researchers began to empirically study this phenomenon. Prior to 1999, studies on volunteering focused mainly on antecedents, attitudes, characteristics, and motivation (Oman, Thoresen, & Mcmahon, 1999; Thoits & Hewitt, 2001). There was literature that explored the association between well-being and variables such as religiosity, organization attendance, or membership in voluntary organizations (Thoits & Hewitt, 2001). These studies found that voluntary association membership reduced stress, and was related to better physical health (Moen, Dempster-McClain, & Williams, 1992; Rietschlin, 1998). While authors were beginning to suggest an association between well-being and volunteering, there was not any empirical evidence (Musick et al., 1999; Oman & Thoresen, 2000). The Americans’ Changing Lives survey series (House, 2014) was a major catalyst to begin testing the volunteering/well-being relationship. Beginning in 1986 and continuing to present day, the Americans’ Changing Lives survey series is a nationally representative, longitudinal study that explores the lives of middle-aged and older Americans. The series explores: “(1) the ways in which a wide range of activities and social relationships that people engage in are broadly ‘productive,’ (2) how individuals adapt to acute life events and chronic stresses that threaten the maintenance of health, effective functioning, and productive activity, and (3) sociocultural variations in the nature, meaning, determinants, and consequences of productive activity and relationship” (House, 2014). Between 1999 and 2003, five studies used the data from Americans’ Changing Lives to examine the well-being/volunteering relationship 14 (Morrow-Howell, Kinnevy, & Mann, 1999; Musick et al., 1999; Musick & Wilson, 2003; Thoits & Hewitt, 2001; Van Willigen, 2000). These studies seemed to unlock a broader interest that included debates about methods and causation, and eventually lead to a large, nationally- representative survey focused on generosity, the Science of Generosity Initiative housed at Notre Dame (Herzog & Price, 2016; Smith & Davidson, 2014). MAIN LITERATURE REVIEW FINDINGS There is nearly complete consensus within the sources of this literature review that well- being and volunteering are positively related. While there is almost no disagreement that the association exists, there are some nuances in the reported findings depending on each study’s design and research questions. Some studies reported that the relationship between volunteerism and well-being depends on the intensity or consistency of the volunteering (Windsor et al., 2008), while others report that increases in well-being are similar among all levels of volunteering (Son & Wilson, 2012). There are others who found that giving help to others is more strongly related to well-being than receiving help from others (Nelson et al., 2016; Schwartz, Meisenhelder, Ma, & Reed, 2003). As will be discussed later, many authors agree that although there is likely a reciprocal relationship to some degree, volunteering has an effect on well-being. The most prominent finding was a positive relationship between volunteering and psychological well-being with 70% (26) of sources reporting this finding. As mentioned above, psychological well-being is operationalized in a variety of ways, from decrease in depression symptoms (e.g, Smith & Davidson, 2014) to global, standardized measures of psychological well-being (e.g., Okun et al., 2011; Piliavin & Siegl, 2007; Schwartz et al., 2003). Twenty studies (54%) reported a positive relationship between volunteering and physical health. 15 Table 2. Summary of Literature Review Source Design Mixed Method Secon -dary Data Nat. Rep. CS L L L L L L L CS R L L R X X X X X X X X X X X Luks, 1988 Musick et al., 1999 Oman et al., 1999 Schwartz & Sendor, 1999 Van Willigen, 2000 Thoits & Hewitt, 2001 Morrow-Howell et al., 2003 Musick & Wilson, 2003 Schwartz et al., 2003 Post, 2005 Wink & Dillon, 2007 Piliavin & Siegl, 2007 Post, 2007 X 1,211 65+ 1,972 55+ 132 Adults X 2,867 25+ X 2,867 25+ 900 60+ X 2,348 25+ 2,016 13-98 Adults Older Adults 184 4,000 Adults Adults 16 X X X X X X X N Age of Sample 1,746 Adults Explores Mediators Moderators & Volunteerism related to Life Satis- faction Phys. Health Psych. Health X X X X X X X X X X X X X X X X X X X X X X X Explores Curvi -linear X X X X X X Happ -iness X X X X Table 2 (cont’d) Source Design Mixed Method Secon -dary Data Nat. Rep. N Age of Sample Explores Mediators & Moderators Volunteerism related to Phys. Health Psych. Health Life Satis- faction Happ -iness Explores Curvi -linear Wink & Dillon, 2007 Piliavin & Siegl, 2007 Post, 2007 L L R Borgonovi, 2008 CS Windsor et al., 2008 Yuen et al., 2010 O’Brien, Townsend, & Ebden, 2010 Pillemer, et al., 2010 Theurer & Wister, 2010 CS E/L CS L CS Okun et al., 2011 CS Brown, Hoye, & Nicholson, 2012 CS X X X X X X X 184 Older Adults 4,000 Adults Adults X 29,200 Adults 2,136 64-68 39 88 60+ 16-76 2,730 Older Adults 4,486 Older Adults 4,161 18+ 3,318 18-98 X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X 17 Table 2 (cont’d) Source Design Mixed Method Secon -dary Data Nat. Rep. N Age of Sample Explores Mediators & Moderators Volunteerism related to Phys. Health Psych. Health Life Satis- faction Happ- iness Explores Curvi -linear Son & Wilson, 2012 Thoits, 2012 Kwok, Chui, & Wong, 2013 Sneed & Cohen, 2013 van Campen et al., 2013 Vecina & Chacón, 2013 Anderson et al., 2014 MacIlvaine, Nelson, Stewart, & Stewart, 2014 Matz-Costa et al., 2014 McDougle, Handy, Konrath, & Walk, 2014 L CS CS L CS CS R CS CS CS X X X X X 3,257 25-74 X 458 41-91 443 Adults 1,654 Older Adults 336 Adults 251 16-78 Older Adults 309 18-65+ 330 50-83 X X X X X X X X X 1,805 Adults X X X X X X X X X X X X X X X X X X X X X X X 18 Volunteerism related to Phys. Health Psych. Health Life Satis- faction Happ- iness Explores Curvi -linear X X X X X X X X X X X X Table 2 (cont’d) Source Design Mixed Method Secon- dary Data Nat. Rep. N Age of Sample Explores Mediators & Moderators Smith & Davidson, 2014 Gimenez-Nadal & Molina, 2015 Herzog & Price, 2016 Nelson et al., 2016 Stukas, Hoye, Nicholson, Brown, & Aisbett, 2016 Klein, 2017 X X CS CS CS E/L CS CS X X X X X X 1,997 Adults 8,746 21-65 1,997 Adults 472 17-67 4,085 18-89 X X X 1,473 18-96 Note: R = Review; CS = Cross-sectional; L = Longitudinal; E = Experimental 19 For example, Oman et al. (1999) found that high levels of volunteering decreased mortality among older adults by 44%, which made it almost as protective as quitting smoking (49% reduction in mortality) and more protective than higher physical mobility, exercising four times per week, and attending weekly religious services. Another example is Piliavan and Siegl (2007) who found that self-reported health is significantly related to consistency and diversity of volunteering, and Smith and Davidson (2014) who found that volunteers report better health than their non-volunteer counterparts. Happiness was the next most prevalent well-being variable with an association with volunteering (43%), followed by Life Satisfaction (38%). As can be seen in Table 2, many studies found positive relationships between volunteering and several of these types of well-being (e.g., Herzog & Price, 2016; O’Brien et al., 2010; Smith & Davidson, 2014; Theurer & Wister, 2010; Thoits & Hewitt, 2001). Fourteen studies explored curvilinear relationships with varied results. Most of these studies found that the benefits of volunteering tapered off after a certain intensity or diversity, indicating a possible form of pathological altruism (Matz-Costa et al., 2014; Morrow-Howell et al., 2003; Musick et al., 1999; van Campen et al., 2013; Van Willigen, 2000; Windsor et al., 2008). Others found no curvilinear relationship, demonstrating that the main effect of volunteerism on well-being may be volunteer status, which confirms the Role Identity theory that simply identifying oneself as a volunteer is what causes most of the benefit (Piliavin & Siegl, 2007; Son & Wilson, 2012). DATA SOURCE AND SAMPLES The literature uses a balanced combination of secondary and primary data. Fifty-one percent of sources used primary data. Of the 18 that used secondary data, three were from non- 20 U.S. countries and five analyzed data from the Americans’ Changing Lives study (House, 2014). No other dataset was used more than once. The studies drew mostly larger samples of adults from the United States with the remaining from Canada, the United Kingdom, Spain, Hong Kong, and Australia. Nearly 75% of the sources included samples of adult age (18+), and the remainder were studies that focused primarily on older adults. Sample sizes range from 39 (Yuen, Huang, Burik, & Smith, 2008) to 29,200 (Borgonovi, 2008), with a mean of 2,765 and a median of 1,889. The sample size range for primary data sources was 39 to 3,318, and the sample size from secondary data sources was 900 to 29,200. Thirteen of the samples were nationally representative. A significant portion of the research investigating the health benefits of volunteerism has used samples of older adults because it is easier to detect health changes among this population (Schwartz et al., 2003). However, most studies focusing solely on older adults were not included in this review because, as much of the gerontological research points out, the experience of older adults is unique. Many older adults are facing greater risk of illness, and injury, as well as a decrease in mental and physical functioning (Anderson et al., 2014; Kwok et al., 2013; Musick et al., 1999; Ramos et al., 2016; Schwartz et al., 2003; Theurer & Wister, 2010; Vecina & Chacón, 2013). Older adults are also navigating a time of profound role upheaval because of major changes in employment and family demands. The experience of retirees, grandparents, and widows and widowers should not be the primary foundation upon which new research for broader age groups is established. It is, however, worth mentioning that Anderson et al., (2014) have recently conducted a critical review on literature that investigates the relationship between volunteering and well-being in older adults. Their review included 73 studies and found that 21 volunteering is associated with reduced symptoms of depression, better self-reported health, fewer functional limitations, and lower mortality. METHODS AND DESIGNS IN THE LITERATURE More than half of the studies (20) used a cross-sectional design and more than a third (14) employed a longitudinal design. Most (9) of the longitudinal studies used secondary panel data, and two studies were longitudinal experiments (Nelson et al., 2016; Yuen et al., 2008). There were no studies that used a quasi-experimental design. Overwhelmingly, the studies were purely quantitative (78%). Eight studies (22%) were mixed-method studies. Because of the nature of this literature review, only empirical sources were included, therefore, there were no qualitative studies. However, several of the mixed-method studies used robust qualitative methods, including home visits and in-depth interviews, as well as analyzing homes, neighborhoods, and photographs (e.g., Herzog & Price, 2016; Schwartz & Sendor, 1999; Smith & Davidson, 2014). To analyze data, most studies used logistic and/or multivariate linear regression, analysis of covariance, and partial correlations for linear relationships; and t-statistics, correlations, and chi- squares for bivariate relationships. Cross products and quadratic terms were added to multivariate regressions to explore moderators and curvilinear relationships (respectively). CONCEPTUALIZATION AND OPERATIONALIZATION OF VOLUNTEERISM IN THE LITERATURE A minority of the studies in this review include a clear definition of volunteerism. Below is a sample of definitions of volunteerism: • a non-obligatory, planned helping activity, sustained over time within an organizational context (Vecina & Chacón, 2013): 22 • an unpaid activity that involves taking actions within an organizational framework that potentially provides some service to one or more other people or to the community at large (Piliavin & Siegl, 2007) • the voluntary giving of time and talents to deliver services or perform tasks with no direct financial compensation expected. Volunteering includes the participation of citizens in the direct delivery of service to others; citizen action groups; advocacy for causes, groups, or individuals; participation in the governance of both private and public agencies; self-help and mutual aid endeavors; and a broad range of informal helping activities (Thoits & Hewitt, 2001) • unpaid work on behalf of those with whom one has no contractual, familial, or friendship obligation (Van Willigen, 2000) • serving others without financial or quid pro quo compensation (Musick et al., 1999) While these definitions vary, they tend to describe mostly the same types of behavior. Volunteerism is operationalized in four primary ways: Status, Intensity, Consistency, and Diversity (Piliavin & Siegl, 2007). Volunteer status is a dichotomous variable that measures only whether a person is or is not engaged in volunteering. Many of the studies that use this operationalization provide a time period (in the past 12 months, currently, etc.). Volunteer intensity refers to the proportion of time applied to the volunteer role. This can be measured over a given time period (hours in the last year) or as an average (hours per month or week). Volunteer consistency is a measure of how long a person has maintained their volunteer status. This measure is used in longitudinal studies by investigating whether a participant maintains 23 their volunteer status across data points (Musick & Wilson, 2003). Finally, volunteer diversity measures the number of different volunteer roles a participant holds. This is usually measured by a simple count of how many organizations for which a person volunteers (Piliavin & Siegl, 2007). OPERATIONALIZATION OF WELL-BEING IN THE LITERATURE Well-being is a broad concept that is defined and operationalized in a variety of ways. Stephen Post, in his 2005 review, cites an APA framework (Anderson, 2003) that defines six dimensions of well-being: biological, psychological and behavioral, environmental and social, economic, spiritual, and emotional. The vast majority of the studies in this review do not provide a specific definition, rather they tend to focus on operationalization. However, four studies (Piliavin & Siegl, 2007; Son & Wilson, 2012; Theurer & Wister, 2010; Vecina & Chacón, 2013) use a framework of hedonic and eudaimonic well-being, drawing heavily from Ryan and Deci (2000). Hedonic well-being refers to an individual’s happiness and overall positive emotional experience. Eudaimonic well-being is more transcendent and describes an overall sense of purpose and Life Satisfaction. While this framework is the most frequently used, it focuses mainly on psychological and emotional well-being, excluding biological, spiritual, social, and economic aspects. Although the actual measurement instruments used vary greatly, well-being is most frequently operationalized in four dimensions: physical, psychological, Life Satisfaction, and happiness. A few of the studies in this review used a global assessment of well-being, with the most common being the Personal Well-Being Index, which includes items like satisfaction with health, social relationships, and standard of living (Brown et al., 2012; O’Brien et al., 2010; Stukas et al., 2016). 24 Physical health is measured objectively in two ways: mortality (Musick et al., 1999; Oman et al., 1999) and hypertension (Sneed & Cohen, 2013). All other studies that measured physical health relied on self-reported measures, which were often single-item questions such as, “How would you rate your health at the present time?” (Piliavin & Siegl, 2007). Scales used include The RAND Health Component Score (Windsor et al., 2008) and the Short Form 36 Health Survey (Schwartz et al., 2003). Psychological health was the most common form of well-being that was measured. Unfortunately, there is little overlap in the methods and instruments used to measure psychological well-being. Some used single-item questions like, “In general, would you say your mental and emotional health is excellent, very good, good, fair, or poor?” (McDougle et al., 2014). Others used validated mental health scales such as the Center for Epidemiological Studies Depression Scale (Dillon & Wink, 2007; Thoits & Hewitt, 2001), the Emotional State Scale (O’Brien et al., 2010), or the Mental Health Continuum Short Form (Nelson et al., 2016). There were at least 17 different validated scales used to measure psychological well-being and five different single-item questions. Also of note is the variance in whether the instruments were used to measure positive emotions or to detect symptomology of mental illnesses. While some simply screened for depression symptoms (e.g., Dillon & Wink, 2007; Musick & Wilson, 2003; Smith & Davidson, 2014), others measured both symptoms and positive emotional states such as worthiness, Self-Mastery, and Self-Esteem (Gimenez-Nadal & Molina, 2015; e.g., O’Brien et al., 2010; Okun et al., 2011; Thoits, 2012). Life Satisfaction was similarly operationalized by both single item questions and validated scales (Kwok et al., 2013; Windsor et al., 2008). An example of a single item question is, “"Now please think about your life as a whole. How satisfied are you with it-are you 25 completely satisfied, very satisfied, somewhat satisfied . . . not at all satisfied?” (Thoits & Hewitt, 2001; Van Willigen, 2000). Additionally, Life Satisfaction was included along with happiness and other well-being concepts on some global scales of well-being, such as the Personal Well-Being Index (O’Brien et al., 2010) and the Quality of Life Index (Schwartz & Sendor, 1999). Happiness and Life Satisfaction are occasionally grouped together into a concept referred to as Subjective Well-Being (Oman & Thoresen, 2000; Thoits, 2012). Nearly all of the studies operationalized with standard single- and multiple-item questions which have been shown to be reliable (van Campen et al., 2013), although there was occasional use of happiness scales or well-being scales that included happiness (Nelson et al., 2016; Schwartz et al., 2003). CONTROLS AND COVARIATES IN THE LITERATURE While nearly every empirical study mentioned control variables, the rate and detail with which they were described was wide-ranging. Some sources only described control variables in general terms (e.g., “demographic variables”), while other gave explicit descriptions. Table 2 displays the frequency of the most common control variables mentioned. Eighteen of the studies (49%) reported including mediating or moderating covariates in their models. These range from basic demographic information to more complex concepts like social integration/capital (Oman et al., 1999; Piliavin & Siegl, 2007; Theurer & Wister, 2010), volunteer satisfaction, religiosity/religious service attendance (McDougle et al., 2014; Musick & Wilson, 2003; Oman et al., 1999), psychological resources (Brown et al., 2012), physical functioning/health (Okun et al., 2011; Sneed & Cohen, 2013), employment (van Campen et al., 2013), motivation (Kwok et al., 2013; Stukas et al., 2016), and perceptions of whether the volunteer activity has value (Piliavin & Siegl, 2007; Thoits, 2012). Many of the studies were 26 able to identify models and pathways that helped to better explain the relationship between volunteering and well-being. Some tested similar covariates, but found that they did not mediate or moderate the relationship between volunteering and well-being (e.g., Morrow-Howell et al., 2003). Table 3. Frequency of Control Variables in the Literature Variable Age Gender/Sex Health Education Social Factors Income/Socioeconomic Status Religious Service Employment/Work Marital Status Race Frequency 27 24 24 20 19 17 15 15 12 8 CRITICAL ANALYSIS OF THE LITERATURE This body of literature is growing. It has developed and matured over its eighteen-year lifespan, and it has room for important improvements. While it has become a robust body, the areas that require a critical analysis are operationalization, external validity, causation, and epistemology. Operationalization in the Literature. As some authors have pointed out, a significant problem in this area of research is consistent operationalization (Bekkers & Wiepking, 2011; Collett & Morrissey, 2007; Herzog & Price, 2016; Smith & Davidson, 2014). When a research study reports a relationship between volunteering and well-being, there is neither a consensus on what “volunteering” means nor a consensus on what “well-being” is. Volunteering seems like a common and obvious concept; however, it needs to be specifically defined in order to ensure the existence of a common research language. As 27 discussed above, only a few of the studies included in this literature review explicitly define volunteering (Herzog & Price, 2016; Nelson et al., 2016; Piliavin & Siegl, 2007; Smith & Davidson, 2014; Thoits & Hewitt, 2001; Van Willigen, 2000; Vecina & Chacón, 2013). Others assume a shared meaning, and while it may seem mostly inconsequential to not have an overt definition, it makes reporting, interpreting, and comparing results more difficult. Volunteering can be a formal role (e.g., being a teacher at a community organization or being the chairperson of a non-profit board of directors) or informal (e.g., helping your child’s teacher on a field trip or doing spring cleaning at a local park); it can be consistent (e.g., tutoring an elementary student once-a-week at a local school), episodic (e.g., collecting donations from neighbors for a food pantry), or seasonal (coaching a youth sports team). Volunteering can also have a direct level of interaction with recipients (serving food at a soup kitchen) or an indirect level of interaction (serving on an advisory committee, or on a political campaign). These examples illustrate the argument that researchers must be clear about what they mean when they refer to volunteering, especially if they are seeking to demonstrate a relationship with well-being or other dependent variables. This is particularly true because all sources in this review ask participants to self- report their volunteering. Self-reporting biases are always a concern because of social desirability, but in this case, self-reporting also creates a reliability concern. When studies ask solely whether a person has volunteered within a given time period (creating a dichotomous variable), the operationalization may seem simple. However, the wording of the question may reduce reliability. For example, Brown et al. (2012) operationalized volunteering by asking, “Do you currently volunteer with any formal organized group?” while several others asked about volunteer work in the past 12 months (Borgonovi, 2008; Okun et al., 2011; Van Willigen, 2000). A person who consistently 28 volunteered until recently would not be considered a volunteer in the study by Brown and colleagues (2012), but would be considered a volunteer in the other studies mentioned. The authors were quite silent on the issue of self-reporting volunteerism, but were more explicit about the reliability concerns of using self-reporting to measure well-being. Van Willigen (2000) points out that previous studies have shown that self-reported health is highly correlated with professional medical assessments while Piliavan and Siegl (2007) provide evidence that self-reports predict observable health measures, and Borgonovi (2008) explains that there is general consensus that self-reports of both health and happiness are reliable. Sneed and Cohen (2013) are the only investigators that used observable data in their longitudinal study by measuring blood pressure. When the goal is to measure volunteer diversity, intensity, or consistency, operationalization increases in complexity and introduces additional risks to reliability. When diversity (measuring how many different volunteer roles a person holds) is the variable of interest, it could be confusing for participants who hold volunteer roles that are very different in formality or other variations of volunteering. For example, if a person regularly volunteers at a local school three days each week and also helped setup at a one-time church event, it is up to the participant whether she would report both roles or only the more consistent role. It is possible that the one-time volunteer role could be forgotten or disregarded for its simplicity. However, if a person had recently volunteered in many one-time roles, it could give the impression that he/she is equally engaged in volunteering with a person who volunteers at several roles with high intensity and consistency. Measuring intensity (proportion of time applied to the volunteer role) through self- reporting also has the potential to be a reliability risk. Participants’ estimates of time spent could 29 be inaccurate because of recall problems or the inability to be precise (Thoits & Hewitt, 2001). Further, asking questions about the number of volunteering hours a person has engaged in during the past week, month, or year could create confusion for participants whose volunteering is episodic or seasonal. For example, an instrument that asks how many hours a person has volunteered in the last month may over- or under-estimate the intensity of a person who works as a volunteer nurse at a week-long summer camp or who coaches youth sports during only one season. Additionally, many studies that asked about intensity did not measure the highest levels of intensity. For example, MacIlvaie and colleagues (2014) operationalized hours per week as 1- 2, 3-5, 6-10, or 11+, and Sneed and Cohen (2013) operationalized hours per year as none, 1–49, 50–99, 100–199, or 200+, which is similar to the categories for all the sources that used the Americans’ Changing Lives data (less than 20 hours, 20-39 hours, 40-79 hours, 80-159 hours, and 160 hours or more) (Musick & Wilson, 2003). These measures have a high level of sensitivity at lower hours, but not at the highest hours. It is very conceivable that a person could volunteer more than ten hours per week and much more than 160 hours per year (volunteering 10 hours per week would result in more than 500 hours per year). This example also demonstrates the difficulty of comparing the dosage of volunteering when the units of measurement are not consistent across studies. If the weekly scale used by MacIlvaie and colleagues (2014) is recalculated to an annual measure by multiplying by 52 (Windsor et al., 2008), the categories become the following: 52-104, 156-260, 312-520, or 572+. By failing to measure volunteer rates at the higher end, studies lose their ability to measure curvilinear relationships and account for pathological altruism or overburden, and they inaccurately group the more extreme volunteering intensities with the high volunteering intensity. Further, some studies (e.g., Klein, 2017) used a 30 volunteer hours scale that includes zero. This implies that the difference between zero hours and one hour is the same as the difference between one and two hours, which will result in biased or skewed data (Son & Wilson, 2012). Some studies accounted for these problems by realigning their scales or using statistical techniques on skewed data (e.g., Son & Wilson, 2012; Windsor et al., 2008). Another important operationalization challenge for this body of literature is its inconsistent operationalization of well-being (Smith & Davidson, 2014). Even when studies measured the same concept, their scales and survey questions varied. Measures of psychological well-being are a good example. As mentioned above, there were at least 17 different validated scales used to measure psychological well-being and five different single-item questions. This is an important limitation in this body of literature because it reduces the ability to make comparisons between studies, and demonstrates that the body of literature is not working in a collaborative or chronological way. External Validity. External validity is another challenge facing the literature. This is primarily due to the lack of control or comparison groups, problems with sampling frames, and type and quality of nationally representative samples. Only two studies employed the use of a true experimental design. Nelson and colleagues (2016) drew a sample from a local community, a public university and a local company. They randomly assigned participants into four groups, including a control group; two groups doing altruistic activities, and one group engaging in self-care. Yuen and colleagues (2008) drew a sample of older adults from five long-term care facilities in one county in South Carolina. Participants were randomly assigned to a control group and a group that engaged in mentoring students. 31 An additional risk to external validity is the frequency with which the studies’ samples were drawn only from groups of existing volunteers. Thirteen of the studies used some form a representative sample, and were therefore able to make comparisons between volunteers and non-volunteers. The remaining observational studies did not use designs that allowed for this type of comparison, and may result in a selection bias (Shye, 2010a; Thoits & Hewitt, 2001). This is often because many of the studies were looking primarily at variables that mediate the volunteer/well-being relationship, and were not interested in the differences between volunteers and non-volunteers (e.g., Brown et al., 2012; Okun et al., 2011; Thoits, 2012; Vecina & Chacón, 2013). However, it is questionable whether these observational studies can be used as evidence to generalize to the broader population. There were 15 studies that used nationally representative samples and were therefore able to compare volunteers and non-volunteers, as well as to make a stronger case for external validity. Seven of these studies used data only from older adults. Of the eight studies that used data from adults of varied ages (Borgonovi, 2008; Gimenez-Nadal & Molina, 2015; Herzog & Price, 2016; Klein, 2017; Musick & Wilson, 2003; Smith & Davidson, 2014; Son & Wilson, 2012; Van Willigen, 2000), only Smith and Davidson (2014) and Herzog and Price (2016) were focused specifically on the relationship between generosity and well-being. The other studies had a much broader focus, and sometimes had instruments that were less precise. Causation. This literature varies in its discussion of causation. Although some sources address the topic at length, others seem to take it for granted or only briefly mention limitations to making causal inferences. The important debates related to causation are directionality and design, as well as theory and the body of evidence. 32 Directionality and design. The primary discussion about causation in this body of literature is whether volunteering promotes an increase in well-being or whether high levels of well-being promote volunteering. Several studies mention this challenge solely as a brief limitation (e.g., Thoits, 2012; Windsor et al., 2008). Only eight of the sources directly address directionality (Borgonovi, 2008; Nelson et al., 2016; Post, 2007; Smith & Davidson, 2014; Son & Wilson, 2012; Thoits & Hewitt, 2001; Van Willigen, 2000; Vecina & Chacón, 2013). The authors use different concepts to comment on making causal inferences: mutual influences (Smith & Davidson, 2014), social causation (Okun et al., 2011), selection effect (Vecina & Chacón, 2013), and self-selection and reverse causation (Borgonovi, 2008), but most either find or argue from theory that the relationship is at least somewhat reciprocal—a phenomenon that Vecina and Chacón (2013) call a “virtuous cycle.” However the phenomenon is described, it remains true that there is real risk of oversimplifying the causal direction of these two variables. It is very possible some or all of the observed effect is the result of a selection effect or reverse causation. This risk and the risk of an omitted variable are a particularly strong risk in the 20 studies that rely on cross-sectional data (See Table 2). Some of the studies attempted to address this risk via their method. The most common approach was through the use of relevant control variables. Nearly every study included control variables, which were included based on theory or a previously discovered relationship. The most common controls are listed in Table 2 and include demographic information, health, social factors, and religiosity. In one cross-sectional study that was particularly focused on causation, Borgonovi (2008) used an advanced statistical method (two-stage least squares estimation) and 33 found that once reverse causation is accounted for, a causal relationship exists between volunteering and happiness, but not between volunteering and health. There is a noteworthy deficit in the number of experimental designs in the literature. Only two studies used experimental designs (Nelson et al., 2016; Yuen et al., 2008), and none used a quasi-experimental design. Of note is one study that became an “accidental” experiment in Germany during the fall of the Berlin Wall (Meier & Stutzer, 2008). Although this study did not meet the search criteria of this literature review, it is relevant to the topic of causation. Meier and Stutzer (2008) assume causation in their longitudinal study, because the study encompasses the collapse of the Berlin Wall and the reunification of Germany. According to the authors, many in East Germany lost their ability to volunteer because of infrastructure changes that prohibited them from continuing in their roles. The authors found that Life Satisfaction decreased generally among East Germans over this time period, but those who lost their volunteer role experienced a much greater decrease in Life Satisfaction, and those who began volunteering during this time also had a less dramatic decrease in Life Satisfaction. Fourteen studies accounted for at least some of these validity risks using longitudinal designs (see Table 2), which allowed the studies to reduce the risk of omitted variable bias and to make a stronger inference that volunteering is an antecedent to increased well-being. Some of the longitudinal studies found success in their attempts to detangle the causal order and pathways of this relationship. For example, Nelson and colleagues (2016) found that volunteering increased well-being by increasing positive emotions and decreasing negative emotions. Sneed and Cohen (2013) reported that volunteering for 200 or more hours per year decreased the risk of hypertension and improved psychological well-being, but did not find similar results among those who volunteered less than 200 hours per year. By analyzing data from a longitudinal study 34 spanning half a century, Dillon and Wink (2007; Wink & Dillon, 2007), demonstrated that prosocial behavior in adolescence was related to physical health in late adulthood. Others reported more complex results: volunteering is related to social well-being and eudaimonic well- being, but not hedonic well-being (Son & Wilson, 2012); the relationship between volunteering and well-being is mediated by “mattering” (Piliavin & Siegl, 2007) and social integration (Thoits & Hewitt, 2001); lower social integration moderates the effect of volunteering on psychological well-being (Piliavin & Siegl, 2007); volunteer status has a significant effect on well-being, and the effect increases until about 100 hours annually (Morrow-Howell et al., 2003); volunteering and Life Satisfaction are linearly related among older adults and curvilinearly related among younger adults (Van Willigen, 2000); volunteering reduces depressive symptoms, but the effect is stronger among older adults (Musick & Wilson, 2003); and church-related volunteering has a greater effect on well-being than secular volunteering (Musick & Wilson, 2003). While these longitudinal studies offer a substantial foundation from which to make a causal inference, there are three important limitations. The first is that nearly all of the longitudinal studies used secondary data from studies that did not focus specifically on the relationship between volunteering and well-being. This creates a scenario in which the measures of well-being are often not the optimal choice for understanding the benefits of volunteering. For example, studies often use scales that measure symptoms of mental illnesses as a way of measuring psychological well-being. Because mental illness and psychological well-being are not polar opposites, drawing conclusions about overall psychological well-being from a measurement of mental illness symptoms can lack reliability. The second is that a noteworthy portion of the studies used the same dataset, the Americans’ Changing Lives (Morrow-Howell et al., 2003; Musick et al., 1999; Musick & Wilson, 2003; Thoits & Hewitt, 2001; Van Willigen, 35 2000). This data was collected over a relatively short period of time (1986-1994), which makes measuring long-term effects impossible. The third limitation is that only five of the studies were published in 2010 or later. Several longitudinal studies are worth noting because their unique data source or rigorous design contribute to the argument for causation. One important longitudinal study to consider is by Piliavin and Siegl (2007), who used data from the Wisconsin Longitudinal Study, which had begun in 1957 with over 10,000 participants. The longevity of this study allowed the authors to provide strong evidence for a causal relationship between volunteering and psychological well- being. Another important longitudinal study is by Son and Wilson (2012) who used nationally representative data from the National Survey of Midlife in the United States, which began in 1995 and ended in 2006. Son and Wilson (2012) tested the relationship between volunteering and three types of well-being (hedonic, social, and eudaimonic) and found a reciprocal relationship between volunteer status and social and eudemonic well-being. No relationship was found with hedonic well-being. Nelson and colleagues (2016) engaged in a longitudinal experiment that investigated the well-being impacts of neutral behavior (control group), behavior that is meant to benefit others, humanity/the world, or the self. They found that both types of prosocial behavior contributed more to improved psychological well-being than the control or self-focused groups. Arguments from theory and body of evidence. The final pieces to the causation discussion are arguments from theory and the body of evidence. None of the studies in this review set out to find whether volunteering was dependent on well-being—they were all looking at whether well-being was dependent on volunteering. Although a minority discussed a reciprocal relationship, none suggested that the causal direction was only from well-being to 36 volunteering. This is due to the theoretical foundation that all of the investigators agree upon: volunteering has at least some effect on well-being. Whether it is role identity theory (Matz- Costa et al., 2014; Musick & Wilson, 2003; Thoits, 2012; van Campen et al., 2013), group selection theory (Post, 2005), social network theory (Borgonovi, 2008), positive psychology (Post, 2005), or some other conceptual framework, there is no theoretical divergence from this assumption. Smith and Davidson (2014) dedicate an entire chapter of their book to the topic of causation and suggest that while there is likely a reciprocal relationship, it is illogical to consider that there is a unilateral pathway from well-being to volunteering. They argue that although healthier people tend to volunteer more, it is obvious that healthy people are not compelled to volunteer by their health. Contrarily, they propose that it is quite probable that for some people, health promotes contentment and possible inaction, rather than acting as an antecedent to volunteering. They also point to some evidence that volunteers are not healthier than non- volunteers. In their analysis of the literature on happiness, Smith and Davidson (2014) report that it is clear that participation in intentional activities is the most important non-genetic factor for influencing levels of happiness. They argue that if this is the case, it does not make sense that happiness and health cause generosity, but that the relationship is at least reciprocal. Both Smith and Davidson (2014) and Post (2005, 2007, 2009), argue from a Positive Psychology perspective, asserting that generous behaviors elicit positive emotions that displace the negative emotions that are known to cause stress-related illnesses. Many of the investigators who have contributed to this literature as well as others have argued that there is now enough evidence to safely draw the conclusion that volunteering effects well-being (Anderson et al., 2014; Herzog & Price, 2016; Morrow-Howell et al., 2003; Post, 2005, 2007; Ramos et al., 2016; Smith & Davidson, 2014; Son & Wilson, 2012; Theurer & 37 Wister, 2010). Although most of the authors making this claim have done so recently, several began making the claim as early as 2003 and 2005 (Morrow-Howell et al., 2003; Post, 2005). The amount of evidence is not the only noteworthy aspect of this discussion. It is also noteworthy that there is no divergence—there is no study with evidence that rejects the positive association between volunteering and well-being, and none that makes a case against the theory that volunteering causes an increase in well-being. Epistemology. Epistemological clarity is noticeably not a major concern of this line of research. Of all the sources used, only Smith and Davidson (2014) explicitly discussed their epistemological approach. Nearly all the sources took a positivist approach to their research design, data analysis, and interpretation. However, this approach was more of an underlying and somewhat tepid presupposition, rather than a clearly described lens for viewing knowledge- creation. The authors operationalized variables, used theory, analyzed data, and described results from this perspective, but they did not openly describe their activities as positivist. The lack of overt epistemologies is concerning in this line of research because the variables being measured could be considered quite elusive to a purely positivist framework. How can one really measure Life Satisfaction, subjective well-being, or mental health? These are not observable phenomena that can be directly operationalized or measured. Other than a few studies focused on observable health outcomes like mortality (Musick et al., 1999; Oman et al., 1999) or hypertension (Sneed & Cohen, 2013), nearly every study in this literature review depends on self-reported health and well-being variables. Twenty-three of the studies used validated measures for well-being variables, and almost all spent some effort to defend their choice of variables. This points to the implicit value that the authors place (or believe their reviewers and readers place) on appearing to produce robust, scientifically verifiable results. 38 The vast majority of the studies presented some form of a theoretical framework for their research approach. Some were very explicit in their descriptions, but all attempted to situate their research within the context of the existing literature. While all the studies described the purpose of their research, only 15 studies included a precise description of the authors’ hypotheses and only one made mention of a null hypothesis. Additionally, sensitivity analyses were very rarely mentioned in order to describe the robustness of findings. RESEARCH GAPS This area of exploration has benefitted from significant contributions from a variety of academic disciplines such as Sociology: (Herzog & Price, 2016; Musick & Wilson, 2003; Smith & Davidson, 2014), Psychology: (Okun et al., 2011; Oman & Thoresen, 2000; Pillemer, Fuller- Rowell, Reid, & Wells, 2010), Medicine: (Post, 2007), Public health: (Schwartz et al., 2003), as well as many other multidisciplinary collaborations. However, the relationship between volunteering and well-being has been underexplored in social work. The rare exceptions to this are two articles published in gerontology journals (Matz-Costa et al., 2014; Morrow-Howell et al., 2003) Social work has generated a plethora of knowledge on child welfare and foster care, but has not made a major contribution to the volunteerism and well-being literature. This is an unfortunate reality because social work is uniquely positioned to translate research into practice due to its well-connected workforce and innovative researchers. It is important to know that volunteerism may lead to improved well-being, but it is another matter to realistically implement changes in policy and practice. Because of the profession’s organization and broad workforce, social workers are uniquely situated to be able translate research into practice. As social workers engage with individual clients, families, groups, and communities, it is imperative that they have 39 access to this growing area of research so that they can make this important translation. It is also important that their perspective and experience is incorporated into how the research is designed and executed. Another significant research gap is exploring whether the relationship between volunteering and well-being varies based on the type of recipients of the volunteer activity. Little research has been conducted in which the beneficiaries of generosity are a specific population (e.g., vulnerable children). The closest that any research has come is to include the type of “cause” or type of volunteering. Several of the studies in this literature review drew their samples of volunteers from organizations focused on specific causes (for example, environmentalism (O’Brien et al., 2010; Pillemer et al., 2010), sick/disabled family members (van Campen et al., 2013), people with chronic illnesses (Schwartz & Sendor, 1999), and heart disease patients (Thoits, 2012). Some studies asked participants in which type of volunteering they engaged (Windsor et al., 2008). The Americans’ Changing Lives study (Morrow-Howell et al., 2003; Musick et al., 1999; Thoits & Hewitt, 2001; Van Willigen, 2000) asked participants about their type of volunteering (for example, religious groups, educational organization, political groups, senior citizen groups, hospitals, etc.). The Science of Generosity project (Herzog & Price, 2016; Smith & Davidson, 2014) also asked about types of causes and issues participants supported (for example, family and neighbors, adult education, children and youth, homelessness, poverty, drug and alcohol use, etc.), but this was a generic question about generosity, not about volunteering in particular. All but three sources declined to report any results regarding specific types of volunteering. Van Willigen (2000) reported that church- and school-based volunteering were the first and second most beneficial types of volunteering to 40 psychological well-being. Windsor and colleagues (2008) and Musick and colleagues (1999) reported no changes in well-being based on different domains of volunteering. The current literature has demonstrated that the association between volunteerism and well-being is complex and has multiple pathways. It has begun to untangle the pathways by operationalizing volunteerism in four dimensions: status, intensity, consistency, and diversity. However, results from these explorations differed across studies. In addition, there are mixed results on whether the effect of volunteer intensity on well-being is linear or curvilinear. While some of this difference is due to study design and operationalization of variables, it also suggests that the association may be nuanced along other factors such as volunteer type. The next step in this line of research should test whether the intended recipients of the volunteering have an effect on well-being. For example, do those who work with animals differ from those who work with people? Or do those who work with the vulnerable (for example, the sick, the poor, children) differ from those who volunteer in community organizations and institutions (schools, libraries, government, etc.)? While many studies have examined the relationship between well-being and volunteerism, no studies have investigated the relationship between well-being and volunteering to care for children who are at risk of child-welfare system involvement. This dissertation addresses this gap by engaging a sample of adults who are already volunteering to care for these children through a faith-based, international organization called Safe Families for Children (SFFC; see below for a more detailed description). The aim will be to better understand who they are, what motivates them, and how their volunteering is related to their well-being. My research questions will address gaps in the child welfare literature as well as inform child welfare policy and practice by drawing attention to innovations in child maltreatment prevention and 41 foster care deflection. This is because promoting child well-being is not solely about how many vulnerable children are cared for, or what their outcomes are, it is also about who cares for them and the impact of being a caregiver. Better understanding these caregivers and their experiences can inform methods for attracting, training, and retaining high quality caregivers into child welfare roles and highlight the need for involvement before a child has been placed into foster care. Additionally, this may help to transform society’s perception of vulnerable children from that of a burden to a blessing. RESEARCH QUESTIONS The following research questions inform the overall aim of this exploratory study, which is to understand the relationship between well-being and volunteering among people who voluntarily care for children at risk of maltreatment. 1) What are the characteristics and background of adults who volunteer to care for children who are at risk of maltreatment (i.e., volunteer Host Families for Safe Families for Children)? 2) Among these adults, what is the relationship between volunteering and well-being? 3) What motivates these adults to volunteer to care for children at risk of maltreatment? 4) Among these volunteers, is there an association between motivation and well-being? 42 METHOD This exploratory, cross-sectional, observational study employs a convenience sample of SCCF volunteers called “Host Families”. During March and April of 2019, participants completed an anonymous online survey that included demographic information, control variables as well as measures for volunteerism, other generosity, well-being, and motivation. Standardized measures were used when appropriate and available including measures for well-being variables and motivation. A goal of the survey design was consistency with the literature to increase the opportunity for inter-study comparisons. PROGRAM DESCRIPTION According to the SFFC website (safe-families.org; accessed 10/18/2019), SFFC is a faith- based, international organization that engages volunteers to accomplish the following stated goals: 1. Keep children safe during a family crisis such as homelessness, hospitalization, or domestic violence in an effort to prevent child abuse and/or neglect; 2. Support, and stabilize families in crisis by surrounding them with caring, compassionate community. 3. Reunite families and reduce the number of children entering the child welfare system. The SFFC vision is “Creating a world where children are safe and families transformed through radically compassionate communities”. The mission is to “host vulnerable children and create extended family–like supports for desperate families through a community of devoted volunteers who are motivated by compassion to keep children safe and families intact”. SFFC acknowledges three values: Radical Hospitality, Compassion fueled by Mercy and Disruptive Generosity. 43 SFFC was founded in Chicago in 2003 by Dr. David Anderson, Executive Director of SFFC and Lydia Home Association (safe-famlies.org; lydiahome.org). After starting as a program of Lydia Home Association, SFFC is currently also offered as a program through Bethany Christian Services and Olive Crest. Lydia Home Association operates in the Chicagoland region describes itself as an organization that provides “Hope, Healing and Home to children in foster care” (http://www.lydiahome.org/who-we-are/). It began in 1916, and currently provide foster care services, residential facilities, and in-home care to families along with SFFC. The mission of Olive Crest is to transform “the lives of at-risk children through the healing power of family” (https://www.olivecrest.org/about/). It operates in California, Nevada, and Washington, providing foster care, adoption, mental health care, and a variety of educational and prevention services, including SFFC. Bethany Christian Services is a “global nonprofit that supports children and families with world-class social services, all designed to help families thrive” (https://bethany.org/about-us). In addition to SFFC, it offers a variety of services in thirty states and eleven countries, including foster care, domestic and international adoption, refugee and immigrant services, and emergency care. Through these three organizations, SFFC works in over 100 locations around the United States. SFFC chapters were opened chapters in the United Kingdom in 2013 and in Canada in 2014. The SFFC model works by establishing a relationship between a Placing Family and a Host Family. The Placing Family is a family in crisis whose children are at risk of maltreatment and entering the child welfare system. Most of these families are experiencing a potentially devastating combination of stress and social isolation which significantly increases the risk of child maltreatment (Tucker & Rodriguez, 2014). Common crises affecting these families are unstable housing, medical emergencies, mental health and substance abuse issues, domestic 44 violence, incarceration of one parent, and unemployment. Because of their social isolation, parents have no one to help care for the children while the crisis is managed. This creates an increased risk of neglect for the children, which is the most common type of treatment investigated by protective services (U.S. Department of Health and Human Services, Administration for Children and Families, Children’s Bureau, 2018). Families referred to SFFC are able to voluntarily place their children into a Host Family while maintaining legal custody of and contact with their children. Families can directly request SFFC services, or they can also be referred by non-profits and child protective services. Each local chapter of SFFC establishes community partnerships as referral sources. In some places, this is other non-profits that provide services to families who are in crisis (e.g., homeless shelters, domestic violence programs, healthcare facilities, substance use treatment programs, etc.). In some chapters, the referrals come directly from state child welfare agencies when the presence of maltreatment or safety concerns exist but do not rise to the level of removal from the home (Nolan, 2013). Host Families are screened and trained by the coordinating agency (Lydia Homes, Olive Crest, Bethany Christian Services) to care for these children and provide support to the Placing Family while stabilization is achieved. Nearly all volunteer recruitment is done through Christian churches, with the goal of a considerable amount of volunteer involvement and program ownership within a church community. The ideal is for the church community to not only care for the children during a family’s crisis (via the Host Family), but for other church members to provide support for the Placing Family during their crisis. The support provided is based on the expertise and interests of the church community, and can include emotional support, problem solving, logistical support, financial support, and networking. The use of volunteers is one way that SFFC keeps 45 overhead costs low, claiming that a SFFC hosting costs only about one-fifth of foster care (Nolan, 2013). Although most recruitment happens through churches, church attendance is not required to be a Host Family. The criteria provided by SFFC are the following: “At least 25 years old, emotionally and financially stable, mature, law abiding, healthy, and active…willing to engage in a long-term friendship with a family in need, agree to not use drugs, and refrain from using profanity or engaging in other negative behavior that may impact a child” (https://safe-families.org/involvement/host-family/). An application, a home visit, and background and reference checks are required for new Host Family volunteers. Training is provided via an online portal or in-person. When the Placing Family has stabilized, the children are able to return home in almost all cases (Nolan, 2015). According to the SFFC website (https://safe-families.org/about/impact/), more than 35,000 children have been hosted since its inception, with an average hosting of 45 days. SFFC reports the majority of children who are hosted are age five or less, and ninety-three percent of hosted children return to the Placing Family at the end of the hosting. The primary outcome for the remaining children is a referral to child protective services and/or the local child welfare system. The program’s effectiveness and outcomes for children are currently being evaluated in the U.S. by faculty from the University of North Carolina and in the U.K. by the Social Research Unit at Dartington, which is a “research and design charity dedicated to improving outcomes for children and young people” (dartington.org/uk). Both studies are using randomized control trials to investigate whether SFFC diverts or simply delays children from entering the child welfare system and whether children in SFFC have different outcomes than those who are referred directly into the child welfare system. 46 SFFC lists six underlying assumptions, seven short-term outcomes, and four long-term outcomes (Nolan, 2013, p. 17) that help to clarify their strategy and goals for accomplishing their mission and implementing their vision: Assumptions. • Children at-risk for maltreatment may be served best by pathways outside Child Welfare • Children at-risk for maltreatment have better outcomes with family preservation & empowerment • Some parents are temporarily not able to care for their children • Crises may be temporary or episodic, not permanent; a temporary solution may be needed • • If children are well cared for, parents are more likely to focus on personal change If safe alternative placements are available for the families of children at-risk for maltreatment, the Child Welfare system may not need to be involved Short-term Outcomes. • Increase in percentage of Placing Families’ children who return to the Placing Families • Decrease in percentage of Placing Families’ children who are screened into the Child Welfare system • Increase in percentage of Placing Families that develop trusting & supportive relationships with Host Families • Increased ability of Placing Families to safely care for their children • Reduction of stress &/or depressive symptoms among Placing Families 47 • Enhanced ability of Placing Families to successfully navigate & manage crises • Increase in percentage of Placing Families who have an established social network upon exiting SFFC Long-term Outcomes. • • • Increased child safety for children of Placing Families Increased well-being for children of Placing Families Increased social networks for Placing Families • Reduction of child maltreatment risk factors for Placing Families SAMPLE A convenience sample consisting of SFFC volunteers called “Host Families” is utilized for this study. All 790 Host Families in the United States who are coordinated through the Bethany Christian Services SFFC program were invited to participate by email. The email invitation was sent to 1073 email addresses from these households with the instruction for only one member per household to complete the survey in order to maintain the assumption of independence.1 This instruction was repeated again as the final sentence in the consent form, which immediately preceded the beginning of the survey. To attain a 5% margin of error and a confidence level of 95% with a sampling frame of 790, the required sample size is greater than or equal to 259 (https://www.qualtrics.com/blog/calculating-sample-size/). The final sample consists of a total of 302 participants, which is a 38% response rate, and provides a margin of error of 4.44% at a confidence level of 95%. 1 SFFC was unable to produce a list with only one email address per household, so the invitation to participate was sent to 1073 total people with the instructions to complete only one survey per household. 48 Most participants in the sample are female (89.8%) and white (97.2%), and ages ranged from 42-93 years old (M = 61, Mdn = 59). More than three-quarters (77.1%) had completed a bachelor’s degree or higher, and 68% were employed at the time of data collection. Additional demographic data is reported in the Results section below. OPERATIONALIZATION AND MEASURES The survey protocol (see Appendix A) is an anonymous, cross-sectional, online survey developed using Qualtrics. The survey includes demographic information, control variables, and measures for volunteerism, other generosity, well-being, and motivation. Standardized measures for well-being and volunteer motivation were used in order to increase opportunities for comparisons between other studies that have explored the volunteerism/well-being relationship (see Table 4). Variables for volunteerism have been customized to the SFFC program, following the pattern from the volunteerism literature wherever possible. Demographics and Program Information. Demographics include gender, race, ethnicity, age, and location. Background and program information include unique background sub-status (e.g., veteran, homelessness, foster care, etc.; Unrau, Sherwood, & Postema, 2019) and questions related directly to the participants experience of the program (e.g., training, support, etc.) that were requested by key stakeholders at SFFC. Results of the questions requested by SFFC will be reported privately and are not included in analysis for this study. Dependent Variable: Well-being. Well-being is a broad and holistic term that describes many concepts related to health and one’s ability to survive and flourish. In this study, the operationalization of well-being is modeled closely after Thoits (2012), whose work resembled the American Changing Lives survey, one of the primary data sources in the volunteerism and 49 Table 4. Standardized Measures Concept Physical Health Self-Mastery Self-Esteem Psychological Distress: Depression Psychological Distress: Anxiety Life Satisfaction Social Support Motivation Standardized Measure Physical Health Index Pearlin Master Scale Rosenberg Self-Esteem Scale Patient Health Questionnaire - 9 Generalized Anxiety Disorder - 7 Personal Well-Being Index Lubben Social Network Scale (short form) Number of items 4 7 10 9 7 8 6 Citation Shavitt et al., 2016 Pearlin et al., 1981 Rosenberg, 1965 Kroenke, Spitzer, & Williams, 2001 Kroenke, Spitzer, & Williams, 2001 International Wellbeing Group, 2013 Lubben et al., 2006 Systematic Quality of Life Model 16 Shye, 2010 well-being literature (House, 2014; Morrow-Howell et al., 2003; Musick et al., 1999; Musick & Wilson, 2003; Thoits & Hewitt, 2001; Van Willigen, 2000). One of the weaknesses in the literature is that there is no uniform operationalization of well-being. Thoits’s model was chosen because of the potential for comparison to other studies that it provides and because of its inclusion of multiple constructs that provide a holistic conceptualization of well-being. Thoits (2012) measured six self-reported dimensions: Physical Health, Self-Esteem, Self-Mastery, Psychological Distress, Happiness, and Life Satisfaction. Thoits (2012) also takes the step of transforming participants’ scores on the scales used into mean scores. The benefit of using means rather than summed scores (which are commonly used in clinical settings) is that means take into account missing data by dividing the summed score by the number of valid answers given by each participant, eliminating the bias of missing data from an item in a given scale. In 50 this study, mean scores for each scale are only calculated for participants who answered at least 50% of the items in a scale. Physical health is measured utilizing the Physical Health Index (PHI), which includes four self-report questions that have been found to be reliable among diverse populations (alpha = .79) and associated with objective physical health measures (Shavitt et al., 2016). These questions, which are very similar to the four self-report questions employed by Thoits (2012), include general health, health satisfaction, a visual/graphic question, and physical health in the past 30 days. Because the four items in this scale do not use a consistent response pattern, responses from each item are transformed into a scale of 0-1 to ensure each item is weighted equally. These new item scores are summed to create a 0-4 point overall physical health score where higher scores equate to better health (Shavitt et al., 2016). Self-Mastery is a concept that describes the degree to which people believe that they have control over the course of their own life (Pearlin & Schooler, 1978). It is a mechanism that can help people cope with stress and can lead to a greater sense of well-being (Pearlin, Menaghan, Lieberman, & Mullan, 1981; Pearlin & Schooler, 1978; Thoits, 2012). The Pearlin Mastery Scale (PM) measures mastery by asking participants to rate their level of agreement to seven items using a four-point Likert scale ranging from Strongly Disagree to Strongly Agree. The PM is a very widely used measure in health and psychological research, including national studies in Canada, the U.S., and Europe (Clench-aas, Nes, & Aarø, 2017). The scale includes five negatively worded items that must be reverse-coded before an average score can be determined (Pearlin et al., 1981; Thoits, 2012). Mean scores will range from low Self-Mastery (score of 1) to high Self-Mastery (score of 4) (Thoits, 2012). 51 Similar to Thoits (2012), Self-Esteem is evaluated in this study using the Rosenberg Self- Esteem Scale (RSE; Rosenberg, 1965), which includes ten items measuring self-worth through the use of both positive and negative views of the self. Items are rated on a four-point Likert scale ranging from Strongly Disagree to Strongly Agree. Five of the items must be reverse coded before an average score can be calculated. The RSE has been used very widely and been the object of many psychometric evaluations, with alphas ranging from .72 to .88 (Gray-Little, Williams, & Hancock, 1997). In the literature, psychological distress is a very common inverse proxy for measuring psychological well-being. Thoits (2012) conceptualized psychological distress as the degree of depression and anxiety symptoms experienced and used a modified version the Brief Symptom Inventory-18, which includes questions regarding symptoms of anxiety and depression (BSI-18). However, because the BSI-18 is not freely available and the publisher would not give permission for use with this project, the Patient Health Questionnaire-9 (PHQ-9) and the Generalized Anxiety Disorder -7 (GAD-7), both very common and well-validated measures, are used to measure depressive and anxiety symptoms, respectively (Kroenke, Spitzer, & Williams, 2001; Pfizer, n.d.). These measures ask about symptomology over the past two months and are scored on a 4-point scale ranging from Not at All (scored as 0) to Nearly Every Day (scored as 3). The Personal Well-Being Index (PWI; International Wellbeing Group, 2013) is used to measure Life Satisfaction globally and across eight domains such as standard of living, personal relationships, and personal safety. Each item receives a score on an end-defined scale ranging from No Satisfaction at All (0) to Completely Satisfied (10). According the instructions, each of the domain items can be analyzed as a separate variable, or the scores can be combined to create an average “subjective wellbeing” score (International Wellbeing Group, 2013). While Thoits 52 (2012) used only the single, global question of the PWI, the complete PWI is used in this study, which is similar to several recent studies on volunteerism and well-being (Brown et al., 2012; O’Brien et al., 2010; Stukas et al., 2016). The PWI has strong convergent validity with the Satisfaction with Life Scale (r = .78), reliability (alpha ranges from .70 to .85), and test-retest reliability (r = .84) (International Wellbeing Group, 2013). Happiness is measured on a single-item scale that has been widely used in happiness research (Veenhoven, n.d.) and utilized in recent studies on happiness and volunteering (Thoits, 2012; van Campen et al., 2013). Very similar single-item scales were also used in several comparable volunteerism studies (Borgonovi, 2008; Herzog & Price, 2016; Smith & Davidson, 2014; Thoits & Hewitt, 2001). Using a five-point scale (from unhappy to very happy) the question asks respondents to rate the extent to which they consider themselves to be a happy person. Similar to van Campen and colleagues (2013), this variable will be treated as an interval variable for analysis. Independent Variable: Volunteerism. Volunteerism, defined as contributing to a charitable cause via time, talents, or services (Herzog & Price, 2016), is the independent variable of interest in this study. As described above, one unique aspect of this study is that it narrows the type of volunteerism to working with children at risk of child welfare involvement through SFFC. In the literature, volunteerism is operationalized in four dimensions (Piliavin & Siegl, 2007): status, intensity, consistency and diversity. Because all participants in the sample are volunteers, status is excluded from this study. Intensity is primarily measured as the number of days with a child placed in the home over the past twelve months. This is measured as a continuous variable, and also used to create a dummy variable for “active” (at least one placement in the past 12 months) and “inactive” (no placements in the past 12 months). 53 Additionally, the number of children hosted at the time of the survey as well as the number of children placed in the home in the past month will also be measured as alternative measures of intensity. Volunteer consistency refers to duration as a SFFC Host Family, which is measured by asking participants how long they have held their volunteer role as a Host Family. Volunteer diversity (the number of other volunteer roles) was included only as a potential control variable so as to account for the effect of other volunteer activities on well-being (see section below on Covariates). Covariate: Motivation. Several studies have sought to better understand whether volunteer motivation is a covariate in the association between well-being and volunteerism (Kwok et al., 2013; O’Brien et al., 2010; Shye, 2010b; Stukas et al., 2016; Vecina & Chacón, 2013). This study seeks to understand whether there is a relationship between well-being and motivation. As Shye (2010) demonstrates, there does not seem to be a clear distinction between motivation, demographic antecedents, and triggers/opportunities in the literature. Shye (2010) argues that most studies incorrectly include both demographic antecedents (resources that make an activity possible, for example, income, education or religion) and triggers/opportunities (environments that lead to volunteering, for example, being asked by a friend or participation in a social group) in their research on motivation. Shye (2010) explains that motivation is a psychological term that refers to a “state of tension that seeks relief or equilibrium through action” and leads to increased well-being (p. 188). In order to operationalize this concept, Shye (2010) created a measure called the Systematic Quality of Life Model (Shye, 2010b). This measure includes 16 broad types of needs and asks participants to rate their importance as a motivation for volunteering on a five-point 54 scale ranging from Not Important to Very Important. The 16 needs can be grouped in four subcategories of well-being, which are used for analysis: personal (items 1-4), physical (items 5- 8), social (items 9-12), or cultural (items 13-16). Items in each category are averaged to create a category score. While Shye’s work received the highest marks in a review of scales that measured quality of life (Taillefer, Dupuis, Roberge, & LeMay, 2003), no quantitative psychometrics are reported as a result of the author’s factor analysis (Shye, 2010b). Covariates. Covariates that are common in the literature are included: race, marital status, employment status, religious service attendance, income, and social support. All of these variables use standard questions from the literature, except social support, which has been found to moderate the relationship between volunteering and well-being (Piliavin & Siegl, 2007). Social Support is measured with the short form of the Lubben Social Network Scale (LSNS), which has been validated for use with diverse groups (alpha = .83) (Fuhrer & Stansfeld, 2002; Lubben et al., 2006). The total score on the LSNS is a mean of the six items, with a higher score indicating increased social engagement (range of 0-5). Additional control variables are other generous behaviors that have been shown to contribute to increased well-being, including financial giving and volunteer role diversity. Volunteer role diversity measures the total number of volunteer roles held (Thoits, 2012). The final control variable is caregiving burden, a measure of the number of caregiving roles a person has in their life. This question was created for this study and will produce a total number of children and adults that participants are caring for on a regular basis. PROCEDURES After gaining full IRB approval from the Michigan State University Human Research Protection Program, the online survey (Appendix A) was distributed to the sample via email 55 (Appendix B) along with three reminder emails. The data collection period was four weeks during March and April of 2019. Participants are anonymous and all responses were confidential and stored on a password-protected account. Before beginning the survey, potential participants were required to give informed consent, which included a clear description that participation was optional. There is no foreseen risk to the participants. ANALYSIS PLAN Based on previous literature, this exploratory study sought to answer the research questions using the following approaches. Research Question 1. What are the characteristics and background of adults who volunteer to care for children who are at risk of maltreatment (i.e., volunteer Host Families for Safe Families for Children)? • Variables such as well-being, generosity, and volunteerism, as well as demographic variables will be measured using descriptive statistics. Research Question 2. Among these adults, what is the relationship between volunteering and well-being? • This study explores whether there is a relationship between intensity and consistency of volunteering as a Host Family and all seven dimensions of well-being. Using an independent samples t-test, the difference between well-being means will be compared between those having a hosting in the past year and those who have not had a hosting in the past year. Correlation and multivariate linear regression are used to answer this research question because the independent and dependent variables are continuous. 56 Regressions are run by deleting cases listwise to ensure that all variables in each model used the same sample of participants. • For concepts that are measured with the use of a scale, Confirmatory Factor Analysis factor scores will be explored to account for the varied loadings each item has on its corresponding construct. Structural Equation Modeling was not used because too many independent and control variables had missing data. Missing data were not imputed because data are missing by design, not at random: participants who have not hosted at least one child in the past month were not given the option to answer questions about volunteer intensity, and excluding cases with missing data would result in a substantial reduction in sample size. • Several previous studies have found the benefits of volunteering tapered off after a certain intensity or diversity, indicating a possible form of pathological altruism (Matz- Costa et al., 2014; Morrow-Howell et al., 2003; Musick et al., 1999; van Campen et al., 2013; Van Willigen, 2000; Windsor et al., 2008). This study explores this phenomenon to better understand if the relationship between volunteering as a Host Family and well- being is curvilinear, such that the relationship between volunteering and well-being will increase as the volunteer intensity increases until a threshold when well-being will plateau and/or decrease as participants experience overburden. This analysis is conducted using Polynomial Regression (van Campen et al., 2013) using the quadratic Curve Estimation feature of SPSS. 57 Research Question 3. What motivates these adults to volunteer to care for children at risk of maltreatment? • Motivation items and subscales from the Systemic Quality of Life Model (Shye, 2010) are reported using descriptive statistics. Research Question 4. Among these volunteers, is there an association between motivation and well-being? • Similar to previous studies, this study explores whether there is a relationship between well-being and motivation among Host Families to better understand the covariates that exist in the volunteering/well-being relationship (Kwok et al., 2013; Shye, 2010b; Stukas et al., 2016). This analysis will include Correlation and Multivariate Regression (Stukas et al., 2016). Regressions were run by deleting cases listwise to ensure that all variables in each model used the same sample of participants. 58 RESULTS The final sample size is 302, which is 38% of the sample frame of 790 households. Analysis was completed using SPSS version 22. R was used for the Confirmatory Factor Analysis. Results of reliability analyses are reported below, followed by results organized by Research Question. RELIABILITY ANALYSIS Table 5 displays the results of a reliability analysis of the data in this study in terms of Chronbach’s alpha. All scales have a Chronbach’s alpha of .73 or greater, including the subscales used to determine motivation on the Systematic Quality of Life Model. Table 5. Chronbach’s alpha. Scale Personal Well-Being Index Rosenberg Self-Esteem Scale Pearlin Self- Mastery Scale Patient Health Questionnaire - 9 Generalized Anxiety Disorder - 7 Physical Health Index Lubben Social Network Scale Motivation - Person Motivation - Physical Motivation - Social Motivation - Cultural Chronbach's alpha 0.84 0.86 0.74 0.74 0.86 0.79 0.83 0.76 0.80 0.73 0.86 Despite its widespread use, Chronbach’s alpha has important weaknesses, including its assumption that all items in a scale load equally on the latent factor being measured and the assumption that the error scores between any pair of items on a scale are not correlated (Raykov, 2001; Yang & Green, 2011). Because these assumptions are likely to be untrue in real world settings, a confirmatory factor analysis (CFA) was performed to provide a more accurate assessment of reliability (Brown, 2015). In CFA models, the observed score for each item in a 59 scale is assumed to be the combination of a portion of a true score (i.e., a given participant’s real- world score on a scale) and some amount of error due to unknown bias (Raykov, 2001). This can be expressed in the following formula, where y is the observed score, b is the loading or proportion, T is the true score and E is the error (Raykov, 2001). Equation 1. Observed Score for CFA Yi = biT1+ Ei In contrast with Chronbach’s alpha, a CFA model does not assume that the portion of the true score is identical in each item of a scale, and it accounts for the possibility that the error score of any pair of items may be correlated (Brown, 2015). Table 6 demonstrates the fit indices of each scale as well as the overall model fit when all scales are combined into one model. Indices reported are root mean square error of approximation (RMSEA- 90% CI Lower Bound), standardized root mean square residual (SRMR), comparative fit index (CFI), and the Tucker-Lewis (TLI, also known as the non- normative fit index) (Hooper, Coughlan, & Mullen, 2008). For the purposes of conducting tests of model fit, missing data is accounted for using the Full Information Maximum Likelihood, which uses all available data and does not delete cases listwise (Cham, Reshetnyak, Rosenfeld, & Breitbart, 2017). To account for non-normality of data, Robust Maximum Likelihood was used as the estimator. Reliability scores are computed using the following formula where b is the loading or proportion of each item on a scale, T is the true score (variance held constant at one) and E is the error (Raykov, 2001). 60 Equation 2. CFA Reliability Table 6. Confirmatory Factor Analysis for Well-Being and Motivation Scales Factor Well-Being Scales Personal Well Being Index (PWI-A) Rosenberg Self-Esteem Scale (RSE) Pearlin Self-Mastery (PSM) Scale Patient Health Questionnaire - 9 General Anxiety Disorder - 7 Physical Health Index Lubben Social Network Scale (LSNS) Full Well-Being Model Systematic Quality of Life Model (SQL) Motivation - Person Motivation - Physical Motivation - Social Motivation - Cultural Full Motivation Model Items w/ Correlated Errors - 2-6, 1-2, 2-5, 5-7, 3-4 2-6 2-9 - - 2-3, 1-2, 1-3 1-3 6-7 11-12 - 1-5, 2-7 SRMR RMSEA TLI CFI Reliability 0.04 0.04 0.04 0.04 0.04 0.02 0.03 0.07 0.00 0.01 0.01 0.02 0.07 0.04 0.93 0.95 0.03 0.01 0.01 0.05 0.00 0.08 0.94 0.96 0.94 0.97 0.94 0.96 0.92 0.95 0.91 0.95 0.99 0.97 0.048 0.81 0.83 0.00 0.00 0.00 0.04 0.06 1.03 0.99 0.99 0.96 0.90 1.00 1.00 1.00 0.99 0.92 0.84 0.82 0.73 0.76 0.87 0.80 0.70 0.84 0.87 0.76 0.86 All scales used in the study have a good reliability score (.70 or greater) and demonstrate good model fit (See Table 6). Cut-off criteria used to establish good model fit are the following: 61 SRMR < .08, RMSEA < .08, TLI > .9, CFI > .9, Reliability > .7 (Raykov & Shrout, 2002; Yang & Green, 2011). An advantage of confirmatory factor analysis as a method for establishing reliability is that it takes into account that individual items on a scale may not just correlate with the factor being measured, but also with each other. These correlations are measured by analyzing the correlation of the errors of the items (See Equation 1). This gives a more accurate representation of how an entire scale works to measure its intended concept, and is a technique used to improve model fit when necessary (Brown, 2015). Several scales in this study show good model fit without the need to improve the fit by exploring the correlations of errors: PWI-A, GAD-7, PHI, and the Cultural scale of the Systemic Quality of Life scale (SQL). The remaining factor models were improved by adding error correlations, and are reported in the second column of Table 6. The third item in the PWI (“How satisfied are you with your health?”) was removed from the Full Well-Being Model because of the redundancy with the third item in the PHI (“How satisfied are you with your present health in general?”). The Full Well-Being Model shows good model fit with the SRMR and RMSEA indices, and scores lower on TLI and CFI. According to Hu and Bentler (1999), scores of .8 can be acceptable. Additionally, Kenny (n.d.) demonstrates that in some instances, certain RMSEA scores mathematically prohibit high TLI and CFI scores, which is the case with the present dataset. The Full Motivation Model shows good model fit, which was improved by including two pairs of correlated errors. RESEARCH QUESTION 1 What are the characteristics and background of adults who volunteer to care for children who are at risk of maltreatment? 62 Demographics. Demographics and background information are displayed in Table 7. Most participants are female (89.8%), white (97.2%), and married (93%). The sample includes a high percentage of older adults, with the mean age of 61 (Mdn = 59, Mode = 57, SD = 10.78). Figure 1 displays histograms of participants’ ages, including a histogram actual ages and a histogram with ages by decade. The youngest participant was 42, and the oldest 93. Less than a third (28.8%) were under the age of 54, about half (50.9%) were between the ages of 55 and 69, and twenty percent were 70 or older. Table 7. Characteristics and Background Employed Female Non-White Married Income Midpoints Age Highest Ed (Bach or Higher) Dependent children in home Dependent adults in home Caregiver at work Foster child in home Social Support (LSNS) Personal Foster Care Personal Adopted Personal Ward Personal Veteran Personal Spouse of Vet Personal Previously Homeless Personal Homeless Percent 68.60 89.80 2.80 93.00 77.1 73.2 (1 or more) 32.1 (1 or more) 33.33 4.70 1.30 2.00 0.70 1.00 7.90 1.00 0.00 M 0.69 0.90 0.03 0.93 Mdn 1.00 1.00 0.00 1.00 SD 0.46 0.30 0.16 0.26 Min 0.00 0.00 0.00 0.00 Max 1.00 1.00 1.00 1.00 106,275 110,000 37,944 10,000 150,000 61.31 93.00 5.00 4.04 1.85 4.00 59.00 4.00 2.00 42.00 1.00 0.00 10.78 0.87 1.44 0.51 0.00 0.85 0.00 4.00 3.67 0.11 0.16 0.05 0.08 0.63 0.08 0.00 3.67 0.00 0.00 0.00 0.00 1.00 0.00 0.00 .79 0.31 0.37 0.23 0.27 0.49 0.27 0.00 1.17 0.00 0.00 0.00 0.00 0.00 0.00 0.00 5 1.00 1.00 1.00 1.00 1.00 1.00 0.00 Figure 2 demonstrates that participants are highly educated. Seventy-seven percent reported completing a bachelor’s or graduate degree (44% and 33.1%, respectively), and less than six percent did not complete at least some college. 63 Figure 1. Age Histograms As expected, this is a highly religious group. Eighty-six percent attend a religious service at least once per week, and more than two-thirds (35.9%) attend more than one time per 64 week. Less than two percent attend religious services less than once per month. Safe Families for Children focuses nearly all their recruitment efforts on Christian churches, which explains this finding. Income data was collected using $20,000 bins. For the sake of analysis (see Table 7), the midpoint of each bin is used. However, the Income Histogram (see Figure 3) uses the original bins to display the income. Participants reported a mean income of approximately $106,000 (Mdn = 110,000, SD = $37,944). Nearly one-third (30.9%) reported an income of $140,000 or more, while only one-eighth reported earning less than $60,000 (See Figure 3). A large majority (68.6%) are currently employed. Figure 2. Education Histogram 65 Figure 3. Income Histogram Because the role of Host Family is a role based on caregiving, other types of caregiving roles in the lives of participants were explored, including dependent children, dependent adults, foster care children, and caregiving at work (See Table 7). The sample is highly engaged in caregiving roles. Nearly one-third (30.7%) reported at least one dependent adult in the home, of which 56% reported having two or more. Almost three-quarters (73.2%) reported having at least one dependent child in the home. Fifty-eight percent of those who reported having dependent children in the home selected that they had two or more. Figure 4 shows the cumulative Caregiving Burden, which includes both dependent children and adults. The mean Caregiving Burden is 2.34 (Mdn = 2.00), with a minimum of zero and a maximum of eight. Only 4.7% reported that at least one of their dependent children was a foster child, which shows very little overlap between SFFC Host Families and Foster Parenting roles despite their similarity. This is 66 likely due to SFFC’s stated identity as an alternative to and prevention from foster care. Exactly one-third reported that their work included a caregiver role in some way. Participants were also asked about their personal histories with roles that could be considered adverse based on the work of Unrau and colleagues (2019). Table 7 displays that few participants reported being in these categories. Only four reported being foster care alumni (1.3%), six reported being adopted (2%), two reported being a ward of the court (0.7%), three reported previously being homeless (1%, 0 reported current homelessness), three (1%) are veterans and 24 are spouses of veterans (7.9%). While SFFC maintains a database with detailed records on the children and Placing Families, they do not maintain a database that can report demographic information on the Host Families. Because of this, the sample in this study cannot be compared to the sampling frame. When comparing sample characteristics to published data on foster parents (Gibbs, 2005), there are obvious differences. Foster parents are more diverse, with 30.2% identifying as a minority/non-white race compared to only 2.8% in the current sample. More than eight out of ten foster parents reported an annual income of less than $50,000, while only 12.6 percent the current sample reported an annual income of less than $60,000. The current sample is also much older, with a mean age of over sixty, compared to the mean age of forty-four among foster parents. Ninety-three percent of the current sample report being married compared to only 75.1% of foster parents. Finally, twenty-three percent of foster parents report having a bachelor’s degree or higher. This is considerably lower than the 77.1 percent of participants in the current study who reported earning a bachelor’s degree or higher. 67 Figure 4. Caregiving Burden Histogram Participants indicated a moderately high degree of social support on the LSNS (M = 3.67 out of 5, SD = .79). The LSNS measures social support by asking participants to select the number of people who provide support in six different ways on the following scale: 0 = zero, 1 = one, 2 = two, 3 = three or four, 4 = five through eight, 5 = nine or more. A weakness of the LSNS is that interpretation is challenging because of the categorical nature of the scale. For example, the mean of 3.67 in this sample represents a response of somewhere between three and eight supportive people, which equates to a moderately high social support score. In this sample, the histogram is positively skewed (See Figure 5). More than eighty-five percent averaged a LSNS score of 3 or higher. 68 Figure 5. Social Support Histogram Other Generosity. Participants reported a high degree of other forms of generosity including financial giving and engaging in other volunteer roles (See Table 8). Nearly all the participants reported donating money in the past year. There is a large range of reported giving ($0-300,000) with a high degree of variance in the giving amount (M = $11,556; Mdn = $7,150; SD = $21,752). About three-quarters (74.8%) reported giving $13,000 or less. Even the top ten percent had a wide range ($21,000-300,000). Nearly ninety percent of participants (89.5%) reported having a volunteer role other than their role as a SFFC Host Family, with the most frequent number of roles being two (33.6%), and 28.2% reporting three or more additional volunteer roles. The number of volunteer hours spent in these roles varied greatly from one to the equivalent of 40 hours per week (M = 144.32, Mdn = 69 60, SD = 266.37).2 When asked how many of the volunteer roles were with at-risk children, 46.9% indicated at least one. Table 8. Volunteerism and Other Generosity Hosted at least one in past 12 mo Children hosting now Hostings in past 12 mo Hosting days in past 12 mo Perception of hosting experience Volunteer Consistency (years) Volunteer roles other than SFFC Volunteer at-risk kids Total volunteer hours not SFFC Donate money Money donated 12 months Percent 68.80 23.10 89.50 46.90 99.00 M 0.69 0.31 3.00 56.15 2.01 2.56 1.93 0.63 144.32 0.99 11,556 Mdn 1.00 0.00 2.00 35.50 2.00 2.00 2.00 0.00 60.00 1.00 7,150 SD 0.46 0.66 2.15 59.20 0.64 1.86 1.18 0.85 266.37 0.10 21,752 Min 0.00 0.00 0.00 0.00 1.00 0.00 0.00 0.00 0.00 0.00 0 Max 1.00 3.00 10.00 300.00 3.00 10.00 4.00 4.00 2,100.00 1.00 300,000 Volunteerism. Almost seventy percent of the sample hosted at least one child in the past year. Participants who indicated they had hosted at least one in the past twelve months were also asked how many hostings they had in the past year, how many days they hosted a child in the past year, and whether they were hosting a child at the time of the survey. Participants reported a mean of three hostings in the past year (Mdn = 2, SD = 2.15), with 20.5% reporting five or more hostings (See Figure 7). The mean number of days a child was hosted in the past year (See Figure 8) was 56.15 (Mdn = 35.5). About one-third indicated they had hosted sixty days or more. Forty-three were hosting at least one child at the time of the survey, with ten reporting hosting two or more children. 2 The responses of two participants who whose answers were so high they were logically impossible were excluded from analysis. 70 Figure 6. Money Donated in Past 12 Months Figure 7. Hostings in the Past 12 Months Histogram 71 Figure 8. Hosting Days in the Past 12 Months Histogram About three-fifths reported that their hostings were equally as stressful as the typical hosting, while the other two-fifths were split evenly between being more stressful and less stressful than the typical hosting. Participants reported being SFFC Host Families for a mean of 2.56 years (Mdn = 2), with 31.9% indicating one year or less and about 25% reporting four or more years. Figure 9 demonstrates the variation in volunteer consistency. Slightly more than half indicated that the decision to become a Host Family was a joint decision with their spouse, followed by 46% who indicated that they were the main initiator or that they did not have a spouse (4.2%). 72 Figure 9. Years Served as a Host Family Histogram Well-being. Table 9 displays the descriptive statistics of participants’ self-reported well- being. Overall, the well-being scores demonstrate a high degree of well-being. Table 9. Well-Being Scores Life Satisfaction Anxiety Depression Physical Health Self-Mastery Self-Esteem Happiness M 8.68 0.32 0.26 3.47 3.25 3.34 4.36 Mdn 8.86 0.14 0.22 3.62 3.29 3.30 4.00 SD 1.03 0.40 0.28 0.49 0.42 0.40 0.53 Min 2.29 0 0 1.02 2.29 2.4 3 Max 10 2.14 1.67 3.97 4 4 5 The Personal Well-Being Index (PWI) score, which is an 8-item scale that measures the concept of Life Satisfaction, has a mean of 8.63. The PWI uses an end-defined scale ranging 73 from No Satisfaction at All (0) to Completely Satisfied (10). The median score is just slightly higher than the mean (Mdn = 8.86, SD = 1.03). Figure 10 demonstrates the vast majority of participants (90.2%) scored a 7.5 or higher and about 87% scored eight or higher in Life Satisfaction. Only two participants scored below a five. The PWI also includes a general question about satisfaction with life as a whole. The sample had a mean of 8.54 (Mdn = 9.00, SD = 1.19). Participants PWI score and the satisfaction with life as a whole question had a Pearson correlation coefficient of .665 (p < .001). On the average, participants reported low levels of psychological distress. The mean Anxiety score was 0.32 with an even lower median of 0.17 (SD = 1.03). Anxiety scores, measured as a mean of the seven-question GAD-7, can range from 0 (Not at All) to 3 (Nearly Every Day), however, only one respondent scored above a two (2.14). Anxiety scores fell well Figure 10. Life Satisfaction Histogram 74 Figure 11. Anxiety Score Histogram below clinically significant levels, with less than five percent indicating clinical symptoms of moderate anxiety. More than one-third of participants (34.3%) reported no anxiety whatsoever, and nearly 95% reported a score of one or less. Depression, which is measured on the PHQ-9 as a mean score ranging from 0 (Not at All) to 3 (Nearly Every Day), was even lower than anxiety in this sample. Less than three percent reported scores that could be considered clinically moderate depression. The mean score was 0.26 (Mdn = .22, SD = .28), and the highest score was 1.67. Figure 12 shows that just a handful of participants (9) scored a one or higher, while 96.8% scored less than one, and a quarter (25.3%) reported no depression symptoms at all. 75 Figure 12. Depression Score Histogram Participants reported a high level of physical health, with a mean score of 3.47 out of a possible 4.00 on the Physical Health Index (Figure 13). Eight out of ten (80.5%) rated their health to be Very Good or Excellent. More than two-fifths (41.1%) had zero unhealthy days in the past thirty days and nearly three-quarters (72.1%) had two or less unhealthy days in the past 30 days. Eight participants reported fifteen or more unhealthy days in the past 30 days. All but nine participants reported that they were Pretty Well Satisfied or More or Less Satisfied with their health. 76 Figure 13. Physical Health Index Histogram Self-Mastery and Self-Esteem were each measured on a scale of 1-4 and are also quite high (M = 3.25 and 3.34, respectively). Only 26.2% of participants scored below a 3.0 on the Self Mastery scale, and only 16.9% scored less than 3.0 on the Self-Esteem scale (see Figures 14 and 15). Nearly all participants indicated they are Happy (59.9%) or Very Happy (37.8%) with a mean of 4.36 on a five-point scale (See Figure 16). Only six participants rated their happiness Neither Happy nor Unhappy, and none selected Not Very Happy or Unhappy. 77 Figure 14. Self-Mastery Histogram Figure 15. Self-Esteem Histogram 78 Figure 16. General Happiness Histogram The histograms in Figures 10-16 suggest that all seven well-being variables do not have a normal distribution. This is also noticeable in the small standard deviations and ranges listed in Table 9. Table 10 confirms the non-normality assumption through both the Kolmogorov- Smirnova and Shapiro-Wilk tests of normality, which are significant for each variable. As mentioned above, the non-normality is due to the vast majority of participants reporting a high degree of well-being in all seven of the well-being domains included in this study. 79 Table 10. Tests of Normality for Well-Being Variables Happiness Self-Mastery Self-Esteem Physical Health Life Satisfaction Depression (PHQ-9) Anxiety (GAD-7) Skewness Kurtosis -1.07 0.15 Sig. Statistic 0.00 0.68 Shapiro-Wilk Sig. 0.00 Kolmogorov- Smirnova Statistic 0.38 -0.10 0.08 -2.00 -1.92 1.72 2.15 -.92 -1.03 5.24 6.84 3.97 5.37 0.12 0.11 0.20 0.14 0.21 0.22 0.00 0.00 0.00 0.00 0.00 0.00 0.97 0.95 0.81 0.86 0.83 0.75 0.00 0.00 0.00 0.00 0.00 0.00 Research question one explored the background and characteristics of Host Families for SFFC in this study. In summary, they are a highly religious and educated group. The vast majority identify as white and female. They demonstrate a high degree and variety of generosity. They tend to be older and to have higher than average incomes. They are happy and physically healthy. They report very low levels of anxiety and depression, and they demonstrate a high degree of Self-Esteem, Self-Mastery, and Life Satisfaction. RESEARCH QUESTION 2 Among these adults, what is the relationship between volunteering and well-being? When the sample is grouped by those who had a placement in the past year, Life Satisfaction showed a significant mean difference (See Table 11). Those who had a placement in the last year reported a .38 higher Life Satisfaction score than those who did not have a placement in the past year (p = .01). 80 Table 11. Difference in Well-Being Means for Placement in Past Year Sig. (2- tailed) Mean Difference Std. Error Difference t df Happiness Anxiety Depression Life Sat (PWI) Self-Mastery Physical Health -0.06 Index Self-Esteem -0.08 aLevene's test for equality of variance significance < .05, equal variances not assumed 0.47 265.00 -0.83 126.88 -1.31 266.00 2.53 124.47 1.30 265.00 0.64 0.41a 0.19 0.01 a 0.20 1.09 127.74 0.43 265.00 0.03 -0.05 -0.05 0.38 0.07 0.28 a 0.67 0.08 0.02 0.07 0.06 0.04 0.15 0.06 0.07 0.05 95% CI of the Difference Lower Upper 0.17 -0.10 -0.17 0.07 0.02 -0.12 0.68 0.08 -0.04 0.18 0.23 0.13 As demonstrated in Table 10, all measures of well-being are non-parametric. To account for this, Spearman’s rho is used (Table 12) for bivariate analysis. For comparison, Pearson correlations can be viewed in Appendix D. Self-Mastery has a small negative relationship with both the number of hostings in the past year and the number of children hosting at the time of the survey. Hosting at least one child in the past 12 months is positively correlated with Life Satisfaction. No other relationships are significant at the bivariate level. Table 13 demonstrates the bivariate relationships of control variables to well-being. Social Support is significantly related to all well-being variables. Higher income and greater education are significantly related to several of the well-being variables, although higher income is negatively related to physical health. Donating money is related to both happiness and Personal Well-Being. Age is negatively related to anxiety. Being employed is positively related to Self-Mastery, however, being a caregiver at work is related to decreased happiness and increased anxiety and depression. Religious service attendance is negatively related to physical 81 health. Overall caregiving burden, race, gender, and other volunteer roles have no significant relationships with well-being. Table 12. Bivariate Analysis – Spearman’s rho Hosted at least one in past 12 mo Hostings in past 12 mo Hosting days in past 12 mo Happy Self Mastery Self Esteem Physical Health Life Sat. Depres- sion Anxiety 0.04 0.08 0.03 0.03 .18** -0.05 0.01 -0.01 -0.16* -0.14 -0.03 -0.13 0.10 -0.08 0.06 -0.07 -0.03 0.01 0.00 -0.07 -0.02 Children hosting now 0.05 -0.16* -0.05 -0.10 0.06 0.07 0.12 Volunteer Consistency *p < .05, **p < .01 -0.12 -0.08 -0.02 -0.05 -0.07 -0.06 -0.10 Additional subgroups in the sample were explored to further investigate the well-being and volunteerism of the sample. No significant differences in well-being were found between men and women in the sample or between non-white and white participants. When comparing means, those who were seventy and older had no significant differences in well-being compared with those under seventy except for anxiety. Consistent with correlational analyses above, those who were seventy and older had less anxiety than those who were under seventy (mean difference = .14; p = .03). Consistent with correlational analyses above as well as linear and curvilinear regression analyses below, those with very high Volunteer Intensity levels (measured as Hosting Days in Past 12 Months) do not have a significantly different well-being scores than those with lower levels of volunteer intensity. This remained true when the cut-point was 100, 150, and 200 hosting days in the past twelve months. Those who reported an annual household 82 income of less than $60,000 reported slightly lower Self-Mastery scores (mean difference = -.18, p = .03). No other differences were found for the group with lower income. Table 13. Bivariate Analysis Covariates: Spearman's rho Social Support Score Religious Service Attend Employed Income Caregiver at work Marital Status Highest Ed Money donated 12 months Age Caregiving Burden Non-White Female Volunteer at risk kids Volunteer roles other than SFFC *p < .05, **p < .01 Happy Self Mastery Self Esteem Physical Health Life Sat. Depres- sion Anxiety .28** .15* .16* .25** .32** -.22** -.13* 0.05 0.04 0.11 -.15* 0.05 -0.01 .17* 0.04 -0.02 0.02 0.01 -0.11 .18** .18** 0.02 -0.08 .17** -0.03 0.00 -0.03 0.09 -0.07 -0.05 0.10 .16** -0.04 -0.06 .20** 0.06 0.07 -0.03 0.02 -0.01 -0.18** 0.12 -.15* 0.01 -0.09 .16** 0.06 -0.09 0.00 0.11 0.02 0.03 -0.05 .27** -0.04 0.07 0.02 .20** 0.07 0.05 0.05 0.06 -0.02 -0.05 -0.11 .16** 0.03 -0.01 -0.10 -0.11 0.03 -0.02 0.11 -0.10 0.04 0.03 .19** .17** 0.10 -0.07 -.28** 0.06 -0.03 0.06 -0.05 -0.01 -0.05 -0.02 0.04 0.05 0.07 -0.07 -0.07 -0.03 -0.05 0.05 0.03 0.03 Multivariate Regression was used to explore the relationship between the seven dimensions of well-being and volunteerism (intensity and consistency). Only variables that demonstrated a significant bivariate relationship were included as covariates in multivariate regressions: social support, religious service attendance, employment status, income, age, money 83 donated in the past twelve months, and being a caregiver at work. Marital status was not included because of the homogeneity of the sample. Table 14 is presented at the end of this section and displays results for the models that include all covariates. Bivariate models and models with only demographic variables as covariates were run, but not included in the table. Table 14 is organized by the seven well-being variables, each shown with volunteer intensity followed by volunteer consistency. As demonstrated in Table 14, there were no significant relationships between the seven dimensions of well-being and intensity or consistency of volunteering in multivariate models. Social Support was the most common covariate that was significantly related to well-being, which demonstrated a significant, positive relationship in all but four analyses (Self- Mastery/Intensity p = .06; Self-Mastery/Consistency p = .2; Self-Esteem/Consistency p = .06; Anxiety/Consistency p = .06). Income showed a significant, positive relationship with Physical Health, Self-Esteem, and Life Satisfaction. Being a caregiver at work was positively related to happiness, but appears to be related to slight increases in depression and anxiety. Being employed had a significant and positive relationship with Self-Mastery. For multivariate regression analyses, factor scores were also used for all variables that were measured using a scale. The factor scores are generated from the Confirmatory Factor Analysis model. They may provide a more accurate representation of the strength of each participant’s results from each scale because they take into account inter-item correlations as well as the true item loadings on each factor. However, factor scores are fixed with a mean of zero and a standard deviation of one, which can complicate interpretation. Regression tables with factor scores are reported in Appendix E, and do not vary substantially from multivariate linear regression models in Table 14. 84 Curvilinear analyses were run to explore whether a quadratic regression better explained the relationships between volunteering and the seven dimensions of well-being. Scatterplots with both linear and curvilinear lines are shown in Appendix F. No curvilinear models were significant. In summary, Research Question two investigated whether there is a relationship between volunteering and well-being. Analyses show that Life Satisfaction was higher for those who had at least one child placement in the last year. While bivariate analyses indicated some significant relationships between volunteering and well-being, multivariate analyses demonstrated no significant relationships. However, multivariate analyses revealed significant relationships between well-being and some covariates. Social Support is positively related to Happiness, Life Satisfaction, Self-mastery, and Physical Health, and negatively related to Anxiety and Depression. Additionally, higher incomes and greater education seem to support several well- being variables. Table 14. Regression Models with All Covariates Included Happiness and Intensity (Constant) Hosting days in past 12 mo Social Support Score Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Model is significant at p < .01 B 2.99 .00 .16 .03 .09 .00 .01 .00 -.23 SE .45 .00 .06 .04 .09 .00 .00 .00 .09 ß .07 .24 .07 .08 .08 .12 .07 -.21 t 6.72 .85 2.93 .91 .98 .92 1.45 .80 -2.56 Sig. .00 .40 .00 .36 .33 .36 .15 .42 .01 85 (Constant) Volunteer Consistency Social Support Score Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Model is significant at p < .01 Happiness and Consistency ß -.06 .25 .07 .10 .15 .08 -.07 -.13 B 3.11 -.02 .16 .03 .12 .00 .00 .00 -.14 SE .39 .02 .05 .03 .08 .00 .00 .00 .08 Table 14 (cont’d) Self-Mastery and Intensity (Constant) Hosting days in past 12 mo Social Support Score Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Model is significant at p < .05 B 2.92 .00 .09 -.04 .15 .00 .00 .00 -.03 SE .37 .00 .05 .03 .08 .00 .00 .00 .08 ß -.13 .16 -.10 .17 .16 .03 -.06 -.03 (Constant) Volunteer Consistency Social Support Score Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Model is significant at p < .05 Self-Mastery and Consistency ß -.06 .09 -.11 .17 .14 .06 -.02 .00 B 3.00 -.01 .05 -.04 .16 .00 .00 .00 .00 SE .32 .02 .04 .03 .07 .00 .00 .00 .06 86 t 8.01 -.88 3.51 1.01 1.47 2.02 1.05 -.94 -1.79 t 7.94 -1.52 1.92 -1.24 1.98 1.71 .39 -.70 -.38 t 9.25 -.80 1.28 -1.57 2.39 1.89 .79 -.20 -.01 Sig. .00 .38 .00 .32 .14 .05 .30 .35 .08 Sig .00 .13 .06 .22 .05 .09 .70 .49 .71 Sig .00 .43 .20 .12 .02 .06 .43 .84 .99 (Constant) Hosting days in past 12 mo Social Support Score Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Y=1 Model is significant at p < .05 Self-Esteem and Intensity ß -.04 .22 -.05 .12 .19 .05 -.11 -.03 B 2.72 .00 .11 -.02 .10 .00 .00 .00 -.02 SE .35 .00 .04 .03 .07 .00 .00 .00 .07 Table 14 (cont’d) Self-Esteem and Consistency ß -.02 .14 -.05 .10 .17 .10 -.04 -.01 (Constant) Volunteer Consistency Social Support Score Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Y=1 Model is approaching significance at p = .06 B 2.77 -.01 .07 -.02 .08 .00 .00 .00 -.01 SE .31 .02 .04 .03 .06 .00 .00 .00 .06 Health and Intensity (Constant) Hosting days in past 12 mo Social Support Score Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Model is significant at p < .01 B 3.10 .00 .22 -.06 -.03 .00 .00 .00 .04 SE .41 .00 .05 .03 .09 .00 .00 .00 .08 ß .03 .35 -.14 -.03 .09 -.06 -.01 .04 87 t 7.70 -.42 2.61 -.57 1.41 2.10 .54 -1.26 -.30 t 8.94 -.33 1.90 -.63 1.32 2.21 1.35 -.58 -.13 t 7.50 .38 4.20 -1.72 -.30 .96 -.74 -.10 .48 Sig. .00 .68 .01 .57 .16 .04 .59 .21 .76 Sig. .00 .75 .06 .53 .19 .03 .18 .57 .89 Sig. .00 .71 .00 .09 .77 .34 .46 .92 .63 Table 14 (cont’d) Health and Consistency (Constant) Volunteer Consistency Social Support Score Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Model is significant at p < .01 B 2.96 -.03 .16 -.05 .00 .00 .00 .00 -.02 SE .41 .02 .05 .03 .08 .00 .00 .00 .08 ß -.09 .24 -.11 .00 .15 .03 -.02 -.02 (Constant) Hosting days in past 12 mo Social Support Score Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Model is significant at p < .01 Life Satisfaction and Intensity ß .01 .31 .10 -.07 .28 .04 .05 -.07 B 6.15 .00 .36 .07 -.14 .00 .00 .00 -.14 SE .73 .00 .09 .06 .15 .00 .01 .00 .15 Life Satisfaction and Consistency (Constant) Volunteer Consistency Social Support Score Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Model is significant at p < .01 B 5.51 -.03 .43 .10 -.14 .00 .00 .00 -.02 SE .72 .04 .09 .06 .15 .00 .01 .00 .14 ß -.04 .33 .11 -.06 .30 .04 -.02 -.01 88 t 7.25 -1.19 3.30 -1.51 .05 2.01 .42 -.23 -.29 t 8.40 .12 4.01 1.23 -.90 3.19 .51 .58 -.94 t 7.68 -.65 5.05 1.67 -.94 4.39 .59 -.28 -.15 Sig. .00 .24 .00 .13 .96 .05 .68 .82 .78 Sig. .00 .90 .00 .22 .37 .00 .61 .56 .35 Sig. .00 .52 .00 .10 .35 .00 .56 .78 .89 Table 14 (cont’d) Depression and Intensity (Constant) Hosting days in past 12 mo Social Support Score Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Model is significant at p < .01 B .88 .00 -.09 -.01 -.05 .00 .00 .00 .07 SE .20 .00 .03 .02 .04 .00 .00 .00 .04 ß -.06 -.29 -.06 -.09 -.08 -.12 .09 .14 (Constant) Volunteer Consistency Social Support Score Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Model is significant at p < .01 Depression and Consistency ß -.03 -.26 -.08 -.05 -.09 -.07 -.02 .13 B .94 -.01 -.09 -.02 -.03 .00 .00 .00 .08 SE .21 .01 .03 .02 .04 .00 .00 .00 .04 Anxiety and Intensity (Constant) Hosting days in past 12 mo Social Support Score Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Model is significant at p = .05 B 1.00 .00 -.07 -.01 .06 .00 -.01 .00 .09 SE .32 .00 .04 .03 .07 .00 .00 .00 .07 ß -.03 -.15 -.03 .07 .00 -.20 .14 .12 89 t 4.40 -.72 -3.49 -.72 -1.09 -.87 -1.45 1.01 1.73 t 4.51 -.46 -3.69 -1.15 -.74 -1.18 -.89 -.25 1.84 t 3.11 -.37 -1.80 -.35 .86 -.02 -2.39 1.58 1.40 Sig. .00 .47 .00 .48 .28 .38 .15 .31 .09 Sig. .00 .65 .00 .25 .46 .24 .37 .80 .07 Sig. .00 .72 .07 .73 .39 .99 .02 .12 .16 Table 14 (cont’d) Anxiety and Consistency (Constant) Volunteer Consistency Social Support Score Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Model is significant at p < .01 B 1.32 .00 -.07 -.03 .02 .00 -.01 .00 .15 SE .32 .02 .04 .03 .07 .00 .00 .00 .06 ß .00 -.13 -.09 .02 .01 -.22 -.02 .17 t 4.10 -.06 -1.87 -1.20 .26 .07 -3.03 -.22 2.40 Sig. .00 .96 .06 .23 .80 .94 .00 .83 .02 RESEARCH QUESTION 3 What motivates these adults to volunteer to care for children at risk of maltreatment? Participants reported a wide range of motivations; however, patterns were clearly evident (See Table 15) in the four sub-categories of Personal, Physical, Social, and Cultural. Each sub- category contains four items that are scored on a five-point scale: Not Important (1), Slightly Important (2), Moderately Important (3), Important (4), and Very Important (5). Subcategory scores are the mean of its four items and can range from 1-5. Physical motivations were the least important of the subcategories (M = 1.39), with “Can improve my economic position” scoring the lowest of any single item (M = 1.13), followed by “Strengthens my feelings of physical security” (M = 1.35), “Enables me to express my power and control over the environment” (M = 1.49), and “May contribute to my health” (M = 1.63). 90 Table 15. Motivation Motive - Personal Express personality Rest from routine Relieves worries Self confidence Motive - Physical Express power/control Improve economic Contribute to health Strengthen security Motive - Social Social status Connect to organization Develop friendships Belonging Motive - Cultural Express beliefs Improve value compatibility Value agreement Relationship to culture/religion M Mdn 1.50 1.78 2.23 2 1 1.48 1 1.44 2 1.98 1.39 1.00 1 1.49 1 1.13 1.63 1 1 1.35 2.25 2.24 1.24 1 2 1.92 3 2.83 3 2.96 3.68 4.00 4 3.9 4 3.38 3.54 4 4 3.89 SD Min Max 4.5 0.81 1.249 5 5 0.913 5 0.892 5 1.127 0.66 4.75 5 0.898 5 0.562 0.983 5 5 0.838 4.5 0.84 0.598 4 5 1.111 5 1.321 5 1.27 1.05 5 5 1.16 5 1.36 1.302 5 5 1.228 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 So many respondents (n = 133; 52.8%) chose “Not Important” on all four items, that the median and mode were 1 (See Figure 17). There is limited variation in this scale (SD = .66), as more than four out of five (84.5%) indicated that Physical Motivations were less than Slightly Important (a score of 2) on this scale. 91 Figure 17. Physical Motivation Histogram Personal motivations were also relatively unimportant to the participants, with a score of 1.78. The lowest item was “Relieves me from personal worries” (M = 1.44), followed by “Allows me to rest from my routine occupations” (M = 1.48) and “Strengthens my self- confidence” (M = 1.98). “Allows me to express my personality” stands out in the Personal Motivations category with a mean of 2.23 and a standard deviation of 1.25, both of which are higher than the other items in this sub-category. However, as Figure 18 demonstrates, the majority of scores were quite low. More than three-fifths (61.2%) scored below two on the Personal Motivation scale, with 74 participants choosing “Not Important” for all four items. 92 Figure 18. Personal Motivation Histogram The mean Social Motivation score (2.24) indicates that on the average, participants find Social motivation to be somewhere between Slightly and Moderately Important. There is more range in the items of this sub-category of motivation than of the previous two. The least important motivation in the sample is “Improves my social status,” which has a mean of 1.24. Next is “Enables me to maintain contacts with institutions and organizations” with a mean of 1.92, followed by “Enables me to develop friendships” (M = 2.83) and “Can strengthen my feeling of belonging to my society or community” (M = 2.96). The histogram for the item (Figure 19), shows a consistent number of participants scoring between 1 and 3 (85%). 93 Figure 19. Social Motivation Histogram Finally, participants reported that Cultural Motivations were much more important (M = 3.68) than the other types of motivation, with “Allows me to express my beliefs” and “Strengthens my relationship to my culture or religion” having the highest individual mean scores (3.9 and 3.89, respectively). The other two items in the Cultural Motivation sub-category (“Enhances agreement among my values” and “Improves compatibility between my values and those of my environment”) had higher means than any other items in the motivation scale (3.54 and 3.38, respectively). More than three-quarters (76.8%) of participants had a score 3.0 (Moderately Important) or higher. The histogram of this sub-category (Figure 20) is a distinctly different shape than the previous three motivation histograms. While the others have greater 94 frequencies at the low end of the scale and drop off quickly, this histogram has small frequencies at the low and increases more gradually. Table 15 displays this numerically by showing that the standard deviation of the sub-categories increases with the same pattern as the category means. Figure 20. Cultural Motivation Histogram The top seven individual types of motivation are the following (listed from highest to lowest with sub-category): 1. Allows me to express my beliefs (Cultural) 2. Strengthens my relationship to my culture or religion (Cultural) 3. Enhances agreement among my values (Cultural) 4. Improves compatibility between my values and those of my environment (Cultural) 5. Can strengthen my feeling of belonging to my society or community (Social) 95 6. Enables me to develop friendships (Social) 7. Allows me to express my personality (Personal) Overall, the Cultural motivation sub-category stands out as the major source of motivation for the sample. Social motivations also appear to play a role, but the Personal and Physical motivations do not demonstrate a strong influence on participants’ involvement in SFFC. RESEARCH QUESTION 4 Among these volunteers, is there an association between motivation and well-being? Table 16 reports the bivariate relationship between the motivation subscales and all seven dimensions of well-being. While the relationship between Cultural Motivation and Life Satisfaction, Self-Esteem, and Happiness are approaching significance (p ~ 0.1), no correlations are statistically significant, and all relationships are quite small. When included in a multiple regression model with all motivation sub-categories, Self-Esteem is no longer approaching significance with Cultural Motivation. However, the Cultural Motivation and Life Satisfaction relationship continues to have a p-value approaching significance (p = .10), and the relationship between Cultural Motivation and Happiness becomes significant (ß = .16; p = .04). When control variables are included in the models, neither of these relationships are significant or approaching significance. Table 16. Motivation Bivariate: Spearman's rho Anxiety Depression Life Sat (PWI) Health Self-Esteem Self-Mastery Happiness Motive - Cultural 0.07 0.02 0.10 0.05 0.10 0.01 0.11 Motive - Social 0.10 0.02 -0.04 0.03 0.04 0.02 0.02 Motive - Physical 0.03 -0.02 -0.08 0.06 -0.02 0.05 -0.03 Motive - Personal 0.02 -0.10 -0.08 0.05 -0.06 -0.04 0.02 96 Overall, there is no general indication that motivation is a key factor in predicting well- being among these volunteers. There are no significant bivariate relationships, and any significance in multivariate models diminishes when control variables are included. 97 DISCUSSION This study uses a convenience sample of volunteers from Safe Families for Children (SFFC), a program across the United States, Canada, and the United Kingdom that utilizes volunteer Host Families to care for children who are at risk of maltreatment. The goals of SFFC are to keep children safe during a family crisis such as homelessness, hospitalization, or domestic violence in an effort to prevent child abuse and/or neglect; support, and stabilize families in crisis by surrounding them with caring, compassionate community; and reunite families and reduce the number of children entering the child welfare system (safe-famlies.org). In March and April of 2019, 790 Host Families were invited to participate in this study, and 38% responded. The participants tend towards being older, wealthier, and highly educated, and most have dependents still in the home. They are religious, which is an expected trait from a religious organization like SFFC. The group is very financially generous and very engaged in diverse volunteer roles. Below, their well-being, volunteerism, and motivation are discussed, followed by a discussion of the main findings of this study. WELL-BEING Participants in the sample report a very high degree of well-being. They report high Life Satisfaction, Self-Esteem, and Self-Mastery, while reporting very low levels of Anxiety and Depression. They are physically healthy and happy. This is consistent with the body of literature that has demonstrated a strong link between volunteerism and well-being. Because of the observational, cross-sectional design of this study, it is not feasible to determine if this high degree of well-being was affected by volunteering or an antecedent to volunteering or both. The lack of a comparison or control group in this study’s design limits analysis that can distinguish what influence the Host Family volunteer role might have on participants’ high well-being 98 scores. This is in contrast to several previous studies that used nationally representative samples, which resulted in larger variation in reported well-being and volunteer engagement (Herzog & Price, 2016; House, 2014; Morrow-Howell et al., 1999; Musick et al., 1999; Musick & Wilson, 2003; Smith & Davidson, 2014; Thoits & Hewitt, 2001; Van Willigen, 2000). Figures 10-16 demonstrate the high degree of well-being in the sample, and also display the lack of variation in the well-being variables. For example, the Happiness histogram (Figure 16) shows that nearly the entire sample selected Happy (4) or Very Happy (5) to describe their degree of happiness, and both the Depression and Anxiety histograms show nearly all participants at clinically insignificant levels (Figures 11-12). The Personal Well-Being Index (PWI), which measure Life Satisfaction across 8 different domains, uses an 11-point scale that could conceivably measure Life Satisfaction with a higher degree of sensitivity than the common four- or five-point scales. However, the participants in this study reported a mean of 8.61, with 87% scoring at eight or more on the scale. This is higher than the substantial body of literature that demonstrates that the average PWI score in Western nations is approximately 7.5 (Cummins, 2008). VOLUNTEERISM An important part of understanding this study and interpreting its results is to understand that, by design, all participants are volunteers Safe Families for Children. This represents a risk of selection bias (see Limitations below). There is no one in this study who does not have the role or status of “volunteer”. As several previous studies using the framework of Role Theory have found, it is often the role, status, or identity of being a volunteer that is the key predictor of increases in well-being (Connolly & O’shea, 2015; Matz-Costa et al., 2014; Morrow-Howell et al., 2003; Musick et al., 1999; Thoits, 2012; Thoits & Hewitt, 2001; Van Willigen, 2000). 99 Not only are all participants volunteers for SFFC, but nearly 90% have other volunteer roles as well. Almost half have other volunteer roles with at-risk children. This is an important limitation in this study: While this study focused on volunteer intensity and consistency, there is no variance on the status/identity of being a volunteer, which is one likely explanation for this study’s inability to detect relationships between volunteering and well-being (see Volunteering and Well-Being below). Several of the most important previous studies have used nationally representative samples, and have therefore been able to compare the differences between those with the identity of “volunteer” and those without (Herzog & Price, 2016; House, 2014; Morrow- Howell et al., 1999; Musick et al., 1999; Musick & Wilson, 2003; Smith & Davidson, 2014; Thoits & Hewitt, 2001; Van Willigen, 2000). One somewhat comparable study by Thoits (2012) drew a sample of only volunteers. Thoits’ sample had a similarly high degree of well-being with low variability. Interestingly, she also found very little or no relationship between volunteer intensity and well-being. A unique aspect of her study was that she measured actual hours spent volunteering, and also asked volunteers how much time they perceived that they spent volunteering. While participants averaged 3.2 hours per week volunteering, they perceived that they spent 5.4 hours per week volunteering. When Thoits tested whether perceived volunteering time was related to well- being, she found that perceived time was related to happiness, life satisfaction, self-esteem, and self-mastery. Although the age range was similar to this study (41-91 in Thoits, 42-93 in this study), Thoits’ sample was older on average (M = 72.9 vs. M = 61.3) and had less variance (SD = 8.8 vs. SD = 10.8). A point made frequently in the literature is that older populations tend to have a higher variability in well-being as a function of the aging process (Anderson et al., 2014). It is also conceivable that the well-being of former cardiac patients had substantial room for 100 improvement based on their physical health, which would allow for easy detection of changes in well-being. VOLUNTEERING AND WELL-BEING Research Question 2 for this study investigated whether there is a relationship between volunteering and well-being among those who volunteer with children at risk of maltreatment. Contrary to most previous studies investigating the relationship between general volunteering and well-being, the results of this study do not clearly answer this question. The primary finding that showed a relationship between volunteerism and well-being in this study is the Life Satisfaction mean difference among those who are active and inactive Host Family volunteers. Those who had been active in the preceding twelve months showed a significant increase of 3.8% in Life Satisfaction scores when compared to those who were inactive in the preceding twelve months. While bivariate analysis demonstrated some limited associations between volunteering and Life Satisfaction and Self-Mastery, these associations did not persist in multivariate linear regression or curvilinear modeling. Furthermore, these relationships existed only when the number of children hosted in the past year or number of children hosting at the time of data collection were used as the volunteerism variables. As mentioned above, these measures are likely not as relevant to the concept of volunteer intensity as other measures of volunteerism. An important note is that having a volunteer role other than being a Host Family in SFFC did not have a significant relationship with any of the well-being variables measured in this study (See Table 13). This may support Role Identity Theory, which posits that it is the identity of being a volunteer that is associated with well-being, not the intensity of the volunteering (Matz- 101 Costa et al., 2014; Morrow-Howell et al., 2003; Musick et al., 1999; Musick & Wilson, 2003; Piliavin & Siegl, 2007; Van Willigen, 2000). The methodological aspect of this study that is unique in the literature is the focus on one particular type of volunteerism. Where nearly all other studies included any type of volunteer role, this study only included people who volunteered in the role of Host Families within the organizations, Safe Families for Children. This role is an unusual type of volunteer role, because of its demand on volunteers. While many roles require volunteers to donate relatively small, episodic amounts of time (e.g., coaching a youth sports team, tutoring, cleaning a park, or teaching a community class), the Host Families role is one in which the volunteers become caregivers for one or more children twenty- four hours per day, every day during the duration of their hosting time. The demand on a volunteer’s time is greater and challenging to operationalize in a comparable way. Other studies (MacIlvaine et al., 2014; Musick & Wilson, 2003; Sneed & Cohen, 2013) operationalize volunteer intensity as hours per week or year with the highest maximum measurements being the equivalent of ten hours per week. Those scales would be irrelevant for Host Families who volunteer for all the hours in a week. This reality required volunteer intensity to be measured in days rather than the typical unit of hours. This disparity deserves to be highlighted, because it represents an important diversion from what is typical in the literature. What should be emphasized is not just that the operationalization is unique in this study, but that the volunteer role itself is unique. This, of course, is why the analysis in this study included a curvilinear model. It is logical to be curious about whether the high time-demand of this role would have a meaningful and significant effect on the volunteer’s well-being. If the concept of pathological altruism, also known as the Burden Assumption (van Campen et al., 2013) and Role Strain (Son 102 & Wilson, 2012), exists in volunteering roles, it is rational to assume it might exist among these volunteers. The very interesting finding in this study is that increased volunteer intensity does not appear to be related to decreases in any of the well-being concepts included in this study. Furthermore, the role of Host Family could be logically considered to be an emotionally taxing volunteer role. Afterall, the children who are hosted are coming from families who are in crisis, and likely bring some of the emotion of that crisis into the host home. It is also logical to assume that the experience of being separated from parents is challenging to the children and that the emotions felt may result in behaviors that demand attention and emotional endurance from the Host Families. In this study, no information was collected on the children who were hosted, so this claim cannot be substantiated in the data, nevertheless, it is not irrational to expect the emotional intensity of the Host Family role to have some negative impact on the well-being of the volunteers. As mentioned above, there is not a negative relationship between volunteer intensity and well-being in this study. In fact, the opposite appears to be true. When the mean well-being scores of active Host Families (active within the past twelve months) were compared to the mean well-being scores of non-active Host Families, the active families had a significantly higher level of Life Satisfaction than the non-active families. MOTIVATION The sample in this study reports being motivated primarily by Cultural and Social motivations. Overall, Physical and Personal motivations were not reported as being important. As noted above, the means and standard deviations of the four sub-types of motivation increased in the same order (from lowest to highest: Physical, Personal, Social, and Cultural), showing that the sample is generally in greatest agreement about the least important motivations and in least agreement about the most important motivations. This makes sense given the physical and 103 financial burden that is assumed by the Host Families. It would, of course, be counter-intuitive to engage in this type of volunteer role in order to satisfy a desire to improve one’s economic situation or sense of physical security. It is quite logical that there is variability among Social motivations. For example, while some engage in volunteering hoping to improve their social connectedness, others may be so motivated by their beliefs (Cultural motivation) that social connectedness plays a smaller role. MOTIVATION AND WELL-BEING Research Question 4 investigated whether there was a relationship between motivation and well-being, which had been explored in previous studies (Kwok et al., 2013; Shye, 2010b; Stukas et al., 2016). Although initial analyses suggested a small relationship between Cultural Motivation and some aspects of well-being, no relationships remained significant in full multivariate regression models. To some degree, this can likely be explained by the lack of variability in the motivation subscales as well as the well-being variables used in this study. LIMITATIONS In addition to limitations noted in the discussion of key variables above, the risk of bias in self-reported measures, selection bias, diversity concerns, attrition, and the challenge of determining causation via cross-sectional data should also be considered. Other limitations of note are the way in which inconsistencies in operationalization and measurement in the literature limit inter-study comparisons and the complexities of missing data. All measures are self-reported in this study, which involves an inherent risk to reliability and social desirability bias. It is conceivable that participants could want to appear happier and healthier than they feel. However, the literature is in agreement that self-reported data is the 104 most reliable for well-being variables, as it would not be logical to try to observe concepts such as happiness and Life Satisfaction in any other way. The greater risk created by self-reporting in this study exists for data regarding volunteer roles, which rely on participants’ memory, which represents a possible source of bias. Social desirability bias also introduces a risk when measuring motivation, given that participants may be inclined to describe their motivations in more altruistic terms, rather than in ways that may appear to be motivated by self-interest. This study uses a convenience sample, which introduces a selection bias and an overall risk to the external validity of the study. Additionally, the sample is drawn from a single organization, SFFC, because of its unique approach to preventing children from entering the foster care system. The slightly low response rate (38%) in this study may also contribute to a selection bias. Because SFFC was unable to provide demographic descriptions of their Host Families, it was not possible to compare the sample in this study with overall numbers. However, the sample size in this study is large enough to ensure a 4.44% margin of error and a confidence level of 95% (https://www.qualtrics.com/blog/calculating-sample-size/), which suggests the sample is adequately representative of the sample frame. Future research should compare results with others involved in caring for children, such as foster parents. This would also allow for a valid distinction to be made between the well-being of those who volunteer in SFFC and other types of volunteers. The invitation to participate in this study asked only one member of each household to participate in the survey. It should be noted that nearly 90% of participants in the study identified as female. Ninety-three percent of participants reported being married, and one limitation of the study is that it is unclear about whom the participants were responding. While the well-being, volunteering, and motivation questions are likely accurate because of their wording and the 105 concepts being measured, there is no way of ensuring that participants were not responding on behalf of their partner or household. This study is also limited by a lack of diversity. Both the sampling frame and the sample for this study are not representative of the broader U.S. population. SFFC Host Family volunteers lack religious and racial diversity. This should be considered when the results of this study are compared with the broader literature. Researchers and practitioners reflecting on the implications of this study must take into account that SFFC recruits nearly all of its volunteers from Christian churches, which could limit the relevance of findings in this study. Attrition is a minor limitation in the study. Analysis of the survey process through Qualtrics demonstrated that there were two points in the survey where approximately five percent of respondents discontinued their participation in the survey. The first was after the scales measuring depression and anxiety. The second was after the question regarding whether a participant had had at least one placement in the past twelve months. Future research should consider moving questions that could illicit an emotional response (for example, about depression and anxiety) to the end of the survey. The literature has addressed the causation limitation through the body of evidence, experimental/longitudinal designs, models that include known covariates, and theory. The overwhelming conclusion has been that volunteering causes increases in well-being and that well-being may also precipitate volunteering to some degree (Smith & Davidson, 2014). However, the design of this study does not allow for speculation on whether the high degree of well-being in the sample can be attributed to volunteerism. The presence of a robust list of control variables and a sensitivity analysis were planned to account for and explain this 106 limitation. Based on the limited relationship between volunteerism and well-being found in this study, the sensitivity analysis was not applicable. There is little consistency in measurement scales used in the literature that can be employed as a specific reference point. Comparisons between studies can be made generally, but cannot be compared quantitatively. To account for this limitation, the framework of a seminal study (Thoits, 2012) was used to inform the operationalization and measurement. This will allow some specific comparisons to be made and offers a possible and reliable framework for future studies. Missing data prohibited the use of Structural Equation Modeling, which would have more robustly accounted for the associations between all dependent, independent, and control variables. Factor scores were computed for regression analyses (see Appendix E) to offer an alternative representation of the data and to include the covariance between well-being factors that have typically not been accounted for in previous studies. Data were deleted casewise for Confirmatory Factor Analysis, which allowed for the greatest amount of raw data to be considered for reliability. Regressions were run by deleting cases listwise to ensure that all variables in each model used the same sample of participants. 107 IMPLICATIONS AND CONCLUSION Within its limitations, this study can offer several important suggestions for future research and practice. The suggestions for future research include the importance of a control group or time element and the use of instruments that can detect all levels of volunteer intensity and well-being. Additionally, this study makes a contribution to the reliability of measuring well-being across several dimensions. Other implications include serving as a foundation for future qualitative studies and for additional research into similar roles such as foster parenting. The most pragmatic next step in the research is to conduct a similar study with foster parents. Implications for practice include suggestions for those who recruit, manage, and train volunteers as well as implications for the child welfare sector. IMPLICATIONS FOR FUTURE RESEARCH Future research should strive for an experimental or quasi-experimental design in order to include a control group of non-volunteers that will allow the researchers to detect changes in well-being between the control and experimental groups. Including non-volunteers would theoretically allow for the important distinction in identity/role as mentioned in the Limitations section, and result in an increase to external validity. However, it may be difficult to randomly select people into the groups of volunteer and non-volunteer. Because the independent variable of interest is volunteering, the non-volunteer group would need to be composed of people who were uninterested in volunteering. This would be inconceivable in a sample of people who had volunteered to take part in a study on volunteering. Based on Role Identity Theory, a theory that claims that it is the status or role of being a volunteer that effects a person’s well-being (Matz- Costa et al., 2014; Morrow-Howell et al., 2003; Musick et al., 1999; Musick & Wilson, 2003; Piliavin & Siegl, 2007; Van Willigen, 2000), it could be argued that even the people who were 108 randomly selected into the non-volunteer group could retain the identity of a volunteer and thus bias the results. This problem could potentially be solved by either a retrospective design or the inclusion of a robust set of covariates. While a retrospective design could not guarantee random selection into volunteer and non-volunteer groups, it could collect the case histories of people who identify as volunteers or non-volunteers and compare the differences between groups. Another approach for researchers wishing to increase external validity could be the inclusion of a robust set of covariates that would attempt to account for unmeasured bias. This study provides a good template for possible covariates for future studies. A sensitivity analysis such as the approach by Frank, Maroulis, Duong, and Kelcey (2013), which quantifies what it would take to nullify a causal inference, would further help to support possible generalizations. A longitudinal or panel study would be an ideal design for any future studies hoping to accurately identify the effect of volunteering on well-being. Future research could collect baseline data on people who have expressed interest in a particular type of volunteer role, and then collect data on multiple future occasions in order to measure the differences between baseline data and data collected after volunteering has been sustained for a substantial amount of time. Along with the inclusion of a robust set of covariates, the addition of a time element would help to increase the probability that any changes in well-being were due to the inclusion of the new volunteer role. This would provide an opportunity to isolate the change in volunteer identity/role as well as volunteer intensity and consistency as predictor variables for well-being. Collecting data over the course of several years would increase the chances of being able to detect Pathological Altruism, which describes a type of generosity or altruism that overburdens the giver in such a way as to become unhealthy or detrimental to well-being (Oakley, 2011, 2013, 2014; Rubin, 2014; Smith, 2015; Smith & Davidson, 2014). Additional 109 findings could result from data collected from those who expressed interest in volunteering, but never began any volunteer activity or those who began, but later dropped out. These people would offer reference points for comparison to those who engaged or more fully engaged in volunteer activities. It would be important to note, however, that their initial desire to volunteer would differentiate them from people who never had interest in volunteering. It is possible that even the initial desire to volunteer would have some positive impact on well-being. Future studies should ensure that their measurement instruments can detect changes in primary variables at a fine-grain level. This study provides a template for those wishing to ensure that all variations in volunteer intensity are taken into account. As mentioned above, previous studies have tended to group people with high and very high levels of volunteer intensity into one group, and have risked biasing their results. The sample in this study demonstrates that it is possible for many people to volunteer and very high levels and in multiple volunteer roles. This study does not provide a good template for fine-grain measurement of well-being variables such as happiness and physical health. The instruments used to operationalize these concepts were reliable, but did not use scales that could detect small differences in well-being. While it could be argued that this is not necessary, future researchers should consider that people who volunteer are likely to have a high degree of well-being and that instruments that can detect variations among the higher levels of well-being may provide more precise results. The Confirmatory Factor Analysis used in this study is a unique contribution to the literature. It demonstrates a very reliable method for operationalizing and measuring seven well- being dimensions as well as four dimensions of motivation. The analyses and results in this 110 study go beyond typical measurements of reliability and offer strong evidence for reliably measuring well-being in future studies. The findings of this study are not in complete agreement with the vast majority of the literature, which has consistently found significant and meaningful relationships between volunteering and many facets of well-being. As mentioned above, the most important difference in this study is the focus on one type of volunteering with greater-than-usual demands on time and emotional energy. A qualitative inquiry may be an important next step to better understanding the well-being and motivations of people volunteering in the Host Family role. This study used reliable and valid instruments for measuring seven different well-being concepts and four different types of motivations, but it did not have the ability to explore the meaning of the well-being or motivational concepts to the volunteers. It is quite likely that the richness and triangulation that a qualitative study could provide would open up new pathways toward understanding the experience of Host Families and other similar types of volunteer roles. The role that may be most likely to be compared to the Host Family role is the role of foster parent. Similar to Host Families, foster parents accept children into their homes and families. Although the role is not technically a volunteer role (foster parents in the United States receive some resources and compensation), the role is not considered or compensated as a career or job. Because Safe Families for Children offers itself as an alternative to or prevention from foster care, it is likely that the results of this study may become more clear or important if they can be compared to the results of a similar study among foster parents. Safe Families for Children is currently engaged in studies that measure the outcomes for children when compared to foster care, but there is not a current study that compares the outcomes for the Host Families and foster parents. The most pragmatic next step in future research is to complete a study that 111 investigates the same research questions among foster parents. This will help to provide insight on the relationship between generosity and well-being when vulnerable children are the beneficiaries of the generosity. It will also provide a population that could include a more diverse sample. It will also shed light on the experiences of foster parents and impact those experiences have on their well-being. Those who recruit for both roles may find valuable information about the well-being and motivation of those they lead, train, and organize. IMPLICATIONS FOR PRACTICE While the results of this study found limited evidence of the volunteerism/well-being relationship several implications for practice can be offered. In particular, several items are important for those who recruit, manage, and train volunteers. As mentioned above, the sample in this study lacks racial and religious diversity. As SFFC seeks to provide quality care for children and their families, it should seek to recruit a more diverse pool of Host Families that will better reflect the reality that children in the child welfare system are disproportionally people of color (Child Welfare Information Gateway, 2016a). SFFC could potentially increase the reach and effectiveness of its work by from recruiting from communities of other religions and in ways that would attract people with no religious affiliation. It is important to notice the high volunteering rates among this sample. They are involved in multiple volunteer roles and opportunities. Volunteer managers should take this into account when engaging and scheduling volunteers. The fact that the participants in this sample are involved in multiple volunteer opportunities demonstrates that they either have extra time to volunteer or the make time for volunteering. Volunteer managers may be able to help these types of people focus their time in order to increase productivity and efficiency. 112 The volunteers in this study are middle-age and older adults, who are well-educated, and have at least an average socioeconomic status. Volunteer recruiters can take note of this demographic and develop specific strategies for reaching this group. Conversely, volunteer recruiters for Safe Families for Children may want to further evaluate whether there are better ways to engage young and middle-age adults. At first glance, the assumption could be that people are waiting until they are “empty-nesters” to begin in the Role as Host Families, but the results of this study show that nearly three-quarters of Host Families have at least one dependent child in the home and about one-third have a dependent adult in the home. Further evaluation may help SFFC to determine who is responding best to their messaging and how their messaging could shift to target groups that are less represented in their current volunteer ranks. Those who recruit for similar roles, such as foster parents, should take note of the types of people who are engaged in volunteering with Safe Families for Children. They are much older than the typical foster parent. They are wealthier and more educated. It may be beneficial to develop strategies that can help to specifically invite this demographic to consider becoming foster parents. The generally high degree of well-being among this group should also be considered as recruiters develop messaging about the experience of being a foster parent. It should be noted, however, that there are some distinct differences between foster parents and Host Families. Most obvious is the financial support that foster parents are given. A difference that is perhaps even more important is the experiences of the children who are involved in foster care. Compared to the children engaged in SFFC, some foster children are the victims of intense maltreatment, and may take a heavier emotional toll on foster parents. Another important aspect for volunteer managers to understand from this study is the importance of social support. Social support was very frequently related to well-being variables, 113 and although social support was not the focus of this study, it does provide evidence for the importance of social support in the lives of volunteers. This may be particularly true for those who volunteer in roles that can be emotionally taxing. In fact, social network theory explains the effect of volunteering on well-being as an effect of increased social support (Musick et al., 1999; Musick & Wilson, 2003). Agencies and organizations who depend on the generosity of volunteers should consider the importance of using resources to develop strategies that can weave social support into the volunteer experience. This could play an important role in the well-being of their volunteers, and may be a key factor for inspiring quality volunteer work, consistency, commitment, and retention. Building social support into agencies and programs should be a high priority for anyone that oversees or recruits people to work with children at risk of maltreatment. Volunteer managers can also learn about motivation in this study. This sample of volunteers who work with children at risk of maltreatment are not concerned about their finances, safety, or trying something new. They are also not motivated by improving their health, social status, or improving their sense of control in life. Instead, they are motivated by a chance to belong and build friendships. Ultimately, they are most motivated by their beliefs and values. Volunteer managers in child welfare programs can use these findings to strategize about how to stimulate interest among potential volunteers, focusing more on beliefs systems and connection than on reducing the appearance of physical or personal risk or cost. IMPLICATIONS FOR POLICY Two-thirds of the participants in this study who hosted at least one child in the past twelve months reported that they were hosting a child for two months or less in that time period. The mean hosting days in the past twelve months was 56.15 (Mdn = 35.5). This could be one 114 explanation for why this type of potentially emotionally intense volunteer role does not decrease the well-being of the volunteers: they are engaged for only a small proportion of each year. The length of hostings is considerably shorter than foster care placements, which typically last longer than a year (M = 20.4 months; Mdn = 12.6 months) (Child Welfare Information Gateway, 2016b). Additionally, the good outcomes of SFFC may also be another factor that explains the high well-being of people who volunteer as Host Families. According to SFFC, more than 90% of the children return home after their stay in SFFC. It is plausible that this outcome influences their well-being. This may not be the case for foster parents. Policymakers should note these differences between SFFC and the child welfare system. The Family First Prevention Services Act is a good step in placing an emphasis in helping families meet their needs in a crisis. Providing support for Safe Families for Children or other programs that depend on volunteers in the funds designated by this act could be a cost-effective method for increasing positive outcomes for children and families. While this study does not explore those particular outcomes, it does demonstrate that the SFFC model utilizes volunteers in a way that does not require them to sacrifice their own well-being. This could be an important point as policymakers attempt to reform and innovate the child welfare system to meet the increasing demands it faces. IMPLICATIONS FOR EDUCATION Child Welfare is a staple of social work education. It is important for all social workers to have an understanding of the child welfare system and its structures, strengths, and weaknesses. Because it will likely be social workers who will have the opportunity and challenge to address the problems of the child welfare system, it is important for their education 115 to include the concept of being innovative in child welfare. This study can help to draw attention to current innovations in the way society addresses the needs of children at risk of maltreatment. Another implication for education is the way in which this study can help students think about the “who” of child welfare. They must be able to consider who will do the frontline work of caring for children at risk of maltreatment, and they need to demonstrate an understanding of how that work will impact the well-being of those people. This study also offers an example for macro social work students as they consider the different forms that prevention services can have. SFFC demonstrates how prevention can be implemented through a volunteer-driven program, and how the volunteers’ well-being is an important part of any program that utilizes volunteers. Finally, social work students should be introduced to the relationship between generosity/volunteerism and well-being. There are likely many areas of social work practice that can benefit from innovative utilization of a volunteer workforce or other acts of generosity. Social work depends on and works through many government systems and regulations, and this child welfare study can be an example for how generosity of individuals or groups of ordinary citizens may be able to complement, support, or enhance the work done through the usual systems. CONCLUSION An aim of this study was to apply the Science of Generosity to some of the biggest problems in child welfare: insufficient foster homes, questionable quality of foster care placements, and the threat of negative and life-altering outcomes for children. Safe Families for Children offers a novel solution to these problems, and although several studies on outcomes for the children are presently in process, there are currently no studies investigating the outcomes for 116 the volunteers. Because SFFC utilizes volunteers to care for children at risk of entering the child welfare system, this study built upon relevant literature that sought to understand the relationship between generous behavior (such as volunteerism) and well-being. The literature that investigates the relationship between generosity and well-being has gained a considerable momentum in the past two decades, including many studies that have investigated the relationship between volunteering (one common form of generosity) and well- being. These studies are nearly unanimous in their findings that volunteering is positively related to (and probably causes) increase well-being. However, the body of literature that focus specifically on volunteerism and well-being among adults of all ages, is somewhat limited and more recent. Few previous studies have given attention to a specific type of volunteer role (Thoits, 2012). This study is positioned to be one of the first that intentionally studies a particularly type of volunteer role. However, the aim of the study was not simply to build upon previous research and to generate new knowledge. Instead, it also sought to draw attention to the need for innovation in child welfare by carefully studying those who voluntarily care for children at risk of maltreatment. It also sought to increase engagement in the field of Social Work with the study of volunteers, their well-being, and their motivations. This study investigated the well-being and motivation of adults who volunteer with children who are at risk of entering foster care. It sought to investigate whether there was a relationship between this type of volunteering and seven different dimensions of well-being: Happiness, Physical Health, Self-Mastery, Self-Esteem, Life Satisfaction, Depression, and Anxiety. It also investigated whether an association existed between well-being and four types of motivation: Physical, Personal, Social, and Cultural. 117 The sample was drawn from Safe Families for Children, which is a faith-based organization that places children at risk of maltreatment into the homes of volunteers (Host Families) to prevent them from going into the foster care system. Participants are a highly religious and educated group. They demonstrate a high degree and diversity of generosity. They tend to be older and to have higher than average incomes. Among this group there is limited evidence of significant relationships between volunteering and well-being dimensions. There is also limited evidence of significant relationships between motivation and well-being. However, an important finding of this study is that despite the high time and emotional demands of doing this type of volunteer work, there is no apparent decrease or drop-off in the well-being of the volunteers. Rather, they are happy and physically healthy. They report very low levels of anxiety and depression, and they demonstrate a high degree of Self-Esteem, Self-Mastery, and Life Satisfaction. While some may believe that working with children at risk of maltreatment is stressful and may result in a decrease in well- being (Tyebjee, 2003), the results of this study suggest that it is not the case for Host Families from Safe Families for Children. Social Support appears to be an important covariate to many of these well-being dimensions. This is consistent with social network theory, which explains the relationship between volunteering and well-being as a function of the increased social support that is intrinsic within the volunteering experience (Borgonovi, 2008; Musick et al., 1999; Musick & Wilson, 2003; Piliavin & Siegl, 2007). Participants report being motivated by their beliefs, values, and religion, and somewhat by social factors. However, they are not seeking to satisfy needs for personal gain, safety, or security from their volunteer work. Regardless of their importance, none of these types of 118 motivations were able to predict well-being. Similar to the relationship between well-being and volunteerism, the lack of variability reduces the ability to detect significant relationships. While this study was unable to demonstrate that their volunteering role increased well- being, it was able to demonstrate that the people who volunteer as Host Families have a very high degree of well-being. This reality persists despite the high demand on time and emotional energy, two factors that could imply a decrease in overall well-being. While the average tenure of a foster parent is 8-14 months (Gibbs, 2005), the sample in this study had been SFFC volunteers for a mean of more than two and a half years, raising the question of whether the experience of being a Host Family is more desirable or sustainable than foster parenting. Future research must answer these questions, but the larger picture is clear: Host Family volunteers for Safe Families for Children have a consistently and impressively high level of Happiness, Physical Health, Life Satisfaction, Self-Esteem, Self-Mastery, and Psychological Well-Being. 119 APPENDICES 120 APPENDIX A: SURVEY PROTOCOL PROGRAM INFORMATION 1. Please rate your level of agreement with the following statements (5-point Likert Scale from Strongly Disagree to Strongly Agree) a. The training I received to be a SFFC Host Family was adequate. b. The support I receive from SFFC staff is adequate. c. The support I receive from my church is adequate. d. I receive an appropriate quality and quantity of communication from SFFC staff e. I feel empowered by SFFC to fulfill my role as a SFFC Host Family f. I have an adequate amount of authority to make the decisions I need to make in order to fulfill my role as a SFFC Host Family. g. I feel overwhelmed by my role as a SFFC Host Family. CONTROLS 2. Caregiving burden a. NOT including children from Safe Families, how many dependent children (age 17 or less) live with you or are in your care on a regular basis (for example, biological, adopted, foster, kinship care, etc.)? b. How many adults (18+) do you care for on a regular basis? c. Is caregiving a part of your job? 3. Are you currently a foster parent? 4. Other generous behavior a. Do you donate money to charity or religious institutions? i. Approximately how much money have you donated to charity or religious institutions in the past 12 months? Total amount: ___ b. How many volunteer roles do you have other than Safe Families for Children? c. How many of the above roles are with children who could be considered vulnerable or at-risk? d. NOT INCLUDING your role as a Host Family in Safe Families for Children, how many total hours in the past 12 months have you spent volunteering for roles? WELL-BEING 5. (Life Satisfaction: Personal Well-being Index) [end-defined scale ranging from No Satisfaction at All (0) to Completely Satisfied (10)] a. Thinking about your own life and personal circumstances, how satisfied are you with your life as a whole? b. How satisfied are you with your standard of living? c. How satisfied are you with your health? d. How satisfied are you with what you are achieving in life? e. How satisfied are you with your personal relationships? f. How satisfied are you with how safe you feel? g. How satisfied are you with feeling part of your community? 121 h. How satisfied are you with your future security? i. How satisfied are you with your spirituality or religion? 6. Physical Health Index a. How would you rate your health in general? i. 5 = excellent, 4 = very good, 3 = good, 2 = fair, 1 = poor. b. Using the scale below, which face comes closest to how you feel your health is today? c. How satisfied are you with your present health in general? Would you say you are d. Now thinking about your physical health which includes physical illness or injury, for how many days during the past 30 days was your physical health not good? 7. Psychological Health PHQ-9 and GAD-7 Over the last 2 weeks, how often have you been bothered by any of the following problems? i. Pretty well satisfied ii. More or less satisfied iii. Not satisfied at all i. Range = 0-30 i. 0 -Not at all ii. 1 - Several days iii. 2 - More than half the days iv. 3- Nearly every day b. Little interest or pleasure in doing things c. Feeling down, depressed, or hopeless d. Trouble falling or staying asleep, or sleeping too much e. Feeling tired or having little energy f. Poor appetite or overeating g. Feeling bad about yourself — or that you are a failure or have let yourself or your family down television h. Trouble concentrating on things, such as reading the newspaper or watching i. Moving or speaking so slowly that other people could have noticed? Or the opposite — being so fidgety or restless that you have been moving around a lot more than usual j. Thoughts that you would be better off dead or of hurting yourself in some way k. Feeling nervous, anxious or on edge l. Not being able to stop or control worrying m. Worrying too much about different things n. Trouble relaxing o. Being so restless that it is hard to sit still p. Becoming easily annoyed or irritable q. Feeling afraid as if something awful might happen 122 8. Pearlin’s Self-Mastery Scale (Pearlin et al., 1981; Pearlin & Schooler, 1978) (1) Strongly Disagree (2) Disagree (3) Agree (4) Strongly Agree. a. There is really no way I can solve some of the problems I have. RC b. Sometimes I feel that I’m being pushed around in life. RC c. I have little control over the things that happen to me. RC d. I can do just about anything I really set my mind to. e. I often feel helpless in dealing with the problems of life. RC f. What happens to me in the future mostly depends on me. g. There is little I can do to change many of the important things in my life. RC 9. ROSENBERG SELF-ESTEEM SCALE (Rosenberg, 1965). Below is a list of statements dealing with your general feelings about yourself. Please indicate how strongly you agree or disagree with each statement. (1) Strongly Disagree (2) Disagree (3) Agree (4) Strongly Agree. a. On the whole, I am satisfied with myself. b. At times I think I am no good at all. RC c. I feel that I have a number of good qualities. d. I am able to do things as well as most other people. e. I feel I do not have much to be proud of. RC f. I certainly feel useless at times. RC g. I feel that I'm a person of worth, at least on an equal plane with others. h. I wish I could have more respect for myself. RC i. All in all, I am inclined to feel that I am a failure. RC j. I take a positive attitude toward myself. 10. When you consider everything about your present life, how would you describe yourself? HAPPINESS a. 1 = unhappy b. 2 = not very happy c. 3 = neither happy nor unhappy d. 4 = happy e. 5 = very happy GENEROSITY – Experimental group only a. Yes b. No c. 0 d. 1 e. 2 f. 3 or more home? 123 11. Have you had at least one child placed in your home in the past 12 months? 12. How many children are placed in your house now? 13. How many placements did you have in the past 12 months? 14. How many days in the past 12 months have you had at least one child placed in your 15. Which option best describes your experience in the past 12 months? The placements I had were k. more stressful than the typical placement l. equally stressful as the typical placement m. less stressful than the typical placement 16. How long have you been a host family for Safe Families? n. Range = 0-10+years 17. Please select the statement that most accurately describes your decision to become a Host a. I primarily initiated our involvement as a Host Family b. My partner/spouse primarily initiated our involvement as a Host Family c. My partner and I mutually initiated our involvement as a Host Family d. I do not have a spouse/partner Family MOTIVATION a. I was placed in foster care/kinship care b. I was adopted c. I was a ward of the court d. I am a veteran e. My partner/spouse is a veteran f. I am homeless 124 18. Systemic Quality of Life (Shye, 2010). Below is a list of motivations for volunteering. Please rate their importance for your work with Safe Families. The 16 needs can be grouped in four categories of well-being, which are used for analysis: personal (items 1- 4), physical (items 5-8), social (items 9-12), or cultural (items 13-16). • Very Important • Important • Moderately Important • Slightly Important • Not Important a. Allows me to express my personality b. Allows me to rest from their routine occupations c. Relieves me from personal worries d. Strengthens my self confidence e. Enables me to express my power and control over the environment f. Can improve my economic condition g. May contribute to my health h. Strengthens my feelings of physical security i. j. Enables me to maintain contacts with institutions and organizations k. Enables me to develop friendships l. Can strengthen my feeling of belonging to my society or community m. Allows me to express my beliefs n. Improves compatibility between my values and those of my environment o. Enhances agreement among my values p. Strengthens my relationship to my culture or religion Improves my social status BACKGROUND 19. Which of the following describe your personal life experience? Select all that apply 21. What is the highest level of education you have completed? g. I was previously homeless h. None of the above 20. What is your annual household income? a. Less than $19,999 b. $20,000 to $39,999 c. $40,000 to $59,999 d. $60,000 to $79,999 e. $80,000 to $99,999 f. $100,000 to $119,999 g. $120,000 to $139,999 h. $140,000 or more a. Did not finish high school b. High School or GED c. Some college d. Bachelor’s degree e. Graduate degree baptisms, and funerals)? a. More than once a week b. Once a week c. Three times a month d. Twice a month e. Once a month f. Several times a year g. One to two times a year h. Never (Smith, Herzog, & Beyerlein, 2010) 23. What is your marital status? a. Married b. Not married a. Working full time b. Working part time c. Unemployed d. Retired 22. How often, if ever, do you normally attend religious services (not counting weddings, 24. Select the answers that best describe your current employment status. 25. Social Network (Lubben Social Network Scale- 6) None, One, Two, Three or four, Five thru eight, Nine or more Family: Considering the people to whom you are related by birth, marriage, adoption, etc... a. How many relatives do you see or hear from at least once a month? b. How many relatives do you feel at ease with that you can talk about private c. How many relatives do you feel close to such that you could call on them for matters? help? Friends: Considering all of your friends including those who live in your 125 Neighborhood matters? d. How many of your friends do you see or heard from at least once a month? e. How many friends do you feel at ease with that you can talk about private f. How many friends do you feel close to such that you could call on them for help? 30. Which category best describes your race? Select all that apply. 26. In which state do you reside? 27. In which city is your SFFC chapter located? 28. What year were you born? 29. What is your gender identity? a. Male b. Female c. Other (please specify) d. African American or Black e. White or Caucasian f. Asian g. Native Hawaiian/Pacific Islander h. American Indian/Alaskan Native i. Hispanic j. Other (please specify) -----------------------End of Survey---------------------------- 126 APPENDIX B: INVITATION EMAIL Dear Safe Families Host Families, Thank you for the work you do to care for vulnerable children! Here’s an easy way to continue to support them and the work of Safe Families by participating in a brief, 10-minute survey. As a Host Family, you are invited to share your voice by participating in a research project called “Volunteerism In Child Welfare: The Well-Being Of Adults Who Volunteer With At-Risk Children.” The study aims to better understand your experience as a Host Family and to learn how your volunteering relates to your overall sense of well-being. Your participation will help to inform Safe Families as well as future research about volunteering with at-risk children. Even if you are not currently hosting a child, we want to hear from you! Your responses will be completely anonymous and there are no required questions. If you choose to not participate in this survey, simply ignore this email. Following the link below indicates your interest in learning more about the study. If you choose to participate, please take this survey as soon as possible. > > > Follow this link to the Survey < < < ${l://SurveyLink?d=Take the Survey} Or copy and paste the URL below into your internet browser: ${l://SurveyURL} Please limit responses to one per household. Thank you so much for your time and participation. Please contact me with any questions! Joshua Bishop PhD Candidate | School of Social Work | Michigan State University Follow the link to opt out of future emails: ${l://OptOutLink?d=Click here to unsubscribe} 127 APPENDIX C: INFORMED CONSENT Thank you for your willingness to participate and support Safe Families! Below is information about the study. After reviewing this information, you may start the survey, which will take about 10 minutes. Research Participant Information and Consent Form Study Title: Volunteerism In Child Welfare: The Well-Being Of Adults Who Are Volunteer With At-Risk Children Researcher: Joshua Bishop, PhD Candidate | Michigan State University, School of Social Work BRIEF SUMMARY You are being asked to participate in a research study of Safe Families for Children. Researchers are required to provide a consent form about the research study to convey that participation is voluntary, to explain risks and benefits of participation including why you might or might not want to participate, and to empower you to make an informed decision. You should feel free to discuss and ask the researchers any questions you may have. There are no risks to participating in this study. Your participation in this study may contribute to the understanding the experience of SFFC Host Families and the effect that volunteering has on different aspects of well-being. WHAT YOU WILL BE ASKED TO DO If you agree to participate, you will be asked to complete an online survey regarding your experiences as a volunteer and your well-being. You are free to skip any questions in the survey. The survey is anonymous and confidential, and no identifiable information will be collected. Data will be kept in password-protected accounts. Results will be reported only as overall findings, not according to individual participants. YOUR RIGHTS TO PARTICIPATE, SAY NO, OR WITHDRAW You have the right to say no to participate in the research. You can stop at any time after it has already started. There will be no consequences if you stop and you will not be criticized. You will not lose any benefits that you normally receive. You will not receive money or any other form of compensation for participating in this study. CONTACT INFORMATION If you have concerns or questions about this study, such as scientific issues, how to do any part of it, or to report an injury, please contact the researcher (Joshua Bishop; 901 Eastern NE Grand Rapids, MI 49505; jbishop@bethany.org; 616-303-0222). If you have questions or concerns about your role and rights as a research participant, would like to obtain information or offer input, or would like to register a complaint about this study, you may contact, anonymously if you wish, the Michigan State University’s Human Research Protection Program at 517-355-2180, Fax 517-432-4503, or e-mail irb@msu.edu or regular mail at 4000 Collins Rd, Suite 136, Lansing, MI 48910. By continuing to the survey, you give your consent to participate in the research. Remember, please only one participant per household! 128 APPENDIX D: ADDITIONAL BIVARIATE ANALYSIS Table 17. Bivariate Analysis IV and DV - Pearson Correlation Hosted at least one in past 12 months Hostings in past 12 mo Hosting days in past 12 mo Children hosting now Volunteer Consistency *p < .05, **p < .01 Happy in general Self Mastery Score Self Esteem Score Physical Health Index 0.03 0.10 0.12 0.09 -0.12 0.08 -0.14 -0.10 -0.12 -0.05 0.03 -0.08 -0.01 -0.02 -0.02 0.08 0.01 0.04 -0.05 -0.10 Life Sat. (PWI) Score .17** -0.03 0.06 0.07 -0.02 Depression score Anxiety score -0.08 0.00 -0.10 0.06 -0.03 -0.06 -0.06 -0.05 .16* -0.04 Table 18. Bivariate Analysis Control Variables and Well-being - Pearson Correlation Happy in general .29** Self Mastery Score .13* Self Esteem Score .13* Physical Health Index .25** Life Sat. (PWI) Score .31** Depression score -.30* Anxiety score -.16** 0.02 0.11 -0.01 0.02 0.01 0.03 0.11 -.16** 0.04 0.04 -0.01 -0.09 -0.07 0.01 -.15* .17** -0.04 0.09 -0.06 .17** .17** 0.02 -0.01 -0.08 .16** -0.04 -0.07 0.02 -0.07 .16** -0.02 0.02 0.00 0.09 .16** -0.05 0.07 -0.07 .20** -0.07 -0.04 0.03 -.130* .18** -0.03 0.09 0.05 0.05 .18** -0.01 -0.07 -0.05 .15* -0.10 -0.06 0.06 .26** 0.08 -0.07 0.04 -0.08 .26** -0.10 -0.05 0.10 -0.04 0.00 0.06 -0.03 -0.10 0.00 -0.03 0.06 0.01 -0.10 .14* -0.07 0.04 0.00 .15* 0.04 -0.11 -0.03 0.06 -0.05 0.07 0.09 -0.03 .19** -.21** .13* 0.05 0.12 0.02 0.04 .128* -0.08 -0.07 129 Social Support Score Religious Service Attend Household Income Caregiving Burden Non-White Female Employed IncomeMidpoints Caregiver at work AGE Marital Status Highest Ed Volunteer at risk kids Volunteer roles other than SFFC Money donated 12 months *p < .05, **p < .01 APPENDIX E: MULTIPLE REGRESSION ANALYSES WITH FACTOR SCORES ß 2 (Constant) 1 (Constant) Happiness and Intensity Model Hosting days in past 12 mo Hosting days in past 12 mo LSNS FS Religious Service Attend Employed Income AGE Table 19. Regression Analyses with Factor Scores B SE 4.31 0.06 0.00 0.00 0.14 3.56 0.39 0.00 0.00 0.09 0.09 0.05 0.17 0.03 0.04 0.07 0.04 0.09 0.03 0.00 0.00 0.12 0.01 0.01 0.11 3.63 0.39 0.00 0.00 0.09 0.08 0.05 0.14 0.03 0.04 0.07 0.10 0.10 0.08 0.00 0.00 0.09 0.01 0.01 0.10 0.00 0.00 0.07 -0.26 0.09 -0.23 Model 1 p = .10; Model 2 p = .07; Model 3 p < .05 Happiness and Consistency Model Hosting days in past 12 mo LSNS FS Religious Service Attend Employed Income AGE Money donated 12 months Caregiver at work 3 (Constant) 1 (Constant) 2 (Constant) Volunteer Consistency 3 (Constant) Volunteer Consistency LSNS FS Religious Service Attend Employed Income AGE Volunteer Consistency LSNS FS Religious Service Attend Employed IncomeMidpoints AGE t Sig. 71.03 0.00 0.10 1.64 9.10 0.00 0.28 1.08 0.04 2.06 0.78 0.44 0.69 0.39 0.16 1.42 1.30 0.20 0.00 9.36 0.29 1.06 0.10 1.67 0.86 0.39 0.32 1.00 0.34 0.97 1.24 0.22 0.77 0.44 -2.82 0.01 t Sig. 70.11 0.00 -0.53 0.60 11.43 0.00 -0.57 0.57 0.01 2.57 0.46 0.74 1.27 0.20 0.07 1.81 0.80 0.43 11.34 0.00 -0.84 0.40 0.02 2.35 0.34 0.95 1.47 0.14 0.04 2.04 0.70 0.49 -0.04 ß SE B 4.38 0.06 -0.01 0.02 3.78 0.33 -0.01 0.02 -0.04 0.10 0.04 0.18 0.02 0.03 0.05 0.10 0.08 0.09 0.00 0.00 0.13 0.00 0.00 0.06 3.79 0.33 -0.02 0.02 -0.06 0.09 0.04 0.17 0.03 0.03 0.07 0.12 0.08 0.11 0.00 0.00 0.15 0.00 0.00 0.05 130 Table 19 (cont’d) Models 2 and 3 are significant at p < .05 Money donated 12 months Caregiver at work 0.00 0.00 -0.16 0.08 -0.06 -0.14 -0.74 0.46 -2.02 0.05 Self-Mastery and Intensity Model 1.00 (Constant) 2.00 (Constant) 3.00 (Constant) Hosting days in past 12 mo Hosting days in past 12 mo LSNS FS Religious Service Attend Employed Income Age SE 0.10 0.00 0.63 0.00 0.07 0.06 0.15 0.00 0.01 0.65 0.00 0.08 0.06 0.16 0.00 0.01 0.00 0.15 Model 1 p = .08; Models 2 and 3 are significant at p < .05 Self-Mastery and Consistency Model Hosting days in past 12 mo LSNS FS Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work B 0.19 0.00 0.18 0.00 0.14 -0.09 0.30 0.00 0.00 0.14 0.00 0.14 -0.09 0.31 0.00 0.00 0.00 -0.06 ß -0.15 -0.16 0.15 -0.13 0.16 0.09 0.04 -0.15 0.15 -0.12 0.17 0.10 0.05 -0.04 -0.03 ß -0.06 -0.06 0.08 -0.13 0.16 0.10 0.07 -0.06 0.09 t 1.94 -1.76 0.29 -1.92 1.88 -1.57 1.94 1.05 0.53 0.22 -1.85 1.82 -1.46 1.97 1.12 0.54 -0.43 -0.40 t 0.85 -0.80 -0.06 -0.85 1.21 -1.79 2.21 1.37 1.01 -0.11 -0.82 1.23 Sig 0.05 0.08 0.78 0.06 0.06 0.12 0.06 0.29 0.60 0.83 0.07 0.07 0.15 0.05 0.26 0.59 0.67 0.69 Sig 0.40 0.42 0.95 0.40 0.23 0.08 0.03 0.17 0.31 0.92 0.41 0.22 1.00 (Constant) 2.00 (Constant) Volunteer Consistency Volunteer Consistency LSNS FS Religious Service Attend Employed Income Age 3.00 (Constant) Volunteer Consistency LSNS FS SE 0.11 0.03 0.56 0.04 0.06 0.05 0.13 0.00 0.01 0.57 0.04 0.07 B 0.09 -0.03 -0.03 -0.03 0.08 -0.10 0.29 0.00 0.01 -0.06 -0.03 0.08 131 Table 19 (cont’d) Models 2 and 3 p < .05 Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work -0.10 0.29 0.00 0.01 0.00 0.03 0.06 0.14 0.00 0.01 0.00 0.13 -0.13 0.15 0.10 0.08 -0.01 0.01 -1.75 2.11 1.34 1.03 -0.18 0.19 0.08 0.04 0.18 0.30 0.86 0.85 Self-Esteem and Intensity Model 1 (Constant) Hosting days in past 12 mo 2 (Constant) Hosting days in past 12 mo LSNS FS Religious Service Attend Employed Income Age 3 (Constant) Hosting days in past 12 mo LSNS FS Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Self-Esteem and Consistency Model 1 (Constant) 2 (Constant) Volunteer Consistency Volunteer Consistency LSNS FS Religious Service Attend Employed Income Age ß -0.02 -0.04 0.10 -0.07 0.11 0.14 0.04 -0.03 0.10 -0.05 0.12 0.19 0.04 -0.10 -0.04 ß 0.00 -0.02 0.04 -0.06 0.09 0.16 0.08 SE 0.11 0.00 0.70 0.00 0.08 0.06 0.17 0.00 0.01 0.71 0.00 0.08 0.07 0.17 0.00 0.01 0.00 0.17 SE 0.11 0.04 0.60 0.04 0.07 0.06 0.14 0.00 0.01 B 0.09 0.00 -0.27 0.00 0.10 -0.05 0.22 0.00 0.00 -0.40 0.00 0.10 -0.04 0.24 0.00 0.00 0.00 -0.08 B 0.04 0.00 -0.56 -0.01 0.04 -0.05 0.17 0.00 0.01 132 t 0.86 -0.24 -0.39 -0.51 1.19 -0.82 1.31 1.65 0.47 -0.56 -0.37 1.17 -0.62 1.39 1.95 0.50 -1.11 -0.49 t 0.37 -0.01 -0.94 -0.26 0.60 -0.84 1.22 2.19 1.12 Sig. 0.39 0.81 0.70 0.61 0.24 0.41 0.19 0.10 0.64 0.58 0.71 0.25 0.53 0.17 0.05 0.62 0.27 0.62 Sig. 0.71 0.99 0.35 0.79 0.55 0.40 0.22 0.03 0.26 Table 19 (cont’d) 3 (Constant) Volunteer Consistency LSNS FS Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Model 2 p = .09 Health and Intensity Model 1 (Constant) Hosting days in past 12 mo 2 (Constant) Hosting days in past 12 mo LSNS FS Religious Service Attend Employed Income Age 3 (Constant) Hosting days in past 12 mo LSNS FS Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Model 2 p < .05; Model 3 p= .08 Health and Consistency Model 1 (Constant) 2 (Constant) Volunteer Consistency Volunteer Consistency LSNS FS Religious Service Attend -0.02 0.04 -0.05 0.09 0.17 0.08 -0.04 -0.03 ß 0.05 0.05 0.22 -0.13 -0.03 0.10 -0.08 0.04 0.23 -0.13 -0.04 0.10 -0.08 0.00 0.05 ß -0.09 -0.09 0.23 -0.10 0.61 0.04 0.07 0.06 0.15 0.00 0.01 0.00 0.14 SE 0.10 0.00 0.64 0.00 0.08 0.06 0.15 0.00 0.01 0.65 0.00 0.08 0.06 0.16 0.00 0.01 0.00 0.15 SE 0.12 0.04 0.60 0.04 0.07 0.06 -0.59 -0.01 0.04 -0.04 0.17 0.00 0.01 0.00 -0.05 B -0.02 0.00 0.86 0.00 0.20 -0.09 -0.06 0.00 -0.01 0.85 0.00 0.21 -0.09 -0.08 0.00 -0.01 0.00 0.10 B 0.09 -0.05 0.21 -0.05 0.23 -0.08 133 -0.97 -0.33 0.58 -0.74 1.19 2.23 1.12 -0.46 -0.38 0.33 0.75 0.56 0.46 0.23 0.03 0.27 0.64 0.71 t -0.24 0.63 1.35 0.54 2.67 -1.52 -0.39 1.16 -1.01 1.31 0.52 2.71 -1.54 -0.52 1.06 -0.99 0.02 0.63 t 0.76 -1.23 0.35 -1.23 3.30 -1.38 Sig. 0.81 0.53 0.18 0.59 0.01 0.13 0.70 0.25 0.32 0.19 0.60 0.01 0.13 0.61 0.29 0.33 0.99 0.53 Sig. 0.45 0.22 0.73 0.22 0.00 0.17 Table 19 (cont’d) Employed Income Age 3 (Constant) Volunteer Consistency LSNS FS Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Models 2 and 3 are significant at p < .01 Life Satisfaction and Intensity Model 1 (Constant) 2 (Constant) Hosting days in past 12 mo Hosting days in past 12 mo LSNS FS Religious Service Attend Employed Income Age 3 (Constant) Hosting days in past 12 mo LSNS FS Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Models 2 and 3 are significant at p < .01 Life Satisfaction and Consistency Model 1 (Constant) 2 (Constant) Volunteer Consistency 0.14 0.00 0.01 0.62 0.04 0.07 0.06 0.15 0.00 0.01 0.00 0.14 SE 0.09 0.00 0.56 0.00 0.07 0.05 0.13 0.00 0.01 0.56 0.00 0.07 0.05 0.14 0.00 0.01 0.00 0.13 SE 0.11 0.04 0.54 -0.01 0.14 0.02 -0.09 0.23 -0.10 -0.01 0.15 0.02 -0.01 -0.01 ß 0.10 0.03 0.28 0.06 -0.08 0.30 0.04 0.03 0.27 0.06 -0.06 0.28 0.04 0.05 -0.09 ß -0.03 -0.12 2.02 0.22 0.34 -1.23 3.23 -1.34 -0.11 1.95 0.22 -0.10 -0.10 t 0.41 1.24 -1.73 0.39 3.61 0.79 -1.06 3.73 0.49 -1.59 0.34 3.40 0.77 -0.78 3.16 0.46 0.56 -1.15 t 0.57 -0.41 -2.09 0.91 0.05 0.83 0.74 0.22 0.00 0.18 0.91 0.05 0.83 0.92 0.92 Sig. 0.69 0.22 0.09 0.70 0.00 0.43 0.29 0.00 0.62 0.11 0.74 0.00 0.44 0.44 0.00 0.65 0.58 0.25 Sig. 0.57 0.68 0.04 -0.02 0.00 0.00 0.21 -0.05 0.23 -0.08 -0.02 0.00 0.00 0.00 -0.02 B 0.04 0.00 -0.96 0.00 0.23 0.04 -0.14 0.00 0.00 -0.90 0.00 0.22 0.04 -0.11 0.00 0.00 0.00 -0.15 B 0.06 -0.02 -1.12 134 Table 19 (cont’d) Volunteer Consistency LSNS FS Religious Service Attend Employed Income Age 3 (Constant) Volunteer Consistency LSNS FS Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Models 2 and 3 are significant at p < .01 Depression and Intensity Model 1 (Constant) Hosting days in past 12 mo 2 (Constant) Hosting days in past 12 mo LSNS FS Religious Service Attend Employed Income Age 3 (Constant) Hosting days in past 12 mo LSNS FS Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Models 2 and 3 are significant at p < .05 Depression and Consistency Model 1 (Constant) -0.02 0.26 0.06 -0.11 0.00 0.00 -1.11 -0.02 0.26 0.06 -0.10 0.00 0.00 0.00 -0.06 B 0.02 0.00 0.92 0.00 -0.19 -0.01 -0.11 0.00 -0.01 1.03 0.00 -0.19 -0.02 -0.15 0.00 -0.01 0.00 0.18 B 0.13 135 0.03 0.06 0.05 0.13 0.00 0.01 0.55 0.04 0.06 0.05 0.13 0.00 0.01 0.00 0.13 SE 0.09 0.00 0.55 0.00 0.06 0.05 0.13 0.00 0.01 0.55 0.00 0.06 0.05 0.14 0.00 0.01 0.00 0.13 SE 0.11 -0.04 0.27 0.08 -0.06 0.31 0.02 -0.04 0.27 0.08 -0.05 0.31 0.02 -0.01 -0.03 ß -0.10 -0.06 -0.25 -0.01 -0.07 -0.04 -0.15 -0.08 -0.24 -0.04 -0.09 -0.09 -0.15 0.12 0.12 ß -0.59 4.18 1.15 -0.83 4.57 0.24 -2.03 -0.64 4.06 1.17 -0.76 4.40 0.21 -0.12 -0.44 t 0.24 -1.18 1.67 -0.74 -3.02 -0.15 -0.84 -0.43 -1.79 1.86 -0.92 -2.89 -0.43 -1.13 -0.93 -1.83 1.37 1.39 t 1.23 0.56 0.00 0.25 0.41 0.00 0.81 0.04 0.52 0.00 0.24 0.45 0.00 0.84 0.90 0.66 Sig. 0.81 0.24 0.10 0.46 0.00 0.88 0.40 0.67 0.08 0.07 0.36 0.00 0.67 0.26 0.36 0.07 0.17 0.17 Sig. 0.22 Table 19 (cont’d) Volunteer Consistency 2 (Constant) Volunteer Consistency LSNS FS Religious Service Attend Employed Income Age 3 (Constant) Volunteer Consistency LSNS FS Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Models 2 and 3 are significant at p < .01 Anxiety and Intensity Model 1 (Constant) Hosting days in past 12 mo 2 (Constant) Hosting days in past 12 mo LSNS FS Religious Service Attend Employed Income Age 3 (Constant) Hosting days in past 12 mo LSNS FS Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Model 3 p = .08 Anxiety and Consistency 0.04 0.57 0.04 0.07 0.06 0.14 0.00 0.01 0.58 0.04 0.07 0.06 0.14 0.00 0.01 0.00 0.14 SE 0.10 0.00 0.65 0.00 0.08 0.06 0.16 0.00 0.01 0.65 0.00 0.08 0.06 0.16 0.00 0.01 0.00 0.16 -0.07 -0.06 -0.26 -0.06 -0.01 -0.09 -0.07 -0.05 -0.25 -0.07 -0.03 -0.09 -0.06 -0.01 0.10 ß -0.05 -0.03 -0.17 0.00 0.10 0.08 -0.17 -0.05 -0.16 -0.02 0.07 0.02 -0.18 0.14 0.11 -0.97 1.86 -0.81 -3.81 -0.88 -0.14 -1.20 -0.95 1.70 -0.65 -3.56 -0.94 -0.36 -1.16 -0.81 -0.14 1.38 t 0.34 -0.64 1.03 -0.37 -2.08 0.03 1.15 0.88 -2.06 1.25 -0.57 -1.98 -0.27 0.84 0.20 -2.12 1.54 1.30 0.34 0.07 0.42 0.00 0.38 0.89 0.23 0.35 0.09 0.52 0.00 0.35 0.72 0.25 0.42 0.89 0.17 Sig. 0.73 0.53 0.31 0.71 0.04 0.98 0.25 0.38 0.04 0.21 0.57 0.05 0.79 0.40 0.84 0.04 0.13 0.20 -0.03 1.06 -0.03 -0.25 -0.05 -0.02 0.00 -0.01 0.98 -0.02 -0.24 -0.05 -0.05 0.00 -0.01 0.00 0.19 B 0.04 0.00 0.67 0.00 -0.16 0.00 0.18 0.00 -0.02 0.82 0.00 -0.15 -0.02 0.13 0.00 -0.02 0.00 0.20 136 Table 19 (cont’d) Model 1 (Constant) 2 (Constant) Volunteer Consistency Volunteer Consistency LSNS FS Religious Service Attend Employed Income Age B 0.15 -0.04 1.62 -0.01 -0.22 -0.06 0.10 0.00 -0.02 1.50 0.00 -0.20 -0.07 0.05 0.00 -0.02 0.00 0.32 SE 0.12 0.04 0.62 0.04 0.07 0.06 0.15 0.00 0.01 0.63 0.04 0.07 0.06 0.15 0.00 0.01 0.00 0.15 ß -0.08 -0.02 -0.21 -0.07 0.05 0.03 -0.21 0.00 -0.19 -0.08 0.03 0.03 -0.20 -0.01 0.15 t 1.29 -1.08 2.60 -0.27 -3.04 -1.01 0.70 0.40 -2.96 2.38 -0.03 -2.72 -1.11 0.36 0.34 -2.76 -0.13 2.15 Sig. 0.20 0.28 0.01 0.79 0.00 0.31 0.48 0.69 0.00 0.02 0.98 0.01 0.27 0.72 0.74 0.01 0.90 0.03 3 (Constant) Volunteer Consistency LSNS FS Religious Service Attend Employed Income Age Money donated 12 months Caregiver at work Models 2 and 3 are significant at p < .01 137 APPENDIX F: CURVILINEAR ANALYSES Figure 21. Happiness and Intensity Figure 22. Self-Mastery and Intensity 138 Figure 23. Self-Esteem and Intensity Figure 24. Health and Intensity 139 Figure 25. Life Satisfaction and Intensity Figure 26. Depression and Intensity 140 Figure 27. Anxiety and Intensity 141 REFERENCES 142 REFERENCES Anderson, N. B. (2003). Emotional Longevity: What Really Determines How Long You Live. Viking. Anderson, N. D., Damianakis, T., Kröger, E., Wagner, L. M., Dawson, D. R., Binns, M. A., … Cook, S. L. (2014). The benefits associated with volunteering among seniors: A critical review and recommendations for future research. Psychological Bulletin, 140(6), 1505– 1533. http://dx.doi.org/10.1037/a0037610 Bar-Tal, D. (1986). Altruistic motivation to help: Definition, utility and operationalization. Humboldt Journal of Social Relations, 13(1/2), 3–14. Bekkers, R., & Wiepking, P. (2011). A literature review of empirical studies of philanthropy: Eight mechanisms that drive charitable giving. Nonprofit and Voluntary Sector Quarterly, 40(5), 924–973. https://doi.org/10.1177/0899764010380927 Borgonovi, F. (2008). Doing well by doing good. The relationship between formal volunteering and self-reported health and happiness. Social Science & Medicine, 66(11), 2321–2334. https://doi.org/10.1016/j.socscimed.2008.01.011 Bragg, M. (n.d.). Altruism. In In Our Time - Philosophy. Retrieved from http://www.bbc.co.uk/programmes/p0038x9c Brown, K. M., Hoye, R., & Nicholson, M. (2012). Self-esteem, self-efficacy, and social connectedness as mediators of the relationship between volunteering and well-being. Journal of Social Service Research, 38(4), 468–483. https://doi.org/10.1080/01488376.2012.687706 Brown, T. A. (2015). Confirmatory factor analysis for applied research, second edition. Guilford Publications. Buchanan, V. (2017, January 13). H.R.253 - 115th Congress (2017-2018): Family First Prevention Services Act of 2017 [Webpage]. Retrieved June 25, 2018, from https://www.congress.gov/bill/115th-congress/house-bill/253 Burns, B. J., Phillips, S. D., Warner, H. R., Barth, R. P., Kolko, D. J., Campbell, Y., & Landsverk, J. (2004). Mental health need and access to mental health services by youths involved with child welfare: A national survey. Journal of the American Academy of Child & Adolescent Psychiatry, 43(8), 960–970. https://doi.org/10.1097/01.chi.0000127590.95585.65 143 Centers for Disease Control. (2018, April 12). Child abuse and neglect: Risk and protective factors. Retrieved January 29, 2019, from https://www.cdc.gov/violenceprevention/childabuseandneglect/riskprotectivefactors.html Cham, H., Reshetnyak, E., Rosenfeld, B., & Breitbart, W. (2017). Full information maximum likelihood estimation for latent variable interactions with incomplete indicators. Multivariate Behavioral Research, 52(1), 12–30. https://doi.org/10.1080/00273171.2016.1245600 Child Welfare Information Gateway. (2016a). Racial disproportionality and disparity in child welfare. Retrieved from https://www.childwelfare.gov/pubPDFs/racial_disproportionality.pdf Child Welfare Information Gateway. (2016b). The AFCARS report. Retrieved from http://www.acf.hhs.gov/sites/default/files/cb/afcarsreport23.pdf Clench-aas, J., Nes, R. B., & Aarø, L. E. (2017). The perceived constraints subscale of the sense of mastery scale: Dimensionality and measurement invariance. Quality of Life Research; Dordrecht, 26(1), 127–138. http://dx.doi.org/10.1007/s11136-016-1359-6 Cohen, J. A., Deblinger, E., & Mannarino, A. P. (2006). Treating trauma and traumatic grief in children and adolescents: A clinician’s guide. Retrieved from http://site.ebrary.com/lib/alltitles/docDetail.action?docID=10172288 Collett, J., & Morrissey, C. (2007). The social psychology of generosity: The state of current interdisciplinary research. Science of Generosity Initiative. Retrieved from http://generosityresearch.nd.edu/assets/17634/social_psychology_of_generosity_final.pdf Connolly, S., & O’shea, E. (2015). The perceived benefits of participating in voluntary activities among older people: Do they differ by volunteer characteristics? Activities, Adaptation & Aging, 39(2), 95–108. https://doi.org/10.1080/01924788.2015.1024075 Cook, A., Spinazzola, J., Ford, J., Lanktree, C., Blaustein, M., Cloitre, M., … van der Kolk, B. (2005). Complex trauma in children and adolescents. Psychiatric Annals, 35(5), 390–398. Cummins, R. A. (2008). The wellbeing of Australians: Money, debt and loneliness. Retrieved from https://www.researchgate.net/profile/Melissa_Weinberg/publication/261702245_Australi an_Unity_Wellbeing_Index_Survey_20/links/004635350cfe162cc6000000/Australian- Unity-Wellbeing-Index-Survey-20.pdf Dillon, M., & Wink, P. (2007). In the course of a lifetime: Tracing religious belief, practice, and change (1st ed.). https://doi.org/10.1525/j.ctt1pnrt5 Erikson, E. H. (1964). Childhood and society (2nd ed.). New York: Norton. 144 Feigin, S., Owens, G., & Goodyear-Smith, F. (2014). Theories of human altruism: A systematic review. Vernon Innovative Publishers. Retrieved from http://www.vipoa.org/neuropsychol/1/1/ Felitti, V. J., Anda, R. F., Nordenberg, D., Williamson, D. F., Spitz, A. M., Edwards, V., … Marks, J. S. (1998). Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults. The adverse childhood experiences (ACE) study. American Journal of Preventive Medicine, 14(4), 245–258. 145 Folman, R. D. (1998). “I was tooken.” Adoption Quarterly, 2(2), 7–35. https://doi.org/10.1300/J145v02n02_02 Frank, K. A., Maroulis, S. J., Duong, M. Q., & Kelcey, B. M. (2013). What would it take to change an inference? Using rubin’s causal model to interpret the robustness of causal inferences. Educational Evaluation and Policy Analysis, 35(4), 437–460. Fredrickson, B. L. (2003). The value of positive emotions. American Scientist, 91(4), 330–335. Fuhrer, R., & Stansfeld, S. A. (2002). How gender affects patterns of social relations and their impact on health: A comparison of one or multiple sources of support from “close persons.” Social Science & Medicine, 54(5), 811–825. https://doi.org/10.1016/S0277- 9536(01)00111-3 Gibbs, D. (2005). Understanding foster parenting: Using administrative data to explore retention. Retrieved from Office of the Assistant Secretary for Planning and Evaluation U.S. Department of Health and Human Services website: https://aspe.hhs.gov/report/understanding-foster-parenting-using-administrative-data- explore-retention Gilligan, C. (1977). In a different voice: Women’s conceptions of self and of morality. Harvard Educational Review, 47(4), 481–517. https://doi.org/10.17763/haer.47.4.g6167429416hg5l0 Gimenez-Nadal, J. I., & Molina, J. A. (2015). Voluntary activities and daily happiness in the United States. Economic Inquiry, 53(4), 1735–1750. https://doi.org/10.1111/ecin.12227 Gray-Little, B., Williams, V. S. L., & Hancock, T. D. (1997). An item response theory analysis of the rosenberg self-esteem scale. Personality and Social Psychology Bulletin, 23(5), 443–451. https://doi.org/10.1177/0146167297235001 Greeson, J. K. P., Briggs, E. C., Kisiel, C. L., Layne, C. M., Ake, G. S., III, Ko, S. J., … Fairbank, J. A. (2011). Complex trauma and mental health in children and adolescents placed in foster care: Findings from the National Child Traumatic Stress Network. Child Welfare, 90(6), 91+. (kwok). Harway, M., & Nutt. (2006). Women and giving. In J. Worell & C. D. Goodheart (Eds.), Handbook of girls’ and women’s psychological health. New York: Oxford University Press. Herzog, P. S., & Price, H. E. (2016). American generosity: Who gives and why [iBooks]. New York, NY: Oxford University Press. Hooper, D., Coughlan, J., & Mullen, M. (2008). Structural equation modelling: Guidelines for determining model fit. Electronic Journal of Business Research Methods, 6(1), 53–60. House, J. S. (2014). Americans’ changing lives: Waves I, II, III, IV, and V, 1986, 1989, 1994, 2002, and 2011. https://doi.org/10.3886/ICPSR04690.v7 Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. https://doi.org/10.1080/10705519909540118 International Wellbeing Group. (2013). Personal wellbeing index: 5th edition. Retrieved from http://www.acqol.com.au/instruments Kenny, D. A. (n.d.). Measuring model fit. Retrieved August 30, 2019, from http://davidakenny.net/cm/fit.htm#null Kitzrow, M. A. (1998). An overview of current psychological theory and research on altruism and prosocial behavior. In R. G. [Ed Bringle & D. K. [Ed Duffy (Eds.), With service in mind: Concepts and models for service-learning in psychology (pp. 19–34, Chapter v, 217 Pages). Retrieved from http://search.proquest.com/docview/620395296/F12ABAB2A19A419EPQ/28 Klein, N. (2017). Prosocial behavior increases perceptions of meaning in life. The Journal of Positive Psychology, 12(4), 354–361. https://doi.org/10.1080/17439760.2016.1209541 Koltko-Rivera, M. E. (2006). Rediscovering the later version of Maslow’s hierarchy of needs: Self-transcendence and opportunities for theory, research, and unification. Review of General Psychology, 10(4), 302–317. http://dx.doi.org/10.1037/1089-2680.10.4.302 Konow, J., & Earley, J. (2008). The hedonistic paradox: Is homo economicus happier? Journal of Public Economics, 92(1–2), 1–33. http://dx.doi.org/10.1016/j.jpubeco.2007.04.006 Kroenke, K., Spitzer, R. L., & Williams, J. B. W. (2001). The PHQ-9. Journal of General Internal Medicine, 16(9), 606–613. https://doi.org/10.1046/j.1525- 1497.2001.016009606.x 146 Kwok, Y. Y., Chui, W. H., & Wong, L. P. (2013). Need satisfaction mechanism linking volunteer motivation and life satisfaction: A mediation study of volunteers’ subjective well-being. Social Indicators Research, 114(3), 1315–1329. http://dx.doi.org/10.1007/s11205-012-0204-8 Lubben, J., Blozik, E., Gillmann, G., Iliffe, S., von Renteln Kruse, W., Beck, J. C., & Stuck, A. E. (2006). Performance of an abbreviated version of the Lubben Social Network Scale among three European community-dwelling older adult populations. The Gerontologist, 46(4), 503–513. https://doi.org/10.1093/geront/46.4.503 Luks, A. (1988, October 1). Helper’s high. Psychology Today; Washington D.C., Wash., 22(10), MacDonald, M., Phipps, S., & Lethbridge, L. (2005). Taking its toll: The influence of paid and unpaid work on women’s well-being. Feminist Economics, 11(1), 63–94. https://doi.org/10.1080/1354570042000332597 39. MacIlvaine, W. R., Nelson, L. A., Stewart, J. A., & Stewart, W. C. (2014). Association of strength of community service to personal wellbeing. Community Mental Health Journal, 50(5), 577–582. https://dx.doi.org/10.1007/s10597-013-9660-0 Matz-Costa, C., Besen, E., Boone James, J., & Pitt-Catsouphes, M. (2014). Differential impact of multiple levels of productive activity engagement on psychological well-being in middle and later life. The Gerontologist, 54(2), 277–289. https://doi.org/10.1093/geront/gns148 McDougle, L., Handy, F., Konrath, S., & Walk, M. (2014). Health outcomes and volunteering: The moderating role of religiosity. Social Indicators Research, 117(2), 337–351. https://dx.doi.org/10.1007/s11205-013-0336-5 Meier, S., & Stutzer, A. (2008). Is volunteering rewarding in itself? Economica, 75(297), 39–59. Moen, P., Dempster-McClain, D., & Williams, R. M. (1992). Successful aging: A life-course perspective on women’s multiple roles and health. American Journal of Sociology, 97(6), 1612–1638. https://doi.org/10.1086/229941 Morrow-Howell, N., Hinterlong, J., Rozario, P. A., & Tang, F. (2003). Effects of volunteering on the well-being of older adults. The Journals of Gerontology: Series B: Psychological Sciences and Social Sciences, 58(3), S137–S145. https://dx.doi.org/10.1093/geronb/58.3.S137 Morrow-Howell, N., Kinnevy, S., & Mann, M. (1999). The perceived benefits of participating in volunteer and educational activities. Journal of Gerontological Social Work, 32(2), 65– 80. 147 Musick, M. A., Herzog, A. R., & House, J. S. (1999). Volunteering and mortality among older adults: Findings from a national sample. The Journals of Gerontology: Series B, 54B(3), S173–S180. https://doi.org/10.1093/geronb/54B.3.S173 Musick, M. A., & Wilson, J. (2003). Volunteering and depression: The role of psychological and social resources in different age groups. Social Science & Medicine, 56(2), 259–269. https://doi.org/10.1016/S0277-9536(02)00025-4 National Association of Social Workers. (2017). Code of ethics of the national association of social workers. Retrieved from https://www.socialworkers.org/About/Ethics/Code-of- Ethics/Code-of-Ethics-English Nelson, S. K., Layous, K., Cole, S. W., & Lyubomirsky, S. (2016). Do unto others or treat yourself? The effects of prosocial and self-focused behavior on psychological flourishing. Emotion, 16(6), 850–861. https://dx.doi.org/10.1037/emo0000178 Nolan, T. (2013). Safe Families for Children: RFI Social Impact Bond. Retrieved from Bethany Christian Services website: unpublished Nolan, T. (2015, October 19). National Director of Safe Families for Children. Oakley, B. A. (2011). Pathological altruism. Oxford University Press. Oakley, B. A. (2013). Concepts and implications of altruism bias and pathological altruism. Proceedings of the National Academy of Sciences of the United States of America, 110, 10408–10415. Oakley, B. A. (2014). Empathy in academe: On the origins of pathological altruism. Academic Questions, 27(1), 48. https://doi.org/10.1007/s12129-013-9403-2 O’Brien, L., Townsend, M., & Ebden, M. (2010). “Doing something positive”: Volunteers’ experiences of the well-being benefits derived from practical conservation activities in nature. Voluntas: International Journal of Voluntary and Nonprofit Organizations, 21(4), 525–545. https://dx.doi.org/10.1007/s11266-010-9149-1 Okun, M. A., Rios, R., Crawford, A. V., & Levy, R. (2011). Does the relation between volunteering and well-being vary with health and age? The International Journal of Aging & Human Development, 72(3), 265–287. http://dx.doi.org/10.2190/AG.72.3.f Oman, D., & Thoresen, C. E. (2000). Role of volunteering in health and happiness. Career Planning and Adult Development Journal, 15(4), 59–70. Oman, D., Thoresen, C. E., & Mcmahon, K. (1999). Volunteerism and mortality among the community-dwelling elderly. Journal of Health Psychology, 4(3), 301–316. https://doi.org/10.1177/135910539900400301 148 Oord, T. J. (Ed.). (2008). The altruism reader: Selections from writings on love, religion, and science. West Conshohocken, PA: Templeton Foundation Press. Pearlin, L. I., Menaghan, E. G., Lieberman, M. A., & Mullan, J. T. (1981). The Stress Process. Journal of Health and Social Behavior, 22(4), 337–356. https://doi.org/10.2307/2136676 Pearlin, L. I., & Schooler, C. (1978). The Structure of Coping. Journal of Health and Social Behavior, 19(1), 2–21. https://doi.org/10.2307/2136319 Perry, B. L. (2006). Understanding social network disruption: The case of youth in foster care. Social Problems, 53(3), 371–391. https://doi.org/10.1525/sp.2006.53.3.371 Pfizer. (n.d.). Manual and scoring instructions for patient health questionnaire (PHQ) and GAD- 7 measures. Retrieved from https://www.phqscreeners.com/sites/g/files/g10016261/f/201412/instructions.pdf Piliavin, J. A., & Siegl, E. (2007). Health benefits of volunteering in the Wisconsin Longitudinal Study. Journal of Health and Social Behavior, 48(4), 450–464. https://dx.doi.org/10.1177/002214650704800408 Pillemer, K., Fuller-Rowell, T. E., Reid, M. C., & Wells, N. M. (2010). Environmental volunteering and health outcomes over a 20-year period. The Gerontologist, 50(5), 594– 602. https://doi.org/10.1093/geront/gnq007 Post, S. G. (2005). Altruism, happiness, and health: It’s good to be good. International Journal of Behavioral Medicine, 12(2), 66–77. https://dx.doi.org/10.1207/s15327558ijbm1202_4 Post, S. G. (Ed.). (2007). Altruism and health: Perspectives from empirical research. Oxford ; New York: Oxford University Press. Post, S. G. (2011). The hidden gifts of helping: How the power of giving, compassion, and hope can get us through hard times (1 edition). San Francisco: Jossey-Bass. Racusin, R., Maerlender, A. C., Sengupta, A., Isquith, P. K., & Straus, M. B. (2005). Psychosocial treatment of children in foster care: A review. Community Mental Health Journal, 41(2), 199–221. Ramos, R., Güntert, S., Brauchli, R., Bauer, G., Wehner, T., & Hämmig, O. (2016). Exploring the interaction between volunteering status, paid job characteristics and quality of volunteers’ motivation on mental health. Voluntas; Baltimore, 27(2), 790–809. https://dx.doi.org/10.1007/s11266-015-9647-2 Raykov, T. (2001). Estimation of congeneric scale reliability using covariance structure analysis with nonlinear constraints. British Journal of Mathematical and Statistical Psychology, 54(2), 315–323. https://doi.org/10.1348/000711001159582 149 Raykov, T., & Shrout, P. E. (2002). Reliability of scales with general structure: Point and interval estimation using a structural equation modeling approach. Structural Equation Modeling: A Multidisciplinary Journal, 9(2), 195–212. https://doi.org/10.1207/S15328007SEM0902_3 Rhodes, M. L. (1985). Gilligan’s theory of moral development as applied to social work. Social Work, 30(2), 101–105. Rietschlin, J. (1998). Voluntary Association Membership and Psychological Distress. Journal of Health and Social Behavior, 39(4), 348–355. https://doi.org/10.2307/2676343 Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press. Rubin, P. H. (2014). Pathological altruism and pathological regulation. Cato Journal, 34(1), 171. Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist; Arlington, Va., 55(1), 68. Salazar, A. M. (2013). The value of a college degree for foster care alumni: Comparisons with general population samples. Social Work, 58(2), 139+. Schwartz, C., Meisenhelder, J. B., Ma, Y., & Reed, G. (2003). Altruistic social interest behaviors are associated with better mental health. Psychosomatic Medicine, 65(5), 778–785. Schwartz, C., & Sendor, R. M. (1999). Helping others helps oneself: Response shift effects in peer support. Social Science & Medicine, 48(11), 1563–1575. https://doi.org/10.1016/S0277-9536(99)00049-0 Shavitt, S., Cho, Y. I., Johnson, T. P., Jiang, D., Holbrook, A., & Stavrakantonaki, M. (2016). Culture moderates the relation between perceived stress, social support, and mental and physical health. Journal of Cross-Cultural Psychology, 47(7), 956–980. https://doi.org/10.1177/0022022116656132 Shye, S. (2010a). The Motivation to Volunteer: A Systemic Quality of Life Theory. Social Indicators Research, 98(2), 183–200. https://doi.org/10.1007/s11205-009-9545-3 Shye, S. (2010b). The Motivation to Volunteer: A Systemic Quality of Life Theory. Social Indicators Research, 98(2), 183–200. https://dx.doi.org/10.1007/s11205-009-9545-3 Smith, C. (2015). American religion and generous financial giving. Retrieved from https://vimeo.com/126502315 Smith, C., & Davidson, H. (2014). The paradox of generosity: Giving we receive, grasping we lose. New York, NY: Oxford University Press. 150 Smith, C., Herzog, P. S., & Beyerlein, K. (2010). Methods report and user’s guide to the 2010 science of generosity survey. Notre Dame, IN: University of Notre Dame. Sneed, R. S., & Cohen, S. (2013). A prospective study of volunteerism and hypertension risk in older adults. Psychology and Aging, 28(2), 578–586. https://dx.doi.org/10.1037/a0032718 Son, J., & Wilson, J. (2012). Volunteer work and hedonic, eudemonic, and social well-being. Sociological Forum, 27(3), 658–681. https://dx.doi.org/10.1111/j.1573- 7861.2012.01340.x Stukas, A. A., Hoye, R., Nicholson, M., Brown, K. M., & Aisbett, L. (2016). Motivations to volunteer and their associations with volunteers’ well-being. Nonprofit and Voluntary Sector Quarterly, 45(1), 112–132. Taillefer, M.-C., Dupuis, G., Roberge, M.-A., & LeMay, S. (2003). Health-related quality of life models: Systematic review of the literature. Social Indicators Research; Dordrecht, 64(2), 293–323. http://dx.doi.org/10.1023/A:1024740307643 Theurer, K., & Wister, A. (2010). Altruistic behaviour and social capital as predictors of well- being among older Canadians. Ageing and Society, 30(1), 157–182. http://dx.doi.org/10.1017/S0144686X09008848 Thoits, P. A. (2012). Role-identity salience, purpose and meaning in life, and well-being among volunteers. Social Psychology Quarterly, 75(4), 360–384. http://dx.doi.org/10.1177/0190272512459662 Thoits, P. A., & Hewitt, L. N. (2001). Volunteer work and well-being. Journal of Health and Social Behavior, 42(2), 115–131. Tucker, M. C., & Rodriguez, C. M. (2014). Family Dysfunction and Social Isolation as Moderators Between Stress and Child Physical Abuse Risk. Journal of Family Violence, 29(2), 175–186. https://doi.org/10.1007/s10896-013-9567-0 Tyebjee, T. (2003). Attitude, interest, and motivation for adoption and foster care. Child Welfare, 82(6), 685. Unrau, Y. A., Sherwood, D. A., & Postema, C. (2019). Financial and educational hardships among social work students during their program of study. Journal of Social Work Education. U.S. Department of Health and Human Services, Administration for Children and Families, Children’s Bureau. (2017). Child maltreatment 2015. Retrieved June 15, 2017, from https://www.acf.hhs.gov/sites/default/files/cb/cm2015.pdf#page=20 151 U.S. Department of Health and Human Services, Administration for Children and Families, Children’s Bureau. (2018). Child Maltreatment 2016: Summary of Key Findings. Retrieved from Child Maltreatment website: https://www.childwelfare.gov/pubPDFs/canstats.pdf van Campen, C., de Boer, A. H., & Iedema, J. (2013). Are informal caregivers less happy than noncaregivers? Happiness and the intensity of caregiving in combination with paid and voluntary work. Scandinavian Journal of Caring Sciences, 27(1), 44–50. https://doi.org/10.1111/j.1471-6712.2012.00998.x Van Willigen, M. (2000). Differential benefits of volunteering across the life course. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 55(5), S308–S318. https://doi.org/10.1093/geronb/55.5.S308 Vecina, M. L., & Chacón, F. (2013). Volunteering and well-being: Is pleasure-based rather than pressure-based prosocial motivation that which is related to positive effects? Journal of Applied Social Psychology, 43(4), 870–878. https://doi.org/10.1111/jasp.12012 Veenhoven, R. (n.d.). World Database of Happiness. Retrieved October 15, 2018, from World Database of Happiness Archive of research findings on subjective enjoyment of life website: https://worlddatabaseofhappiness.eur.nl/hap_cor/desc_hind.php?ind=654 Windsor, T. D., Anstey, K. J., & Rodgers, B. (2008). Volunteering and psychological well-being among young-old adults: How much is too much? The Gerontologist, 48(1), 59–70. Wink, P., & Dillon, M. (2007). Do generative adolescents become healthy older adults? In S. G. Post (Ed.), Altruism & health: Perspectives from empirical research (pp. 45–55). Oxford University Press. Yang, Y., & Green, S. B. (2011). Coefficient alpha: A reliability coefficient for the 21st century? Journal of Psychoeducational Assessment. Retrieved from https://journals-sagepub- com/doi/10.1177/0734282911406668 Yuen, H. K., Huang, P., Burik, J. K., & Smith, T. G. (2008). Impact of participating in volunteer activities for residents living in long-term-care facilities. American Journal of Occupational Therapy, 62(1), 71–76. http://dx.doi.org/10.5014/ajot.62.1.71 152