STRESS, WELL-BEING, RETENTION, AND SOCIAL SUPPORT AMONG LICENSED FOSTER PARENTS By Elizabeth Sharda A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Social Work—Doctor of Philosophy 2020 ABSTRACT FOSTER PARENTS Elizabeth Sharda By STRESS, WELL-BEING, RETENTION, AND SOCIAL SUPPORT AMONG LICENSED Foster parents are the largest group providing care for some of our most vulnerable children- those in the foster care system. In the course of caring for children who have experienced significant trauma, loss, and uncertainty, foster parents face unique stressors specific to their role. As a result, many leave fostering after only a short period. Though the need for licensed foster parents is high, relatively little is known about the factors that impact their ability to remain in their role, and even less is known about what impacts their well-being within it. The purpose of this study was to use a conceptual model grounded in social support theory to examine the impact that stress and support have on foster parent retention and well-being. Specifically, I investigated the relationships between stressors of fostering and parenting stress, and between parenting stress and two outcome variables: well-being and the intent to continue fostering. Additionally, I examined social support as a potential moderator, or buffer, of each of these relationships. The study utilized a cross-sectional, non-experimental design and web-based survey methodology with a sample of foster parents from one Michigan county (N=139). In addition to descriptive and bivariate analysis, multiple linear regression and binary logistic regression were used to analyze the identified research questions. Results indicated that there was no relationship between the total number of stressors reported by foster parents and their levels of parenting stress. However, parenting stress was significantly higher among foster parents who reported certain stressors: foster child behavior problems, difficulty obtaining services, and disagreement with a licensing rule or policy. A significant, negative relationship was found between parenting stress and well-being. Additionally, social support did moderate this relationship, suggesting that social support serves as a protective factor for foster parents experiencing parenting stress. No significant relationship was found between parenting stress and the intent to still be fostering in 18 months. However, parenting stress was a significant predictor of past thoughts about giving up fostering. This study contributes to existing literature by adding to the small (but growing) body of quantitative studies conducted with U.S. foster parent populations. Further, the study uses a conceptual framework grounded in social support theory to increase understanding related to the impact of stress and support on foster parent well-being and retention. Among other notable findings, it offers the first known evidence that social support serves in a buffering role for foster parents, protecting their well-being even in the presence of high parenting stress. This has several important implications for this population and the professionals and systems they encounter, including improved foster parent training, assessment, and mentoring. Future studies should be conducted with larger and more representative samples of foster parents, particularly in terms of gender and race. In addition, research dedicated to developing a multidimensional measure of social support specific to foster parents would be particularly valuable. Copyright by ELIZABETH SHARDA 2020 This dissertation is dedicated to the foster family of Baby J, who walked through unspeakable tragedy and incredible stress with the love of their people and the strength of their God. v ACKNOWLEDGEMENTS In reflecting upon this process of researching social support of foster parents, I am profoundly grateful for my own web of support. I did not realize this goal on my own, but was helped along by family members, friends, mentors, and colleagues. In particular, I would like to thank: • My husband, Brian, for the support, selflessness, and reassurance that kept me going throughout this process. I couldn’t ask for a better teammate. • My daughter, who started kindergarten as I started my doctoral program. Watching her learn and grow is one of my greatest joys. May we both continue to be life-long learners. • My dissertation committee members, for their commitment, encouragement, and guidance. • My dissertation chair, Dr. Gary Anderson, whose compassion, wisdom, and confidence in me were always generously shared. • My fellow doctoral students, for their genuine friendship and laughter. • My colleagues at Hope College, for their belief in me and patience in my process. Lastly, I am grateful for the children and families in the foster care system whom I have had the privilege to know. Your resilience inspires and propels me. May we continually strive to build and re-build a system that truly results in your well-being. vi TABLE OF CONTENTS Uniqueness of the Foster Parent Role Stressors of Fostering Impacts of Fostering Foster Parent Resilience Factors Individual Stressors System Stressors The Foster Care System Foster Parenting Statement of the Problem Purpose of the Study Research Questions Significance of the Study LIST OF TABLES LIST OF FIGURES CHAPTER 1: INTRODUCTION CHAPTER 2: LITERATURE REVIEW Theoretical and Conceptual Frameworks Foster Parent Population Stressors of Fostering The Experience of Stress Among Foster Parents Outcomes of Stress Foster Home Retention Foster Parent Well-being Social Support of Foster Parents Types of Social Support Social Support Measurement Quantitative Findings Conclusion Individual Stressors System-related Stressors Ecological Framework Resilience Theory Social Support Definition of Social Support Types of Social Support Measurement of Social Support Theories of Social Support Foster Parent Demographics Motivations and Benefits of Fostering vii x xi 1 1 3 5 8 8 9 10 11 11 13 13 14 16 17 17 19 22 23 23 26 28 31 31 32 33 33 35 37 40 41 45 46 47 48 50 53 Study Design Procedures Age Gender Race Parent Relationship Education Employment Income Religious Participation Study Design and Procedures Participants Measures Analysis Sociodemographic Characteristics Foster Home Characteristics Total Time as a Foster Parent Total Number of Foster Children Current Number of Foster Children Kinship Status Stressors of Fostering Parenting Stress Well-being Retention Social Support Data Screening Preliminary Analysis Analysis of Research Questions CHAPTER 3: METHODS CHAPTER 4: DESCRIPTIVE RESULTS Sociodemographic Characteristics Foster Home Characteristics Stressors Parenting Stress Well-being Retention Social Support Conclusion CHAPTER 5: BIVARIATE AND MULTIVARIATE RESULTS Research Question 1 Research Question 2 Research Question 1, A & B Research Question 1, C Research Question 2, A & B 54 54 54 54 58 59 59 59 59 59 59 60 60 60 60 60 60 60 60 61 61 62 63 64 65 66 66 70 70 73 73 74 76 78 78 79 80 81 83 84 87 88 89 92 viii Stressors of Fostering Parenting Stress Well-being Retention Social Support Buffering Hypothesis Types of Social Support Research Question 1 Research Question 2 Research Question 3 Research Question 2, C Research Question 3 Conclusion Research Question 3, A & B Research Question 3, C & D Key Findings Limitations of the Study Conclusion CHAPTER 6: DISCUSSION CHAPTER 7: IMPLICATIONS AND CONCLUSION APPENDICES REFERENCES Implications for Practice Implications for Policy Implications for Social Work Education Implications for Research Conclusion 95 97 99 101 102 103 104 104 106 109 110 111 113 114 115 116 118 120 121 123 123 127 128 130 132 135 136 143 146 152 174 APPENDIX A: MSU IRB Determination Letter APPENDIX B: MDHHS IRB Determination Letter APPENDIX C: Invitation to Participate and Reminder Messages APPENDIX D: Survey Instrument ix Table 1 Typologies of social support LIST OF TABLES Table 2 Operational definitions of foster parent retention Table 3 Measures of foster parent support Table 4 Operational definitions of study variables Table 5 Scale reliability Table 6 Sociodemographic characteristics of the sample (N=139) Table 7 Foster home characteristics of the sample (N=139) Table 8 Stressors of fostering (N=139) Table 9 Well-being scores of the sample (N=134) Table 10 Retention results of the sample (N=138) Table 11 Social support results of the sample (N=139) Table 12 Pearson correlation results; Research question 1 Table 13 Independent samples t-test and ANOVA results for parenting stress Table 14 Multiple linear regression results for parenting stress Table 15 Independent samples t-test for parenting stress by individual stressors Table 16 Pearson correlation results; Research question 2 Table 17 Independent samples t-test and ANOVA results for well-being Table 18 Multiple linear regression results for well-being Table 19 Independent samples t-test for retention (intent to continue) Table 20 Chi square test of independence for retention (intent to continue) 24 41 48 61 69 73 75 77 79 79 81 84 85 88 89 90 91 94 97 98 Table 21 Binary logistic regression results for retention (intent to continue) 100 x Figure 1 Proposed buffering model of foster parenting Figure 2 Compensatory model of resilience Figure 3 Buffering model of resilience Figure 4 Buffering hypothesis of social support Figure 5 Proposed buffering model of foster parenting Figure 6 Proposed buffering model of foster parenting Figure 7 Interaction effect of parenting stress and social support on well-being Figure 8 Proposed buffering model of foster parenting 13 20 21 29 30 83 95 103 LIST OF FIGURES xi CHAPTER 1: INTRODUCTION Foster parents provide care for some of society’s most vulnerable children- those living in the foster care system. Foster parents voluntarily assume the incredibly challenging task of parenting children who have experienced significant trauma, instability, and loss and whose futures are uncertain. They encounter multiple sources of stress (Oke, Rostill-Brookes, & Larkin, 2011), both in their parenting duties (Beuhler, Cox, & Cuddeback, 2003; Cavazzi, Guilfoyle, & Sims, 2010) and in their dealings with the foster care system (Blythe, Wilkes, & Halcomb, 2014) and the many professionals involved. They receive training and support that are inadequate given the needs they face (Eaton & Caltabiano, 2009; Murray, Tarren-Sweeney, & France, 2011). As a result, many leave fostering after only a short period (Gibbs & Wildfire, 2007). The need for licensed foster homes is abundant (U.S. Children’s Bureau, 2019); therefore, it is vital that this population be understood, particularly in terms of what might promote well-being and longevity in their role. The Foster Care System Each year, the U.S. foster care system serves more than half a million of our country’s children, working to promote their safety, well-being, and permanency in circumstances of abuse, neglect, and other forms of maltreatment (U.S. Children’s Bureau, 2019). On September 30, 2018, there were 437,283 children (age 0-20) in the foster care system (U.S. Children’s Bureau, 2019). A quarter of a million additional children encountered the foster care system at some point during the previous year (687,345 total children served by the foster care system in FY2017-2018; U.S. Children’s Bureau, 2019). Within that period, 262,956 children entered foster care, and 250,103 exited. In fact, more children have entered care than have left care in every year since 2011, though the gap between the two figures has somewhat narrowed recently 1 (U.S. Children’s Bureau, 2013). As a result, the total foster care population has swelled since reaching a historic low in 2012 (397,153; U.S. Children’s Bureau, 2015; U.S. Children’s Bureau, 2019). Children enter foster care for various and multiple reasons related to their safety and wellbeing. By far, the most common reason for removal and placement in foster care is neglect (62% of children in care on 9/30/17; U.S. Children’s Bureau, 2019). Children and families also experience parental drug abuse (36%), caretaker inability to cope (14%), physical abuse (13%) and housing problems (10%), among other challenges. Each of these can be a source of trauma, as can removal and placement in foster care. As a result, foster children experience post- traumatic stress disorder (PTSD) at alarmingly high rates, with one study of former foster youth finding that 25.2% had experienced PTSD in the previous year, a rate nearly twice that of U.S. war veterans (Pecora et al., 2005). More recently, Haselgruber, Sölva, & Lueger-Schuster, (2020) found that 31.6% of foster children in the study sample demonstrated symptoms of PTSD, while an additional 22.8% were categorized as having Complex PTSD. Many children in foster care have experienced complex trauma, defined as “recurrent interpersonal trauma perpetrated by caregivers early in life” (Greeson et al., 2011, p. 92). In a large study of foster children in trauma treatment through the National Child Traumatic Stress Network, complex trauma was operationalized as exposure to two or more interpersonal trauma types (sexual abuse, physical abuse, emotional abuse, neglect, domestic violence; Greeson et al., 2011). Researchers found that 70.4% of the sample (Total N=2,251) met the definition of complex trauma, with an astonishing 11.7% endorsing all five interpersonal trauma types. Children exposed to complex trauma experience impairment across numerous domains, including attachment, biology, cognition, behavioral control, and affective regulation (Cook et 2 al., 2005). In the Greeson et al. study (2011), children in foster care who had experienced complex trauma had significantly higher likelihood of having internalizing problems (OR = 1.6), posttraumatic stress (OR = 1.5), and clinical diagnoses (OR = 1.2), when compared with children who had experienced other types of trauma. The needs of children in foster care are heightened, due to the developmental, behavioral, mental health, learning, and physiological concerns they face. At least a quarter of foster children experience developmental delays (Leslie, Gordon, Ganger, & Gist, 2002). Children in foster care are more likely than same-aged peers to display behavioral problems (Lohaus, Chodura, Möller, Symanzik, Ehrenberg, Job, & Heinrichs, 2017; Tarren-Sweeney, 2008; Turney & Wildeman, 2016). In terms of mental health, foster children display high prevalence of conduct disorder, attention deficit/hyperactivity disorder, depression, and generalized anxiety disorder (Tarren- Sweeney, 2008; Turney & Wildeman, 2016). Learning disabilities are nearly twice as common among foster children versus children in the general population (Turney & Wildeman, 2016). Medical problems are also more common among children in care: one study found that 30% of foster children had three or more health conditions, while fewer than 20% had none (Halfon, Mendoca, & Berkowitz, 1995). Specifically, foster children are more likely than non-placed peers to experience asthma, obesity, speech problems, hearing problems, and vision problems (Turney & Wildeman, 2016). In a variety of domains, the needs of foster children are greater than children in the general population. Foster Parenting Considering the swelling rolls of foster care and the considerable needs of the children in care, it is important to examine where these children go when they enter care. The goal of the child welfare system is to attain the safety, permanency, and well-being of these children within 3 the least restrictive environment, that is, the most family-like setting that can suitably meet the child’s needs (Child Welfare Information Gateway, 2018). While children are placed in a variety of settings (group homes, residential facilities, supervised independent living, etc.), the largest percentage of children (46%) are placed in non-relative licensed foster homes (U.S. Children’s Bureau, 2019). Additionally, 32% are placed in relative foster homes, resulting in 78% of children in out-of-home care residing in foster homes (337,757 children in September 2018; U.S. Children’s Bureau, 2019). There is an increasing need for licensed foster families (both relative and non-relative) to meet the needs of an expanding child welfare system (U.S. Children’s Bureau, 2019). However, most states experienced a decline in foster home capacity between 2012 and 2017 (Chronicle of Social Change, 2017). While the histories, needs, and outcomes of children in care are widely studied, comparatively little is known about the experiences of the people who are most likely to be providing care for them- foster parents (Blythe et al., 2014). Foster parents are “relative or non-relative adults who have been approved by the State, or by an agency licensed by the State, to provide [children in foster care] with shelter and care” (Child Welfare Information Gateway, n.d.). The foster parent role lies somewhere between that of a parent and a professional (Blythe et al., 2014), spanning both public and private spheres (Kirton, 2001). Foster parents perform the tasks faced by all parents, whether those be late night feedings for a two-month old, soccer practice shuttling of a ten-year-old, or teenage curfew negotiations. However, their role is also a unique one. They face additional duties inherent in the charge of caring for children who have experienced considerable abuse, neglect, and loss. Further, they carry out these duties as temporary caregivers operating under the supervision of the child welfare system. 4 Uniqueness of the Foster Parent Role Foster parenting goes beyond normative experiences of parenting, often being referred to as more of a “re-parenting” role (Murray et al., 2011; Oke et al., 2011; Tarren-Sweeney, 2008), and being performed by foster parents who lack the familiarity and history with the children that most parents possess (Lanigan & Burleson, 2017; Lohaus et al., 2017). Orme and Combes-Orme (2014) speak to this element of fostering: “Frequently, foster parents do not even have complete knowledge of their foster children’s previous lives and may need to adjust their parenting as they go to deal with children who have been scarred by sexual and physical abuse, deprivation, and chronic insecurity” (p. 125). Certainly foster parents are not the only parents who lack knowledge of their children’s early lives. This may also be true of adoptive parents and step- parents. However, even among these families, there are elements of the foster parent role that are unique, most notably, the complex trauma exposure among children, operating within the oversight of the social service system, the temporary nature of foster care, and the limited authority of foster parents. As noted above, foster children have been exposed to traumatic experiences. The maltreatment leading to their contact with the foster care system can be traumatic, as can the experience of removal from parents and, all too frequently, separation from siblings upon placement in care. Additionally, foster children can be further traumatized within the system, through abuse or neglect in foster homes, loss of connection through the termination of parental rights, and the constant unpredictability inherent in the pursuit of permanency. Almost as a rule, foster parents operate within a context of trauma that is much less common among parents in the general population. They must not only handle the emotional and behavioral consequences of this trauma, but also must consider how to help children in their homes feel safe, a difficult task 5 considering the lack of safety they have likely experienced in the past. This pursuit of psychological safety for children guides decisions about physical touch, food, discipline, and more (Priesler, 2013). For example, many foster parents make boundaries around bedrooms clear, establishing space “where they can go and be by themselves and, just give them that safe place right away where this is yours, nobody else needs to be in here, has to come in” (Lanigan & Burleson, 2017, p. 908). Foster parents make explicit statements such as, “This is your safe place. Nobody touches you inappropriately.” (Lanigan & Burleson, 2017, p. 908). While all parents wish for and work for their children to feel safe, this task is particularly salient and especially challenging for those parents (including foster parents) caring for children whose safety has been violated in the past. Fostering is unique among parenting roles because foster parents operate within and as part of the foster care system, comprised of case workers, licensors, child protection workers, attorneys, judges, and therapists, among others (Lanigan & Burleson, 2017). Foster parents must be aware of the expectations, timelines, policies, and politics of the system. They must learn to “work the child welfare system” (Marcellus, 2010, p. 18), from ensuring that their homes meet state licensing requirements to obtaining needed services for the children in their care. They also must consider that their parenting actions and decisions are observed, documented, and sometimes critiqued by others in the system. Marcellus (2010) called this role one of “public parenting” (p. 19), in that foster parents operate under the scrutiny of the child welfare system and the public eye, often being held to expectations higher than those faced by other parents. Foster parenting is also unique in terms of its impermanence. Foster care is, by definition, temporary (Pickin, Brunsden & Hill, 2011). It is intended to serve as a stopgap solution to conditions within families that endanger children’s safety and well-being. It is meant to be short- 6 term, and as a result, foster parents are, necessarily, short-term parents, parenting within a “context of impermanency” (Lanigan & Burleson, 2017, p. 913). Foster parents are, in theory, prepared for this dynamic through pre-service training and licensure. However, the reality of ‘temporary parenting’ can be more challenging than anticipated. In her study on the experiences of foster parents, Marcellus (2010) acknowledged the “emotional double bind,” in which foster parent are expected to invest emotionally in the children in their care, and yet be prepared to release them, either to return to their families of origin, to another relative, or to adoptive parents. Sometimes this release happens at a moment’s notice and, occasionally, without any further contact. Hebert and Kulkin (2016) also discuss the difficulty of temporary parenting: “The current child welfare system expects certified non-relative foster parents to take vulnerable children into their own homes and love and nurture them as if they were their own, yet be willing to give them up when the agency or the court system decides that the birth parents are able to resume looking after them” (p. 129-130). This “context of impermanency,” though necessary, can also be unsettling for the children in foster care. It can be difficult to establish a sense of belonging, particularly for the many children who experience multiple placements. As one foster parent stated, “They were always kind of expecting us to return them for any given reason at any given time” (Lanigan & Burleson, 2017, p.909). Another unique aspect of fostering is the limited authority held by foster parents. As noted, foster parents share responsibility for the children in their care with a host of child welfare professionals, as well as the parents from whom the children were removed. Since they are filling a temporary role, they retain only partial authority over decisions related to their foster children. Birth parents retain some rights, such as the right to approve psychotropic medications, non-emergency surgeries, interstate travel, and even haircuts. Judges ultimately hold power over 7 long-term decisions in the lives of foster children. While this division of authority is a critical part of the foster care process, it can result in a feeling of “powerless responsibility” among foster parents (Marcellus, 2010, p. 19), in which they hold sole responsibility over day-to-day decisions (e.g. what to make for dinner, bedtime, how to discipline, etc.) and little to no input over long-term decisions (e.g. Will the child return home or not?). As stated by one foster parent, “I had all the care and responsibility but no power” (Thomson & McArthur, 2009, p. 74). Stressors of Fostering Given the unique duties of the fostering role and the context in which it is performed, there are, understandably, significant stressors experienced by foster parents. Sources of stress include children’s needs and behaviors, navigating the complicated child welfare system, financial strain, and stigmatizing societal views of fostering (Denby, Rindfleisch, & Bean, 1999; Farmer, Lipscombe, & Moyers, 2005; Geiger, Hayes, & Leitz, 2013; Miller, Green, and Lambros, 2019). Stressors related to fostering can be grouped as individual stressors and system- related stressors (Adams, Hassett, & Lumsden, 2018). Individual Stressors. Child-related stressors can include difficult behaviors (Buehler et al., 2003; Cavazzi et al., 2010), physical health problems (Lanigan & Burleson, 2017; Lietz et al., 2016), or mental health needs (Barnett et al., 2017; Lietz et al., 2016). Children entering the home can be a source of stress for foster parents (Lanigan & Burleson, 2017; Lietz, Julien-Chinn, Geiger, & Piel, 2016), as can children leaving the home (Buehler et al., 2003; Lietz et al., 2016; MacGregor, Rodger, Cummings, & Leschied, 2006; Samrai, Beinart, & Harper, 2011; Shklarski, 2019; Thomson & McArthur, 2009). Both events include significant changes to the family structure and schedule, while the departure of a child in particular can bring about ambiguous 8 loss (Thomson & McArthur, 2009) and grief (Donachy, 2017; Hebert & Kulkin, 2016; Pickin et al., 2011) in foster parents. System Stressors. Perhaps the greatest source of stress for foster parents is the child welfare system itself. Fostering literature suggests that many foster parents find the system to be much more demanding than the children for whom they provide care (Blythe et al., 2014). Foster parents report challenges in obtaining necessary services for their foster children (Buehler et al., 2003; MacGregor et al., 2006), challenging relationships with birth parents (Buehler et al., 2003; Cavazzi et al., 2010; Lanigan & Burleson, 2017), incomplete information about children in their care (Buehler et al., 2003; Cavazzi et al, 2010; Lanigan & Burleson, 2017; Samrai et al., 2011), and inadequate crisis response from foster care professionals (Cavazzi et al., 2010; MacGregor et al., 2006). They also identify an overall lack of recognition and respect as members of the foster care team (Murray et al., 2011) manifested as a lack of support and training (Murray et al., 2011) and exclusion from the decision-making process (Buehler et al., 2003; Cavazzi et al., 2010; Samrai et al., 2011). Worker turnover is problematic for foster parents (Buehler et al., 2003; Murray et al., 2011), as is poor communication from workers (Cavazzi et al., 2010; Lanigan & Burleson, 2017; Murray et al., 2011). Finally, foster parents report that allegations of abuse or neglect against them are perhaps one of the most stressful experiences of all (Murray et al., 2011; Nixon, 1997). Considering the number and needs of children in the foster care system and the fact that foster homes are the most commonly used placement resource for these children, it is critical to explore foster parents’ experiences of stress as well as their ability to maintain well-being and remain in their roles. 9 Impacts of Fostering While many foster parents report that fostering is intrinsically rewarding (Buehler et al., 2003; Cavazzi et al., 2010; Geiger et al., 2013) and also brings benefits to themselves (Pickin et al., 2011) and their families (Geiger et al., 2010; Samrai et al., 2011), there is also a cost to fostering. As a result of the stressors noted above, foster parents experience greater parenting stress than parents in the general population (Vanschoonlandt, Vanderfaeillie, van Holen, de Maeyer, & Robberechts, 2013). Stressors of fostering are also associated with decreased overall well-being among foster parents (Wilson, Sinclair, & Gibbs, 2000). Foster parent well-being is essential, for the parents as well as for the children in their care. However, research on the well- being of foster parents is very limited (Hannah & Woolgar, 2018; Miller, Green, & Lambros, 2019). Foster parents often do not remain in their role, with as many as 62% leaving within the first year (Gibbs & Wildfire, 2007; Kangas, 2015). One study (Gibbs & Wildfire, 2007) found that the median length of service for foster families is only 8-14 months- less than the amount of time the average foster child will spend in care (mean: 19.2 months, median: 14.7 months; U.S. Children’s Bureau, 2019). This concerning disparity means that many children in care will experience multiple foster home placements, adding further to their experiences of instability and compounding loss. Further, the retention problem among foster parents “creates both a service and an experience gap” (Lanigan & Burleson, 2017, p. 905) that increases pressure on foster care agencies to recruit, train, and license new foster homes and increases the likelihood that children entering care or needing re-placement will be placed with new and inexperienced foster families. 10 Foster Parent Resilience Factors While the stressors facing foster parents are substantial, research points to factors that may help lessen the impact of such stressors. Fostering literature identifies several individual traits or factors that may contribute to foster parent resilience, such as emotional strength (Oke et al., 2011), assertiveness (Lietz et al., 2016; Maclay, Bunce, & Purves, 2006), flexibility (Buehler et al., 2003), humor (Lietz et al., 2016), empathy (Gieger et al. 2016), ability to set boundaries (Lietz et al., 2016), and religious beliefs and participation (Buehler et al., 2003; Lietz et al., 2016). Additionally, factors in the foster parent’s environment may also contribute to their resilience, including family connectedness and communication (Lietz et al., 2016), good relationships with child welfare professionals (Gieger et al., 2017; Murray et al, 2011; Pickin et al., 2011; Rodger et al., 2006; Samrai et al, 2011), and societal understanding of foster care (Piel, Geiger, Julien-Chinn, & Lietz, 2017). An additional environmental factor that may contribute to foster parents’ ability to remain well in their roles is social support (Eaton & Caltabiano, 2009; Farmer et al., 2005). However, support is often lacking. Foster parents report a disparity between the levels of support they receive and the high burden of caring for foster children (Murray et al., 2011). Formal support from professionals within and outside the child welfare system, as well as informal support from friends, family, and others are both key, according to foster parents (Orme et al., 2006; Rodger et al., 2006). However, many describe a lack of support from formal (i.e. professional) sources (Cavazzi et al., 2010; Samrai et al., 2011) and isolation in their own social networks (Nixon, 1997). Statement of the Problem What I have described in the preceding pages is something that I’ve lived as a licensed foster parent. I have personally experienced the stress of the role and felt my own well-being 11 suffer as a result. I also know the critical importance of the role and have seen firsthand the positive impact that stability in one foster home can make in the life of a young person in care. Thanks to the supports in place around us- family, friends, religious community, our foster agency, and other foster parents, my husband and I have been licensed for over a decade. Having navigated the ups and downs of foster parenting, I am especially interested in exploring those factors that help foster parents to remain in their role and remain well in their role. Foster parents are prematurely leaving their roles (Gibbs & Wildfire, 2007), while the number of children in foster care and subsequent need for licensed foster homes remain high (Chronicle for Social Change, 2017; U.S. Children’s Bureau, 2019). More research is needed on the impact of fostering on foster parents, specifically, the relationship between the stress of the role and foster parents’ well-being and longevity in their position. Further exploration is also needed regarding those factors which might promote resilience in this indispensable role. Social support is one factor with some preliminary evidence demonstrating its connection to positive outcomes for foster parents and the children in their care (Eaton & Caltabiano, 2009; Farmer et al., 2005; Richardson, Futris, Mallette, & Campbell, 2018). However, the depth and breadth of research on social support among foster families is lacking, and what studies are available have not approached the topic in a consistent way, theoretically or methodologically. Few studies examining social support and fostering include conceptualizations that acknowledge the multidimensionality of the former. Further, few of these studies include rationalization for why social support is operationalized the way that it is, or precisely why it is examined as an independent variable, moderator, or mediator. Studies that acknowledge the complexities of social support and connect these to methodological choices are largely absent from the fostering literature. 12 Purpose of the Study The current study aims to address these gaps in research by utilizing a social support theoretical framework to examine stress, well-being, retention, and social support among licensed foster parents. The purpose of the current study is to increase knowledge on the impact of stress and support on foster parents’ well-being and retention in their role. The following theoretical model was utilized to frame the research, and is further explicated in Chapter 2: Literature Review. The foster parent outcomes explored in the current study include foster parent well-being and foster home retention. Figure 1 Proposed buffering model of foster parenting Research Questions Based on the proposed model, the following research questions were used to guide the study: RQ1. A. Is there a relationship between stressors experienced by foster parents and their levels of parenting stress? B. Does social support moderate this relationship? C. Is there a relationship between individual stressors and parenting stress? RQ2. A. Is there a relationship between parenting stress and well-being among foster parents? B. Does social support moderate this relationship? C. Do the various types of social support moderate this relationship? RQ3. A. Is there a relationship between parenting stress and retention (intent to continue) among foster parents? 13 B. Does social support moderate this relationship? C. Is there a relationship between parenting stress and thoughts about leaving D. Does social support moderate this relationship? fostering? Significance of the Study Within the context of a child welfare system that increasingly aims to place children in the least restrictive environment, licensed foster homes are indispensable. The current study contributes to the literature on fostering by increasing understanding of the impact of fostering on foster parents. Specifically, this study contributes to a small (but growing) body of quantitative research conducted on a U.S. foster parent population. Further, this study utilizes a conceptual model grounded in social support theory and a multidimensional conception of social support. While a small number of existing studies have examined the role of social support among foster parents, this research does not consistently reflect an understanding of or adherence to social support theory. This study also examines two essential outcomes: foster parent retention and foster parent well-being. While foster parent retention receives more attention within the child welfare system, as well as fostering literature, the well-being of foster parents is of equal, or perhaps even greater, importance. It is critical that we understand not only what contributes to foster parents’ longevity in their roles, but also what keeps them well. The knowledge resulting from this study can be used to inform child welfare policy on foster parent licensure, training, and retention. It can be used to develop both pre-service and in-service training curricula on common stressors of fostering, types and sources of social support, and how to assess and build one’s own support network. Further, it can be used to shift child welfare practice, in terms of support and retention programming, but also in terms of worker-foster parent relationships that are less stressful and more supportive (Lietz et al., 2016). Given the sustained need for licensed 14 foster homes, and for healthy foster parents within those homes, this study offers a valuable contribution to our collective understanding of the impact of stress and support on foster parent retention and well-being. 15 CHAPTER 2: LITERATURE REVIEW Current fostering literature is lacking in theoretically-grounded research on social support. In order to address this gap, it is essential that the current study be conceptualized, methodologically designed, and implemented with acknowledgement of and in accordance with established theory. The following literature review will begin with an overview of the theoretical and conceptual frameworks at the foundation of the current study, including the ecological framework and resilience theory. I will briefly describe these and their contributions to the topic of foster parenting. Building upon these, I will provide an in-depth explication of the primary conceptual framework: social support. As a meta-construct that has been referred to as “fuzzy” due to its incredible breadth, social support requires careful clarification before attempting to understand its relevance to the foster parent population (Hupcey, 1998, p. 1231). To this end, I will examine various social support constructs, types of support, and conclude with the buffering hypothesis of social support (Cohen & Hoberman, 1983; Cohen & McKay, 1984). A theoretical model, based on this foundational knowledge, will be proposed and used to guide the subsequent review of fostering literature. The review of empirical literature will begin with an overview of the foster parent population, including demographics, motivations, and benefits of fostering. Next I will review literature on stressors and stress in the foster parent role, as well as research related to potential outcomes of that stress, specifically foster parent well-being and retention. Lastly, I will include a detailed review of social support as one factor that potentially buffers against foster parent stress. 16 Theoretical and Conceptual Frameworks In seeking to understand the resilience of foster parents, there are at least three theoretical and conceptual frameworks that prove valuable: the ecological framework, resilience theory, and social support. Ecological Framework The ecological framework, as proposed by Bronfenbrenner (1977), is so foundational to the practice of social work that it often does not require explicit mention. It is a lens through which we view our client systems, their strengths, their challenges, and potential targets for change. The individual is not viewed in isolation, but rather in the nested context of his or her surroundings. The framework seems particularly relevant to the present examination of foster parents, and is therefore included here. According to Piel, Gieger, Julien-Chinn, & Lietz (2017), “Considering that foster families interact with a variety of systems, as well as varying levels of family and community, research and practice with this population should be approached from an ecological perspective” (p. 1036). Foster parents perform their roles within a context that includes an array of systems, each of which influence and are influenced by the foster families and by each other: nuclear family, work, extended family and friends, foster care and other social service agencies, the state and federal child welfare systems, broader local and national cultures and laws. Piel and colleagues (2017) utilized Bronfenbrenner’s (1977) framework to describe the levels of a foster family’s ecology: • A microsystem is the complex of relations between the developing person and environment in an immediate setting containing that person (Bronfenbrenner, 1977, p. 514). For foster parents, this can include spouses, partners, children, extended family, 17 friends, neighbors, coworkers, clergy and others in faith communities, and licensing and foster care workers. • A mesosystem includes the linkages and direct interactions that occur within the microsystems (Bronfenbrenner, 1977). For foster parents, as for anyone, these interactions can be healthy or unhealthy, and subsequently contribute to well-being and other outcomes, or detract from them (Piel et al., 2017). • An exosystem consists of interactions between two or more social structures that do not contain an individual, yet impact that individual and his or her interactions (Bronfenbrenner, 1977). Foster parents are affected by policies, decisions, and interactions among structures in which they are not immediately included. For example, foster parents do not necessarily have involvement in the decision to terminate parental rights of a birth parent of a child in their care; however, they are most certainly affected by such a decision. • A macrosystem consists of the “blueprints” that guide the structures and activities occurring at the concrete level in other ecosystem levels (Bronfenbrenner, 1977, p. 515). For foster parents, this includes the political, economic, and social climate of the community, as well as state and federal government. If the political climate is one of scarcity in terms of support and funding for social services, foster parents will experience the effects of this. • Lastly, the chronosystem refers to the impact of events over the course of time upon the interactions across systems. For example, a publicized death of a foster child while in care would, over time, impact the policies and practices that guide interactions between foster care agencies and the foster parents they license. 18 Attention to the social ecology of individuals and families is inherent in social work practice and research. For the purposes of this study, attention to the unique environmental context of foster parenting is crucial, as is the recognition that systems within the environment can serve as sources of both stress and support. Within the ecological framework is found the concept of goodness of fit, which “refers to the degree to which a system is situated in an environment that has the resources needed to allow it to thrive” (Piel et al., 2017, p. 1041). Viewing foster parents through the lens of the ecological framework means not only considering the context of their environments, but examining the extent to which these environments contribute to, or detract from, their overall well-being. Foster parent well-being is not a function of solely individual factors, but also of the resources and stresses in their environments. According to Piel et al. (2017), “Families who foster need to experience an ecology of support that … contributes to their capacity to foster” (p. 1042). This study aims to increase understanding of this ‘ecology of support’ and its relationship to foster parent well-being in the performance of a critical, yet often stressful, role. Resilience Theory The view of one’s environment as a source of stress and support is congruent with resilience theory. In fact, resilience theory is congruent with the ecological framework in that when there is goodness of fit between people and their environments, resilience is enhanced (Greene, 2014). Environments that are nurturing, providing needed support and resources, contribute to individual and system resilience (Greene, 2014). Resilience, according to Masten (2014), is “the capacity of a dynamic system to adapt successfully to disturbances that threaten system function, viability, or development” (p. 10). In contrast to early psychological theories 19 based on a pathology approach, resilience theory examines factors that contribute to positive outcomes even in the presence of considerable stress. Key constructs within resilience theory include risk, or vulnerability, factors and protective factors. Risks are factors that increase the likelihood of negative outcomes following adversity (Greene, 2014). Protective factors are those that serve to decrease risk and increase adaptation (Greene, 2014). Protective factors may be personal traits or environmental conditions (Greene, 2014). Resilience is conceptualized as a process in which these risk and protective factors develop and dynamically interact at multiple ecological levels to result in adaptive outcomes (Greene, 2014; Marcellus, 2010). The processes by which risk and protective factors influence outcomes are illustrated by two models of resilience: the compensatory model and the buffering model. Within the compensatory (or main effect) model, risk and protective factors each have direct relationships to the outcome of interest (Fig. 2; Masten, 2014). Practically speaking this means that the protective factor increases likelihood of a positive outcome whether or not risk is present. In contrast, the buffering model requires the presence of risk. The protective factor, rather than influencing the outcome directly, influences the relationship between risk and outcome (Fig. 3; Masten, 2014). Figure 2 Compensatory model of resilience Note. Adapted from Masten, A. (2014). Ordinary Magic: Resilience in Development. New York, NY: Guilford Publications. 20 Figure 3 Buffering model of resilience Note. Adapted from Masten, A. (2014). Ordinary Magic: Resilience in Development. New York, NY: Guilford Publications. Resilience theory is often applied to the study of individuals, but “the concept of resilience can be applied to any dynamic system, including a family, a school, a community, an organization, an economy, or an ecosystem” (Masten, 2014, p. 11). Family resilience is an important branch of the broader study of resilience. Walsh defines family resilience as “the ability of the family, as a functional system, to withstand and rebound from adversity” (2016, p. 14). Further, “Family resilience is not an outcome, but a culture of coping and adaptation” within the context of change and stress (Lietz et al., 2016, p. 663). This culture of coping and adaptation includes processes within and around a family which enable it to manage and respond to stress, both chronic and crisis-driven. Family-level factors can be protective and contribute to overall resilience. Such factors, according to Walsh (2016), include the ability to make meaning out of challenging circumstances, flexibility, collaborative problem solving, open communication, and social support. Resilience and family resilience theories offer a helpful foundation for understanding how some foster families persist in their roles, within a context of often continuous stress. “Adaptation for [foster] families did not involve reaction to one particular loss or crisis; instead, families were continually facing new challenges with each transition involving a new set of 21 stressors requiring coping and another process of adaptation. Very few family systems require the level of ongoing adaptation to the structure, roles, boundaries, and daily activities as that of families who provide foster care” (Lietz et al., 2016, p. 663). This “ongoing adaptation” refers to the fact that foster families in particular often do not face a singular transition or crisis, but multiple challenges at the same time with new challenges surfacing continually. Considering the impermanence and uncertainty inherent in the foster care system, it is logical that this uncertainty extends to foster families caring for children within that system. A foster family might simultaneously be facing the departure of one child (Challenge #1), the entrance of another foster child (Challenge #2), and a change in foster care case worker (Challenge #3). Each of these transitions alone introduces new stressors and necessitates the adaptation of the foster family and the individuals within it. Further, foster parents must constantly be anticipating the next challenge, whether it be a difficult conversation with a birth parent, a child being sent home from day care, or notification that they have violated a foster home licensing rule. This context of continual, often unpredictable, stressors highlights the need for attention to those factors that promote positive outcomes among foster parents. Social support is one environmental factor that may contribute to resilience within this stressful role. Social Support Despite the fact that social support has been widely studied since the mid-1970s, a consistent and clear definition is elusive (Hupcey, 1998; Underwood, 2000). According to early social support theorists Cohen and McKay (1984), “The investigators do not agree upon a meaning of the term social support, a technique for measurement of support, or a conception of the mechanism(s) by which it presumably operates” (p. 254). This absence of agreement has persisted since that time, and may have hindered the research on social support and its 22 development as a coping resource applicable to myriad populations (Underwood, 2000), including foster parents. Definition of Social Support. Social support may lack a consistent definition because it is not a singular construct, but rather a meta-construct comprising multiple, distinct constructs or dimensions (Sarason & Sarason, 2009; Vaux, 1988). It may be too complex and broad to be captured in a singular definition (Sarason, Sarason, & Pierce, 1990; Vaux, 1992). According to Vaux (1988), social support, as a meta-construct, includes the dimensions of support networks, supportive behaviors, and subjective appraisals of support. Support networks, or social networks, are conceptualized as “the structure of social relationships- the existence, quantity, and type of relationships” (Cohen, 1992, p. 109). Supportive behaviors, or received support, refer to “the mobilization and receipt of behaviors intended to aid persons in the face of stressful events” (p. 109). Subjective appraisals of support, also referred to as perceived social support, are “the function of social relationships- the perception that social relationships will (if necessary) provide resources such as emotional support or information” (p. 109). Types of Social Support. Typologies of social support are varied and reflective of its multiple dimensions. Support can be typed by its source, for example. Social support can originate from formal sources, which would include those in professional helping roles (social workers, therapists, clergy, etc.), or from informal sources (family, friends, neighbors, coworkers, etc.). The main method for classifying types of social support is based on the function of the support- what does it provide (or what is it perceived to provide) to the recipient? Social support theorists differ in the number of types they present; however, there are some consistent 23 categories (Vaux, 1992). Cutrona and Russell (1987, p. 40) provide a summary of various functional support typologies (Table 1). Table 1 Typologies of social support Weiss (1974) Cobb (1979) Kahn (1979) Attachment Affect Emotional Support Network Support Social Integration Reassurance of Worth Reliable Alliance Material Support Aid Esteem Support Affirmation Guidance Opportunity for Nurturance Instrumental Support Active Support Schaefer et al. (1981) Emotional Cohen et al. (1985) Belonging Emotional Tangible Tangible Informational Appraisal Note. Adapted from Cutrona, C. E. and Russell, D. (1987). The provisions of social relationships and adaptation to stress. In W. H. Jones & D. Perlman (Eds.) Advances in Personal Relationships (Vol. 1, pp. 37-67). Greenwich, CT: JAI Press. The current study will utilize the typology proposed by Weiss (1974) and adopted by Cutrona and Russell (1987) for their development of the Social Provisions Scale (SPS). This typology is functional, in that it relates to the intended function of support (e.g. reassurance of worth) as opposed to a structural approach, which would reflect a social network dimension of support and would classify based on amount or density of support sources. The Weiss typology, and subsequent Social Provisions Scale, also utilizes the construct of perceived support (supportive appraisals), rather than received support (supportive behaviors). According to this typology, support can be divided into two groups: assistance related and non-assistance related (Cutrona & Russell, 1987). Assistance-related support types are those most relevant to problem- 24 solving in the face of stress, and include guidance and reliable alliance. Guidance includes advice or information, usually specific to the problem at hand, and is also called informational or appraisal support (Cutrona, 1990). Reliable alliance is the perception that one can depend on others for material assistance. This is also referred to as material or tangible support, and can be in the form of financial assistance, resources, help with tasks, and needed services (Cohen, 2004; Cohen & Wills, 1985). Non-assistance related support types are those which are more related to affectional ties and self-efficacy than to problem-solving (Cutrona & Russell, 1987). Non-assistance related support types include: attachment, reassurance of worth, social integration, and opportunities for nurturance. Attachment, also known as emotional support, is characterized by actions that lead one to believe that one is loved and cared for. (Note: This conceptualization of attachment is notably distinct from the use of the term in the context of developmental psychology). Reassurance of worth is the recognition of one’s skills, competence, and value by others, and is sometimes called esteem support. Social integration is provided by relationships in which individuals share concerns and create a sense of belonging and companionship. Finally, Weiss (1974) included in his typology one type of social support in which the individual gives, rather than receives support. Weiss called this type ‘opportunities for nurturance,’ but it has also been conceptualized as mutual support, as it focuses on the sense that one provides as well as receives support within a relationship. It is also important to consider that while most descriptions of social support and its types are positive, not all social support is beneficial and some can be even harmful (Hupcey, 1998). Relationships within social networks are not solely helpful, and may, in fact, be demanding or draining for an individual. Further, the support offered may be intended as positive, but received 25 as negative. Hupcey (1998) gives the example of a support provider encouraging smoking cessation, which could be perceived as intrusive or judgmental by the recipient. Conversely, support could be inherently negative (i.e. a support provider encouraging harmful behaviors like smoking in the context of the relationship). Finally, social support can be negative when it is not reciprocal (Hupcey, 1998). Persons who are always on the receiving end of support may come to resent such support and view themselves as inadequate. While conceptually distinct, certainly there are overlaps among these types in practice. For example, the provision of tangible support (e.g. bringing a meal to a sick friend) may be received as evidence of care and affection and therefore could also be viewed as attachment support. Since social support is a multidimensional concept, it is essential that studies of social support delineate which aspects of support are being studied and utilize measurement tools that are designed to capture them (Cutrona & Russell, 1987). Measurement of Social Support. Methods for empirical measurement of social support are as varied as its many definitions and types. Therefore, it is essential that measures are selected with deliberate attention to exactly what the researcher means to assess. According to Gottlieb & Bergen (2010), “The measurement of social support requires clarity about the aspects of the social surround that are most relevant to the aims and context of research, and precision in their measurement” (p. 511). However, this clarity is often lacking, even in the development of support measures. “At the theoretical level, the authors of social support measures have rarely articulated the assumptions underlying their instruments” (Cutrona & Russell, 1987, p. 37). Categories of social support measurement tools reflect its varied definitions. Some measures assess for network support, or social integration. These measure the number of relational ties reported by an individual as well as the density or interconnectedness of these relationships 26 (Lakey & Cohen, 2007). Cohen’s (1991) Social Network Index is one example of a social integration measure. Measures of received or enacted support ask individuals to report whether or not they have received certain supportive actions over a designated period of time (e.g. 30 days; Lakey & Cohen, 2007). Some measures ask the frequency at which they’ve received these supportive actions, or the adequacy of the actions. The Inventory of Socially Supportive Behaviors (ISSB; Barrera, Sandler, & Ramsey, 1991) is a widely utilized measure of enacted support. Measures of perceived support capture an individual’s subjective appraisal of the availability of support. Such measures ask respondents to report the extent to which assistance would be available to them if needed. Some perceived support measures are global, asking about support in general. Others specify by source of support, such as the Significant Other Scale (SOS; Weinman, Wright, & Johnston, 1995), or function of support, such as the Social Provisions Scale (SPS; Cutrona & Russell, 1987). Social support measures are typically subjective and utilize self-report methods via survey or interview (Lakey & Cohen, 2007). Other assessment methods, used far less frequently, include observation and diary measures. The Social Support Behavior Code is an observational tool that assesses the extent to which a provider displays specific and observable supportive behaviors in a conversation (Lakey & Cohen, 2007). Diary methods ask respondents to regularly (usually daily) record their experiences of social support, typically specific supportive actions and often from a specific source or person. Compared to other types of social support assessment, perceived support measurement tools are the most closely tied with mental health outcomes, with findings consistently indicating that those who report higher levels of perceived support have better mental health than those who 27 report lower perceived support (Lakey & Cohen, 2007). Similarly, perceived support has been correlated with physical health indicators, though the magnitude of this relationship is smaller (Lakey & Cohen, 2007). Additionally, according to Lakey and Cohen (2007), the construct validity of perceived support is best understood among the categories of social support measures. Lastly, perceived support, rather than received support has consistently demonstrated to buffer, or moderate, the effects of stress on individual health or well-being (Gottleib & Bergen, 2010). Theories of Social Support. While there are myriad methods for classifying and measuring social support, there are two main strands of theory related to social support which correspond to previously mentioned models of resilience: the main effect model and the buffering hypothesis (Cohen & McKay, 1984, p. 253). The main effect model posits that social support has an independent effect on well-being, regardless of the presence or absence of stress. Individuals will reap the benefits of social support, even if they are not experiencing stress. Theorists suggest that certain types of social support, such as social network support, are more likely to operate through a main effect model (Cohen & Wills, 1985). Being a part of a supportive community can be beneficial, regardless of one’s level of stress. The buffering hypothesis of social support is predicated upon the presence of stress and theorizes that stress will have a stronger negative effect on the well-being of those with less social support, or conversely, that stress will have less of an impact on those with greater social support (Cohen & McKay, 1984). In other words, the impact of stress on well-being is moderated (or “buffered”) by the level of social support. In considering the buffering model of social support, it is useful to distinguish between stressful events and the experience of stress, as the two are often conflated. A stressful event, or 28 stressor, is that which has the potential to elicit stress, while stress is an experience of the negative effects (physiological, affective, behavioral) of appraising an event or circumstance as stressful (Cohen & McKay, 1984; Lazarus & Folkman, 1984). It is the experience of stress that is theorized to impact well-being. The buffering hypothesis suggests that social support buffers against the negative effects of stress at potentially two points in the causal linkages between stressor, stress, and outcome (Cohen & McKay, 1984; Cohen & Wills, 1985). As noted in Figure 3, social support may affect one’s appraisal of a potential stressor, such that it is not appraised as beyond one’s means of coping and therefore not considered stressful. Informational and tangible supports may function to change an individual’s understanding of a potential stressor or change the nature of the stressor itself, thus changing the appraisal of that stressor. Social support may also facilitate coping or adjustment following an experience of stress, such that the pathological outcome is lessened or eliminated (Cohen & McKay, 1984; Cohen & Wills, 1985). Emotional support may operate in this way, in that once distress is present, it serves to lessen the experience of stress or strengthen aspects of the self that were perhaps threatened by the stressor (Lazarus & Folkman, 1984, as cited in Cohen, 1992). Figure 4 Buffering hypothesis of social support 29 Figure 4 (cont’d) Note. Adapted from Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310–357. https://doi.org/10.1037/0033-2909.98.2.310 The above theoretical diagram (Figure 4) is potentially valuable for guiding research related to stress, support, and outcomes among foster parents. What are the stressors in the foster parent’s environment? What are pathological outcomes in fostering? What role does social support play in protecting against these outcomes or promoting positive outcomes? Utilizing the theoretical frameworks outlined above, Figure 4 illustrates a proposed model for conceptualizing, measuring, and analyzing the relationship between stress and negative outcomes among foster parents, as well as the potential buffering function of social support. Figure 5 Proposed buffering model of foster parenting This model will serve to organize the following review of key empirical literature on fostering, as well as the proposed study design. The review will include stressors of fostering and the experience of stress by foster parents, outcomes in the fostering role, and what helps foster parents cope. One particular coping factor, social support, will be explored in depth. First, however, it is beneficial to delineate who foster parents are, how and why they enter their role, and the benefits that they experience therein. 30 Foster Parent Population Prior to examining the literature on fostering, it is useful to define what is meant by the term ‘foster parent,’ and examine characteristics of this population. For the purposes of the current study, ‘foster parent’ includes all those (relatives and non-relatives) who have been granted a foster home license by their state child welfare authority. Foster Parent Demographics There are limited data sources which offer demographic details of the foster parents in the United States; however, two studies offer information useful for describing this population. The National Survey of Child and Adolescent Well-Being, Wave II (NSCAW II) included among its sizeable sample (N=5,873) children who are in foster care (N=1,105), and gathered information on their current caregivers. According to an NSCAW II report from 2011, the majority of (non- kin) foster parents (53.0%) are 30-49 years old, White (51.2%; 25.0% African American, 19.0% Hispanic, and 2.6% ‘other’), and married (63.5%; Dolan, Smith, Casanueva, & Ringeisen, 2011). Foster parents tend to have at least a high school education, and have an income above the poverty line (though nearly 20% reported income below the federal poverty level; Dolan, Smith, Casanueva, & Ringeisen, 2011). Kinship foster families differ from non-kin in that they are typically older, have lower income, and are more likely to be African American or Hispanic (Dolan, Smith, Casanueva, & Ringeisen, 2011). A 2009 study of administrative data in Illinois also offers insights on the demographics of the foster parent population (Zinn, 2009). The study used child welfare system data on 15,842 foster families. Results indicate that foster parents (including non-kin and kin) had a mean age of 44.8 years and were primarily African American (57.7%; 34.6% White, 4.2% Hispanic, 3.4% other). Twenty-four percent of foster parents were single parents. The mean income of foster 31 parents was $35,587. Foster parents had been in their roles for an average of 2.5 years. Kinship caregivers within the study differed from non-kin on most characteristics, in that they were older, more likely to be African American or Hispanic, more likely to be single, and had lower income (Zinn, 2009). Motivations and Benefits of Fostering Foster parents come into their role through a variety of paths and motivations. Most foster parents find out about fostering by knowing other foster families (34.7%; Rodger, Cummings, & Leschied, 2006). Others find out through connections to specific foster children (12.8%). Some had parents who were foster parents (11%), or were in foster care as children themselves (3.6%). The remainder of foster parents learned about fostering through radio (6.5%), billboard (2.3%), or Internet (.2%) advertisements. Motivations for fostering are varied, but most reasons center on a theme of altruism and include a strong desire to help, to give back, or to provide something to the children. Many foster parents report being motivated to foster by a general aspiration to serve the community or fulfill a collective responsibility (de Maeyer, Vanderfaeillie, Vanschoonlandt, Robberechts, & Holen, 2014; Geiger, Hayes, & Lietz, 2013). Similarly, many foster parents report religious motivations for fostering (Buehler, Cox, & Cuddeback, 2003; de Maeyer et al., 2014; Rodger et al., 2006). Foster parents also report many child-centered motivations for fostering, such as providing a good home or love for a child (de Maeyer et al., 2014; Rodger et al., 2006), and saving children from further harm (Rodger et al., 2006) or from being placed in institutions (de Maeyer et al., 2014). Much less frequently endorsed are self-interested motivations, such as increasing family income (Rodger et al., 2006), wanting more children in the home (de Maeyer et al., 2014; MacGregor, Rodger, Cummings, & Leschied, 2006; Rodger et al., 2006), or benefitting biological children (MacGregor et al., 2006). 32 Many of the motivations for fostering reported by foster parents are reflected in the benefits they experience. Despite the many challenges of fostering, most foster parents report satisfaction with their role, particularly with the direct care of children (Eaton & Caltabiano, 2009). They report that it is intrinsically rewarding (Buehler et al., 2003; Cavazzi, Guilfoyle, & Sims, 2010; Geiger et al., 2013). Foster parents benefit from seeing ways that children grow and develop, giving them a sense of normality and a secure environment, and saving them from hardship (Buehler et al., 2003). The loving relationships built in foster care not only benefit the children, but impact foster parents positively as well (Buehler et al., 2003; Pickin, Brunsden, & Hill, 2011). There are positive effects on the entire family (Samrai, Beinart, & Harper, 2011), and benefits to biological children, such as increased empathy (Geiger, Piel, Lietz, & Julien- Chinn, 2016). Stressors of Fostering While there are notable benefits to fostering, these are often reaped in the presence of numerous and substantial stressors. According to Kwok and Wong (2000), stressors are sources of stress, or “external conditions that induce stress in the individual” (p.58). Foster parents face various conditions that exceed normative challenges of the parenting role (Tarren-Sweeney, 2008). The stressors experienced by foster parents are above and beyond, yet also include, those of “ordinary parenting” (Oke et al., 2011, p. 16). Stressors within the “exceptional burden of care” (Murray et al., 2011) faced by foster parents include both individual factors as well as system-related factors (Adams et al., 2018). Individual Stressors Children entering foster care often have significant developmental, behavioral, mental health, and physiological concerns. According to Leslie, Gordon, Ganger, and Gist (2002), at 33 least a quarter of children in foster care experience developmental delays. Foster children are much more likely to display behavior problems than same-aged peers (Lohaus et al., 2017; Tarren-Sweeney, 2008; Turney & Wildeman, 2016). In terms of mental health, there is a high prevalence of conduct disorder, ADHD, depression, and generalized anxiety disorder among children in out-of-home care (Tarren-Sweeney, 2008; Turney & Wildeman, 2016). Behavioral and emotional challenges of foster children are one source of stress for foster parents (Buehler et al., 2003; Cavazzi et al., 2010; Donachy, 2017; Farmer et al., 2005; Morgan & Baron, 2011). Physical health problems can also be a source of stress. Nearly 30% of the foster children in one study had three or more health conditions, while fewer than 20% had none (Halfon, Mendonca, & Berkowitz, 1995). In addition to stressors originating with foster children, there are some stressors specific to the individual as a foster parent. As previously noted, the role itself is ambiguous, residing between parent and professional (Blythe et al., 2014; Pickin et al., 2011). This ambiguity can be found in foster parents’ relationships with their licensing agencies, in that they are not employees, nor are they clients. Therefore the agency’s obligations to them and the trust within the relationship are often uncertain (Hudson & Levasseur, 2002). The ambiguity of fostering is also found in relationships to the children (Hudson & Levasseur, 2002). The temporary nature of the role is often a source of stress, in that foster parents are caught in an “emotional double bind” (Marcellus, 2010, p. 16) of emotionally investing in children who they must also be prepared to release, even at a moment’s notice (Pickin et al., 2011). When foster children do leave, there is often significant grief, regardless of the reason for the departure (Cavazzi et al., 2010; Murray et al., 2011; Pickin et al., 2011; Samrai et al., 2011; Thomson & McArthur, 2009). This grief is often compounded by the ambiguous nature of the loss, in that the individual being grieved is 34 alive but not physically present, and the loss is one not typically recognized by society (Donachy, 2017; Pickin et al., 2011; Thomson & McArthur, 2009). System-related Stressors Foster parents report numerous challenges associated with the child welfare system, often reporting that the system is a greater source of stress than their parenting duties (Blythe et al., 2014; Geiger et al., 2013). Navigating the system itself is a source of stress (Buehler et al., 2003; Geiger et al., 2013; Lietz, Julien-Chinn, Geiger, & Piel, 2016; Marcellus, 2010), as is attempting to access services for children within this system (Beuhler et al., 2003; Lanigan & Burleson, 2017; Murray et al., 2011). Relationships with foster care professionals can be stressful, as caseworkers handle large caseloads (Geiger et al., 2017), turnover frequently (Buehler et al., 2003; Farmer et al., 2005; Geiger et al., 2013; Geiger, Piel, & Julien-Chinn, 2017; Murray et al., 2011), and may not be available or communicate as frequently as foster parents need (Cavazzi et al., 2010; Farmer et al., 2005; Gieger et al., 2013; Geiger et al., 2017). One of the most difficult stressors for foster parents is when they are the target of abuse allegations and/or investigations (Murray et al., 2011; Wilson, Sinclair, & Gibbs, 2000), especially when they do not feel supported by their worker in this circumstance (Murray et al., 2011). Foster parents report often feeling powerless over the foster care process, including decisions that significantly affect the lives of their foster children, and consequently their families (Buehler et al., 2003; Cavazzi et al., 2010; Geiger et al., 2013; Marcellus, 2010; Pickin et al., 2011; Randle, Ernst, Liesch, & Dolnicar, 2017; Thomson & McArthur, 2009). Despite the perceived lack of power, foster parents also report feeling a great deal of responsibility, paired with high expectations and oversight of their parenting (Marcellus, 2010). This “public parenting” (Marcellus, 2010, p.14) facet of fostering can be a source of substantial stress for 35 foster parents, who perform their roles under the supervision of the child welfare system and the professionals within it, including case workers, licensors, attorneys, judges, and others, as well as the general public. Marcellus states, “Foster families have the weight of public trust placed on them and are held to a higher standard of parenting than birthparents because of the government’s responsibility for the [child]” (2010, p. 22). Working with birth parents can also lead to stress among foster parents (Beuhler et al., 2003; Cavazzi et al., 2010; Lanigan & Burleson, 2017; Murray et al., 2011; Wilson et al., 2000). Directly, foster parents report stress resulting from verbal abuse and threats from birth parents, their emotions regarding birth parents, cultural differences with the birth family, concerns about safety, and questions about how to communicate with children about birth parents (Murray et al., 2011). Indirectly, foster parents may experience stress when children exhibit problem behaviors following contact with birth parents, don’t want to attend visits, or are disappointed about missed visits (Murray et al., 2011). Many of the system-related stressors reported by foster parents are due to a perception that they aren’t receiving what they need to fulfill their role. Foster parents report that their training is insufficient (Cooley, Thompson, & Wojciak, 2017), and identify specific areas in which they desire more training (e.g. child behaviors, promoting their own well-being; Murray et al., 2011). Other deficits reported by foster parents are information (Buehler et al., 2003; Cavazzi et al., 2010; Gieger et al., 2017; Lanigan & Burleson, 2017; Murray et al., 2011) and concrete supports like respite care (MacGregor et al., 2006). Unfortunately, in the context of these numerous stressors, foster parents also report a lack of acknowledgement or appreciation for their efforts (Cavazzi et al., 2010; Maclay et al., 2006). 36 The Experience of Stress among Foster Parents According to the buffering hypothesis of social support (Figure 4), stressors have the potential to lead to the experience of stress if they are appraised to be taxing beyond the resources available to the individual (Lazarus & Folkman, 1984). Murray and colleagues (2011) note that this is frequently the case among foster parents, who experience a “mismatch between the level of support and training offered or provided foster carers by the children’s agencies, and the burden of caring for children with complex emotional, behavioural and interpersonal difficulties” (p. 156). While there are few studies within foster parenting literature that distinguish between stressors and the experience of stress, or between stress and pathologic outcomes, there are a small number which specifically examine stress. One useful construct for examining stress among foster parents is parenting stress. Parenting stress is the “aversive psychological reaction to the demands of being a parent” (Deater-Deckard, 1998, p. 315). It includes negative feelings towards oneself and one’s child(ren) as a result of the challenges of parenthood (Deater-Deckard, 1998). Such challenges, for all parents, include “managing children’s behavior, establishing and maintaining family routines, and engaging in daily care-giving tasks” (Dunning & Gaillo, 2012, p. 147). Additionally, parenting stress includes a discrepancy between parenting demands and parent resources, both internal and external. This mismatch, or lack of goodness-of-fit, results in the experience of parenting stress, which has been demonstrated to correlate with maladaptive parenting practices, such as authoritarian discipline (Deater-Deckard & Scarr, 1996). Some groups of parents experience increased parenting stress, such as step-parents (Shapiro, 2014), parents experiencing family structural transitions (Cooper & Mclanahan, 2009), and parents of children with Autism Spectrum Disorder (McStay, Dissanayake, Scheeren, Koot & Begeer, 37 2013). Because foster parents also experience additional, unique demands in their role, parenting stress is a valuable construct with this population; however, only a handful of studies on foster parenting have utilized it. The somewhat limited research on parenting stress among foster parents points to a relationship between fostering and elevated levels of parenting stress. Several studies utilized standardized measures of parenting stress to assess this construct. Morgan and Baron (2011) utilized the Parenting Stress Index Short Form (PSI-SF) to measure parenting stress among British foster parents (N=58), and found that 54% scored in the borderline or clinical range. A mixed-methods study of New Zealand foster parents (N=17) by Murray and colleagues (2011) also found elevated scores on the PSI. Specifically, foster parents scored two standard deviations above the normative mean on the Child Domain subscale of the PSI. A 2004 study examined parenting stress among U.S. foster parents whose children were receiving clinical services (N=259), and found that 62% scored in the clinical range on the Parent-Child Dysfunctional Relationship Scale of the PSI (Timmer, Sedlar, & Urquiza, 2004). Using the parental distress subscale of the Parental Stress Questionnaire, Lohaus and colleagues (2017) found that foster mothers (N=72) experienced greater parental stress than mothers caring for biological children (N=130). A study of Flemish foster (N=39) mothers utilized a different tool (Nijmegen Questionnaire for the Parenting Situation) to measure parenting stress and also found that foster mothers experienced greater parenting stress than mothers in the general population, with 71% of foster mothers scoring in the clinical range on at least one parenting stress subscale (Vanschoonlandt, Vanderfaeillie, Holen, de Maeyer, & Robberechts, 2013). In contrast, a recent study of foster mothers (Richardson, Futris, Mallette, & Campbell, 2018) found that on average, they experienced low levels of parenting stress, as measured by a revised, 10-item version of the 38 Parental Stress Scale (Berry & Jones, 1995). This could be due to differences in the various measurement tools used to assess parenting stress. Another study which found no evidence of elevated parenting stress for foster parents was conducted by Gabler and colleagues, who used the Parenting Stress Index (2018). However, this study’s sample population was foster parents with new placements, who may not have had sufficient time with their foster child(ren) to experience detectable parenting stress. Research among foster parents has also demonstrated relationships between parenting stress and other relevant variables. Several studies found a significant, positive relationship between child behavior problems and parenting stress (Gabler et al., 2014; Lohaus et al., 2017; Morgan & Baron, 2011; Vanderfaeillie, van Holen, Trogh, & Andries, 2012; Vanschoonlandt et al., 2013). Further, longitudinal studies indicate that child behavior problems lead to increased parenting stress among foster parents (Gabler et al., 2018; Goemans, Van Geel, & Vedder, 2018; Van Holen, Vanderfaeillie, Vanschoonlandt, De Maeyer, & Stroobants, 2015). McKeough, Bear, Jones, Thompson, Kelly, & Campbell (2017) examined the relationship between challenges specific to the fostering role and parenting stress, finding significant relationships with the following challenges: challenging foster child behaviors, time management, lack of control, and lack of support from their foster care organization. In terms of outcomes related to parenting stress, foster parents who report high levels of parenting stress are significantly more likely to report more negative parenting experiences (Leathers, Spielfogel, Geiger, Barnett, & Vande Voort, 2019) and low parenting satisfaction (Soliday et al., 1994). Parenting stress has also been demonstrated to be related to poor parenting quality among foster parents (Vanderfaeillie et al., 2012). Specific parenting constructs demonstrated to significantly negatively correlate to parenting stress include co-parenting quality 39 (Richardson & Futris, 2019; Richardson et al., 2018) and parental sensitivity (Gabler et al., 2018). Positive parenting quality is essential to successful placements of foster children. Oosterman, Schuengel, Slot, Bullens, and Doreleijers (2007) found that foster placements are more likely to succeed when foster parents are able to demonstrate parental sensitivity by responding to children’s needs and problems. Parenting stress can impede successful foster placements. In addition to research specific to parenting stress among foster parents, other studies measured stress among foster parents more broadly. McKeough, Bear, Jones, Thompson, Kelly & Campbell (2017) measured general stress using the stress subscale of the Depression Anxiety Stress Scale (DASS-21). While the mean score fell within the normal range, the authors note that the data were considerably skewed, in that 60% of foster parents were in the normal range, but 28% scored in mild to moderate range and 13% were in the severe to extremely severe range on stress. Tunno (2015) utilized the Perceived Stress Scale (PSS) to measure general stress in foster parents (N=32), finding only low to moderate levels of stress. One survey of foster parents in Illinois (N=189) utilized a single question to measure stress, asking whether they felt they had enough time for their responsibilities, and found that 77.3% of foster parents reported that they almost always had sufficient time (Cole & Eamon, 2007). Most foster parents (77.3%) reported that they had ‘almost always’ had sufficient time for their responsibilities; however, the use of this single item as a measure of stress may be problematic in terms of validity. Outcomes of Stress Within foster parenting literature, there are key outcomes studied that may result from the stress experienced by foster parents. Some outcomes, such as foster home retention, are related 40 to goals within the child welfare system, while others, such as foster parent well-being, are related to foster parents as individuals. Foster Home Retention Foster home retention is a system-related outcome often included in fostering studies (Denby et al., 1999; Eaton & Caltabiano, 2009; Geiger et al., 2013; Gibbs & Wildfire, 2007; Rodger, Cummings, & Leschied, 2006; Whenan, Oxlad, & Lushington, 2009; Wilson et al., 2000). Very little actual retention data is available, as retention is typically measured as the intent to discontinue fostering in a given time frame or past ideation about discontinuing (Denby et al., 1999; Eaton & Caltabiano, 2009; Geiger et al., 2013; Whenan et al., 2009). Time frames vary from 2 months to three years, with some studies not offering a time frame (e.g. “Do you intend to give up fostering in the future?”). Further complicating the measurement of retention is that some studies utilized a positive framing of the variable, e.g. “How likely are you to continue fostering?”, while others used a negative framing, e.g. “How likely are you to leave fostering?” Table 2 shows the operational definitions of retention among relevant studies. Table 2 Operational definitions of foster home retention Authors (Year) Retention item(s) Cooley et al. (2015); Denby et al. (1999); Hannah & Woolgar (2018) Eaton & Caltabiano (2009) “How likely is it that you’ll still be fostering in one year?” Type of operational definition Future intention, positive “How likely is it that you will give up fostering in the next 18 months” Future intention, negative “In the last [18 months, 2 months], how often have you felt like giving up fostering?” Past ideation, negative 41 Table 2 (cont’d) Geiger et al. (2013) Gibbs & Wildfire (2007) Orme et al. (2006) Randle et al. (2017) Rodger et al. (2006) “What is the likelihood of you giving up fostering in the next 18 months?” Length of foster parent service (longitudinal analysis of administrative data) “Do you intend to continue fostering over the next [six months, year, three years]?” Future intention, negative Actual retention Future intention, positive Discontinuation ideation (six types of discontinuation thoughts in the past four months, based on Suicidal Ideation Questionnaire) “At any one time have you considered withdrawing from fostering?” Past ideation, negative Past ideation, negative Whenan et al. (2003) “I am very likely to continue fostering in the future.” Future intention, positive Wilson et al. (2000) “Have you ever felt like you would like to give up fostering?” “Do you intend to give up fostering in the future?” Past ideation, negative Future intention, negative The methodological concerns with the operationalizing of various retention-related variables, creates a challenge in summarizing the collective findings. However, there are some themes to be found within the results. In terms of retention as intent to continue, several studies asked about the likelihood of remaining in the fostering role (Cooley et al., 2015; Denby et al., 1999; Hannah & Woolgar; 2018). Denby and colleagues (1999) found that about two-thirds (67.3%) of foster parents intend to still be fostering in one year. Hannah and Woolgar (2018) found that a higher proportion (77.9%) thought that they were very likely to continue. Cooley and colleagues (2015) did not 42 report percentages on individual item responses, but found that the mean response was in between “somewhat likely” and “very likely” to continue. Orme (2006) also used positive framing to ask about intent to continue over a range of time periods, and found that 89.8% of foster mothers plan to still be fostering in three years (93.7% at one year; 95.4% at six months). Whenan and colleagues (2003) used positive framing, but did not include a time frame, finding that 84.5% of foster parents considered themselves very likely to continue fostering in the future. Other studies asked about intent to continue, but framed the question negatively, asking about the likelihood of giving up fostering. Two studies used an 18 month time period, finding that 20- 25% were likely or very likely to leave (Eaton & Caltabiano, 2009; Geiger et al., 2013). A third study asked about intent to leave fostering in the future, but did not specify a time frame and did not report univariate results on this variable (Wilson et al., 2000). Four studies operationalized retention as past ideation related to leaving fostering (Eaton & Caltabiano, 2009; Rodger et al., 2006; Randle et al., 2017; Wilson et al., 2000). Eaton and Caltabiano (2009) found that 27.4% of foster parents had thought about giving up fostering often or very often in the last two months, and 31.7% had thought about it in the past 18 months. Rodger and colleagues (2006) did not specify a time frame and found that 58.3% of foster parents had considered quitting fostering. Similarly, Wilson et al. (2000) asked if foster parents had ever felt like giving up fostering, but did not report descriptive data on this variable. Randle et al. (2017) adapted a suicide ideation scale to measure ideation about leaving fostering; however, mean scores on this scale were not published. While most studies operationalize retention as either intent to continue or past ideation, there is limited research on actual retention of licensed foster homes. Gibbs and Wildfire (2007) conducted the most frequently cited study on actual retention, analyzing administrative child 43 welfare data from three states (New Mexico, Oklahoma, and Oregon). They found that between 47% and 62% of foster parents leave their roles within one year of taking their first placement. The figures jump to 68-81% of foster parents who leave within two years. The median length of service for foster parents was between 8 (New Mexico and Oregon) and 14 months (Oklahoma). As a comparison, the median length of stay for a child in foster care is 14.7 months (U.S. Children’s Bureau, 2019). The results of this study indicate a discrepancy between foster parents’ intentions to remain in their roles and the number of foster parents who do remain. There are a number of factors that may contribute to foster parents’ decision to remain in their role or not, including foster parent factors like hardiness (Hendrix & Ford, 2003) as well as family factors, such as finances, adoption of foster children, or needing to focus on their families’ needs (Ahn et al., 2017; Geiger et al. 2013; Kangas, 2015; Rhodes, Orme, Cox, & Buehler, 2003; Triseliotis, Borland, & Hill, 1998). However, the stressors of fostering are often cited as a reason for foster parent departure (Kangas, 2015; Rhodes, Orme, & Buehler, 2001; Rodger et al., 2006; Wilson et al., 2000), a relationship that may be stronger among new foster parents (Wilson et al., 2000). While child-related factors, such as behaviors (Ahn et al., 2017; Denby et al., 1999; Rodger et al., 2006), likely impact the decision to stay or leave, other factors found to correlate with intentions to continue fostering include: social worker availability (Denby et al., 1999; Rhodes et al., 2001), positive relationships with social workers (Rodger et al., 2006), role satisfaction (Eaton & Caltabiano, 2009), and peer support (Denby et al., 1999). Additionally, foster parents who consider leaving are more likely to have received less training, perceive less support, and have poorer perceptions of the foster care system than those foster parents who intend to remain in their role (Randle et al., 2017; Rhodes et al., 2001). Among former foster parents who had exited their role, Ahn and colleagues (2017) found that the most common 44 reason was a change in life situation (58.37%), followed by problems with the foster care agency (28.16%) and problems with children in the home (11.02%). In an earlier study, Rhodes and colleagues (2001) found that the most cited reasons for leaving were a lack of support from the foster care agency and poor communication from the foster care worker, followed closely by foster child behavior problems. Foster Parent Well-being While retention of foster parents is an important outcome for the child welfare system, it is also worthwhile to investigate the impact of stress on foster parent health and well-being. However, few studies examine foster parent well-being. According to Miller, Green, and Lambros (2019), “Although foster parent well-being is critical to the success of foster youth and the foster system, the literature has yet to identify how foster parents maintain their health and well-being in the face of these challenges and stressors” (p. 109). Among the studies that examine foster parent well-being, the construct is typically operationalized as psychological health, either overall or as the presence of specific mental health disorders. Two studies utilized the General Health Questionnaire (GHQ) to measure psychological health (Farmer et al., 2005; Wilson et al., 2000). One found that 29% of foster parents scored outside the normal range for overall psychological health (Farmer et al., 2005). Perhaps most notably, 81% of foster parents in this study had clinical or subclinical scores on the social functioning subscale, indicating that they weren’t doing well in daily functioning, weren’t enjoying activities, and weren’t coping well (Farmer et al., 2005). The other study found that stressful events in fostering, such as strong disagreements with social services, were associated with decreased foster parent well-being (Wilson et al., 2000). 45 Other studies utilized measures of specific mental health conditions (rather than overall psychological health) found that, generally, foster parents do not experience high levels of depression or anxiety symptoms (Cole & Eamon, 2007; Morgan & Baron, 2011; Whenan et al., 2009). Cole & Eamon (2007) found that foster parents who participated in support group and found it helpful had lower depression scores. Whenan and colleagues (2009) found that foster parents with lower parenting self-efficacy and colder parent-child relationships were more likely to score high on a measure of depression, anxiety, and stress. Morgan and Baron (2011) also found that high scores on a depression and anxiety measure were correlated to lower parenting self-efficacy. Additionally, they found that foster child behavior problems were significantly, positively related to parent depression and anxiety. Social Support of Foster Parents Existing literature points to several coping factors that may contribute to resilience among foster parents. Some factors are personal traits, such as emotional strength (Oke et al., 2011), assertiveness (Lietz et al., 2016; Maclay et al., 2006), flexibility (Lietz et al., 2016), humor (Lietz et al., 2016), empathy (Geiger et al., 2016), ability to set boundaries (Lietz et al., 2016), and personal faith or spirituality (Buehler et al., 2003; Lietz et al., 2016). Other factors reside in the foster parent’s environment, ranging from family connectedness and communication at the micro level (Lietz et al., 2016) to good relationships with child welfare professionals at the mezzo level (Geiger et al., 2017; Murray et al., 2011; Pickin et al., 2011; Rodger et al., 2006; Samrai et al., 2011) to societal understanding of foster care (Piel, Geiger, Julien-Chinn, & Lietz, 2017). Many of the factors promoting resilience among foster parents fall under the umbrella of social support. The following section will provide a review of empirical literature on social 46 support among foster parents, including types of support, social support measurement, and key findings. As noted previously, social support consists of support networks, supportive behaviors, and subjective appraisals of support (Vaux, 1988). While each has been studied among foster parents, support networks are studied less frequently, and are not typically used to explore the buffering model of social support (Cohen, Gottlieb, & Underwood, 2000). Therefore, this review will focus primarily on literature examining supportive behaviors (received support) and/or subjective appraisals of support (perceived support). One study, which did use a support network approach (Nixon, 1997), asked foster parents how many people they felt they could talk to during a time of crisis (i.e., allegation of abuse). Over half of foster parents had only one person they could talk to, and 22% felt they had no one, highlighting the isolation often experienced by foster parents. Types of Social Support Foster parenting literature provides examples of multiple types of social support and their utility for this population. In terms of tangible support, foster parents identify several practical measures, including respite (Geiger et al., 2013; MacGregor et al., 2006; Murray et al., 2011; Nixon, 1997; Samrai et al., 2011), child care and other child resources (Geiger et al., 2013; MacGregor et al., 2006; Murray et al., 2011), transportation (Geiger et al., 2013), and financial support (Gieger et al., 2013; Hudson & Levasseur, 2002; MacGregor et al., 2006; Nixon, 1997; Samrai et al., 2011). Foster parents also recognize emotional support as being critical to success in their role (Hudson & Levasseur, 2002; Lietz et al., 2016; MacGregor et al., 2006; Murray et al., 2011; Nixon, 1997). Included in emotional support is the feeling of being respected, recognized, and 47 acknowledged (Geiger et al., 2013; Hudson & Levasseur, 2002), as well as having someone who will listen and offer advice (Farmer et al., 2004). Lastly, foster parents identify aspects of informational support that assist them in their role (Piel et al., 2017), including timely and accurate information on the history and needs of the children in their care (Brown & Campbell, 2007; Geiger et al., 2013; MacGregor et al., 2006; Murray et al., 2011; Piel et al., 2017) and relevant and accessible training (Cooley et al., 2017; Hudson & Levasseur, 2002; Lanigan & Burleson, 2017; MacGregor et al., 2006; Murray et al., 2011; Nixon, 1997; Samrai et al., 2011; Whenan et al., 2009). Social Support Measurement Within fostering literature, social support is often treated as a rather straightforward construct when, as detailed previously, it is in fact a multifaceted meta-construct. Studies typically do not explicate whether social support is operationalized as social network, received support, or perceived support. Additionally, the definition of and distinction between support types is typically not clear. As a variable within studies of foster parents, social support is most commonly operationalized as perceived support- though not always explicitly identified as such. Social support (perceived and otherwise) is measured among foster parents using a variety of instruments. Table 3 Measures of foster parent social support Study Ahn et al., 2017 Instrument Created instrument Operational Definition of Social Support Adequacy of agency support Cole & Eamon, 2007 Two individual items Cooley et al., 2015; Cooley et al., 2019 Protective Factors Survey, Subscale Adult to child ratio and participation in support group Perceived support 48 Table 3 (cont’d) Crum, 2010 Denby et al., 1999 Parent Child Relationship Inventory (PCRI), Parental Support Subscale Created instrument Eaton & Caltabiano, 2009; Geiger et al., 2013 Farmer et al., 2005 Foster Care Significant Other Survey (FCSOS) Individual item Perceived support Received support from other parents and social workers Perceived emotional and tangible support from specified formal and informal sources Perceived availability of supportive friends Leathers et al., 2019 Support Functions Scale Need for various types of support Lohaus et al., 2017 Nixon, 1997 Subscale of Parental Stress Questionnaire Created instrument Perceived support in general and from partner Social network size Orme et al., 2006; Richardson et al., 2018 Randle et al., 2017 Help with Fostering Inventory (HFI) Created instrument Soliday et al., 1994 Vanderfaillie et al., 2016; VanHolen et al., 2015; Vanschoonlandt et al., 2014 Subscales of COPE Inventory Support Needs & Satisfaction Questionnaire Perceived support from specified formal and informal sources Perceived quality of support from foster care agency Perceived instrumental and emotional support Support needs in a variety of domains and perceived support from case worker Table 3 contains details on social support instruments utilized in studies of foster parent populations. Most authors did not explicitly state their approach to measuring social support or delineate why they measured a certain social support construct over others. Therefore, I determined the details (such as perceived versus received support) for most studies by examining the items or descriptions of the items. As indicated, most measured perceived support, and some specified the source of support. Many developed their own instruments for gauging social 49 support. Perhaps the most concerning operationalization of social support was in one study that measured social support as whether or not foster parents participated in an agency-sponsored support group and their adult-to-child ratio in the home (Cole & Eamon, 2007). Quantitative Findings The aforementioned methodological inconsistencies make it somewhat difficult to synthesize the available literature. However, there are results worth highlighting in terms of providing a context for the current study. Soliday and colleagues (1994) found that parenting stress was significantly negatively related to satisfaction with social support in foster parents, specifically within intimate relationships. Similarly, Richardson and colleagues (2018) found a significant, negative relationship between parenting stress and social support among foster mothers. However, another study found that perceived social support from partners was not significantly associated with parental stress in foster parents (Lohaus et al., 2017). This discrepancy may be due to the different measures of social support or different methods of analysis, and points to a need for further investigation into the relationship between parenting stress and social support. In terms of outcomes, social support has been linked to child welfare goals, such as placement stability (Crum, 2010) and foster home retention (Denby et al., 1999; Randle et al., 2017). The two studies that found a relationship between social support and retention both measured social support from foster care agency sources (Denby et al., 1999; Randle et al., 2017). However, Ahn et al. (2017) also measured perceived agency support and found that it was not predictive of higher time in fostering (which is distinct from retention operationalized as intent to continue). Another study measured perceived support in general and found no significant relationship between support and retention as intent to continue (Cooley et al., 2015). 50 Lastly, Eaton and Caltabiano (2009) found that the relationship between (perceived) social support and retention (intent to continue) was mediated by satisfaction with fostering. That is, greater social support leads foster parents to feel more satisfied with their role, and satisfaction leads to greater intent to remain in the role. In summary, findings on support and retention are inconsistent, perhaps due to the wide variety of operationalization of both constructs. Social support has also been demonstrated to correlate with individual factors, such as foster parent satisfaction (Eaton & Caltabiano, 2009; Geiger et al., 2017; Soliday et al., 1994; Vanderfaillie et al., 2016), mental health (Cole & Eamon, 2007), and general health (Farmer et al., 2005). One study found that, among foster parents reporting high levels of disruptive child behaviors, tangible support was protective against fostering challenges (Cooley, Farineau, & Mullis, 2015). According to the buffering hypothesis of social support, discussed previously in this review, social support acts as a moderator between stressors and potential outcomes. Among foster parenting literature, however, there is a lack of studies examining social support as a moderating variable. In fact, only three such studies were located for this review. Cole and Eamon (2007) examined the relationship between stress and depressive symptoms among foster parents, and found that social support did not moderate this relationship. However, worth mentioning is the fact that the measures used for key variables of stress (whether parents felt they had sufficient time for their responsibilities) and social support (participation in a support group and adult-child ratio in the home) were not psychometrically tested and may not be valid measures of the constructs. In contrast, Richardson and colleagues (2018) utilized established scales to measure key variables, and found that social support did moderate the relationship between parental stress and co-parenting relationship quality among foster parents. Cooley and 51 colleagues (2019) found social support to be a moderator between perceived problems with child behaviors and perceived challenging aspects of fostering. Evidence for social support as a moderator among a foster parent population is mixed. Further, the only study to examine social support as a moderator between stress and support utilized measures of questionable validity (Cole & Eamon, 2007). More research is needed to explore the buffering hypothesis among foster parents. While the literature on social support as a buffer against foster parent stress is nearly non- existent, related literature on various caregiving populations contains studies which examine the buffering, or moderating, role of social support. In a study of caregivers for traumatic brain injury (TBI) patients, perceived social support was found to moderate the relationship between reduced executive functioning following TBI and caregiver psychological distress (Ergh, Rapport, Coleman, & Hanks, 2002). Similarly, Gellert and colleagues (2018) found that social support moderated the relationship between perceived stress and quality of life among both dementia patients and their caregiving partners. Among grandparents in a primary caregiving role for grandchildren, perceived social support moderated between parental stress and depression (Hayslip, Blumenthal, & Garner, 2015). However, other studies did not find support for the buffering model. In a study of psychological adjustment of mothers with deaf children, the main effect model of social support was significant, while the buffering model was not (Quittner, Glueckauf, & Jackson, 1990). Though not conclusive, there is support for the buffering role of social support among caregiving populations under increased stress. However, thus far, research on foster parents has not explored this relationship. 52 Conclusion Overall, there is a lack of research on the impact of stress among foster parents and factors, including social support, which may promote coping and protect against pathologic outcomes. The research that does exist in this area has not typically adhered to a strong theoretical framework and thus studies have not been theoretically nor methodologically consistent. This study aims to address this gap by utilizing the buffering model of social support to explore the path from stressors to stress to negative outcomes, and whether social support serves as a buffer along this path. 53 CHAPTER 3: METHODS The following chapter will describe the methods used in the present study and will include four sections. First, I will detail the study design and procedures. Second, I will provide a description of the study sample. Third, I will describe measurement methods for key variables, including a description of each instrument and its psychometric properties. Finally, I will discuss the methods used for data screening and analysis. Study Design and Procedures Study Design This study utilized a cross-sectional design and web-based survey methodology to explore the relationships between stress, social support, well-being, and retention among licensed foster parents in Kent County, Michigan. Kent County was selected because it has a large child welfare system that could provide a sufficient sample and it includes urban, suburban, and rural areas. Additionally, I had existing professional relationships with child welfare professionals in Kent County, which provided access to the study population. Prior to data collection, I sought Institutional Review Board (IRB) approval from the Michigan State University Human Research Protection Program (MSU HRPP) and the Michigan Department of Health and Human Services Institutional Review Board for the Protection of Human Research Subjects (MDHHS IRB). Both IRBs determined that the study was exempt human subjects research, because identifying information was not collected from participants (Appendices A and B). Procedures In order to collect data from the sample population, I contacted each of the five child- placing agencies (CPAs) in Kent County, Michigan via an electronic mail (e-mail) message to 54 program managers or licensing supervisors within the agencies’ foster care programs. The CPAs are contracted by the Michigan Department of Health and Human Services to provide foster home licensing and foster care services. The five agencies include: Wellspring Lutheran Services, Bethany Christian Services, D.A. Blodgett-St. John’s, Samaritas, and Catholic Charities of West Michigan. All five are private, non-profit social service agencies. All but one (D.A.Blodgett-St. John’s) are faith-based organizations. In total, the CPAs oversee approximately 500 foster homes (C. Raftery, Care Coordinator Manager, West Michigan Partnership for Children [WMPC], personal communication, April 16, 2019), which include 652 individual foster parents (M. Pelz, WMPC Data Analytics Lead, personal communication, April 14, 2020). Of the 652 foster parents, 69.6% are female (n=454) and 29.1% are male (n=190). In terms of race, 83.6% of Kent County foster parents are White (n=545), 12.1% are Black/African American, and less than 1% are either Asian or American Indian/Alaska Native. Less than 3% of foster parents are Hispanic in ethnicity (2.8%, n=18). Nearly one-fifth of foster parents are kinship caregivers (18.4%, n=120; M. Pelz, WMPC Data Analytics Lead, personal communication, April 14, 2020). I secured agreements with all five CPAs to participate in the study. While no formal incentives were offered to agencies, I offered to share results and provide training to foster parents and staff based on the findings. The agencies agreed to share the survey with licensed foster homes via e-mail and agency foster parent portal websites. Each of the agencies identified a staff liaison for the study, whose role was to communicate with me and forward the survey invitation and subsequent reminders (Appendix C) to agency foster parents. Prior to dissemination of the survey, I pilot tested the survey. I sent targeted invitations to former foster parents and child welfare professionals with whom I had working relationships. I 55 also posted an open invitation on my personal social media account, requesting feedback from foster parents and child welfare professionals. I obtained feedback on length, ease of use, and clarity from 17 current foster parents (outside of the sampling frame), former foster parents, and child welfare professionals. Following IRB approvals, I sent an invitation to participate via e-mail message to each of the liaisons at the five CPAs (Appendix C). The invitation introduced the study and myself as the researcher. It included a link to the web-based survey within the Qualtrics survey platform. The initial invitation and subsequent messages included several elements consistent with Dillman’s (2014) method of increasing response rates, including highlighting the need for the study and potential contributions of participating, identifying Michigan State University and county CPA’s as collaborators, and citing potential practice implications for study results. Further, the message clarified that responses would go directly to me as the researcher and not to the CPAs, in order to encourage participants to respond freely and honestly. To ensure independence of observations among the data, the message instructed that only one parent from within two-parent foster homes should complete the survey. Finally, in the interest of transparency and increasing participant engagement, I also identified myself as a fellow foster parent and shared my personal investment in increased understanding of the foster parent experience. The initial invitations were sent via e-mail message from the CPA liaisons to all licensed foster parents for whom the agencies had e-mail addresses. Foster parents who clicked on the embedded link within the message were taken to the survey instrument in Qualtrics (Appendix D). The first page of the survey was a Research Participant Information and Consent Form, which ended with the statement, “By clicking the blue arrow below, you indicate your informed consent for participation in this research study.” Individuals who elected to participate were then 56 taken to a series of two screening questions, intended to establish the study sample. The first question asked, “Do you have a foster home license, issued to you by the State of Michigan?” Participants who selected “Yes” were taken to the next question. Those who responded “No” were taken to the end of the survey. This item ensured that the study sample included only those who self-identify as licensed foster parents. The second screening question asked, “How many TOTAL foster children have you had in your home, since you became licensed?” Participants who selected “No placements yet” were directed to the end of the survey. Those who selected any other response continued to the next survey question. This item ensured that the study sample included only those foster parents who have had at least one foster child placed in their home. Following the Statement of Informed Consent and initial screening questions, participants completed items related to their foster home, the stressors related to fostering they had experienced, and the respective stress levels of these stressors. Next, participants completed the three scales detailed within the Measures section below, in the following order: Parental Stress Scale, Social Provisions Scale, and Mental Health Continuum-Short Form. Demographic information was completed last, as recommended by Dillman (2014). Following completion of the survey, participants viewed a message of thanks for their time and efforts. In accordance with MSU HRPP guidelines regarding online surveys and lottery-type drawings, participants were not offered a financial incentive to participate in the study. Survey responses were collected for five weeks, between October and November 2019. Based on the Dillman method (Dillman et al., 2014) for increasing response rate, reminder e- mail messages were sent, via CPA liaisons, at 7 days, 14 days, and 21 days. At one month, CPA liaisons forwarded a final reminder, with a deadline set for one week later. Following the 57 deadline, I closed the survey to new responses and downloaded the data into IBM SPSS Statistics 23 for screening and analysis. Participants The study utilized non-random, convenience sampling methods to obtain the sample. This method was selected based on the lack of a large and accessible sampling frame from which to draw a randomized sample and the confidentiality constraints communicated to me by the participating CPAs. The convenience method of sampling is common among studies within the foster parent population (Gieger et al., 2017; Goemans et al., 2018; Octoman & McLean, 2014; Orme et al., 2006; Randle et al., 2017). The survey was sent to all licensed foster parents for whom CPAs had a valid e-mail address, which included approximately 500 homes. The web- based survey was completed by 148 participants. Given that survey instructions asked that only one foster parent per home complete the survey, the response rate was approximately 29.6%. Seven participants were eliminated because they did not fit study parameters. Specifically, participants must currently be licensed by the State of Michigan to provide family foster care. This parameter was selected to limit the sample to those foster parent who had completed a standard, state-regulated licensing process, which includes pre-service training, home study, and parent interviews, among other requirements. Additionally, participants must have had at least one foster child placed in their home since becoming licensed. This parameter was selected because the key study variables (parenting stress, social support, and well-being) are based on an assumption that participants have actually had the experience of fostering at least one child. Three participants indicated that they were not licensed foster parents, while four others were licensed but had not yet had foster children placed in their homes. Two other cases were excluded from the final sample because they were missing data for more than 10% of items. 58 According to Bennett (2001), statistical analysis may be biased when the amount of missing data is greater than 10%. After removing these nine cases, the final sample for the study included 139 licensed foster parents. Based on an a priori power analysis conducted in G*Power (Faul, Erdfelder, Lang & Buchner, 2007), the estimated sample size needed to detect a medium effect size (.15) within multiple linear regression with 5 total predictors is 89. The final sample size of 139 was determined to be sufficient to proceed with the planned analysis. Measures Sociodemographic Characteristics Age. Age was measured by asking participants to supply their year of birth and then subtracting it from 2019 during the data screening process. Gender. Gender was measured by asking participants to select one of the following options: Male, Female, or Other. Those who selected ‘Other’ had the option to describe how they identified their gender. Race. Participants were asked to select their race and/or ethnicity, with the option of selecting more than one category. Options included: American Indian/Alaska Native, Asian American, Black or African American, Hispanic or Latino, Native Hawaiian or Pacific Islander, White, or Other. Those who selected ‘Other’ had the option to provide a written description of their racial identity. If participants selected more than one category, they were later categorized as Multiracial. Parent Relationship. Parent relationship was measured by asking, “Which of the following best describes your household?” Response options included: Single-parent household, Two-parent household (unmarried), and Two-parent household (married). 59 Education. Education was measured by asking participants to select their highest level of education completed, with the following response options: Less than high school, High school graduate, Some college, Associate’s degree, Bachelor’s degree, and Advanced/Graduate degree. Employment. Employment was measured by asking participants to indicate whether they were employed full-time, employed part-time, or not employed. Income. Income was measured by asking participants to select the response that included their entire household income for the previous year, before taxes. Response options included: Less than $19,999, $20,000 to $39,999, $40,000 to $59,999, $60,000 to $79,999, $80,000 to $99,999, $100,000 to $119,999, $120,000 to $139,999, and $140,000 or more. Religious Participation. Religious participation was measured by asking participants whether they regularly (at least monthly) attend religious services. Response options included Yes and No. Foster Home Characteristics Total Time as a Foster Parent. Total time as a foster parent was measured by asking participants, “Approximately how many years have you been a foster parent?” Participants then responded in years and/or months. Total Number of Foster Children. Participants were asked to indicate how many total foster children they had had placed in their home, since becoming licensed. Response options included: No placements yet, 1-2, 3-5, 5-10, and More than 10. Current Number of Foster Children. Current number of foster children was measured by asking participants how many foster children are currently placed in their home. Response options included: 0, 1, 2, 3, 4, and More than 4. 60 Kinship Status. Respondents were given the following brief definition of a kinship caregiver: “A kinship caregiver is a relative or close friend (sometimes called fictive kin) who steps in to provide care and custody of children when their parents are unable to do so.” Respondents were then asked, “Based on this definition, are you a kinship caregiver?” with response options including Yes and No. Table 4 Operational definitions of study variables Conceptual Variable Stressors of fostering Parenting stress Well-being Retention Social support Stressors of Fostering Operational Definition Total number of stressors reported by each participant Total summed score on Parental Stress Scale (PSS) Total summed score (continuous scoring method) on Mental Health Continuum-Short Form (MHC-SF) Likelihood of still fostering in 18 months Total summed score on Social Provisions Scale (SPS) Potential stressors related to the fostering role were measured as a single item, developed for this study. The item states, “Foster parents experience many challenges in their role. Below is a list of potentially challenging events commonly experienced by foster parents. Which of these events have you experienced in the past two months or so?” Participants then selected as many events as applied to them. The list of potential stressors was developed from a review of fostering literature and includes the following, as well as an ‘Other’ selection, with an option to specify the stressor: • New child placed in the foster home • Foster child moved out of the home • Foster child behavior problems • Foster child medical needs 61 • Foster child mental health needs • Allegations of abuse or neglect against your foster home • Difficulty obtaining services for foster child • Difficulties in relationship with birth parents • Change in foster care case manager • Poor communication from case manager (or other agency staff) • Lack of information related to foster child • Disagreement with decision in foster care case The number of stressors endorsed by each participant was summed for a total score. Stressors of fostering was operationally defined as the total number of stressors reported by each participant. Table 4 provides operational definitions of key variables in the study. Additionally, participants who endorsed one or more stressors were asked to indicate how stressful they experienced each to be. For each endorsed stressor, participants were asked, “On a scale of 1 (not at all) to 4 (extremely), how stressful was [stressor] during the past two months or so?” Participants were not given the option to rate stressors that they had not endorsed. The responses on each individual’s endorsed stressors were totaled for an overall score. Parenting Stress Parenting stress was measured using the Parental Stress Scale (PSS; Berry & Jones, 1995). This scale measures positive and negative aspects of parenting and is intended for use with both mothers and fathers of children with and without clinical problems. The PSS is a self- report measure consisting of 18 Likert-type items that are represented on a 5-point scale which includes Strongly disagree, Disagree, Undecided, Agree, and Strongly agree. Sample items include “I feel overwhelmed by the responsibility of being a parent” and “Having child(ren) 62 leaves little time and flexibility in my life.” The PSS was adapted for this study by inserting the word “foster” before any reference to child(ren) or parent, (e.g., “Caring for my foster child(ren) sometimes takes more time and energy than I have to give” and “I am happy in my role as a foster parent”). After reverse-scoring 8 items, individual item scores are summed for a total possible score ranging from 18 to 90, with higher scores indicating greater parenting stress. The PSS has demonstrated good reliability, both in terms of internal reliability (α=.83) and test-retest reliability (r=.81) over a period of six weeks (Berry & Jones, 1995). The PSS also showed adequate concurrent validity (r=.75, p<.01) with the Parenting Stress Index (PSI; Abidin, 1986; Berry & Jones, 1995). Additionally, the scale demonstrated validity in discriminating between parents of typically developing children and parents of children with developmental and behavioral problems (Berry & Jones, 1995). Well-being The dependent variable, well-being, was assessed using the short form version of the Mental Health Continuum (MHC-SF; Keyes, 2009). The MHC-SF is a self-report measure of emotional, social, and psychological well-being. The scale consists of 14 items on which participants are asked to indicate how often they have felt or experienced each item (e.g. “happy” and “that your life has a sense of direction or meaning to it”) during the past month, with responses on a 6-point Likert-type scale including Never, Once or Twice, About Once a Week, About 2 or 3 Times a Week, Almost Every Day, and Every Day. The scale includes 3 items measuring emotional well-being, 5 items measuring social well-being, and 6 items measuring psychological well-being. The scale can be scored in two ways. The first scoring method is continuous scoring, in which individual item scores are summed for a total overall score ranging from zero to 70, with higher scores indicating greater overall well-being. The second scoring 63 method is Categorical Diagnosis, in which participants are categorized as flourishing, moderately mentally healthy, or languishing. A diagnosis of flourishing is made if someone feels 1 of the 3 emotional well-being symptoms "every day" or "almost every day" and feels 6 of the 11 positive functioning (social and psychological well-being) symptoms "every day" or "almost every day" in the past month. Languishing is the diagnosis when someone feels 1 of the 3 emotional well- being symptoms "never" or "once or twice" and feels 6 of the 11 positive functioning symptoms "never" or "once or twice" in the past month. Individuals who are neither “languishing” nor “flourishing” are coded as “moderately mentally healthy.” The continuous scoring method was used in this study, as a continuous variable was most useful to the proposed analysis. However, categorical diagnosis scoring was also performed and reported with descriptive results in Chapter 4. The MHC-SF has strong psychometrics and has been normed on general populations (Keyes, 2002). In terms of reliability, the MHC-SF demonstrated strong internal consistency (α>.80) and test-retest reliability over three months (r=.69) and nine months (r=.65). Construct validity of the MHC-SF is based on Ryff’s (1989) model of psychological well-being and Keyes’ (1998) model of social well-being. Retention The dependent variable of retention was operationalized as the intent to continue fostering, consistent with literature in this area (Denby et al., 1999; Eaton & Caltabiano, 2009; Geiger et al., 2013; Whenan et al., 2009). This was measured using a single item, which asked, “How likely is it that you will still be fostering 18 months from now?” Participants responded on a 4-point Likert-type scale, with possible responses including Very Unlikely, Somewhat Unlikely, Somewhat Likely, Very Likely. Additionally, it may be possible that foster parents may consider leaving their roles and even desire to do so, but may remain regardless of these 64 desires. In order to assess for thoughts about leaving, participants were asked, “In the past 2 months, how often have you thought about giving up fostering?” Participants responded on a 5- point Likert-type scale with the following response options: Never, Hardly ever, Sometimes, Often, Very often. Social Support Social support was measured with the Social Provisions Scale (SPS; Cutrona & Russell, 1987), a self-report measure of perceived social support from a functional approach. The scale consists of 24 items with responses on a 4-point Likert-type scale, including Strongly Disagree, Disagree, Agree, & Strongly Agree. The SPS includes six subscales, consisting of four items each and representing various functional types of social support: guidance (“e.g. There is someone I could talk to about important decisions in my life.”), reassurance of worth (e.g. “I have relationships where my competence and skills are recognized.”), social integration (e.g. “I feel part of a group of people who share my attitudes and beliefs.”), attachment (e.g. “I feel a strong emotional bond with at least one other person.”), nurturance (e.g. “There are people who depend on me for help.”), and reliable alliance (e.g. “There are people I can depend on to help me if I really need it.”). Twelve items in the scale are reverse-scored, and then items are summed for a total score ranging from 24 to 96, with higher scores indicating greater perceived social support. The SPS has demonstrated strong internal consistency for the total scale (α=.915) and adequate reliability for the various subscales (α=.653-.760; Cutrona & Russell, 1987). Cutrona and Russell (1987) also reported on support for the concurrent validity of the SPS, based on findings on the relationship between social support and measures of loneliness and interpersonal relationships. Construct validity of the SPS is based on Weiss’s (1974) model of social 65 provisions, which includes the six types of provisions needed for an individual to feel adequately supported, reflected in the subscales of the SPS (Cutrona & Russell, 1987). Analysis I downloaded study data from the online survey platform (Qualtrics) to IBM SPSS Statistics 23 for analysis. The following section details the data screening and analysis procedures completed for this study. First, I will discuss data screening, including checks of data integrity (post hoc power analysis, missing values analysis, and normality of distributions) and reliability of the instruments (Mental Health Continuum- Short Form, Parental Stress Scale, and Social Provisions Scale). Next, I will explain the preliminary analysis procedures, followed by a description of bivariate and multivariate analyses within each of the three primary research questions. Data Screening As an initial step in data screening, I removed all cases that did not fit study criteria. The total number of responses to the survey was 148. As detailed in the Participants section above, I removed nine total cases for a final sample size of 139. (Three participants indicated that they were not licensed foster parents. Four participants were licensed but had not yet had foster children placed in their homes. Two cases were excluded from the final sample because they were missing data for more than 10% of items.) I continued the data screening process by removing unnecessary variables added by Qualtrics and not contained in the survey instrument (e.g. survey start date). I then named and labeled all variables and recoded several variables to include the correct values. I changed any words in open-ended responses to the numbers that they represent (e.g., ‘one’ was changed to 1, ‘½’ to .5, etc.). I created new variables for total time fostering in years (participants gave 66 responses in years and months), total number of stressors (sum of stressors endorsed), total stress experienced (sum of stress levels for stressors endorsed), and age (birth year subtracted from 2019). I conducted a frequency analysis of all categorical variables. Due to low (or no) responses in some categories, I collapsed some of the categorical variables into fewer categories. The variable, Parent Relationship, was recoded from three categories (single parent, two-parent unmarried, and two-parent married) into two (single parent, two-parent). Income was recoded from 8 categories to 6. Additionally, I recoded individual race variables (e.g. Hispanic/Latino, White/Caucasian, etc.) into one categorical race variable with all response options included. Those participants who selected more than one race (N=6) were recoded into a Multiracial category. Due to the low number of participants within non-White racial/ethnic categories, I also created a dichotomous race variable with White participants in one category and all other races/ethnicities in the other. Finally, I collapsed the dependent variable of retention (How likely is it that you will still be fostering 18 months from now?) from four categories (Very Unlikely, Somewhat Unlikely, Somewhat Likely, and Very Likely) into two categories (Unlikely and Likely). Frequency analysis also revealed an error in the categories for the Total Number of Foster Children variable. The survey response options on this item included: No placements yet, 1-2 children, 3-5 children, 5-10 children, and More than 10 children. Participants who responded “No placements yet” were not included in the final sample, so the variable included four categories. However, there was overlap between two of the categories, as someone who had had five total foster children could select the ‘3-5 children’ response or the ‘5-10 children’ response. This error was addressed by combining the responses from these two categories into a new one: 67 3-10 children. The resulting variable, Total Number of Foster Children, has three categories: 1-2 children, 3-10 children, and More than 10 children. Next, I checked for data integrity. I conducted a Power Analysis to confirm statistical power of the sample. The power analysis indicated that 139 participants would produce sufficient power (1-β=.99) to detect a medium effect size (f²=0.15) in a multiple linear regression model with five predictor variables and a probability of a Type I error of α=.05. I performed a Missing Values Analysis, to check for amount and patterns of missing data. Most variables had 0% missing data, and all but one had less than 3% missing. The variable of Income was missing data in 5% of cases. According to Little’s Missing Completely at Random (MCAR) test, the data are likely to be missing at random. (The results indicated a failure to reject the null hypothesis [p=.648] that data are missing completely at random.) Therefore, missing values were not imputed or replaced with mean values. Missing data were handled via pairwise deletion within each set of statistical tests. I conducted a descriptive analysis for continuous variables in order to check for the assumption of normality of distribution. I examined skewness and kurtosis of all continuous variables. The variable, Total Time as a Foster Parent, was out of range on both skewness (4.17) and kurtosis (20.35). In order to address this, I converted the continuous variable, Total Time as a Foster Parent, into a categorical variable with the following categories: Less than 1 year, 1-2 years, 2-5 years, and More than 5 years. The four categories each have roughly 25% of participants. I also created boxplots of continuous variables to identify potential outliers. Analysis indicated that the well-being variable (MHC-SF total score) had one potential extreme outlier. The extreme outlier value was within the range of possible values for this variable. Additionally, the 5% trimmed mean (51.45) was not far from the actual mean (50.67), and was 68 within the 95% Confidence Interval for the mean (48.61-52.72). Based on this analysis, I elected to leave the potential outlier in the dataset. I completed reverse scoring on the indicated items of the PSS and SPS instruments and then created total score variables for the PSS, SPS, and MHC-SF as directed in each instrument’s scoring guidelines. In addition, I scored and created new variables for the subscales of the SPS (guidance, reassurance of worth, social integration, attachment, nurturance, and reliable alliance) and for the categorical diagnosis scoring of the MHC-SF (flourishing, moderately mentally healthy, or languishing). I then conducted reliability analyses for the scales and SPS subscales. As illustrated in Table 5, the reliability statistic for each of the scales and subscales is comparable to (or greater than) those found in each instrument’s respective validation studies. Table 5 Scale reliability Mental Health Continuum- Short Form Parental Stress Scale Social Provisions Scale Guidance Reassurance of worth Social integration Attachment Nurturance Reliable alliance 69 Cronbach’s α .931 .917 .943 .852 .815 .819 .810 .733 .810 Preliminary Analysis I generated descriptive statistics for sociodemographic variables, foster home characteristic variables, and key study variables. This included frequencies of categorical variables (number and percentage) and measures of central tendency (mean, median) and measures of dispersion (standard deviation, range) for continuous variables. Bivariate analysis was then conducted to identify those variables that have a significant relationship to the dependent variables within the proposed theoretical model (parenting stress, well-being, and retention), in order to determine which variables to include in the final multivariate analyses. For the dependent variables of parenting stress and well-being (continuous), this included independent samples t-tests for dichotomous variables (gender, kinship, single/two-parent), one- way ANOVA for categorical variables (race, agency, employment, income, education level, total number of children, current number of children), and Pearson’s correlation for continuous variables (total stressors, total stress level, time as a foster parent, age, parenting stress, social support). For the dependent variable of retention (categorical), this included chi-square goodness of fit tests for categorical variables and independent samples t-tests for continuous variables. Analysis of Research Questions Lastly, I examined the following research questions using various multivariate analysis procedures: RQ1. A. Is there a relationship between stressors experienced by foster parents and their levels of parenting stress? B. Does social support moderate this relationship? C. Is there a relationship between individual stressors and parenting stress? RQ2. A. Is there a relationship between parenting stress and well-being among foster parents? B. Does social support moderate this relationship? 70 C. Do the various types of social support moderate this relationship? RQ3. A. Is there a relationship between parenting stress and retention (intent to continue) among foster parents? B. Does social support moderate this relationship? C. Is there a relationship between parenting stress and thoughts about leaving fostering? D. Does social support moderate this relationship? Prior to multivariate analysis, I centered the continuous variables of parenting stress and social support (as well as social support subscales) on their respective mean scores. I examined research question 1A using bivariate correlation (Pearson’s r) of the total number of stressors (IV) and parenting stress total score (DV). I then used multiple linear regression to further analyze the relationship between the two variables, as well as potential covariates and the moderator, social support. For covariates with three or more categories, I created dummy variables prior to entering them into the regression model. For research question 1B, an interaction term between social support and total stressors was added to the regression model. For research question 1C, I used independent samples t-tests to investigate whether there were significant differences on mean parenting stress scores between foster parents who reported experiencing each individual stressor and those who did not. Following bivariate analysis, I used multiple linear regression to examine research question 2A. The independent variable (parenting stress) and moderating variable (social support) were entered into the regression analysis, along with covariates, and the dependent variable (well-being). For research question 2B, an interaction term including parenting stress and social support was added to the regression. For research question 2C, I repeated the multiple linear regression analyses for each type of social support, operationalized as subscales of the SPS measure. 71 Lastly, I utilized logistic regression to explore the final set of research questions. The dichotomous retention variable was used as the dependent variable. The independent variable (parenting stress), moderator (social support), and covariates were entered into the regression analysis. For research question 3B, an interaction term (parenting stress x social support) was also entered. This procedure was repeated for research questions 3C and 3D, using thoughts about leaving fostering as the dependent variable, in place of retention (intent to continue). 72 CHAPTER 4: DESCRIPTIVE RESULTS Following the data screening process outlined in Chapter 3, I generated descriptive statistics for sociodemographic variables, foster home characteristic variables, and key study variables. This chapter details the results of this analysis. Sociodemographic Characteristics The sample, which comprised 139 foster parents, was overwhelmingly female (n=126, 90.6%) and white (n=127, 92.7%). The mean age of participants was 42.73 years (Range= 23.0- 79.0 years). The majority of participants represented two-parent foster homes (n=113, 82.5%). Education level of participants ranged from high school completion to advanced or graduate degree, with the largest percentage of participants reporting a Bachelor’s degree (n=39, 45.5%) as their highest completed level of education. Most participants were employed outside the home, with 44.2% (n=61) employed full-time and 28.3% (n=39) employed part-time. Annual income ranged from below $40,000 to over $120,000, with the largest percentage of participants (n=32, 24.2%) reporting $60,000-79,999. The majority of study participants (n=98, 70.5%) reported regular (at least monthly) religious participation. Table 6 details the demographic characteristics of the sample. Table 6 Sociodemographic characteristics of the sample (N = 139) Age Gender Mean (SD) 42.73 (11.69) Frequency (%) 13 (9.4) 126 (90.6) Male Female 73 Table 6 (cont’d) Race/Ethnicity American Indian/Alaska Native Asian Black or African American Hispanic or Latino White Multi-racial Other Household type Single parent Two-parent Employment Income Not employed Employed, part-time Employed, full-time Less than $40,000 $40,000-59,999 $60,000-79,999 $80,000-99,999 $100,000-119,999 $120,00 or more Education Less than high school High school Some college Associate’s degree Bachelor’s degree Advanced/graduate degree Religious participation Yes No 0 (0) 1 (0.7) 4 (2.9) 0 (0) 4 (2.9) 1 (0.7) 127 (92.7) 24 (17.5) 113 (82.5) 38 (27.5) 39 (28.3) 61 (44.2) 13 (9.8) 26 (19.7) 32 (24.2) 24 (18.2) 14 (10.6) 23 (17.4) 0 (0) 13 (9.4) 29 (21.0) 15 (10.9) 49 (35.5) 32 (23.2) 98 (70.5) 41 (29.5) Foster Home Characteristics In addition to sociodemographic characteristics, information on foster home characteristics was also collected from study participants. The range of fostering experience among participants was quite large, with the time since initial foster home licensure ranging from 2 months to 50 years. The mean length of time fostering was 4.58 years, while the median was 2.08 years. (To account for non-normal distribution [Skewness=4.17, Kurtosis=20.35], this 74 variable was transformed from continuous to categorical for analysis. Descriptive statistics are reported below for both the continuous and categorical variables which capture ‘Total time as a foster parent.’) Foster parent participants also varied in terms of the total number of children fostered since becoming licensed, with the most reporting between 3 and 10 total children fostered (n=61, 43.9%). In terms of current foster placements, most participants had one foster child currently placed with them (n=45, 32.4%), though a sizable proportion had no children placed at the time of survey completion (n=41, 29.5%). Eighteen participants (13.1%) indicated that they were kinship caregivers, defined within the survey as “a relative or close friend (sometimes called fictive kin) who steps in to provide care and custody of children when their parents are unable to do so.” The sample included participants from each of the five child- placing agencies participating in the study, with the greatest number (n=51, 36.7%) licensed through Agency C. Table 7 details the foster home characteristics of the sample. In terms of representativeness of the sample in light of Kent County foster parent population data, the sample has higher proportions of female foster parents (90.6%; Kent County: 69.6%) and White foster parents (Sample: 92.7%; Kent County: 83.6%), and a somewhat lower proportion of kinship caregivers (13.1%; Kent County: 18.4%). Table 7 Foster home characteristics of the sample (N = 139) Total time as a foster parent (continuous, in years) Total time as a foster parent (categorical) Less than 1 yr. 1-2 yrs. 2-5 yrs. More than 5 yrs. Mean (SD) 4.58 (7.07) Frequency (%) 34 (24.5) 35 (25.2) 36 (25.9) 34 (24.5) 75 Table 7 (cont’d) Total number of foster children since licensure Current number of foster children 1-2 3-10 10+ 0 1 2 3 or more Foster care agency Agency A Agency B Agency C Agency D Agency E Kinship caregiver Yes No 50 (36.0) 61 (43.9) 28 (20.1) 41 (29.5) 45 (32.4) 39 (28.1) 14 (10.1) 14 (10.1) 19 (13.7) 51 (36.7) 32 (23.0) 23 (16.5) 18 (13.1) 119 (86.9) Stressors Participants were given a list of common challenges of fostering and asked to indicate which they had experienced ‘in the past two months or so.’ Participants were able to select multiple stressors. The most frequently reported stressor of fostering was foster child behavior problems, experienced by 41.7% of participants. Similarly, 39.6% of foster parents reported experiencing foster child mental health needs. The third and fourth most common stressors were poor communication from agency staff (35.3%) and change in case manager (31.7%). The least frequently experienced stressors included disagreement with a licensing rule or policy (10.1%) and allegation of abuse or neglect against the foster home (10.1%). Six participants selected ‘Other’ as a type of stressor experienced, with five participants elaborating on this selection with open-ended responses related to difficulty in getting return calls, separation of siblings, adoption of foster child, ‘getting in trouble’ for foster child behavior, and conflicting guidance from various professionals involved in a case. 76 Foster parents who indicated that they had experienced one or more of the challenges in the past two months were then asked to rate how stressful they experienced each one to be. Response options included: (1) Not at all stressful, (2) Mildly stressful, (3) Moderately stressful, and (4) Extremely stressful. Those stressors with the highest mean stress levels included allegation of abuse or neglect against the foster home (M = 3.29), disagreement with a decision in the case (M = 3.23), poor communication from agency staff (M = 3.19), and foster child behavior problems (M = 3.11), all of which were rated somewhere between moderately stressful and extremely stressful. The stressors with the lowest mean stress levels were foster child medical needs (M = 2.18) and change in case manager (M = 2.26), rated just above mildly stressful. None of the stressors had mean scores at or below mildly stressful (2.0). Table 8 provides detailed results on the frequency of stressors experienced and the stress level reported for each. Table 8 Stressors of fostering (N = 139) Foster child behavior problems Foster child mental health needs Poor communication from agency staff Change in case manager Foster child medical needs Difficulty obtaining services for foster child Foster child moved from the home New foster child Lack of information on foster child Difficulty with birth parents Disagreement with decision in case Allegation of abuse or neglect against foster home 77 Stressor Experienced* Frequency (%) 58 (41.7) 55 (39.6) 49 (35.3) 44 (31.7) 39 (28.1) 31 (22.3) 30 (21.6) 29 (20.9) 28 (20.1) 22 (15.8) 22 (15.8) 14 (10.1) Stress Level** Mean (SD) 3.12 (.774) 2.84 (.764) 3.19 (.891) 2.26 (.902) 2.18 (.790) 2.97 (.850) 2.72 (.960) 2.61 (.956) 2.71 (.854) 2.91 (.921) 3.23 (.813) 3.29 (.914) Table 8 (cont’d) Disagreement with licensing rule or policy Other Total number of stressors experienced Total stress level *Frequency of participants who experienced each stressor ‘in the past 2 months or so.’ **Average stress level for those who reported experiencing each stressor. 2.79 (.802) 2.86 (1.464) Mean (SD) 3.21 (2.45) 8.84 (7.59) 14 (10.1) 6 (4.3) Parenting Stress Parenting stress was measured using the Parental Stress Scale (PSS; Berry & Jones, 1995). After reverse-scoring indicated items, item scores are totaled for a possible overall score ranging from 18 to 90, with higher scores indicating greater levels of parenting stress. The overall score of participants in the sample ranged from 19.00 to 80.00. The mean total score of participants in the sample was 45.48, with a standard deviation of 12.53. Well-being Well-being was measured using the Mental Health Continuum, Short Form (MHC-SF; Keyes, 2009). This measure of emotional, social, and psychological well-being can be scored in two ways. The first is a total score, found by summing the item responses and ranging from a possible score of zero to 70. Higher overall scores are indicative of greater well-being. Participants in the sample had overall scores ranging from 5 to 70, with a mean overall score of 50.67 (SD = 12.26). The second scoring method is a categorical diagnosis method in which participants are categorized as flourishing, moderately mentally healthy, or languishing. The majority of study participants were categorized as flourishing (74.6%), while approximately a quarter of participants were either moderately mentally healthy or languishing. Table 9 includes detailed results on the MHC-SF. 78 Table 9 Well-being scores of the sample (N = 134) Well-being (Total) Well-being (Categorical diagnosis) Languishing Moderate Flourishing Mean (SD) 50.67 (12.26) Frequency (%) 4 (3.0) 30 (22.4) 100 (74.6) Retention Retention was measured by asking participants to indicate the likelihood that they would still be licensed 18 months from now, with response options including Very unlikely, Somewhat unlikely, Somewhat likely, and Very likely. Approximately two-thirds of participants indicated that it was either very likely (31.2%) or somewhat likely (37.4%) that they would still be fostering in 18 months. Participants were also asked how often they have thought about giving up fostering in the past two months, with response options including Never, Hardly ever, Sometimes, Often, and Very often. The largest percentage of participants (33.3%) indicated that they sometimes thought about giving up fostering, though a similar proportion (28.3%) of the sample reported never thinking about it. The smallest percentage (5.8%) reported that they thought about it very often. Detailed results for both retention-related items are in Table 10. Table 10 Retention results of the sample (N = 138) 79 Frequency (%) Table 10 (cont’d) Likelihood of fostering 18 months from now Thought about giving up fostering Very likely Somewhat likely Somewhat unlikely Very unlikely Never Hardly ever Sometimes Often Very often 43 (31.2) 52 (37.4) 27 (19.6) 16 (11.6) 39 (28.3) 18 (13.0) 46 (33.3) 27 (19.4) 8 (5.8) Social Support Social support was measured using the Social Provisions Scale (SPS; Cutrona & Russell, 1987), a measure of perceived support. Possible total scores on the measure range from 24 to 96, with higher scores indicating greater social support. After reverse-scoring indicated items, participant scores were summed for an overall score. The mean overall score within the sample was 81.04 (SD = 10.16; Min = 45, Max = 96). The SPS also includes six subscales, representing various functional types of social support: • Guidance: Advice or information, usually specific to the problem at hand • Reassurance of worth: Recognition of one’s skills, competence, and value by others, and • Social integration: Relationships in which individuals share concerns and create a sense • Attachment: Relationships characterized by actions that lead one to believe that one is is sometimes called esteem support of belonging and companionship loved and cared for within a relationship • Opportunities for nurturance: The sense that one provides as well as receives support • Reliable alliance: The perception that one can depend on others for material assistance 80 Each subscale has a possible score ranging from 4 to 16.The mean subscale scores within the study sample all fell between 13 and 14. The highest mean score was on the Guidance subscale (M = 13.82), while the lowest was on the Social Integration subscale (13.22). Table 11 provides detailed results on the SPS and its subscales. Table 11 Social support results of the sample (N = 139) Social support Guidance Reassurance of worth Social Integration Attachment Nurturance Reliable alliance Conclusion Mean (SD) 81.04 (10.16) 13.82 (2.08) 13.30 (1.99) 13.22 (2.16) 13.42 (2.31) 13.54 (1.88) 13.72 (2.10) In summary, descriptive analyses revealed that the study sample (N=139) was primarily female (90.6%), white (92.7%), and from a two-parent household (82.5%). The mean age was approximately 43 years and most participants (70.5%) reported regular religious participation. In terms of foster home characteristics, half of participants had been fostering for two years or less and most (60.5%) were currently fostering 1 or 2 children. Kinship caregivers were 13% of the sample. The most frequently experienced stressors of fostering were foster child behavior problems, foster child mental health needs, and poor communication from agency staff. The stressors with the highest reported levels of stress were allegations of abuse or neglect, disagreement with a decision in the case, and poor communication from agency staff. 81 Participants had a mean parenting stress score of 45.48 (on a scale of 18 to 90), and 75% of participants were in the flourishing category of well-being. In response to retention items, two- thirds indicated that it was likely that they would still be fostering 18 months from now. A quarter of participants said that they think about giving up fostering often or very often. In Chapter 6, I will offer a detailed discussion of these results. First, in Chapter 5, I will present bivariate and multivariate results in response to the identified research questions. 82 CHAPTER 5: BIVARIATE AND MULTIVARIATE RESULTS This chapter presents the results of statistical analyses used to answer the following research questions: RQ1. A. Is there a relationship between stressors experienced by foster parents and their levels of parenting stress? D. Does social support moderate this relationship? E. Is there a relationship between individual stressors and parenting stress? RQ2. A. Is there a relationship between parenting stress and well-being among foster parents? B. Does social support moderate this relationship? C. Do the various types of social support moderate this relationship? RQ3. A. Is there a relationship between parenting stress and retention (intent to continue) among foster parents? E. Does social support moderate this relationship? F. Is there a relationship between parenting stress and thoughts about leaving fostering? G. Does social support moderate this relationship? Following the descriptive analyses detailed in Chapter 4, I used bivariate analysis procedures within each set of research questions to assess the relationships between dependent variables and independent variables and potential covariates. Finally, I utilized multivariate analyses (linear and logistic regression) to respond to each research question and to test the proposed theoretical framework (Figure 6). Figure 6 Proposed buffering model of foster parenting 83 Research Question 1 The first set of research questions was designed to assess the relationship between potential stressors of fostering and parenting stress, and the potential buffering (or moderating) role of social support. RQ1. A. Is there a relationship between stressors experienced by foster parents and their levels of parenting stress? B. Does social support moderate this relationship? C. Is there a relationship between individual stressors and parenting stress? In order to test these research questions, I first completed bivariate analyses between the dependent variable (parenting stress) and the independent variables (total number of stressors, social support) and potential covariates (including sociodemographic and foster home characteristics). Pearson correlation was used to test the relationships between continuous variables and the dependent variable, parenting stress. As shown in Table 12, a significant negative relationship was found between parenting stress and the moderating variable, social support (r = -.442, p<.001). This suggests that greater levels of social support are correlated with lower levels of parenting stress. Additionally, there was a significant, negative relationship between parenting stress and age (r = .272, p=.001), indicating that higher age is related to lower levels of parenting stress. The relationship between parenting stress and the independent variable, total number of stressors, was not found to be significant. Table 12 Pearson correlation results, Research question 1 Variable 1. Parenting stress 2. Total number of stressors 3. Social support 4. Age *p<.05, **p<.01, ***p<.001 -.442*** -.272** 1 - .130 2 - .033 .055 3 - -.110 4 - 84 Independent samples t-tests were used to assess the differences in mean on the dependent variable, parenting stress, among dichotomous variables. Results, as displayed in Table 13, demonstrated that there were significant differences on parenting stress scores between foster parents who are single parents and those who are from two-parent households (t = -3.111, p = .002). However, the nature of the difference is the opposite of what one might expect. Foster parents from single parent households reported lower levels of parenting stress (M = 38.29, SD = 11.84) than those from two-parent households (M = 46.78, SD = 12.21). Though there were differences on mean scores of parenting stress within other dichotomous variables, none were significant at the p<.05 level. Table 13 Independent samples t-test and ANOVA results for parenting stress Gender Mean (SD) Male Female Race White Non-white Parent relationship Single parent Two-parent Kinship Religious Yes No Yes No Income Employment $59,999 or less $60,000-99,999 $100,000 or more Not employed Employed, part-time Employed, full-time 41.62 (10.83) 45.88 (12.67) 45.83 (12.51) 39.40 (12.65) 38.29 (11.84) 46.78 (12.21) 40.39 (15.57) 46.24 (12.00) 46.23 (12.22) 43.68 (13.22) Mean (SD) 40.69 (11.29) 46.64 (11.85) 47.87 (12.47) 44.32 (11.57) 46.36 (12.32) 45.92 (13.29) -3.111** t -1.170 -1.565 1.851 -1.095 F 1.76 .29 85 Less than 1 yr 1-2 yrs 2-5 yrs More than 5 yrs 1-2 3-10 More than 10 Total number of foster children Table 13 (cont’d) Education High school or less Some college/Associate’s Degree Bachelor’s Degree or higher Total time as a foster parent Current number of foster children 0 1 2 3 or more Foster care agency Agency A Agency B Agency C Agency D Agency E *p<.05, **p<.01, ***p<.001 44.69 (8.52) 40.23 (13.42) 48.04 (11.25) 47.47 (14.65) 46.57 (12.49) 42.53 (9.56) 45.50 (12.53) 46.50 (13.85) 44.93 (10.99) 44.86 (13.54) 46.12 (12.47) 42.80 (13.38) 47.13 (11.52) 47.64 (12.51) 48.50 (14.05) 41.26 (11.89) 44.75 (11.77) 44.44 (12.62) 50.22 (12.91) 6.28** 1.04 .26 1.09 1.70 The next set of bivariate tests performed for the first research question was One-Way Analysis of Variance (ANOVA), in order to assess for differences on mean parenting stress scores within categorical variables with three or more categories. Table 13 includes mean scores by category and the corresponding F statistic for each variable. Of the sociodemographic variables, income (F(2,106) = 1.76, p = .176) and employment (F(2,135) = .29, p = .749) did not demonstrate significant differences on mean scores of parenting stress. Additionally, none of the following foster home characteristic variables were found to have significant differences on parenting stress scores: time as a foster parent (F(3,135) = 1.04, p = .377), total number of children fostered (F(2,136) = .26, p = .775), current number of foster children (F(3,135) = 1.09, p = .357), foster care agency (F(4,134) = 1.70, p = .155). The only categorical variable found to have significant differences on mean scores of parenting stress was education (F(2,135) = 6.28, p 86 = .002). Following the significant results in One-way ANOVA, a post-hoc test was performed. Post hoc comparisons using the Tukey HSD test indicated that the mean parenting stress score for participants with some college or Associate’s degree (M = 40.23, SD = 13.42) was significantly lower than those with a Bachelor’s degree or higher (M = 48.04, SD = 11.25). The mean score of participants with a high school education (M = 44.69, SD = 8.52) did not significantly differ from the other two groups, however. Research Question 1, A & B In order to assess the relationship between total number of stressors and parenting stress, while controlling for the moderator and covariates, I utilized multiple linear regression analysis. Parenting stress was entered as the dependent variable, while total number of stressors (IV), social support (moderator), and identified covariates (age, parent relationship, and education level) were all entered as predictor variables. Dummy coding was used for the categorical variable education level, and “High school or less” was left out of the regression analysis as the reference category. As shown in Table 14, the overall model was significant at the p<.001 level. Multicollinearity diagnostic coefficients (tolerance and VIF) were well within acceptable ranges (Pallant, 2013). The R2coefficient indicated that approximately 35% of the variance in parenting stress scores among the sample could be explained by the proposed model. However, the only significant predictors of parenting stress within the model were parent relationship and social support. These results indicate that, in response to Research Question 1A, there is no significant relationship between the total number of stressors reported by foster parents and parenting stress. A second regression model, to test the moderating effect of social support (RQ 1B), was not necessary, because there was no significant relationship between the total number of stressors 87 and parenting stress. In response to Research Question 1B, social support is not a moderator of this relationship, but rather has a direct effect on the dependent variable of parenting stress. Table 14 Multiple linear regression results for parenting stress Variable Constant Age Parent relationship Education (some college/Assoc. degree) Education (Bachelor’s degree or higher) Social support Total number of stressors R2 *p<.05, **p<.01, ***p<.001 -.16 (.08) 8.23 (2.41) -4.77 (3.28) .52 (3.08) -.53 (.09) .61 (.37) .356*** B (SE) 79.97(9.44)*** β -.15 .25** -.18 -.02 -.43*** .12 Research Question 1, C Following the main analyses for the first set of research questions, I explored the relationship between individual stressors and parenting stress scores. For this analysis, I utilized independent samples t-tests to investigate whether there were significant differences on mean parenting stress scores between foster parents who reported experiencing each individual stressor and those who did not. Results, detailed in Table 15 indicated that mean differences on parenting stress were significant for four of the stressors: child behavior problems, difficulty obtaining services, disagreement with a decision in the case, and disagreement with a licensing rule or policy. For three of the stressors (child behavior problems, difficulty obtaining services, and disagreement with a licensing rule or policy), foster parents who reported experiencing the stressor had higher mean parenting stress scores. Foster parents who experienced a disagreement with a decision in the foster care case had parenting stress scores that were significantly lower than those who did not. In terms of magnitude, each of these significant differences evidenced a small (child behaviors: d=.409; disagreement with decision: d=-.480) or medium (difficulty 88 obtaining services: d=.741; disagreement with licensing policy: d=.559) effect size (Cohen, 1988). These results suggest that while the cumulative amount of stressors experienced does not relate to parenting stress, the presence of specific stressors does impact the experience of parenting stress. Table 15 Stressor experienced SD 12.71 M 47.59 43.37 13.03 45.21 12.10 12.94 45.80 SD 12.49 M 44.93 Stressor not experienced Independent samples t-test for parenting stress for individual stressors Stressor New foster child Child moved from home Child behavior problems Child medical needs Child mental health needs Allegations of abuse/neglect Difficulty obtaining services Difficulty with birth parents Change in case manager Poor communication from agency Lack of information on foster child Disagreement with a decision in the case Disagreement with a licensing rule or policy *p<.05, **p<.01, ***p<.001 44.05 45.31 43.58 45.95 44.58 46.39 44.78 12.42 12.45 12.57 12.59 12.40 11.99 12.49 12.18 44.82 12.51 45.39 46.47 48.43 44.67 47.67 47.00 52.10 45.95 44.48 47.14 48.11 40.64 51.71 t -1.016 -.485 -2.387* .478 -1.678 -.477 10.66 12.63 11.54 12.85 17.13 10.41 -3.464** 12.94 12.87 12.65 12.25 11.32 12.41 -.192 .642 -1.154 -1.243 1.998* -1.983* Research Question 2 The second set of research questions was designed to assess the relationship between parenting stress and well-being among foster parents, and the potential buffering (or moderating) role of social support. RQ2. A. Is there a relationship between parenting stress and well-being among foster parents? 89 B. Does social support moderate this relationship? C. Do the various types of social support moderate this relationship? In order to test these research questions, I first completed bivariate analyses between the dependent variable (well-being) and the independent variables (parenting stress, social support) and potential covariates (including sociodemographic and foster home characteristics). I utilized Pearson correlation to test the bivariate relationships between the dependent variable (well-being) and the independent variables (parenting stress, social support), as well as a continuous potential covariate (age). Significant relationships were found between the dependent variable and each of the other continuous variables, as noted in Table 16. Higher levels of well- being were found to be associated with lower parenting stress, greater social support, and higher age. Additionally, as indicated for RQ1, parenting stress and social support were significantly negatively correlated. Age and parenting stress were also significantly negatively correlated. Table 16 Pearson correlation results; Research question 2 1. Well-being 2. Parenting stress 3. Social support 4. Age *p<.05, **p<.01, ***p<.001 -.477*** .587*** .177** 1 - 2 - -.442*** -.272** 3 - .055 4 - Independent samples t-tests were used to test for significant differences on mean scores of the dependent variable (well-being) within dichotomous variables. Results, as displayed in Table 17, demonstrated that there were no significant differences on well-being scores within the dichotomous variables. 90 Table 17 Independent samples t-test and ANOVA results for well-being Gender Male Female Race White Non-white Parent relationship Single parent Two-parent Kinship Religious participation Yes No Yes No Income Employment $59,999 or less $60,000-99,999 $100,000 or more Not employed Employed, part-time Employed, full-time Education High school or less Some college/Associate’s Degree Bachelor’s Degree or higher Total time as a foster parent Total number of foster children Current number of foster children Less than 1 yr 1-2 yrs 2-5 yrs More than 5 yrs 1-2 3-10 More than 10 0 1 2 3 or more Mean (SD) 50.31 (9.31) 50.71 (12.55) 50.59 (12.50) 51.30 (10.51) 51.91 (14.78) 50.72 (11.59) 53.06 (14.51) 50.31 (12.00) 52.01 (11.01) 47.76 (14.46) Mean (SD) 48.15 (10.05) 49.03 (12.81) 52.26 (12.84) 51.84 (11.78) 52.56 (8.60) 48.54 (14.25) 52.62 (6.96) 52.64 (13.20) 49.41 (12.35) 50.06 (11.87) 48.71 (13.15) 51.83 (9.30) 52.06 (14.47) 49.12 (11.66) 51.48 (11.42) 51.68 (14.96) 51.56 (9.64) 50.76 (11.04) 51.05 (13.84) 46.71 (17.80) t -.111 .175 .438 -.876 -1.806 F .931 1.568 1.157 .574 .623 .566 91 Table 17 (cont’d) Foster care agency Agency A Agency B Agency C Agency D Agency E *p<.05, **p<.01, ***p<.001 42.79 (14.77) 55.32 (9.79) 51.35 (10.58) 51.56 (11.56) 48.87 (48.87) 2.434* One-Way Analysis of Variance (ANOVA) was utilized to assess for differences on mean well-being scores within categorical variables with three or more categories. Table 17 includes mean scores by category and the corresponding F statistic for each variable. None of the sociodemographic variables demonstrated significant differences on mean scores of well-being: income (F[2,106] = .931, p = .397), education (F[2, 135] = 1.157, p = .317), employment (F[2, 135] = 1.568, p = .212). Additionally, the following foster home characteristic variables were found to have no significant differences on parenting stress scores: time as a foster parent (F[3,135] = .574, p = .633), total number of children fostered (F[2,136] = .623, p = .538), current number of foster children (F[3,135] = .566, p = .638). The only categorical variable found to have significant differences on mean scores of well-being was foster care agency (F[4, 134] = 2.434, p = .050). Following the significant results in One-way ANOVA, a post-hoc test was performed. Post hoc comparisons using the Tukey HSD test indicated that the mean well-being score for foster parents from Agency A (M = 42.79, SD = 14.77) was significantly lower than those from Agency B (M = 55.32, SD = 9.79). The mean difference of 12.53 (95% CI: .83- 24.23) was significant at p = .029. None of the other mean differences on well-being between foster care agencies were statistically significant. Research Question 2, A & B Multiple linear regression was used to analyze the relationship between parenting stress and well-being, while controlling for the moderating variable and covariates. Before the 92 regression model was run, the independent variable (parenting stress) and moderating variable (social support) were centered on their respective means. In the first regression model, well- being was entered as the dependent variable, while parenting stress (IV), social support (moderator), and identified covariates (age and foster care agency) were all entered as predictors. Dummy coding was created for the foster care agency variable, and “Agency A” was left out of the regression analysis as the reference category. As shown in Table 18, Model 1 was statistically significant F(7, 127)=14.229, p<.001. Correlations between the independent variables, as well as multicollinearity diagnostic coefficients (tolerance and VIF) were well within acceptable ranges (Pallant, 2013). The R2coefficient indicated that approximately 44% of the variance in well-being scores among the sample could be explained by the proposed model. Parenting stress, social support, and foster care agency were all significant predictors of well- being, with social support being the strongest predictor (β=.471). Age did not make a unique significant contribution. The second regression model was tested in order to assess the role of social support as a moderating variable in the relationship between parenting stress and well-being. After centering both variables on their respective means, I created an interaction term including the IV (parenting stress) and moderating variable (social support). The interaction term was added to Model 2. Results indicated that the overall model was statistically significant F(8,126)=13.625, p<.001. The model explained 46% (R2=.464) of the variance in well-being scores. Social support remained the strongest predictor of well-being (β=.420). The interaction term was found to be significant (B=.014, SE=.006, β=.166, p=.018). The increase in R2 from Model 1 to Model 2 was .024, indicating that the interaction between parenting stress and social support explains an additional 2.4% of the variance in well-being, when the effects of parenting stress and well-being 93 are controlled. Figure 7 illustrates the interaction effect between parenting stress and social support on the dependent variable, well-being. Foster parents who reported high levels of social support experienced a buffering effect, which means that their well-being was less negatively impacted than those with low levels of social support. The model was entered into an ANCOVA procedure, which demonstrated that the interaction term had a small effect on the dependent variable (Partial Eta squared = .038; Cohen, 1988). Social support had a large effect size on wellbeing (PES=.199) and parenting stress had a medium effect size (PES=.068). Foster care agency had a small effect on the dependent variable of well-being. Table 18 B β Model 1 40.637*** Multiple linear regression results for well-being Variable Constant Age Agency (B)a Agency (C) Agency (D) Agency (E) Parenting Stress Social Support Parenting Stress*Social Support R2 ΔR2 *p<.05, **p<.01, ***p<.001 aAgency A was used as the reference group among foster care agencies. .089 7.191* 7.078* 7.110* 6.092 -.222** .568*** .085 .202 .279 .245 .185 -.227 .471 .440 SE 4.221 .073 3.420 2.897 3.079 3.248 .077 .091 B 41.915*** .095 5.727 6.252* 6.174* 5.743 -.242** .507*** .014* .464 .024 Model 2 β .091 .161 .247 .213 .175 -.248 .420 .166 SE 4.015 .072 3.414 2.865 3.049 3.192 .076 .093 .006 94 Figure 7 Interaction Effect of Parenting Stress and Social Support on Well-being 60 55 50 Low Social Support High Social Support g n 45 i e b - l l e W 40 35 30 25 20 Low Parenting Stress High Parenting Stress These results provide evidence in response to the identified research questions. First, results indicate that there is a significant, negative relationship between parenting stress and well-being among foster parents. Second, social support functions as a moderator, or buffer, within the relationship between parenting stress and well-being. Specifically, among foster parents who had greater social support, parenting stress had a less negative impact on their well- being. Additionally, social support appears to have a direct effect on well-being, regardless of the presence of parenting stress. Research Question 2, C Following the analyses for research questions 2A and 2B, I completed additional analyses intended to explore the roles of the various subscales of the Social Provisions Scale, the social support measure. The six subscales reflect the functional types of perceived social support and 95 include: guidance, reassurance of worth, social integration, attachment, nurturance, and reliable alliance. (Further details on the types of social support can be found in Chapter 2.) I ran the hierarchical regression analyses for RQ2 six additional times, each time using a different type of support (SPS subscale) as the moderating variable. Each support type variable was centered prior to analysis. Within the subsequent regression models, five of the subscales were found to be significant predictors of well-being when controlling for the independent variable (parenting stress) and previously identified covariates (age, foster care agency): Guidance (B = 2.694, SE = .444, β = .459, p<.001); Reassurance of Worth (B = 2.510, SE = .511, β = .409, p<.001); Social Integration (B = 2.508, SE = .400, β = .425, p<.001); Attachment (B = 2.256, SE = .380, β = .425, p<.001); and Reliable Alliance(B = 2.034, SE = .481, β = .348, p<.001). One social support type, Nurturance, was not a significant predictor of well-being within its respective model. I then tested individual social support types as moderators within the regression analyses by creating interaction terms with each social support type and parenting stress and including them in the regression models for each. Within the respective regression models, four of the interaction terms were found to be significant: Guidance (β = .179, p=.012, ΔR² = .028), Reassurance of Worth (β = .158, p=.027, ΔR² = .024), Social Integration (β = .142, p=.044, ΔR² = .018), and Attachment (β = .178, p=.013, ΔR² = .027). These results provide support for the moderating role of these specific social support subtypes between parenting stress and well- being. Reliable Alliance did not have a significant moderating effect, suggesting that it has more of a direct effect on foster parent well-being. Nurturance did not have a significant moderating effect and, as noted above, did not have a direct effect on well-being. 96 Research Question 3 The final set of research questions examined the relationship between parenting stress and retention among foster parents, and whether social support moderates this relationship. RQ3. A. Is there a relationship between parenting stress and retention (intent to continue) among foster parents? B. Does social support moderate this relationship? C. Is there a relationship between parenting stress and thoughts about leaving fostering? D. Does social support moderate this relationship? As noted in Chapter 3, retention was operationalized as the intent to continue. Specifically, participants were asked about the likelihood that they would still be fostering in 18 months. Response options included very unlikely, somewhat unlikely, somewhat likely, and very likely. For analysis purposes, responses were recoded into two categories: unlikely and likely. To test this set of research questions, I first completed bivariate analyses between the dependent variable (retention) and the independent variables (parenting stress, social support) and potential covariates (including sociodemographic and foster home characteristics). I utilized independent samples t-tests to assess whether there were differences in scores on continuous variables among foster parents who were unlikely to continue fostering and those who were likely to continue. As seen in Table 19, there were differences between the means, but none of the differences were significant at the p<.05 level. Table 19 Independent samples t-test for retention (intent to continue) Age (in years) Parenting stress Social support Well-being *p<.05, **p<.01, ***p<.001 41.05 (9.77) 47.88 (13.87) 80.26 (10.26) 48.23 (14.79) 97 Unlikely Likely Mean (SD) 43.49 (12.47) 44.67 (11.57) 81.24 (10.09) 51.76 (10.90) t -1.22 1.42 -.53 -1.40 I used the Chi-square test for independence to explore the relationships between categorical sociodemographic and foster home characteristic variables and retention, the dichotomous dependent variable. As illustrated in Table 20, one variable which demonstrated a significant association with retention was religious participation (ꭓ² [1, n = 138] = 5.30, p = .021). The first result indicates that participants who reported regular religious participation were more likely than expected to report that they intended to still be fostering 18 months from now. Similarly, those who did not report regular religious participation were less likely than expected to indicate that they would still be fostering in 18 months. The phi coefficient for this result is .213, indicating a small effect size (Cohen, 1988). Additionally, the variable foster care agency demonstrated a significant association with retention (ꭓ² [4, n = 138] = 9.90, p = .042). Specifically, a higher percentage of foster parents from Agency A (92.9%) reported an intent to continue fostering, particularly when compared to Agency B (52.6%) and Agency D (52.2%). Cramer’s V for this result was .268, indicating a small effect size. Table 20 Chi square test of independence for retention (intent to continue) Gender Frequency (%) Male Female Race White Non-white Parent relationship Single parent Two-parent Kinship Religious No Yes No Yes 10 (76.9) 85 (68.0) 86 (68.3) 9 (90.0) 17 (70.8) 76 (67.9) 80 (67.2) 13 (76.5) 22 (53.7) 73 (75.3) ꭓ2 .120 1.176 .002 .238 5.301* 3 (23.1) 40 (32.0) 40 (31.7) 1 (10.0) 7 29.2) 36 (32.1) 39 (32.8) 4 (23.5) 19 (46.3) 24 (24.7) 98 Table 20 (cont’d) Time as a foster parent Total number of children fostered Current number of foster children <1 yr 1-2 yrs 2-5 yrs >5 yrs 1-2 3-10 >10 0 1 2 3+ Foster care agency Agency A Agency B Agency C Agency D Agency E Income Low (<$60,000) Medium ($60,000-99,999) High ($100,000+) Not employed Employed, part-time Employed, full-time Employment Education High school or less Some college Bachelor’s degree or higher *p<.05, **p<.01, ***p<.001 12 (36.4) 11 (31.4) 10 (27.8) 10 (29.4) 19 (38.8) 16 (26.2) 8 (28.6) 13 (31.7) 14 (31.8) 10 (25.6) 6 (42.9) 1 (7.1) 9 (47.4) 13 (25.5) 9 (29.0) 11 (47.8) 2 (15.4) 18 (31.0) 9 (23.7) 9 (23.7) 12 (30.8) 22 (36.7) 2 (15.4) 14 (32.6) 26 (32.1) 21 (63.6) 24 (68.6) 26 (70.6) 95 (68.8) 30 (61.2) 45 (73.8) 20 (71.4) 28 (68.3) 30 (68.2) 29 (74.4) 8 (57.1) 13 (92.9) 10 (52.6) 38 (74.5) 22 (71.0) 12 (52.2) 11 (84.6) 40 (69.0) 29 (76.3) 29 (76.3) 27 (69.2) 38 (63.3) 11 (84.6) 29 (67.4) 55 (67.9) .658 2.104 1.461 9.900* 1.587 1.83 1.579 Research Question 3, A & B Binary logistic regression was used to analyze the relationship between parenting stress and retention, while controlling for covariates. Before the regression model was run, the independent variable (parenting stress) and moderating variable (social support) were both centered. I also ran a multiple linear regression model with all the variables to assess for multicollinearity. Tolerance and VIF were both within acceptable ranges (Pallant, 2013). Within the logistic regression model, retention was entered as the dependent variable, while parenting stress (IV), social support (moderator), and identified covariates (religious participation, and 99 foster care agency) were all entered as predictor variables. Dummy coding was created for the foster care agency variable, and “Agency B” was left out of the regression analysis as the reference category. The overall model was statistically significant (ꭓ²[7] = 22.208, p<.002), categorizing 74.6% of cases correctly (93.7% of those likely to stay in their roles and 32.6% of those unlikely to stay). Religious participation and foster care agency (A & C) were significant predictors of retention. Odds ratios indicate that participants with regular religious participation were 3.7 times as likely to report intent to continue fostering, compared to those without religious participation (95% CI = 1.54-8.96). Additionally, foster parents from Agency A were 18.4 times as likely to report an intent to continue compared to foster parents from Agency B (95% CI = 1.75-192.75), while Agency C foster parents were 3.7 times as likely (95% CI = 1.13- 12.30). A second regression model to test for a moderating effect was not necessary, since neither of the predictors in the interaction term were found to be significant. Table 21 B Binary logistic regression results for retention (intent to continue) Constant Religious Agency Aa Agency C Agency D Agency E Social support Parenting stress *p<.05, **p<.01, ***p<.001 aAgency B was used as the reference group among foster care agencies. -1.062 1.311** 2.911* 1.315* 1.276 .267 -.002 -.033 SE .609 .449 1.199 .610 .662 .668 .022 .019 Exp(B) .346 3.712 18.378 3.723 3.581 1.306 .998 .967 95% CI 1.54-8.96 1.75-192.75 1.13-12.30 .98-13.11 .35-4.84 .96-1.04 .93-1.00 These results provide evidence in response to the identified research questions. Results indicate that there is not a significant relationship between parenting stress and retention, as operationalized as foster parents’ intent to remain fostering 18 months from now. The only 100 significant predictors of retention were religious participation and the foster care agency through which foster parents are licensed. Research Question 3, C & D Following the above analyses for the research questions 3A and 3B, I repeated the analyses using retention operationalized as ‘Thoughts about giving up fostering’ (retention as past ideation) for the dependent variable. For this variable, participants were asked “In the past two months or so, how often have you thought about giving up fostering?” Response options included Never, Hardly ever, Sometimes, Often, and Very often. For the purposes of the regression analysis, the variable was transformed into a dichotomous variable by coding ‘Never’ and ‘Hardly ever’ responses as 0, and ‘Often’ and ‘Very often’ responses as 1. Participants who responded ‘Sometimes’ were excluded from the dichotomous variable and subsequent analysis. Bivariate analyses were used to identify potential covariates for inclusion in the binary logistic regression model. Among sociodemographic and foster home variables, religious participation was the only one to demonstrate a significant relationship to the dependent variable, ‘Thoughts about giving up fostering.’ The Chi square test for independence indicated that participants who reported regular religious participation were significantly less likely to consider leaving (ꭓ² [1, n = 92] = 4.57, p = .032). Binary logistic regression analysis was then used to analyze the relationship between parenting stress and thoughts about giving up fostering while controlling for the identified covariate, and the potential moderating role of social support. ‘Thoughts about giving up fostering’ was entered as the dependent variable, while parenting stress (IV), social support (moderator), and religious participation (covariate) were all entered as predictor variables. The overall model was statistically significant (ꭓ²[3] = 26.192, p<.001), categorizing 75.0% of cases 101 correctly (87.7% of those who never or hardly ever consider giving up fostering and 54.3% of those who often or very often consider it). Parenting stress and religious participation were significant predictors of thoughts about giving up fostering. Social support was not a statistically significant predictor; therefore, a second regression model to test the moderating effect (Research Question 3D) was not necessary. These results suggest that parenting stress and religious participation are both related to the frequency that foster parents consider leaving their role. Conclusion The results detailed above are in response to the three sets of research questions developed for the study. In regards to the first set of research questions, I found no relationship between the total number of stressors and levels of parenting stress. However, parenting stress was significantly higher among foster parents who experienced certain stressors, namely: child behavior problems, difficulty obtaining services, and disagreement with a licensing rule or policy. Findings related to the second set of research questions indicated that there was a significant, negative relationship between parenting stress and well-being, and that social support moderated this relationship. Additionally, four types of support (attachment, reassurance of worth, guidance, social integration) were moderators, or buffers, between parenting stress and well-being. Reliable alliance had a direct and positive impact on well-being, while opportunities for nurturance had no relationship to well-being. Finally, in regards to the third set of research questions, there was no relationship between parenting stress and retention as the intent to continue. However, parenting stress was a significant predictor of thoughts about giving up fostering. In the following chapter, I will provide a detailed discussion of these results, as well as descriptive results presented in Chapter 4. 102 CHAPTER 6: DISCUSSION As the most frequently used placement option for children in foster care, licensed foster homes are a critical component of the U.S. child welfare system (U.S. Children’s Bureau, 2019). Foster parents report stress and lack of support, leading many to leave their challenging positions within the first 1-2 years (Gibbs & Wildfire, 2007), yet relatively little is known about factors which contribute to foster parents leaving their role and even less is known about what impacts their well-being within it. The current study contributes to foster care literature by increasing understanding of the impact of stress on foster parents’ well-being and retention in their role. Specifically, this study contributes to a small (but growing) body of quantitative research conducted on a U.S. foster parent population by utilizing a cross-sectional, non-experimental design and web-based survey methodology among a foster parent population in Kent County, Michigan. Further, this study utilized a conceptual model (Figure 8) grounded in social support theory and a multidimensional conception of social support. Figure 8 Proposed buffering model of foster parenting Based on the proposed model, the following research questions were used to guide the study: RQ1. A. Is there a relationship between stressors experienced by foster parents and their levels of parenting stress? B. Does social support moderate this relationship? 103 C. Is there a relationship between individual stressors and parenting stress? RQ2. A. Is there a relationship between parenting stress and well-being among foster parents? B. Does social support moderate this relationship? C. Do the various types of social support moderate this relationship? RQ3. A. Is there a relationship between parenting stress and retention (intent to continue) among foster parents? B. Does social support moderate this relationship? C. Is there a relationship between parenting stress and thoughts about leaving fostering? D. Does social support moderate this relationship? In this chapter, I will discuss the implications of study findings presented in Chapters 4 and 5. First, I will discuss results related to each of the key study variables: stressors of fostering, parenting stress, retention, well-being, and social support. Next, I will discuss each set of research questions. I will conclude by suggesting potential limitations of the study. Key Findings Stressors of Fostering The stressors faced by foster parents are multiple and varied. This study confirms current understanding of what those stressors are, and expands knowledge around how commonly they are experienced and their connection to outcomes like parenting stress, well-being, and foster home retention. Existing literature documents stressors that are common to fostering and which have been explored primarily through qualitative methods (Barnett, Jankowski, Butcher, Meister, Parton, & Drake, 2017; Buehler et al., 2003; Cavazzi et al., 2010; Lanigan & Burleson, 2017; Pickin et al., 2011; Shklarski, 2019). This study strengthens this knowledge by providing empirical support for the types of stressors experienced by foster parents. Each of the 13 104 stressors included in the current study, which were drawn from existing literature, was experienced by at least 10% of participants in the past two months. These stressors are commonplace among foster parents: the mean number of stressors faced in the previous two months was 3.21 and the mode number of stressors was 3. The most commonly reported stressors were foster child behavior problems, foster child mental health needs, and poor communication from agency staff. This study expands knowledge of fostering stressors by exploring how stressful foster parents perceive each stressor to be. For each stressor endorsed, participants were asked to indicate how stressful it was for them (1=Not at all stressful, 2=Mildly stressful, 3=Moderately stressful, 4=Extremely stressful). The stressors rated as most stressful included allegations of abuse and/or neglect (M=3.29), disagreement with a decision in the case (M=3.23), and poor communication from foster agency staff (M=3.19). With the exception of poor communication, the stressors rated as most stressful were not those that were most common. The three stressors rated as most stressful are all system-related, rather than directly relating to care of the foster child (e.g. child’s mental health needs). This finding is consistent with qualitative literature describing the system as being a greater source of stress for foster parents than the parenting of foster children (Geiger et al., 2013). However, not all system-related stressors were rated high in terms of stress. Case manager turnover, which has received considerable attention in child welfare workforce literature (Hopkins, Cohen-Callow, Kim, & Hwang, 2010; Middleton & Potter, 2015; Wilke, Rakes, & Randolph, 2019), was rated just above mildly stressful by foster parents. This suggests that a change in case manager is not as stressful to foster parents as many other stressors they experience, perhaps because it is experienced so frequently. Finally, worth 105 noting is the finding that none of the 13 individual stressors had a mean rating below mildly stressful, emphasizing the stressful nature of the unique role foster parents fulfill. Further, this study adds to fostering literature by distinguishing between stressful events, or stressors, and the experience of stress. Stressors are those circumstances or events which have the potential to elicit stress, whereas stress is the experience of the negative effects of appraising such a circumstance or event as stressful (Cohen & McKay, 1984; Lazarus & Folkman, 1984). Previous fostering research has not typically made this distinction explicit. Parenting Stress Parenting stress is conceptually defined as the “aversive psychological demands of being a parent” (Deater-Deckard, 1998, p. 315). Among foster parents, greater levels of parenting stress is related to lower quality of parenting (Vanderfaeillie et al., 2011), lower parental sensitivity (Gabler et al., 2018), and poor co-parenting quality between partners (Richardson & Futris, 2019; Richardson et al., 2018). Foster parents who are sensitive to the needs of their foster children and display other characteristics of high-quality parenting are critical to the success of foster placements and subsequently, the well-being and permanence of the children (Oosterman et al., 2017). Therefore, an understanding of parenting stress and related factors among foster parents is imperative. Parenting stress was operationalized in this study as total score on the Parental Stress Scale (PSS), with higher scores indicating greater levels of parenting stress. The mean total score for participants in the sample was 45.48, on a scale of 18 to 90. The PSS does not include cutoff scores for normative, risk, or clinical levels of parenting stress. However, studies which use the PSS among foster parents and similar populations offer a context into which the current study’s findings may be placed. 106 Richardson and colleagues (2018) conducted the only known study to use the PSS with a foster parent population. This study utilized an adapted 10-item version of the PSS and adjusted the item responses from a 5-point Likert-style scale to a 7-point scale. Therefore, it is challenging to make direct comparisons between the Richardson et al. study and the current study. Richardson et al. (2018) reported a mean score (rather than a total score) of 3.27 on the 1 to 7 scale. This score is below the midpoint of 4. Following this scoring method, the mean score of participants in the current study was 2.53 on a 1 to 5 scale, which is also below the midpoint of 3. Studies which used the original PSS among other parenting populations provide a more meaningful context for the current study’s results. Coughlin, Sharry, Geurin, & Drumm (2009) studied parenting stress among parents of children ages 6 to 11 with significant behavioral and emotional difficulties, finding an overall mean score of 47.76 on the PSS. This result is comparable to the overall mean score of 45.48 within the current study. Among families in transition, foster parents in this study had higher levels of parenting stress than female spouses of soldiers deployed to Iraq (M=35.80; Everson, Darling, & Herzog, 2013), and slightly lower levels of parenting stress than stepmothers (M=50.89; Shapiro & Stewart, 2011). Bivariate results of the study indicated that certain demographic characteristics, including age, education, and parent relationship, were related to parenting stress. Older foster parents in the study had lower levels of parenting stress, suggesting that age may be somewhat of a protective factor against the stress of foster parenting. Conversely, younger foster parents may be at a greater risk of experiencing parenting stress. In a practice context, this suggests the need for targeted support for younger foster parents, as well as the opportunity to tap into the lived experiences of older foster parents. Foster parents with more education (Bachelor’s degree or 107 higher) also experienced higher levels of parenting stress than those with less education. Single foster parents, surprisingly, had lower levels of parenting stress than foster parents within two- parent households. This is inconsistent with research among the general population that found single mothers to have significantly greater levels of parenting stress than married mothers (Copeland & Harbaugh, 2005). This discrepancy may be due to differences in measurement tools (Parental Stress Scale versus Parenting Stress Index) or the population being studied (foster parents versus first time mothers postpartum). Each of these bivariate findings warrant further exploration in order to more fully understand factors contributing to parenting stress among foster parents. In this study, foster parents who reported certain stressors of fostering had significantly higher levels of parenting stress than those who did not. Specifically, foster parents who endorsed experiencing foster child behavior problems in the past two months had higher levels of parenting stress than those who did not. This finding is consistent with fostering literature that demonstrates a significant relationship between foster child behavior problems and parenting stress among foster parents (Gabler et al., 2014; Goemans et al., 2018; Lohaus et al., 2017, Morgan and Baron, 2011; Vanderfaiellie et al., 2012; Vanschoonlandt et al., 2013). Fostering literature also points to a connection between a lack of agency support and parenting stress (McKeough et al., 2017). Findings from this study provide support for this connection, in that foster parents who reported experiencing difficulties accessing services and disagreement with foster agency licensing rules or policies had higher levels of parenting stress than those who did not report these challenges. Interestingly, foster parents who reported disagreeing with a decision in the foster care case had significantly lower parenting stress than those who did not. This may be due to foster parents experiencing little stress in their parenting role and still being unhappy 108 with an agency or court decision, such as the decision to reunify with birth parents; however, this finding calls for further exploration. Well-being This study adds to a body of literature on foster parent well-being that is “extremely limited” (Hannah & Woolgar, 2018). Literature on foster child outcomes points to the importance of the health and well-being of their caregivers (Fisher & Stoolmiller, 2008); however, only a small number of studies actually examine well-being among foster parents (Cole & Eamon, 2007; Farmer et al., 2005; Morgan & Baron, 2011; Whenan et al., 2003; Wilson et al., 2000) and even fewer investigate the factors that may influence it (Cole & Eamon, 2007; Wilson et al., 2000). While there is practical usefulness in examining foster parent well-being as a contributing factor towards identified child welfare system outcomes, there is also value in studying it for the sake of better understanding and supporting foster parents themselves. “To improve the quantity and consistency of foster parents able and willing to serve, research should consider the well-being of foster parents for the sake of their well-being and not only for how they can serve foster youth” (Miller et al., 2019, p. 112). This study is a noteworthy addition to the scarce literature on foster parent well-being. In terms of findings, the current study demonstrates that in terms of well-being, most foster parents are quite healthy. Three quarters (74.6%) of study participants scored in the ‘flourishing’ category, meaning that they frequently (i.e. every day or almost every day) experienced symptoms of positive mental health, while 22.4% were categorized as having ‘moderate’ mental health. Very few (3%) qualified as ‘languishing,’ indicating that they rarely (i.e. never or once or twice in the past month) experienced symptoms of positive mental health. This finding is consistent with previous studies on foster parent well-being, which suggest that most foster parents report well-being levels in the normal to high range (Farmer et al., 2005; 109 Morgan & Baron, 2011; Whenan et al., 2003). While no previous studies of foster parents utilized the MHC-SF to measure well-being, studies conducted among other groups provide a context for the current study findings. Kasin, Muñoz, Ong, Whicker, & Twohig (2020) conducted research with parents of children who are deaf or hard of hearing, finding that 66% of participants were flourishing, 33% had moderate mental health, and 1% were languishing. Proeschold-Bell and colleagues (2019) found even higher levels of well-being among caregivers in residential facilities for orphaned and vulnerable children in African and Asian nations (77% flourishing, 23% moderate, 0% languishing). What is largely absent from fostering literature, and a notable contribution of this study, is the examination of factors (parenting stress and social support) which may contribute to the foster parents’ levels of well-being. As stated by Miller and colleagues (2019), “Although, foster parent well-being is critical to the success of foster youth and the foster system, the literature has yet to identify how foster parents can maintain their health and well-being in the face of these challenges and stressors” (p. 109). Findings in this area are discussed in the “Buffering Model of Social Support” section below, in response to Research Question 2. Retention Within this study, retention was operationalized in terms of future intention. Specifically, participants were asked “How likely is it that you will still be fostering 18 months from now?” and responded on a 4-point Likert-style scale ranging from Very unlikely to Very likely. Approximately two-thirds of participants indicated that it was either somewhat or very likely that they would still be fostering in 18 months. This finding is similar to other studies that have framed retention as intent to continue (Denby et al., 1999; Hannah & Woolgar, 2018). Denby and colleagues (1999) found that 67.3% of foster parents in their study intended to still be fostering one year in the future. In contrast, Orme et al. (2006) found that well over 90% of 110 foster mothers intended to still be fostering one year in the future, much more than in the current study. Other studies found slightly lower proportions of foster parents who intended to give up fostering when compared to this study. Eaton and Caltabiano (2009) found that 19.9% intended to leave, while Geiger et al. (2013) found that 24.2% were likely to discontinue fostering, compared to 31.5% in the current study. In addition to retention as the intent to continue, this study asked foster parents about their past ideation about quitting fostering. Participants were asked how frequently in the last 2 months they had thought about giving up fostering and responded on a 5-point Likert-style scale ranging from Never to Very Often. A quarter of foster parents (25.2%) considered leaving often or very often, while 33.3% reported thinking about it sometimes, and 41.3% thought about it hardly ever or never. This finding is consistent with Eaton and Caltabiano (2009), who found that 27.4% of foster parents thought about leaving often or very often in the previous 2 months. Social Support Existing literature on foster parenting is lacking in studies that recognize the multidimensionality of social support and connect this knowledge to methodological choices. This study expands fostering literature by delineating the social support construct being studied (perceived support), the types of support (functional typology), and why these make sense given the aims of the study. Based in knowledge of social support theory, I chose to study perceived support, as it is the social support construct most closely tied to mental health outcomes (Lakey & Cohen, 2007) and has consistently demonstrated to buffer, or moderate, the effects of stress on individual health or well-being (Gottleib & Bergen, 2010). I also chose to examine social support from a functional perspective, aiming to understand the various purposes of social support, because fostering literature suggests that support needs of foster parents are not uniform but 111 varied. The use of the Social Provisions Scale (SPS; Cutrona & Russell, 1987), which measures functional types of perceived support, reflects these decisions. This study is the first to use the SPS among foster parents and further, adds to fostering literature by making methodological choices that are theory-driven, and explicitly explaining these choices. Findings from the current study indicate that foster parents perceive a moderate amount of social support. Though the SPS does not include scoring benchmarks, these findings can be placed within the context of research using the SPS, including studies among similar populations. The mean total score of participants in the current sample was 81.04, on a scale of 24 to 96. This is somewhat higher than total scores found in other study populations, such as patients with multiple sclerosis (M=77.40; Chiu, Motl, & Ditchman, 2016) and parents of school-aged children with autism (M=75.06; Robinson & Weiss, 2020), indicating that foster parents in the study perceive greater levels of social support. A study among parents of children enrolled in Early Head Start reported an average score rather than a total score, with a mean of 3.35 (Green, Furrer, & McAlister, 2011), which is nearly identical to the current study (M=3.38). The methodological inconsistencies of studies examining social support with a foster parent population make it difficult to synthesize the resultant body of literature and to place the current study within it. However, there are several findings worth noting when seeking to contextualize the results of the current study. The current study found that social support was significantly, negatively correlated to parenting stress, consistent with Richardson et al. (2018) and Soliday et al. (1994). Conversely, Lohaus et al. (2017) found no significant relationship between parenting stress and social support. This disparity in findings could be due to differences in operationalizing and measurement of social support among the studies. 112 In terms of social support’s relationship to retention, the current study found no significant correlation between perceived social support and retention (operationalized as intent to continue). This is consistent with Cooley et al. (2015), who also measured perceived social support and retention as intent to continue and found no significant relationship between the two variables. The current study did, however, find a correlation between social support and thoughts about leaving. Foster parents who had higher levels of social support reported less frequent thoughts about leaving their fostering role. This supports Randle et al. (2017), who found that social support was significantly lower among foster parents who had the most frequent thoughts about quitting. The disparate findings between retention plans (intent to continue) and retention thoughts suggests that there may be an intervening variable that leads to some foster parents who think about leaving fostering to not actually plan to do so. This phenomenon has not been explored within existing fostering literature. The current study also found that social support was significantly, positively correlated with foster parent well-being. This is the first study to examine this relationship among foster parents, using standardized measures of both variables. The only other research to examine the relationship between social support and well-being among a foster parent population was conducted by Cole and Eamon (2007), who operationalized social support as perceived helpfulness of a support group and found that it was related to lower levels of depression symptoms. The current study advances fostering research by using valid, psychometrically-tested measures to demonstrate that foster parents who perceive greater social support report higher levels of well-being. Types of Social Support. As noted above, another important contribution of this study to the broader body of fostering literature is the conceptualization of social support as 113 multidimensional. There are a variety of types of social support that are not reflected in most studies on the topic among foster parents. This study utilized a multidimensional measure of social support, the SPS, to not only assess levels of overall support, but also levels of various types of support, such as social integration (“a sense of belonging to a group that shares similar interests, concerns, and recreational activities;” Cutrona & Russell, 1987, p. 39) and reliable alliance (“the assurance that others can be counted upon for tangible assistance;” Cutrona & Russell, 1987, p. 39). Foster parents in the current study had mean scores on individual types of social support that were slightly below college student samples in other studies (Ptacek et al., 1999; Vogel & Wei, 2005). The highest support scores were for guidance, which includes advice or information, while the lowest were for social integration, or “a sense of belonging to a group that shares similar interests, concerns, and recreational activities” (Cutrona & Russell, 1987, p. 40). This suggests that while foster parents feel they have the availability of advice and helpful information when it is needed, they are less assured of being a part of a supportive group of peers. Buffering Hypothesis As noted above, one of the main contributions of this study is that it utilizes a conceptual framework that is grounded in social support theory, specifically the buffering hypothesis. This framework was utilized to guide all aspects of the study, including research question development, selection of constructs, and operationalization of key variables. It is therefore necessary, within the present discussion, to use the framework “as a mirror to check whether the findings agree with the framework or whether there are some discrepancies” (Imenda, 2014, p. 188). Figure 8, above, offers a visual diagram of the buffering hypothesis as it relates to foster parents. The buffering hypothesis posits that in the presence of stress, social support serves as a protective factor (or buffer) against the negative effects of stress at potentially two points in the 114 causal linkage between stress and outcomes: the link between possible stressors and the experience of stress, and the link between stress and pathologic outcomes. The former is addressed by Research Question 1, the latter by Research Questions 2 and 3. Research Question 1. The first set of research questions in the current study asked: Is there a relationship between stressors experienced by foster parents and their levels of parenting stress, and does social support moderate this relationship? By posing this question, the study adds to the existing body of foster care literature, within which studies do not typically delineate between potentially stressful events and the experience of stress. A stressor is an event or circumstance with the potential to elicit stress, whereas stress is the experience of the negative effects of appraising an event or circumstance as stressful (Cohen & McKay, 1984; Lazarus & Folkman, 1984). The buffering hypothesis of social support suggests that social support may protect individuals by causing them to experience less stress as a result of potential stressors. Within the multivariate analysis of the current study, there was no significant relationship between the stressors of fostering and parenting stress. The only significant predictors of parenting stress were parent relationship type and social support. This suggests that the amount of stressors experienced is not related to the experience of parenting stress. However, the lack of significant results in this study may be due to the operationalizing of stressors as the total number of potential stressors experienced in the past two months. As noted above, four individual stressors (child behavior problems, difficulty obtaining services, disagreement with a licensing rule or policy, and disagreement with a decision in the case) had significant relationships with parenting stress. This suggests that it is not the amount of stressors that contributes to parenting stress, but the nature of the particular stressors being experienced. Future studies should consider 115 how to most effectively operationalize stressors of fostering, and should investigate further the relationship between individual stressors of fostering and parenting stress. Research Question 2. The second set of research questions asked: Is there a relationship between parenting stress experienced by foster parents and their levels of well-being, and does social support moderate this relationship? This study contributes to fostering literature, first by asking these questions and also by examining them through the use of standardized measures of parenting stress and well-being. The only known study to examine social support as a buffering factor in the relationship between parenting stress and well-being of foster parents was conducted by Cole and Eamon (2007), who did not use a standardized measure to assess parenting stress and used a standardized measure to assess mental health symptoms as an outcome. The current study found a significant, negative relationship between parenting stress and well-being among foster parents. Foster parents who experienced more parenting stress had lower levels of well-being. Additionally, social support functions as a buffering, or protective, factor in this relationship. Practically speaking, this means that foster parents who have high levels of social support are less negatively impacted by their parenting stress. They are able to remain well even in the context of stress in parenting. This is consistent with studies among other caregiving populations, such as grandparents in a primary caregiving role (Hayslip et al., 2015), caregivers of TBI patients (Ergh et al., 2002), and caregiving partners of dementia patients (Gellert et al., 2018). This finding is also consistent with studies conducted among foster parents which found social support to serve as a buffer between other variables, such as parenting stress and co-parenting relationship quality (Richardson et al., 2018) and problem child behaviors and perceived challenging aspects of fostering (Cooley et al., 2019). 116 The current study further added to fostering literature by examining the buffering role of overall social support and then exploring the role of each functional type of social support. Upon further analyses, four types of social support were each found to serve as moderating variables, or buffers, between parenting stress and well-being. These included social integration, attachment, guidance, and reassurance of worth. One type, reliable alliance, had a direct effect on well-being, but did not serve as a buffer between parenting stress and well-being. Reliable alliance refers to tangible, material supports available to foster families, such as meals, transportation assistance, or gifted zoo or museum memberships. This result suggests that this type of support is beneficial to foster parents regardless of their stress levels. One type of support, opportunity for nurturance, did not have a buffering effect nor a direct effect on foster parent well-being. Opportunity for nurturance, according to Weiss’s (1974) theoretical model, is the sense that others rely upon oneself for their well-being. As noted by Cutrona and Russell (1987), this is not, strictly speaking, a type of social support, since the individual is providing rather than receiving assistance. However, the idea that support can be mutual or reciprocal is common in social support literature (Cutrona & Russell, 1987; Hupcey, 1998). The fostering population is essentially defined by its role in providing support and nurturance to others, which is both rewarding and a source of significant stress. This finding suggests that for foster parents, the opportunity to care for others does not necessarily mitigate the impact of the stress caused by this caregiving. In response to the second set of research questions, this study found that overall social support, as well as several types of support, functions as a buffer in the relationship between parenting stress and well-being of foster parents. These findings begin to fill the gap identified by Miller et al. (2019), who stated, “…the literature has yet to identify how foster parents can 117 maintain their health and well-being in the face of these challenges and stressors” (p. 109). In a role that is inherently stressful, social support is a factor that has demonstrated promise in protecting foster parents from the effects of stress and preserving their well-being. Research Question 3. The final set of research questions asked: Is there a relationship between parenting stress experienced by foster parents and retention, and does social support moderate this relationship? Again, the current study expands fostering literature by being the first known study to explore these questions quantitatively. Wilson et al. (2000) examined the relationship between stressful events and intent to continue and found a significant relationship between the two that was particularly strong among newer foster parents. However, as demonstrated within this study, stressful events are conceptually distinct from the experience of parenting stress. This study operationalized retention as the intent to remain licensed 18 months from now. Bivariate analysis revealed that there was no significant relationship between parenting stress and retention, suggesting that the decision to remain in the fostering role is independent of one’s level of parenting stress. Two variables did have significant relationships with retention, however: foster care agency and religious participation. Foster parents from two agencies had significantly lower numbers of foster parents who intended to continue fostering than the other three. This finding indicates that the experience of foster parents is not uniform across child- placing agencies; there may be practice differences between the agencies that lead to disparities in retention rates. These practice differences are an area of exploration in future research. Additionally, participants who reported regular (at least monthly) participation in religious services were more likely to report intentions to remain in fostering. This finding suggests that some aspect of religious participation (e.g. feeling ‘called’ to fostering) may impact foster 118 parents’ decisions on whether to remain in their role, regardless of their stress levels. Few studies have explored religiosity in foster parents; however, DeMaeyer, Vanderfaeillie, Vanschoonlandt, Robberechts , and Van Holen (2014) reported that nearly three quarters of foster parents in their study identified fulfilling religious beliefs as one of their reasons for becoming foster parents. Religiosity and religious participation are areas for further examination within the fostering population. In addition to retention as the intent to continue, I examined these research questions using retention as thoughts about giving up fostering as the dependent variable. Analysis demonstrated a significant relationship between parenting stress and thoughts about giving up. Specifically, those who thought about giving up often or very often significantly higher levels of parenting stress than those who reported having these thoughts never or hardly ever. Social support did not moderate this relationship, but rather, had a direct effect on thoughts about giving up. Religious participation also was higher among foster parents who never or hardly ever thought about leaving fostering. Together these results suggest that thinking about giving up fostering and actually planning to do so are distinct experiences within the foster parent population. There may be factors that contribute to foster parents remaining in their role despite their own thoughts about leaving. Based on its relationship to both retention-related variables in the current study, religious participation may be one such factor. Fostering literature points to others that may warrant further research, including commitment to specific children, which Eaton & Caltabiano (2009) found to contribute to intent to continue. Whenan et al. (2009) found that foster parents with greater parenting self-efficacy were more likely to report intentions to continue fostering. The influence of these variables, as well as the relationship between thoughts about leaving and 119 plans to leave fostering, are areas for future exploration, if we are to understand what helps keep foster parents in their roles. Limitations of the Study It is important to acknowledge the limitations of the study and subsequent findings. First, the study is limited in its methodology. As a cross-sectional, non-experimental design study, its data are limited to one point in time snapshot. Results cannot be used to determine causality between the variables. Further, changes in the variables over time could not be observed. A second limitation of the study is the size and representativeness of the sample. Though power analysis demonstrated that the study sample size of 139 was sufficient to detect a medium effect size at p<.05. However, a larger sample may have resulted in additional significant findings that could not be detected by this sample. Further, the sample was obtained by convenience sampling by emailing licensed foster parents at five private child-placing agencies. Convenience sampling is a common method in fostering research (Cooley et al., 2019; De Maeyer et al., 2015; Geiger et al., 2017; Julien-Chinn et al., 2017), but it can compromise the representativeness of the sample. Men, foster parents of color, and kinship caregivers were underrepresented in the study sample. It is also possible that foster parents with low well-being were underrepresented, considering that persons who are not mentally or emotionally healthy may be less likely to respond to a survey request. Finally, there are measurement limitations within the study. The survey tool was entirely self-reported. While self-report measures are commonly used in fostering research (Adams et al., 2018), they also introduce potential biases to the results, such as social desirability bias. Foster parents may be reluctant to report that they’re experiencing stress in their role or that they are not doing well in terms of their mental health, out of fear that this may jeopardize their relationship with their child-placing agency or their ability to take future placements of foster children. This 120 study aimed to reduce this bias by not collecting identifying information and routing survey responses directly to the researcher, rather than the child-placing agencies. Other measurement limitations are related to specific variables being measured. In regards to stressors of fostering, the study only asked about potential stressors particular to the fostering role. It is plausible that foster parents experience stressors unrelated to their role, such as work-related challenges or health concerns, which also impact their well-being and longevity in their role. The variable of retention is also a potential measurement limitation. Retention was operationalized as the likelihood that one would still be fostering 18 months from now. Retention in this study, therefore, is anticipated and not actual. There is no research demonstrating that anticipated retention is a valid measure of actual retention. Lastly, this study gathered minimal information about the foster children placed in the homes of the foster parent participants. Factors such as foster child age and length of time in the home may impact foster parents’ experiences, but for reasons of parsimony were not included in the study. Conclusion Within a role teeming with potential stressors, social support has shown some promise in helping foster parents to manage and cope. However, social support has not been studied among foster parents in a way that is methodologically consistent, nor in a way that reflects knowledge of and adherence to social support theory. The current study expands fostering literature by utilizing a conceptual model that is theoretically grounded and reflected in the methodology of the study. It sheds light on foster parents’ experiences of stressors and parenting stress. It offers evidence for social support as a buffer against stress for foster parents, helping to preserve their well-being. It highlights a distinction between foster parents’ thoughts about leaving their role and their plans to actually do so. It also raises several questions and areas for potential 121 exploration through further research. These directions for future research, as well as implications for policy and practice, are detailed in the following chapter. 122 CHAPTER 7: IMPLICATIONS AND CONCLUSION This study has several implications for foster families and the individuals, organizations, and systems in their social environments. The findings increase understanding of those factors, namely stress and support, which impact foster parents’ willingness to remain in their role and capacity to remain well in their role. Such knowledge has the potential to meaningfully impact social work practice, child welfare policy, social work education, and future research. The following chapter proposes some key implications in these areas. Implications for Practice The findings have implications for practice with foster families, in terms of training, assessment, and intervention. Improved procedures in these areas have potential to decrease foster parent stress, increase well-being, and improve retention of licensed foster homes. Foster parent training policies vary widely from state to state, both in terms of quantity and required topics (U.S. Children’s Bureau, n.d.-b). Most states, however, require foster parents to receive both pre-service (prior to licensure) and in-service (as a licensed foster parent) training. Through pre-service training, foster parents receive information that is intended to orient them to the foster care system and their role within it. Based on these findings, as well as previous fostering research, this initial training should also include information on common stressors of foster parenting and the types and sources of support to alleviate the impact of these stressors. This would be especially impactful if delivered by foster parents speaking from their lived experience. Further, prospective foster parents should have the opportunity to consider their current systems of social support, identifying areas of strength as well as gaps to address. Prospective foster parents could be given learning opportunities through case scenarios to identify their existing supports and local resources that could address service needs. As noted by Piel et al., “Training and orientation for new foster families should move beyond introduction to 123 potential resources, to actually exploring with foster parents the context and interaction of supports that collectively impact their ability to care for children” (2017, p. 1041). Exploration of supports could extend to in-service training, in which already-licensed foster parents have opportunities to assess their support resources and needs in light of their current fostering situations. Stress and support needs fluctuate over time and with changes in household composition; therefore, foster parents should have ongoing opportunities to learn about and assess their own support. Assessment of foster parent support must come from a strength-based perspective, focusing first on identifying existing sources and types of supports, including formal services and foster parents’ own informal networks. This could be accomplished through an existing tool such as an ecomap, which identifies systems in a family’s social environment (Hartman, 1995). A social support assessment tool specific to foster parenting would be especially helpful. Such a tool would need to be developed, as discussed in the research implications section below. Subsequently, gaps, barriers, and strategies for increasing support could be identified. This assessment of support could begin during the initial licensing process, as part of individual meetings with a licensing worker. It could also be incorporated into other existing practices, such as annual reviews and re-licensing meetings, in which information is already gathered from foster parents on a regular basis. Individual assessment is critical to a tailored, responsive approach to foster parent support. Each foster family has a unique constellation of stressors, skills, and resources, all of which can change over time and with shifting circumstances. Additionally, individual assessment is key for those groups of foster parents who may be lacking in supports or who face unique, additional stressors, such as kinship caregivers or foster parents who are LGBTQ. 124 Lastly, understanding of the stressors of fostering and the protective role of social support should be incorporated into interventions with foster families. Many of the needed changes in this area are related to child welfare workers and their relationships with individual foster parents. According to Denby et al., “It behooves agency officials to concentrate their efforts towards developing staff persons who embrace a commitment toward maintaining working relationships with foster parents” (1999, p. 301). Certainly, there are many child welfare workers who excel at building and maintaining collaborative and mutually respectful relationships with foster families. However, fostering research, including the current study, consistently points to relationships with foster care agencies and workers as the biggest source of stress for foster families (Blythe et al., 2014; Gabler et al., 2014; Geiger et al., 2013). Among stressors common to fostering, poor communication from agency staff was found to be one of the most common and most stressful for foster parents in this study. Additionally, results indicate that there are disparities in retention between child-placing agencies, suggesting that agency culture and procedures may play a part in foster parent retention. The aim of practice change is not to create additional work for an already overburdened child welfare workforce, but to increase understanding of the foster parent experience and, subsequently, the approach to working relationships with them. While changing foster parents’ experience of the system may seem daunting, many potential shifts in practice would require little additional effort or money. Child welfare workers could be trained on effective support practices by foster parents themselves, who are able to speak to the role-specific stressors they experience, as well as what has been helpful. Workers could also increase efforts to celebrate effective foster parenting, check in with foster parents who’ve recently had a child leave, or return phone calls and emails in a timely 125 manner. Each of these represents a small practice shift that could have significant benefits, in terms of perceived support among foster parents. An intervention that has shown promise in supporting and retaining child welfare workers is that of peer mentoring (Strand & Bosco-Ruggiero, 2009). Fostering literature suggests that individuals who left fostering were less likely to have had a foster parent mentor than those who remained in their roles (Rhodes, Orme, & Buehler, 2001). Peer mentoring interventions could provide foster parents with social support starting at the time of licensure and continuing through the duration of their fostering. Peer mentors could serve as single sources of multiple types of support, including tangible (respite care), informational (training and resource recommendations), esteem (validation of skills), among others. Further, foster parents serving as mentors would benefit from the opportunity to share their expertise and contribute to new foster parents’ development. Peer mentor positions should be somewhat formalized and include preparatory training, clear expectations, and financial compensation. Though research on formal foster parent mentoring programs is lacking, one study by Miller and colleagues (2017) offers suggestions from foster parents for developing successful mentoring programs. Themes include recommendations for program structure, ongoing training and support, and mentor-mentee matching practices. Additionally, this research also points to a need for support that is individualized and responsive, rather than the traditional approach, which has tended to be one-size-fits all. Building on the recommendation for ongoing assessment of supports made above, workers could use this information to tap into existing supports (both formal and informal) in response to specific stressors. For example, a foster parent who has welcomed a new child into her home might be in particular need of tangible supports, such as meals. A worker with knowledge of that foster 126 parent’s support system could contact an informal source of support, such as a church group, to arrange meals for the family. Similarly, a foster parent who is having trouble addressing problem behaviors of his foster child might be in need of informational support, or advice, from another foster parent. A worker could link the two foster parents in a peer mentoring arrangement. To be optimally effective, foster parent support should be tailored to the nature of the presenting need. Implications for Policy In order to be successful, changes in practice must be supported and sustained through changes in policy, from the federal level to the local child welfare agency. Existing policy structures at the federal level confirm that the system does a poor job of maintaining an adequate supply of foster homes to meet the placement needs of children in care. The Child and Family Services Review (CFSR) provides federal oversight of state child welfare systems. In the most recent CFSR, only 14 states achieved the desired ‘substantial conformity’ in regards to the systemic factor of ‘Foster and Adoptive Parent Licensing, Recruitment, and Retention’ (U.S. Children’s Bureau, 2020). This highlights the broad nature of the need for reform in terms of foster parent recruitment and retention. However, it is also important to recognize that there are currently no items within CFSR that specifically assess foster parent retention (U.S. Children’s Bureau, n.d.-a). Retention is an essential strategy for maintaining an adequate number of foster homes to meet the needs of children in care. Adding an item which assesses foster parent support and retention to future CFSR’s is a policy change that would have far-reaching implications, because all state child welfare systems are held to these standards and required to gather and report data on the extent to which they are achieved. On the state level, child welfare systems should work to identify, implement, and scale up innovative practices for foster parent support and retention. Fostering literature suggests that in the absence of adequate support from formal sources (i.e. child welfare and other professionals), 127 foster parents build their own informal networks of support (Maclay et al., 2006). State child welfare departments could convene focus groups of experienced foster parents to learn their self- initiated support practices and how they are beneficial. These foster parents could also help identify gaps in the state’s existing support efforts and aid in the design and implementation of new support interventions. The child welfare system would benefit from creating and sustaining structures for foster parent involvement and leadership, such as advisory boards and paid training positions. Changes in policy must be accompanied by increasing dedicated funding at the local level for foster parent support and retention. “Most social service agencies desire to provide optimal services to parents, and workers do attempt to be accessible. However, the reality of budget constraints sometimes thwart the best of intentions” (Denby, 1999, p. 301). If child welfare agencies are expected to improve in these areas, then they must be given the resources with which to do this effectively and sustainably. Such funds could be dedicated to establishing and increasing access to support groups, creating paid peer support roles for experienced foster parents, and implementing support-related training for parents and workers. Increased funding could also enable child welfare systems to decrease worker caseloads, which could free workers to concentrate more on relationships with foster parents. Financial support could also be offered to foster parents more directly, in the form of increased foster care stipends or income tax credits for foster parents. Implications for Social Work Education This study has implications for schools of social work, specifically in the areas of undergraduate and graduate education, as well as continuing education for social workers and foster parents. First, schools of social work, including those with Title IV-E child welfare 128 certificate programs, serve as the initial training ground for many future child welfare professionals at the BSW and MSW levels. As such, they are well-positioned to promote supportive relationships and practices between child welfare workers and foster parents through a variety of means, including course content and field placement experiences. Within courses, particularly child welfare courses, educators can emphasize the importance and uniqueness of the foster parent role, the critical nature of foster parent support, and strategies for engagement and partnership with foster parents as members of the child welfare team. Schools of social work can utilize foster parents in the guest speaker role so that students can hear firsthand about the experience of fostering. Field placements within child welfare agencies offer valuable experience practicing and reflecting upon course content related to fostering. Students in child welfare field placements should be ensured of time spent learning from foster parents themselves. Students in field placement could shadow seasoned foster parents, gaining helpful insight into the role and how it is experienced. Many schools of social work, particularly at large public universities, have continuing education initiatives that offer professional development to social workers in the form of in- person and online training sessions. These initiatives provide an existing structure through which child welfare staff could receive training on the topics such as common stressors of fostering and matching support types to specific stressors. Additionally, partnerships between schools of social work and child welfare agencies offer opportunities for foster parent training. Through the educational resources and broad reach of schools of social work, researchers can partner with practitioners and foster parents to develop, implement, and evaluate trainings on fostering stressors and supports to foster parents themselves. 129 Implications for Research Lastly, the current study has several research implications. It adds to existing research on fostering in several important ways. First, it offers quantitative findings on the impact of fostering on foster parents, which is currently a small, but growing, body of literature. It is the first known study to use standardized measures to demonstrate the relationship between parenting stress and well-being in the fostering population. It also offers a theoretically-grounded examination of the buffering role of social support for foster parents. Finally, by using multidimensional measures of parenting stressors and support, it offers a step toward a more nuanced approach to foster parent support, in which specific types of support can be matched to specific stressors for optimal impact. Future research can build on these contributions by continuing to test the buffering hypothesis of social support among foster parents. Studies which include samples that are larger and more representative, in terms of gender and race, would be beneficial. Longitudinal studies would potentially offer evidence of causality that is not possible in the current study. Future studies could also explore questions raised by the current study, particularly in regards to foster parent retention. This study demonstrated that foster parents with high stress levels think about leaving fostering more often than those with lower stress levels; however, they are not more likely to plan on leaving their role. This discrepancy between thoughts about leaving and plans to leave fostering raises important questions and opportunities for future research. Why would a foster parent think about leaving, but not actually plan to do so? Is the intent to continue fostering a valid indicator of actual retention? What are the factors that impact the decision to leave fostering? The study points to two factors that are related to the intent to leave: foster care agency and religious participation. Both are areas for exploration in further research. Why do certain agencies have much higher rates of foster parents who intend to 130 continue? Are there agency practices that promote the intent to continue and could be replicated? Religious participation was also linked to less intent to leave and less thoughts about giving up fostering. Might religious participation be a proxy for one or more other factors, such as social network size or altruism? This and other retention-focused questions raised here should be explored in future research. Additionally, the practice recommendations outlined above point to a need for more research on valid, reliable measures of support for foster parents. This study advances knowledge in this area by being the second known study to use the Social Provisions Scale (SPS; Cutrona & Russell, 1987), a multidimensional measure of perceived support. Future research could expand on this by adapting the SPS or other measures to reflect the supportive actions specific to the fostering role. A fostering-specific social support assessment could be developed through future research. Such a tool could be utilized to identify foster parents with low levels of support (overall or in particular areas) early and make plans for strengthening support systems. This could be done prior to licensure, as prospective foster parents prepare for their future role, or during annual licensing reviews or monthly home visits with foster care workers. Finally, future research can build on this study’s multidimensional examination of both stressful events and supports in order to create a response to foster parent stress that is nuanced and responsive. Within social support literature, this idea is known as the stressor-support specificity model (Cohen & McKay, 1984) or the optimal matching model (Cutrona, 1990). These models are refinements of the buffering hypothesis used in the current study, and propose that supports are most effective “when the type of support provided matches the coping requirements elicited by a particular stressor or stress experience” (Cohen & McKay, 1984, p. 261). For example, when a stressful event is perceived to be controllable, tangible assistance is 131 most applicable (Cutrona, 1990). In terms of foster parenting, foster parents generally have control over a new child being placed in their home. This stressful experience can be ameliorated by help with meals, transportation, clothing, and other concrete supports. In contrast, when an event or circumstance results in feelings of failure or inadequacy, it is reassurance of worth, also known as esteem support, which is most helpful (Cohen & McKay, 1984). Foster parents who struggle with managing problem behaviors of foster children may feel that they’re failing. Hearing from others, especially foster workers and fellow foster parents, that they are doing their best and that their efforts are noticed can be beneficial. Future research that explores the ‘optimal matching’ between stressors of fostering and types of support could hold incredible potential for foster parents and foster care agencies. Conclusion As a social worker, researcher, educator, and licensed foster parent, I have both a professional and personal investment in the ability of the child welfare system to promote foster parent well-being and retention. As the most commonly used placement type, licensed foster homes are a critical resource for children in out-of-home care in the United States (U.S. Children’s Bureau, 2019). Yet, foster parents frequently leave their roles within the first two years (Gibbs & Wildfire, 2007) and most states have seen a decrease in foster home capacity over the past several years (Chronicle of Social Change, 2017). The foster parent role is inherently and uniquely stressful. Foster parents care for children who’ve experienced complex trauma (Greeson et al., 2011) and have significant and multiple needs as a result (Turney & Wildeman, 2016). They do so as temporary parents within a “context of impermanency” (Lanigan & Burleson, 2017, p. 913) and under public scrutiny (Marcellus, 2010). However, literature on fostering is lacking in research on what may contribute to foster parent well-being and longevity in this challenging role. 132 This study contributes to foster care literature by increasing knowledge related to the impact of stress and support on foster parents’ well-being and retention in their role. Specifically, this study adds to a small (but growing) body of quantitative research conducted on a U.S. foster parent population. Further, this study is grounded in social support theory and a multidimensional conceptualization of social support. While a number of existing studies have examined social support among foster parents, the research does not consistently reflect the multidimensional nature of social support or attention to social support theory. This study adds to the body of fostering literature in several important ways, specifically by: • utilizing an inventory of common stressors of fostering which measured both presence of the stressors and foster parents perceptions on the stress caused by each stressor experienced. Findings indicated that stressors of fostering are common, and that those experienced as most stressful are system-related. • making a distinction, in accordance with social support theory, between stressors and the experience of stress. These were measured as separate variables, and results identified specific stressors that were related to increased levels of parenting stress. • distinguishing between thoughts about leaving fostering and the intent to leave in the next 18 months. Parenting stress was found to be significantly related to the former, but not the latter, indicating that foster parents who have high levels of stress are more likely to think about leaving fostering, but no more likely to actually plan to do so. • producing the first known evidence that social support serves in a buffering role for foster parents, preserving their well-being even in the presence of high levels of parenting stress. 133 • laying a foundation for a more nuanced approach to foster parent support. By expanding the understanding of social support to be multidimensional, the study points to the need for foster parent support to be targeted and appropriate to the nature of the stressors they face. Overall, the findings of this study increase understanding of the experience of foster parents, specifically in terms of stress, retention, well-being, and social support. In the context of a role that is characterized by unique and multiple stressors, and will continue to be so, this study points to the critical importance of support for foster parent well-being. It lays a foundation for a multidimensional understanding of foster parent support that stands to benefit not only foster parents, but the broader child welfare system, strengthening its ability to more fully meet its goals of safety, permanence, and well-being for the children in its care. 134 APPENDICES 135 APPENDIX A MSU IRB Determination Letter 136 137 138 139 140 141 142 APPENDIX B MDHHS IRB Determination Letter 143 144 145 APPENDIX C Invitation to Participate and Reminder Messages 146 147 148 149 150 151 APPENDIX D Survey Instrument 152 Foster Parent Stress and Support Survey Start of Block: Consent to Participate Q138 Q16 Research Participant Information and Consent Form Study Title: Stress, Support, and Well-being Among Foster Parents Researcher and Title: Liz Sharda, LMSW Department and Institution: School of Social Work, Michigan State University Contact Information: shardael@msu.edu BRIEF SUMMARY You are being asked to participate in a research study. Researchers are required to provide a consent form to inform you 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. You are being asked to participate in a research study of foster parenting, specifically, the well-being and retention of foster parents and the factors that impact them. Your participation in this study will take about 15-20 minutes. You will be asked to complete a series of questions about your experience as a foster parent, as well as information about yourself and your foster home. The most likely risk of participating in this study is discomfort resulting from thinking about the difficulties associated with being a foster parent... You will not directly benefit from your participation in this study. However, your participation in this study may benefit other foster parents by contributing to the understanding of what foster parents experience in their role, and how fostering impacts them and their well-being. PURPOSE OF RESEARCH The purpose of this research study is to examine the relationships between stress, social support, well-being, and retention among foster parents. WHAT YOU WILL BE ASKED TO DO You will be asked to answer a series of questions related to your experiences and opinions as a foster parent. You 153 will also be asked questions about yourself and your foster home. You may skip any questions that you prefer not to answer. The survey will take approximately 15-20 minutes to complete. PRIVACY AND CONFIDENTIALITY You will not be asked to share any identifying information as a part of this survey. The researchers will not have access to your name, birthdate, address, email address, or any other identifying information. Other information that you share through your responses to this survey will be stored in a password- protected account on an online survey platform, and then downloaded to a password-protected computer. Your individual responses will only be accessible to researchers, who will combine them with others’ responses before sharing findings with your foster care agency and other stakeholders. 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. COSTS AND COMPENSATION FOR BEING IN THE STUDY You will not receive money or any other form of compensation for participating in this study. RESEARCH RESULTS If you are interested in receiving a copy of the study findings, please email researcher Liz Sharda at shardael@msu.edu. 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, Liz Sharda, LMSW, at shardael@msu.edu, or Baker Hall, 655 Auditorium Rd, East Lansing, MI 48824 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. Documentation of Informed consent. By clicking the blue arrow below, you indicate your informed consent for participation in this research study. End of Block: Consent to Participate Start of Block: Foster Home Information Q55 Do you currently have a foster home license? ▼ Yes (1) ... No (2) Skip To: End of Survey If Do you currently have a foster home license? = No 154 Q57 Approximately how many years have you been a foster parent? (If less than one year, list the number of months you've been a foster parent.) o Years (1) ________________________________________________ o Months (2) ________________________________________________ Q137 How many TOTAL foster children have you had in your home, since you became licensed? o No placements yet (1) o 1-2 (2) o 3-5 (3) o 5-10 (4) o More than 10 (5) Skip To: End of Survey If How many TOTAL foster children have you had in your home, since you became licensed? = No placements yet Page Break 155 Q1 Foster parents experience many challenges in their role. Below is a list of potentially challenging events commonly experienced by foster parents. Which of these events have you experienced in the past two months or so? (Select all that apply.) of that age) (3) ▢ New child placed in foster home (1) ▢ Foster child moved out of home (2) ▢ Foster child behavior problems (i.e. Behaviors that are beyond those typical for children ▢ Foster child medical needs (4) ▢ Foster child mental health needs (5) ▢ Allegations of abuse or neglect against your foster home (6) ▢ Difficulty obtaining services (e.g., counseling) for foster child (7) ▢ Difficulties in relationship with birth parents (8) ▢ Change in foster care case manager (9) ▢ Poor communication from foster care case manager (or other foster care agency staff) ▢ Lack of information related to foster child (e.g., trauma history or behavior problems) ▢ Disagreement with decision in foster care case (e.g., decision to terminate parental ▢ Disagreement with a foster care licensing rule or policy (14) rights) (12) (10) (11) 156 ▢ Other (13) ________________________________________________ Display This Question: If If Stressors q://QID1/SelectedChoicesCount Is Greater Than or Equal to 1 Q18 For those events that you experienced in the past two months or so, please indicate how stressful they were for you. Display This Question: If Stressors = New child placed in foster home Q2 On a scale of 1 (not at all) to 4 (extremely), how stressful was the new placement of a foster child in your home during the past two months or so? ▼ 1 - Not at all stressful (1) ... 4 - Extremely stressful (4) Display This Question: If Stressors = Foster child moved out of home Q5 On a scale of 1 (not at all) to 4 (extremely), how stressful was the foster child moving out of your home during the past two months or so? ▼ 1 - Not at all stressful (1) ... 4 - Extremely stressful (4) Display This Question: If Stressors = Foster child behavior problems (i.e. Behaviors that are beyond those typical for children of that age) Q6 On a scale of 1 (not at all) to 4 (extremely), how stressful were the foster child behavior problems during the past two months or so? ▼ 1 - Not at all stressful (1) ... 4 - Extremely stressful (4) 157 Display This Question: If Stressors = Foster child medical needs Q7 On a scale of 1 (not at all) to 4 (extremely), how stressful were the foster child medical needs during the past two months or so? ▼ 1 - Not at all stressful (1) ... 4 - Extremely stressful (4) Display This Question: If Stressors = Foster child mental health needs Q8 On a scale of 1 (not at all) to 4 (extremely), how stressful were the foster child mental health needs during the past two months or so? ▼ 1 - Not at all stressful (1) ... 4 - Extremely stressful (4) Display This Question: If Stressors = Allegations of abuse or neglect against your foster home Q9 On a scale of 1 (not at all) to 4 (extremely), how stressful were the allegations of abuse or neglect against your home during the past two months or so? ▼ 1 - Not at all stressful (1) ... 4 - Extremely stressful (4) Display This Question: If Stressors = Difficulty obtaining services (e.g., counseling) for foster child Q10 On a scale of 1 (not at all) to 4 (extremely), how stressful was the difficulty in obtaining services (e.g., counseling) for your foster child during the past two months or so? ▼ 1 - Not at all stressful (1) ... 4 - Extremely stressful (4) Display This Question: If Stressors = Difficulties in relationship with birth parents 158 Q11 On a scale of 1 (not at all) to 4 (extremely), how stressful were the difficulties in relationship with birth parents during the past two months or so? ▼ 1 - Not at all stressful (1) ... 4 - Extremely stressful (4) Display This Question: If Stressors = Change in foster care case manager Q12 On a scale of 1 (not at all) to 4 (extremely), how stressful was the change in foster care case manager during the past two months or so? ▼ 1 - Not at all stressful (1) ... 4 - Extremely stressful (4) Display This Question: If Stressors = Poor communication from foster care case manager (or other foster care agency staff) Q13 On a scale of 1 (not at all) to 4 (extremely), how stressful was the poor communication from foster care case manager (or other foster care agency staff) during the past two months or so? ▼ 1 - Not at all stressful (1) ... 4 - Extremely stressful (4) Display This Question: If Stressors = Lack of information related to foster child (e.g., trauma history or behavior problems) Q14 On a scale of 1 (not at all) to 4 (extremely), how stressful was the lack of information related to your foster child during the past two months or so? ▼ 1 - Not at all stressful (1) ... 4 - Extremely stressful (4) Display This Question: If Stressors = Disagreement with decision in foster care case (e.g., decision to terminate parental rights) 159 Q15 On a scale of 1 (not at all) to 4 (extremely), how stressful was the disagreement with a decision or policy in the foster care case during the past two months or so? ▼ 1 - Not at all stressful (1) ... 4 - Extremely stressful (4) Display This Question: If Stressors = Disagreement with a foster care licensing rule or policy Q145 On a scale of 1 (not at all) to 4 (extremely), how stressful was the disagreement with foster care licensing rule or policy during the past two months or so? ▼ 1 - Not at all stressful (1) ... 4 - Extremely stressful (4) Display This Question: If If Stressors Other Is Not Empty Q144 On a scale of 1 (not at all) to 4 (extremely), how stressful was the following challenge that you identified: ${Q1/ChoiceTextEntryValue/13} ▼ Not at all stressful (1) ... Extremely stressful (4) Page Break 160 Q58 How many foster children are currently placed in your home? ▼ 0 (1) ... More than 4 (6) Q59 Through which foster care agency are you licensed? ▼ Bethany Christian Services (1) ... Other (6) Q56 A kinship caregiver is a relative or close friend (sometimes called fictive kin) who steps in to provide care and custody of children when their parents are unable to do so. Based on this definition, are you a kinship caregiver? ▼ Yes (1) ... No (2) Q78 In the past 2 months or so, how often have you thought about giving up fostering? ▼ Never (1) ... Very often (5) Q55 How likely is it that you will still be fostering 18 months from now? ▼ Very unlikely (1) ... Very likely (4) End of Block: Foster Home Information Start of Block: Parenting Stress Q22 The following statements describe feelings and perceptions about the experience of being a foster parent. Think of each of the items in terms of how your relationship with your foster child or children typically is. Please indicate the degree to which you agree or disagree with the following items. 161 Q21 I am happy in my role as a foster parent. ▼ Strongly disagree (1) ... Strongly agree (5) Q23 There is little or nothing I wouldn't do for my foster child(ren) if it was necessary. ▼ Strongly disagree (1) ... Strongly agree (5) Q24 Caring for my foster child(ren) sometimes takes more time and energy than I have to give. ▼ Strongly disagree (1) ... Strongly agree (5) Q25 I sometimes worry whether I am doing enough for my foster child(ren). ▼ Strongly disagree (1) ... Strongly agree (5) Q26 I feel close to my foster child(ren). ▼ Strongly disagree (1) ... Strongly agree (5) Q27 I enjoy spending time with my foster child(ren). ▼ Strongly disagree (1) ... Strongly agree (5) 162 Q28 My foster child(ren) is an important source of affection for me. ▼ Strongly disagree (1) ... Strongly agree (5) Q29 Having foster child(ren) gives me a more certain and optimistic view for the future. ▼ Strongly disagree (1) ... Strongly agree (5) Q30 The major source of stress in my life is my foster child(ren). ▼ Strongly disagree (1) ... Strongly agree (5) Q31 Having foster child(ren) leaves little time and flexibility in my life. ▼ Strongly disagree (1) ... Strongly agree (5) Q32 Having foster child(ren) has been a financial burden. ▼ Strongly disagree (1) ... Strongly agree (5) Q33 It is difficult to balance different responsibilities because of my foster child(ren). ▼ Strongly disagree (1) ... Strongly agree (5) Q34 The behavior of my foster child(ren) is often embarrassing or stressful to me. ▼ Strongly disagree (1) ... Strongly agree (5) 163 Q35 If I had it to do over again, I might decide not to have foster child(ren). ▼ Strongly disagree (1) ... Strongly agree (5) Q36 I feel overwhelmed by the responsibility of being a foster parent. ▼ Strongly disagree (1) ... Strongly agree (5) Q37 Having foster child(ren) has meant having too few choices and too little control over my life. ▼ Strongly disagree (1) ... Strongly agree (5) Q38 I am satisfied as a foster parent. ▼ Strongly disagree (1) ... Strongly agree (5) Q39 I find my foster child(ren) enjoyable. ▼ Strongly disagree (1) ... Strongly agree (5) Page Break 164 End of Block: Parenting Stress Start of Block: Social Support Q134 In answering the next set of questions about your experiences as a foster parent, please think about your current relationship with friends, family members, coworkers, community members, and so on. Please indicate to what extent you agree that each statement describes your current relationships with other people, particularly within your role as a foster parent. If you feel a statement is very true of your current relationships, you would select “strongly agree”. If you feel a statement clearly does not describe your relationships, you would select “strongly disagree”. Q109 There are people I can depend on to help me if I really need it. ▼ Strongly disagree (1) ... Strongly agree (4) Q111 I feel that I do not have close personal relationships with other people. ▼ Strongly disagree (1) ... Strongly agree (4) Q112 There is no one I can turn to for guidance in times of stress. ▼ Strongly disagree (1) ... Strongly agree (4) Q113 There are people who depend on me for help. ▼ Strongly disagree (1) ... Strongly agree (4) 165 Q114 There are people who enjoy the same social activities I do. ▼ Strongly disagree (1) ... Strongly agree (4) Q115 Other people do not view me as competent. ▼ Strongly disagree (1) ... Strongly agree (4) Q116 I feel personally responsible for the well-being of another person. ▼ Strongly disagree (1) ... Strongly agree (4) Q117 I feel part of a group of people who share my attitudes and beliefs. ▼ Strongly disagree (1) ... Strongly agree (4) Q118 I do not think other people respect my skills and abilities. ▼ Strongly disagree (1) ... Strongly agree (4) Q119 If something went wrong, no one would come to my assistance. ▼ Strongly disagree (1) ... Strongly agree (4) 166 Q120 I have close relationships that provide me with a sense of emotional security and well-being. ▼ Strongly disagree (1) ... Strongly agree (4) Q121 There is someone I could talk to about important decisions in my life. ▼ Strongly disagree (1) ... Strongly agree (4) Q122 I have relationships where my competence and skills are recognized. ▼ Strongly disagree (1) ... Strongly agree (4) Q123 There is no one who shares my interests and concerns. ▼ Strongly disagree (1) ... Strongly agree (4) Q124 There is no one who really relies on me for their well-being. ▼ Strongly disagree (1) ... Strongly agree (4) Q125 There is a trustworthy person I could turn to for advice if I were having problems. ▼ Strongly disagree (1) ... Strongly agree (4) 167 Q126 I feel a strong emotional bond with at least one other person. ▼ Strongly disagree (1) ... Strongly agree (4) Q127 There is no one I can depend on for aid if I really need it. ▼ Strongly disagree (1) ... Strongly agree (4) Q128 There is no one I feel comfortable talking about problems with. ▼ Strongly disagree (1) ... Strongly agree (4) Q129 There are people who admire my talents and abilities. ▼ Strongly disagree (1) ... Strongly agree (4) Q130 I lack a feeling of intimacy with another person. ▼ Strongly disagree (1) ... Strongly agree (4) Q131 There is no one who likes to do the things I do. ▼ Strongly disagree (1) ... Strongly agree (4) Q132 There are people I can count on in an emergency. ▼ Strongly disagree (1) ... Strongly agree (4) 168 End of Block: Social Support Start of Block: Well-being Q80 During the past month, how often did you feel happy? ▼ Never (1) ... Every day (6) Q81 During the past month, how often did you feel interested in life? ▼ Never (1) ... Every day (6) Q82 During the past month, how often did you feel satisfied with life? ▼ Never (1) ... Every day (6) Q83 During the past month, how often did you feel that you had something important to contribute to society? ▼ Never (1) ... Every day (6) Q84 During the past month, how often did you feel that you belonged to a community (like a social group, school, neighborhood, etc.)? ▼ Never (1) ... Every day (6) Q85 During the past month, how often did you feel that our society is a good place, or is becoming a better place, for all people? ▼ Never (1) ... Every day (6) 169 Q86 During the past month, how often did you feel that people are basically good? ▼ Never (1) ... Every day (6) Q87 During the past month, how often did you feel that the way our society works made sense to you? ▼ Never (1) ... Every day (6) Q88 During the past month, how often did you feel that you liked most parts of your personality? ▼ Never (1) ... Every day (6) Q89 During the past month, how often did you feel good at managing the responsibilities of your daily life? ▼ Never (1) ... Every day (6) Q90 During the past month, how often did you feel that you had warm and trusting relationships with others? ▼ Never (1) ... Every day (6) Q91 During the past month, how often did you feel that you had experiences that challenged you to grow and become a better person? ▼ Never (1) ... Every day (6) 170 Q92 During the past month, how often did you feel confident to think or express your own ideas and opinions? ▼ Never (1) ... Every day (6) Q93 During the past month, how often did you feel that your life has a sense of direction or meaning to it? ▼ Never (1) ... Every day (6) End of Block: Well-being Start of Block: Demographics Q41 What is your year of birth? ________________________________________________________________ Q51 What is your gender? o Male (1) o Female (2) o Other (3) ________________________________________________ Q52 Which of the following best describes your household? ▼ Single parent household (1) ... Two-parent household, not married (3) 171 Q46 Choose one or more races or ethnicities that you consider yourself to be: ▢ American Indian or Alaska Native (3) ▢ Asian (4) ▢ Black or African American (2) ▢ Hispanic/Latino (7) ▢ Native Hawaiian or Pacific Islander (5) ▢ White (1) ▢ Other (6) ________________________________________________ Q53 Are you employed? ▼ Yes- Full-time (1) ... No (3) Q48 Please indicate the answer that includes your entire household income (in the previous year) before taxes. ▼ Less than $19,999 (2) ... $140,000 or more (12) Q54 What it the highest level of education you have completed? ▼ Less than high school (1) ... Advanced/Graduate degree (6) 172 Q141 Do you regularly (at least monthly) attend religious services? o Yes (1) o No (2) Q50 What is your ZIP code? ________________________________________________________________ End of Block: Demographics 173 REFERENCES 174 REFERENCES Abidin, R.R. (1986). Parenting Stress Index Manual (2nd ed.). Charlottesville, VA: Pediatric Psychology Press. Adams, E., Hassett, A. R., & Lumsden, V. (2018). What do we know about the impact of stress on foster carers and contributing factors? Adoption & Fostering, 42(4), 338–353. https://doi.org/10.1177/0308575918799956 Ahn, H., Greeno, E. J., Bright, C. L., Hartzel, S., & Reiman, S. (2017). A survival analysis of the length of foster parenting duration and implications for recruitment and retention of foster parents. Children and Youth Services Review, 79, 478–484. https://doi.org/10.1016/j.childyouth.2017.06.069 Barnett, E. R., Jankowski, M. K., Butcher, R. L., Meister, C., Parton, R. R., Drake, R. E. (2017). Foster and adoptive parent perspectives on needs and services: A mixed methods study. The Journal of Behavioral Health Services & Research, 45(1), 74–89. https://doi.org/10.1007/s11414-017-9569-4 Barrera, M., Jr., Sandler, I.N., & Ramsay, T.B. (1981). Preliminary development of a scale of social support: Studies on college students. American Journal of Community Psychology, 9, 435-447. https://doi.org/10.1007/BF00918174 Belsky, J. (1984). The determinants of parenting: A process model. Child Development, 55(1), 83–96. https://doi.org/10.2307/1129836 Bennett, D. A. (2001). How can I deal with missing data in my study? Australian and New Zealand Journal of Public Health, 25(5), 464–469. https://doi.org/10.1111/j.1467- 842X.2001.tb00294.x Berrick, J., & Skivenes, M. (2012). Dimensions of high quality foster care: Parenting Plus. Children and Youth Services Review, 34, 1956–1965. https://doi.org/10.1016/j.childyouth.2012.05.026 Berry, J. O., & Jones, W. H. (1995). The Parental Stress Scale: Initial psychometric evidence. Journal of Social and Personal Relationships, 12, 463-472. Blythe, S. L., Wilkes, L., & Halcomb, E. J. (2014). The foster carer’s experience: An integrative review. Collegian, 21(1), 21–32. https://doi.org/10.1016/j.colegn.2012.12.001 Bolger, N., Zuckerman, A. & Kessler, R. C. (2000). Invisible support and adjustment to stress. Journal of Personality and Social Psychology, 79, 953-61. https://doi.org/10.1037/0022- 3514.79.6.953 175 Bronfenbrenner, U. (1977). Toward an experimental ecology of human development. American psychologist, 32, 513.. https://doi.org/10.1007/s10648-006-9029-9 Brown, J. D. (2008). Foster Parents’ Perceptions of Factors Needed for Successful Foster Placements. Journal of Child and Family Studies, 17, 538–554. https://doi.org/10.1007/s10826-007-9172-z Brown, J., & Calder, P. (2000). Concept mapping the needs of foster parents. Child Welfare, 79, 729–746. Retrieved from http://ezproxy.msu.edu.proxy1.cl.msu.edu/login?url=https://search-proquest- com.proxy1.cl.msu.edu/docview/213810977?accountid=12598 Brown, J. D., & Campbell, M. (2007). Foster parent perceptions of placement success. Children and Youth Services Review, 29, 1010–1020. https://doi.org/10.1016/j.childyouth.2007.02.002 Buehler, C., Cox, M. E., & Cuddeback, G. (2003). Foster parents’ perceptions of factors that promote or inhibit successful fostering. Qualitative Social Work, 2, 61–83. https://doi.org/10.1177/1473325003002001281 Card, N. A., & Barnett, M. A. (2015). Methodological Considerations in Studying Individual and Family Resilience. Family Relations, 64(1), 120–133. https://doi.org/10.1111/fare.12102 Castillo, J. T., & Fenzl-Crossman, A. (2010). The relationship between non-marital fathers’ social networks and social capital and father involvement. Child & Family Social Work, 15, 66–76. https://doi.org/10.1111/j.1365-2206.2009.00644.x Cavazzi, T., Guilfoyle, A., & Sims, M. (2010). A phenomenological study of foster caregivers’ experiences of formal and informal support. Illinois Child Welfare, 5, 125–141. Retrieved from http://www.illinoischildwelfare.org/archives/volume5/icw5-cavazzi.pdf Child Welfare Information Gateway. (n.d.) Family foster care. https://www.childwelfare.gov/topics/outofhome/foster-care/fam-foster/ Child Welfare Information Gateway. (2017). Foster care statistics 2015. Washington, DC: U.S. Department of Health and Human Services, Children’s Bureau. Child Welfare Information Gateway. (2018). Placement of Children with Relatives. Washington, DC: Department of Health and Human Services, Children’s Bureau. Retrieved from https://fas.org/sgp/ Chiu, C.-Y., Motl, R. W., & Ditchman, N. (2016). Validation of the Social Provisions Scale in people with multiple sclerosis. Rehabilitation Psychology, 61, 297–307. https://doi.org/10.1037/rep0000089 176 Chronicle of Social Change. (2017). The Foster Care Housing Crisis. Retrieved from https://chronicleofsocialchange.org/wp-content/uploads/2017/10/The-Foster-Care- Housing-Crisis-10-31.pdf Cobb, S. (1979). Social support and health throughout the life course. In M. W. Riley (Ed.), Aging From Birth to Death: Interdisciplinary Perspectives (pp. 93-106). Boulder, CO: Westview Press. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd edition). New York: Academic Press. Cohen, S. (1991). Social supports and physical health: Symptoms, health behaviors, and infectious diseases. In Greene, A. L., Cummings, E. M., & Karraker, K. H. (Eds.) Life- Span Developmental Psychology: Perspectives on Stress and Coping (pp. 213-234). Hillsdale, NJ: Erlbaum. Cohen, S. (1992). Stress, social support, and disorder. In H.O.F. Veiel & U. Baumann (Eds.), The Meaning and Measurement of Social Support, (pp. 109-124). New York, NY: Hemisphere Publishing Corporation. Cohen, S. (2004). Social relationships and health. The American Psychologist, 59, 676–84. https://doi.org/10.1037/0003-066X.59.8.676 Cohen, S., Gottlieb, B. H., & Underwood, L. G. (2000). Social relationships and health. In Cohen, S., Underwood, L. G., & Gottlieb, B. H. (Eds.), Social Support Measurement and Intervention (pp. 3-25). New York, NY: Oxford University Press. Cohen, S., & Hoberman, H. (1983). Positive events and social supports as buffers of life change stress. Journal of Applied Social Psychology, 13, 99-125. https://doi.org/10.1111/j.1559- 1816.1983.tb02325.x Cohen, S. & McKay, G. (1984). Social support, stress, and the buffering hypothesis: A theoretical analysis. In A. Baum, S.E. Taylor, & J.E. Singer (Eds.), Handbook of Psychology and Health (pp. 253-267). Hillsdale, NJ: Laurence Erlbaum. Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310–357. https://doi.org/10.1037/0033-2909.98.2.310 Cole, S. A., & Eamon, M. K. (2007). Predictors of depressive symptoms among foster caregivers. Child Abuse & Neglect, 31, 295–310. https://doi.org/10.1016/j.chiabu.2006.06.010 Cook, A., Spinazzola, J., Ford, J., Lanktree, C., Blaustein, M., Cloitre, M., DeRosa, R., Hubbard, R., Kagan, R., Liautaud, J., Mallah, K., Olafson, E., van der Kolk, B. (2005). Complex Trauma in CHildren and Adolescents. Psychiatric Animal, 35.5(May), 390–398. 177 Retrieved from http://0- search.proquest.com.bianca.penlib.du.edu/docview/217061735?accountid=14608 Cooley, M. E., Farineau, H. M., & Mullis, A. K. (2015). Child behaviors as a moderator: Examining the relationship between foster parent supports, satisfaction, and intent to continue fostering. Child Abuse & Neglect, 45, 46–56. https://doi.org/10.1016/j.chiabu.2015.05.007 Cooley, M. E., Thompson, H. M., & Newell, E. (2019). Examining the influence of social support on the relationship between child behavior problems and foster parent satisfaction and challenges. Child & Youth Care Forum, 48, 289–303. https://doi.org/10.1007/s10566- 018-9478-6 Cooley, M. E., Thompson, H. M., & Wojciak, A. S. (2017). Risk, resilience, and complexity: Experiences of foster parents. Children and Youth Services Review, 76, 35–41. https://doi.org/10.1016/j.childyouth.2017.02.030 Cooper, C. E., & Mclanahan, S. S. (2009). Family structure transitions and maternal parenting stress. Journal of Marriage and Family, 71, 558–574. Copeland, D., & Harbaugh, B. L. (2005). Differences in parenting stress between married and single first time mothers at six to eight weeks after birth. Issues in Comprehensive Pediatric Nursing, 28(3), 139–152. https://doi.org/10.1080/01460860500227556 Coughlin, M., Sharry, J., Guerin, S., & Drumm, M. (2009). A controlled clinical evaluation of the Parents Plus Children’s Programme: A video-based programme for parents of children aged 6 to 11 with behavioural and developmental problems. Clinical Child Psychology and Psychiatry, 14(4), 541–558. https://doi.org/10.1177/1359104509339081 Creswell, J.W. (2013). Qualitative inquiry and research design: Choosing among five approaches (3rd ed.). Thousand Oaks, CA: Sage Publications, Inc. Crum, W. (2010). Foster parent parenting characteristics that lead to increased placement stability or disruption. Children and Youth Services Review, 32, 185–190. https://doi.org/10.1016/j.childyouth.2009.08.022 Cutrona, C. E. (1990). Stress and social support- in search of optimal matching. Journal of Social and Clinical Psychology, 9, 3–14. https://doi.org/10.1521/jscp.1990.9.1.3 Cutrona, C. E., Hessling, R. M. & Suhr, J. A. (1997). The influence of husband and wife personality on marital social support interactions. Personal Relationships, 4, 379-93. https://doi.org/10.1111/j.1475-6811.1997.tb00152.x Cutrona, C. E. and Russell, D. (1987). The provisions of social relationships and adaptation to stress. In W. H. Jones & D. Perlman (Eds.) Advances in Personal Relationships (Vol. 1, pp. 37-67). Greenwich, CT: JAI Press. 178 De Maeyer, S., Vanderfaeillie, J., Vanschoonlandt, F., Robberechts, M., & Holen, F. Van. (2014). Motivation for foster care. Children and Youth Services Review, 36, 143–149. https://doi.org/10.1016/j.childyouth.2013.11.003 De Maeyer, S., Vanderfaeillie, J., Robberechts, M., Vanschoonlandt, F., & Holen, F. Van. (2015). Foster parents’ coping style and attitudes toward parenting. Children and Youth Services Review, 53, 70–76. https://doi.org/10.1016/j.childyouth.2015.03.023 Deater-Deckard, K. (1998). Parenting stress and child adjustment: Some old hypotheses and new questions. Clinical Psychology: Science and Practice, 5, 314–332. https://doi.org/10.1111/j.1468-2850.1998.tb00152.x Denby, R., Rindfleisch, N., & Bean, G. (1999). Predictors of foster parents’ satisfaction and intent to continue to foster. Child Abuse & Neglect, 23, 287–303. https://doi.org/10.1016/S0145-2134(98)00126-4 De Maeyer, S., Vanderfaeillie, J., Vanschoonlandt, F., Robberechts, M., & Van Holen, F. (2014). Motivation for foster care. Children and Youth Services Review, 36, 143–149. https://doi.org/10.1016/j.childyouth.2013.11.003 DeVooght, K., & Blazey, D. (2013). Family Foster Care Reimbursement Rates in the U.S.: A Report from a 2012 National Survey on Family Foster Care Provider Classifications and Rates. Retrieved from https://www.childtrends.org/wp-content/uploads/2013/04/Foster- Care-Payment-Rate-Report.pdf Dillman, D. A., Smyth, J. D., & Christian, L. M. (2014). Internet, Phone, Mail, and Mixed Mode Surveys: The Tailored Design Method (4th ed.). Hoboken, NJ, US: John Wiley & Sons Inc. Dolan, M., Smith, K., Casanueva, C., & Ringeisen, H. (2011). NSCAW II Baseline Report. Washington, DC: Office of Planning, Research, and Evaluation, Administration for Children and Families, U.S. Department of Health and Human Services. Donachy, G. S. (2017). The caregiving relationship under stress: Foster carers’ experience of loss of the sense of self. Journal of Child Psychotherapy, 43, 223–242. https://doi.org/10.1080/0075417X.2017.1323943 Dunning, M. J., & Giallo, R. (2012). Fatigue, parenting stress, self-efficacy and satisfaction in mothers of infants and young children. Journal of Reproductive and Infant Psychology, 30, 145–159. https://doi.org/10.1080/02646838.2012.693910 Eaton, A., & Caltabiano, M. (2009). A four factor model predicting likelihood of foster carer retention. Australian Journal of Social Issues, 44, 215–229. https://doi.org/10.1002/j.1839-4655.2009.tb00141.x 179 Eisenberger, N. I., Taylor, S. E., Gable, S. L., Hilmert, C. J., & Lieberman, M. D. (2007). Neural pathways link social support to attenuated neuroendocrine stress responses. NeuroImage, 35, 1601–1612. https://doi.org/10.1016/j.neuroimage.2007.01.038 Ekas, N. V, Lickenbrock, D. M., & Whitman, T. L. (2010). Optimism, social support, and well- being in mothers of children with Autism Spectrum Disorder. Journal of Autism and Developmental Disorders, 40, 1274–1284. https://doi.org/10.1007/s10803-010-0986-y Ergh, T. C., Rapport, L. J., Coleman, R. D., & Hanks, R. a. (2002). Predictors of caregiver and family functioning following traumatic brain injury: social support moderates caregiver distress. The Journal of Head Trauma Rehabilitation, 17, 155–174. https://doi.org/10.1097/00001199-200204000-00006 Everson, R. B., Darling, C. A., & Herzog, J. R. (2013). Parenting stress among US Army spouses during combat-related deployments: the role of sense of coherence. Child & Family Social Work, 18, 168–178. https://doi.org/10.1111/j.1365-2206.2011.00818.x Farmer, E., Lipscombe, J., & Moyers, S. (2005). Foster carer strain and its impact on parenting and placement outcomes for adolescents. British Journal of Social Work, 35, 237–253. https://doi.org/10.1093/bjsw/bch181 Farmer, E., Moyers, S., & Lipscombe, J. (2004). Fostering Adolescents. London, England: Jessica Kingsley Publishers. Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39, 175-191. Festinger, T., & Baker, A. J. L. (2013). The quality of evaluations of foster parent training: An empirical review. Children and Youth Services Review, 35(12), 2147–2153. https://doi.org/10.1016/j.childyouth.2013.10.009 Fisher, P. A., & Stoolmiller, M. (2008). Intervention effects on foster parent stress: associations with child cortisol levels. Development and Psychopathology, 20(3), 1003–1021. https://doi.org/10.1017/S0954579408000473 Foster Parent’s Bill of Rights Law. MCL Act 203 of 1994, Section 722.958a (2015). Retrieved from http://www.legislature.mi.gov/(S(i3rs4i2eh4kg3nqzbdpvy4dw))/mileg.aspx?page=print& objectname=mcl-722-958a Gabler, S., Kungl, M., Bovenschen, I., Lang, K., Zimmermann, J., Nowacki, K., Kliewer- Neumann, J. & Spangler, G. (2018). Predictors of foster parents’ stress and associations to sensitivity in the first year after placement. Child Abuse & Neglect, 79, 325–338. https://doi.org/10.1016/j.chiabu.2018.02.009 180 Geens, N., & Vandenbroeck, M. (2014). The (ab)sense of a concept of social support in parenting research: a social work perspective. Child & Family Social Work, 19, 491–500. https://doi.org/10.1111/cfs.12048 Geiger, J. M., Hayes, M. J., & Lietz, C. A. (2013). Should I stay or should I go? A mixed methods study examining the factors influencing foster parents’ decisions to continue or discontinue providing foster care. Children and Youth Services Review, 35, 1356–1365. https://doi.org/10.1016/j.childyouth.2013.05.003 Geiger, J. M., Piel, M. H., Lietz, C. A., & Julien-Chinn, F.J. (2016). Empathy as an essential foundation to successful foster parenting. Journal of Child and Family Studies, 25, 3771– 3779. https://doi.org/10.1007/s10826-016-0529-z Geiger, J. M., Piel, M. H., & Julien-Chinn, F. J. (2017). Improving relationships in child welfare practice: Perspectives of foster care providers. Child and Adolescent Social Work Journal, 34, 23–33. https://doi.org/10.1007/s10560-016-0471-3 Gellert, P., Hä Usler, A., Suhr, R., Gholami, M., Rapp, M., Kuhlmey, A., & Nordheim, J. (2018). Testing the stress-buffering hypothesis of social support in couples coping with early- stage dementia. PLoS ONE, 13(1). https://doi.org/10.1371/journal.pone.0189849 Gerin, W., Pieper, C., Levy, R., & Pickering, T. G. (1992). Social support in social interaction: A moderator of cardiovascular reactivity. Psychosomatic Medicine, 54, 324–336. https://doi.org/10.1097/00006842-199205000-00008 Giacomini, M. (2010). Theory matters in qualitative health research. In I. Bourgeault, R. Dingwall, R., & R. deVries (Eds.), The SAGE Handbook of Qualitative Methods in Health Research (pp. 125-156). Thousand Oaks, CA: SAGE Publications. Gibbs, D., & Wildfire, J. (2007). Length of service for foster parents: Using administrative data to understand retention. Children and Youth Services Review, 29, 588–599. https://doi.org/10.1016/j.childyouth.2006.11.002 Goemans, A., Van Geel, M., & Vedder, P. (2018). Foster children’s behavioral development and foster parent stress: testing a transactional model. Journal of Child and Family Studies, 27, 990–1001. https://doi.org/10.1007/s10826-017-0941-z Gottlieb, B. H., & Bergen, A. E. (2010). Social support concepts and measures. Journal of Psychosomatic Research, 69, 511–520. https://doi.org/10.1016/j.jpsychores.2009.10.001 Green, B. L., Furrer, C. J., & Mcallister, C. L. (2011). Does attachment style influence social support or the other way around? A longitudinal study of Early Head Start mothers. Attachment & Human Development, 13(1), 27–47. https://doi.org/10.1080/14616734.2010.488121 181 Greene, R. R. (2014). Resilience as effective functional capacity: An ecological-stress model. Journal of Human Behavior in the Social Environment, 24, 937–950. https://doi.org/10.1080/10911359.2014.921589 Greeson, J. K. P., Briggs, E. C., Kisiel, C. L., Layne, C. M., Ake, G. S., Ko, S. J., Gerrity, E. T., Steinberg, A. M., Howard, M. L., Pynoos, R. S., 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), 91–108. Retrieved from http://works.bepress.com/johanna_greeson/9/ Grimm, B. (2003). Child & Family Services Reviews: Part II in a series; Foster parent training: What the CFS Reviews do & don’t tell us. Journal for the National Center for Youth Law, XXIV(2), 3–31. Hacsi, T. (1995). From indenture to family foster care: A brief history of child placing. Child Welfare, 74(1), 162. Retrieved from http://ezproxy.msu.edu.proxy2.cl.msu.edu/login?url=https://search-proquest- com.proxy2.cl.msu.edu/docview/213805834?accountid=12598 Halfon, N., Mendonca, A., & Berkowitz, G. (1995). Health status of children in foster care: The experience of the Center for the Vulnerable Child. Archives of Pediatric and Adolescent Medicine, 149, 386–392. https://doi.org/10.1001/archpedi.1995.02170160040006 Hannah, B., & Woolgar, M. (2018). Secondary trauma and compassion fatigue in foster carers. Clinical Child Psychology and Psychiatry, 23, 629–643. https://doi.org/10.1177/1359104518778327 Hartman, A. (1995). Diagrammatic assessment of family relationships. Families in Society: The Journal of Contemporary Social Services, 76(2), 111–122. https://doi.org/10.1177/104438949507600207 Haselgruber, A., Sölva, K., & Lueger-Schuster, B. (2020). Validation of ICD-11 PTSD and complex PTSD in foster children using the International Trauma Questionnaire. Acta Psychiatrica Scandinavica, 141(1), 60–73. https://doi.org/10.1111/acps.13100 Hayslip, B., Blumenthal, H., & Garner, A. (2015). Social support and grandparent caregiver health: One-year longitudinal findings for grandparents raising their grandchildren. Journals of Gerentology: Social Sciences, 70, 804–812. https://doi.org/10.1093/geronb/gbu165 Hebert, C. G., & Kulkin, H. (2016). Attending to foster parent grief: exploring the use of grief awareness training for child welfare workers. Adoption & Fostering, 40(2), 128–139. https://doi.org/10.1177/0308575916644169 Hendrix, S. K., & Ford, J. (2003). Hardiness of foster families and the intent to continue to foster. Journal of Family Social Work, 7(2), 25–34. https://doi.org/10.1300/J039v7n02_03 182 Holland, P., & Gorey, K. M. (2004). Historical, Developmental, and Behavioral Factors Associated with Foster Care Challenges. Child and Adolescent Social Work Journal, 21, 117–135. https://doi.org/10.1023/B:CASW.0000022727.40123.95 Hopkins, K. M., Cohen-Callow, A., Kim, H. J., & Hwang, J. (2010). Beyond intent to leave: Using multiple outcome measures for assessing turnover in child welfare. Children and Youth Services Review, 32(10), 1380–1387. https://doi.org/10.1016/j.childyouth.2010.06.006 Hornstein, E. A., & Eisenberger, N. I. (2017). Unpacking the buffering effect of social support figures: Social support attenuates fear acquisition. PLoS ONE, 12(5). https://doi.org/10.1371/journal.pone.0175891 Horton, T. V., Wallander, J. L., Noojin, A., Moore, S., Morris, S., Cole, E., & Lowe, W. (2001). Hope and social support as resilience factors against psychological distress of mothers who care for children with chronic physical conditions. Rehabilitation Psychology, 46, 382–399. https://doi.org/10.1037//0090-5550.46.4.382 Hudson, P., & Levasseur, K. (2002). Supporting foster parents: Caring voices. Child Welfare, 81, 853–877. Retrieved from https://search-ebscohost- com.proxy2.cl.msu.edu/login.aspx?direct=true&db=ssf&AN=507786326&scope=site Hupcey, J. E. (1998). Clarifying the social support theory-research linkage. Journal of Advanced Nursing, 27, 1231–1241. https://doi.org/10.1046/j.1365-2648.1998.01231.x Imenda, S. (2014). Is there a conceptual difference between theoretical and conceptual frameworks? Journal of Social Science, 38, 195-105. Julien-Chinn, F. J., Cotter, K. L., Piel, M. H., Geiger, J. M., & Lietz, C. A. (2017). Examining risk, strengths, and functioning of foster families: Implications for strengths-based practice. Journal of Family Social Work, 20, 306-321. https://doi.org/10.1080/10522158.2017.1348111 Kahn, R. L. (1979). Aging and social support. In M. W. Riley (Ed.), Aging from Birth to Death: Interdisciplinary Perspectives (pp. 77-91). Boulder, CO: Westview Press. Kangas, S. (2015). FY2015 boilerplate report section 583: Foster care homes report. [Memorandum]. Lansing, MI: Department of Human Services. Retrieved from https://www.michigan.gov/documents/dhs/Section_583_480505_7.pdf Kasin, C., Muñoz, K., Ong, C., Whicker, J., & Twohig, M. (2020). Well-being of parents of children who are deaf or hard of hearing. The Journal of Early Hearing Detection and Intervention, 2020(1), 86–97. Keyes, C. L. M. (1998). Social well–being. Social Psychology Quarterly, 61, 121–140. 183 Keyes, C. L. M. (2002). The Mental Health Continuum : From languishing to flourishing in life. Journal of Health and Social Behavor, 43(2), 207–222. Keyes, C. L. M. (2009). Brief description of the mental health continuum short form (MHC-SF). Available: http://www.sociology.emory.edu/ckeyes/. [On–line, retrieved March 18, 2019]. Kirton, D. (2001). Love and money: payment, motivation and the fostering task. Child and Family Social Work, 6, 199–208. https://doi.org/10.1046/j.1365-2206.2001.00208.x Kwok, S., & Wong, D. (2000). Mental health of parents with young children in Hong Kong: The roles of parenting stress and parenting self-efficacy. Child and Family Social Work, 5, 57– 65. Lakey, B., & Cohen, J. L. (2007). Social support assessment. In S. Ayers, A. Baum, C. McManus, & et. al. (Eds.), Cambridge handbook of psychology, health and medicine (2nd ed.). Cambridge, UK: Cambridge University Press. Retrieved from https://login.ezproxy.hope.edu/login?url=https://search.credoreference.com/content/entry/ cupphm/social_support_assessment/0?institutionId=3628 Lanigan, J. D., & Burleson, E. (2017). Foster parent’s perspectives regarding the transition of a new placement into their home: An exploratory study. Journal of Child and Family Studies, 26, 905–915. https://doi.org/10.1007/s10826-016-0597-0 Lazarus, R.S. & Folkman, S. (1984). Stress, Appraisal, and Coping. New York, NY: Springer Publishing Company, Inc. Leathers, S. J., Spielfogel, J. E., Geiger, J., Barnett, J., & Vande Voort, B. L. (2019). Placement disruption in foster care: Children’s behavior, foster parent support, and parenting experiences. Child Abuse & Neglect, 91, 147–159. https://doi.org/10.1016/j.chiabu.2019.03.012 Leslie, L. K., Gordon, J. N., Ganger, W., & Gist, K. (2002). Developmental delay in young children in child welfare by initial placement type. Infant Mental Health Journal, 23, 496–516. https://doi.org/10.1002/imhj.10030 Lietz, C. A., Julien-Chinn, F. J., Geiger, J. M., & Piel, M.H. (2016). Cultivating resilience in families who foster: Understanding how families cope and adapt over time. Family Process, 55(4), 660–672. https://doi.org/10.1111/famp.12239 Lohaus, A., Chodura, S., Möller, C., Symanzik, T., Ehrenberg, D., Job, A., Heinrichs, N. (2017). Children’s mental health problems and their relation to parental stress in foster mothers and fathers. Child and Adolescent Psychiatry and Mental Health, 11, 43. https://doi.org/10.1186/s13034-017-0180-5 184 MacGregor, T.E., Rodger, S., Cummings, A.L., & Leschied, A.W. (2006). The needs of foster parents: A qualitative study of motivation, support, and retention. Qualitative Social Work, 5, 351–368. https://doi.org/10.1177/1473325006067365 Maclay, F., Bunce, M., & Purves, D. (2006). Surviving the system as a foster carer. Adoption & Fostering, 3, 29–38. https://doi.org/10.1177/030857590603000105 Marcellus, L. (2010). Supporting resilience in foster families: A model for program design that supports recruitment, retention, and satisfaction of foster families who care for infants with prenatal substance exposure. Child Welfare, 89, 7–29. ProQuest. Web. 27 Sep. 2018. Masten, A. (2014). Ordinary Magic: Resilience in Development. New York, NY: Guilford Publications. Mckeough, A., Bear, K., Jones, C., Thompson, D., Kelly, P. J., & Campbell, L. E. (2017). Foster carer stress and satisfaction: An investigation of organisational, psychological and placement factors. Children and Youth Services Review, 76, 10–19. https://doi.org/10.1016/j.childyouth.2017.02.002 McStay, R. L., Dissanayake, C., Scheeren, A., Koot, H. M., & Begeer, S. (2013). Parenting stress and autism: The role of age, autism severity, quality of life and problem behaviour of children and adolescents with autism. Autism. 18(5), 502-510. https://doi.org/10.1177/1362361313485163 Michigan Bureau of Children and Adult Licensing. (2015). Licensing Rules for Child Placing Agencies. State of Michigan Department of Health and Human Services. Retrieved from https://www.michigan.gov/documents/dhs/BCAL-PUB-11_216515_7.pdf Michigan Department of Health and Human Services. (2014). Children’s Foster Care Manual. State of Michigan Department of Health and Human Services. Retrieved from https://dhhs.michigan.gov/OLMWEB/EX/FO/Public/FOM/905- 3.pdf#pagemode=bookmarks Middleton, J. S., & Potter, C. C. (2015). Relationship between vicarious traumatization and turnover among child welfare professionals. Journal of Public Child Welfare, 9, 195–216. https://doi.org/10.1080/15548732.2015.1021987 Miller, A. E., Green, T. D., & Lambros, K. M. (2019). Foster parent self-care: A conceptual model. Children and Youth Services Review, 99, 107–114. https://doi.org/10.1016/j.childyouth.2019.01.014 Miller, J. J., Benner, K., Pope, N., Dumas, T., Damron, L. J., Segress, M., Slone, M., Thrasher, S., Niu, C. (2017). Conceptualizing effective foster parent mentor programs: A participatory planning process. Child and Youth Services Review, 73, 411–418. https://doi.org/10.1016/j.childyouth.2017.01.004 185 Mitchell, R. E., & Trickett, E. J. (1980). Social networks as mediators of social support: An analysis of the effects and determinants of social networks. Community Mental Health Journal, 16(1), 27–44. https://doi.org/10.1007/BF00780665 Morgan, K., & Baron, R. (2011). Challenging behaviour in looked after young people, feelings of parental self-efficacy and psychological well-being in foster carers. Adoption & Fostering, 35, 18–32. https://doi.org/10.1177/030857591103500104 Moyers, S., Farmer, E., & Lipscombe, J. (2006). Contact with family members and its impact on adolescents and their foster placements. British Journal of Social Work, 36, 541–559. https://doi.org/10.1093/bjsw/bch270 Murray, L., Tarren-Sweeney, M., & France, K. (2011). Foster carer perceptions of support and training in the context of high burden of care. Child and Family Social Work, 16, 149– 158. https://doi.org/10.1111/j.1365-2206.2010.00722.x National Foster Parent Association. (n.d.). History of Foster Care in the United States. Retrieved from https://nfpaonline.org/page-1105741 National Survey of Child and Adolescent Well-Being. (2007). Foster Children’s Caregivers and Caregiving Environments. Washington, DC.: Office of Planning, Research and Evaluation, Administration for Children and Families, Department of Health and Human Services. Retrieved from https://www.acf.hhs.gov/sites/default/files/opre/caregivers.pdf Nixon, S. (1997). The limits of support in foster care. British Journal of Social Work, 27, 913– 930. https://doi.org/10.1093/oxfordjournals.bjsw.a011285 Octoman, O., & Mclean, S. (2014). Challenging behaviour in foster care: what supports do foster carers want? Adoption & Fostering, 38, 149–158. https://doi.org/10.1177/0308575914532404 Oke, N., Rostill-Brookes, H., & Larkin, M. (2011). Against the odds: Foster carers’ perceptions of family, commitment and belonging in successful placements. Clinical Child Psychology and Psychiatry, 18, 7–24. https://doi.org/10.1177/1359104511426398 Oosterman, M., Schuengel, C., Slot, N. W., Bullens, R. A. R., & Doreleijers, T. A. H. (2006). Disruptions in foster care: A review and meta-analysis. Children and Youth Services Review, 29, 53–76. https://doi.org/10.1016/j.childyouth.2006.07.003 Orme, J. G., Cherry, D. J., & Rhodes, K. W. (2006). The Help with Fostering Inventory. Children and Youth Services Review, 28, 1293–1311. https://doi.org/10.1016/j.childyouth.2006.02.001 Orme, J. G., & Combs-Orme, T. (2014). Foster parenting together: Foster parent couples. Children and Youth Services Review, 36, 124–132. https://doi.org/10.1016/j.childyouth.2013.11.017 186 Pallant, J. (2013). SPSS Survival Manual: A Step-by-Step Guide to Data Analysis Using IBM SPSS, 5th ed. New York, NY: McGraw-Hill. Pecora, P. J., Kessler, R. C., Williams, J., O’Brien, K., Downs, A. C., English, D., White, J., Hiripi, E., White, C. R., Wiggins, T., & Holmes, K. (2005). Improving Family Foster Care: Findings from the Northwest Foster Care Alumni Study. Seattle, WA: Casey Family Programs. Retrieved from https://caseyfamilypro-wpengine.netdna- ssl.com/media/AlumniStudies_NW_Report_FR.pdf Pickin, L., Brunsden, V., & Hill, R. (2011). Exploring the emotional experiences of foster carers using the Photovoice technique. Adoption & Fostering, 35(2), 61–75. https://doi.org/10.1177/030857591103500207 Piel, M. H., Geiger, J. M., Julien-Chinn, F. J., & Lietz, C. A. (2017). An ecological systems approach to understanding social support in foster family resilience. Child & Family Social Work, 22, 1034–1043. http://doi.org/10.1111/cfs.12323 Preisler, J. (2013). Being safe vs. feeling safe. Fostering Perspectives, 17(2), 1–6. Retrieved from http://fosteringperspectives.org/fpv17n2/psychological- safety.html%5Cnhttp://fosteringperspectives.org/fpv17n2/psychological-safety.html Proeschold-Bell, R. J., Molokwu, N. J., Keyes, C. L. M., Sohail, M. M., Eagle, D. E., Parnell, H. E., Kinghorn, W. A., Amanya, C., Vann, V., Madan, I., Biru, B. M., Lewis, D., Dubie, M. E., & Whetten, K. (2019). Caring and thriving: An international qualitative study of caregivers of orphaned and vulnerable children and strategies to sustain positive mental health. Children and Youth Services Review, 98, 143–153. https://doi.org/10.1016/j.childyouth.2018.12.024 Quittner, A. L., Glueckauf, R. L., & Jackson, D. N. (1990). Chronic parenting stress: Modarating versus mediating effects of social support. Journal of Personality and Social Psychology, 59, 1266–1278. https://doi.org/10.1037/0022-3514.59.6.1266 Randle, M., Ernst, D., Leisch, F., & Dolnicar, S. (2017). What makes foster carers think about quitting? Recommendations for improved retention of foster carers. Child & Family Social Work, 22, 1175–1186. https://doi.org/10.1111/cfs.12334 Rhodes, K. W., Orme, J. G., & Buehler, C. (2001). A comparison of family foster parents who quit, consider quitting, and plan to continue fostering. Social Service Review, 75(1), 84–114. Rhodes, K. W., Orme, J. G., Cox, M. E., Buehler, C., Rhodes, K. W., Orme, J. G., & Cox, M. E. (2003). Foster family resources, psychosocial functioning, and retention. Social Work Research, 27(3), 135–150. https://doi.org/10.1093/swr/27.3.135 Richardson, E. W., & Futris, T. G. (2019). Foster caregivers’ marital and coparenting relationship experiences: A dyadic perspective. Family Relations, 68(2), 185–196. https://doi.org/10.1111/fare.12354 187 Richardson, E. W., Futris, T. G., Mallette, J. K., & Campbell, A. (2018). Foster mothers’ parenting stress and coparenting quality: An examination of the moderating role of support. Children and Youth Services Review, 89, 77–82. https://doi.org/10.1016/j.childyouth.2018.04.024 Ridings, L. E., Beasley, L. O., Bohora, S., Espeleta, H. C., & Silovsky, J. F. (2018). The role of social support on depression among vulnerable caregivers reporting bidirectional physical violence. Journal of Interpersonal Violence, 1–23. https://doi.org/10.1177/0886260518767913 Robinson, S., & Weiss, J. A. (2020). Examining the relationship between social support and stress for parents of individuals with autism. Research in Autism Spectrum Disorders, 74, 1–10. https://doi.org/10.1016/j.rasd.2020.101557 Rodger, S., Cummings, A., & Leschied, A. W. (2006). Who is caring for our most vulnerable children?: The motivation to foster in child welfare. Child abuse & neglect, 30, 1129-1142. https://doi.org/10.1016/j.chiabu.2006.04.005 Ryff, C. D. & Keyes, C. L. M. (1995). The structure of psychological well–being revisited. Journal of Personality and Social Psychology, 69, 719–727. Samrai, A., Beinart, H., & Harper, P. (2011). Exploring foster carer perceptions and experiences of placements and placement support. Adoption & Fostering, 35(3), 38–49. https://doi.org/10.1177/030857591103500305 Sarason, I. G., & Sarason, B. R. (2009). Social support: Mapping the construct. Journal of Social and Personal Relationships, 26(1), 113-120. https://doi.org/10.1177/0265407509105526 Schaefer, C., Coyne, J. C., & Lazarus, R. S. (1981). The health-related functions of social support. Journal of Behavioral Medicine, 4, 381-406. Schofield, G., & Beek, M. (2005). Risk and Resilience in Long-Term Foster-Care. British Journal of Social Work, 35, 1283-1301. https://doi.org/10.1093/bjsw/bch213 Schofield, G., Beek, M., Ward, E., & Biggart, L. (2013). Professional foster carer and committed parent: role conflict and role enrichment at the interface between work and family in long-term foster care. Child & Family Social Work, 18(1), 46–56. https://doi.org/10.1111/cfs.12034 Shapiro, D. N., & Stewart, A. J. (2011). Parenting stress, perceived child regard, and depressive symptoms among stepmothers and biological mothers. Family Relations, 60(5), 533-544. Shklarski, L. (2019). Foster parent skills and dilemmas: A qualitative study. Child Welfare, 97, 41–62. Retrieved from https://search-proquest- com.proxy2.cl.msu.edu/docview/2264892910/fulltextPDF/51D64175A53347C3PQ/1?acco untid=12598 188 Soliday, E., McCluskey-Fawcett, K., & Meck, N. (1994). Foster mothers’ stress, coping, and social support in parenting drug-exposed and other at-risk toddlers. Children’s Health Care, 23, 15–32. https://doi.org/http://dx.doi.org/10.1207/s15326888chc2301_2 Strand, V. C., & Bosco-Ruggiero, S. (2009). Initiating and sustaining a mentoring program for child welfare staff. Administration in Social Work, 34(1), 49–67. https://doi.org/10.1080/03643100903432941 Tarren-Sweeney, M. (2008). The mental health of children in out-of-home care. Current Opinion in Psychiatry, 21, 345–349. https://doi.org/10.1097/YCO.0b013e32830321fa Thomson, L., & McArthur, M. (2009). Who’s in our family: An application of the theory of family boundary ambiguity to the experiences of former foster carers. Adoption & Fostering, 33, 68–79. https://doi.org/10.1177/030857590903300107 Tilden, V. P. (1986). New perspectives on social support. The Nurse Practitioner, 11, 60–61. https://doi.org/10.1097/00006205-198608000-00007 Tilden, V. P., & Galyen, R. D. (1987). Cost and conflict: The darker side of social support. Western Journal of Nursing Research, 9(1), 9–18. https://doi.org/10.1177/019394598700900103 Timmer, S. G., Sedlar, G., & Urquiza, A. J. (2004). Challenging Children in Kin Versus Nonkin Foster Care: Perceived Costs and Benefits to Caregivers. Child Maltreatment, 9, 251– 262. https://doi.org/10.1177/1077559504266998 Tunno, A. M. (2015). Assessing foster parent factors that predict placement disruption of youth in foster care (Doctoral dissertation). University of Kansas. Turney, K., & Wildeman, C. (2016). Mental and physical health of children in foster care. Pediatrics, 138(5). https://doi.org/10.1542/peds.2016-1118 Underwood, P.W. (2000). Social support: The promise and the reality. In V.H. Hill (Ed.), Handbook of Stress, Coping, and Health: Implications for Nursing Research, Theory, and Practice. Thousand Oaks, CA: Sage Publications, Inc. U.S. Children’s Bureau. (n.d.-a) Child and family services review quick reference items list. Washington, DC: U.S. Department of Health and Human Services. U.S. Children’s Bureau. (n.d.-b) Home study requirements for prospective foster parents. Washington, DC: U.S. Department of Health and Human Services. U.S. Children’s Bureau. (2013). The AFCARS report, No. 20. Washington, D.C.: U.S. Department of Health and Human Services, Administration for Children and Families, Administration on Children, Youth and Families, Children's Bureau. 189 U.S. Children’s Bureau. (2015). The AFCARS report, No. 22. Washington, D.C.: U.S. Department of Health and Human Services, Administration for Children and Families, Administration on Children, Youth and Families, Children's Bureau. U.S. Children’s Bureau. (2019). The AFCARS report, No. 26. Washington, D.C.: U.S. Department of Health and Human Services, Administration for Children and Families, Administration on Children, Youth and Families, Children's Bureau. U.S. Children’s Bureau. (2020). Child and Family Services Reviews Aggregate Report. Washington, DC: U.S. Department of Health and Human Services, Administration for Children and Families, Administration on Children, Youth, and Families, Children’s Bureau. Retrieved from https://www.acf.hhs.gov/sites/default/files/cb/cfsr_aggregate_report_2020.pdf U.S. Department of Health and Human Services, Administration for Children, Youth and Families. (November 2001). National Survey of Child and Adolescent Well-Being: One Year in Foster Care Report. Washington, D.C. Vanderfaeillie, J., Van Holen, F., De Maeyer, S., Gypen, L., & Belenger, L. (2016). Support Needs and Satisfaction in Foster Care: Differences Between Foster Mothers and Foster Fathers. Journal of Child and Family Studies, 25, 1515–1524. https://doi.org/10.1007/s10826-015-0320-6 Vanderfaeillie, J., van Holen, F., Trogh, L., & Andries, C. (2012). The impact of foster children’s behavioural problems on Flemish foster mothers’ parenting behaviour. Child and Family Social Work, 17(1), 34–42. https://doi.org/10.1111/j.1365-2206.2011.00770.x Van Holen, F., Vanderfaeillie, J., Vanschoonlandt, F., De Maeyer, S., & Stroobants, T. (2015). Explorative study into support needs of caregivers in short-term foster care regarding problem behaviour and dealing with biological parents Exploratief onderzoek naar de ondersteuningsbehoeften betreffende gedragsproblemen en omgang met biologische ouders van pleegouders in korte-termijn pleegzorg. European Journal of Social Work, 18, 97–113. https://doi.org/10.1080/13691457.2014.883365 Vanschoonlandt, F., Vanderfaeillie, J., van Holen, F., De Maeyer, S., & Robberechts, M. (2013). Parenting stress and parenting behavior among foster mothers of foster children with externalizing problems. Children and Youth Services Review, 35, 1742–1750. https://doi.org/10.1016/j.childyouth.2013.07.012 Vanschoonlandt, F., Van Holen, F., Vanderfaeillie, J., De Maeyer, S., & Andries, C. (2014). Flemish foster mothers’ perceptions of support needs regarding difficult behaviors of their foster child and their own parental approach. Child and Adolescent Social Work Journal, 31(1), 71–86. https://doi.org/10.1007/s10560-013-0310-8 Vaux, A. (1988). Social Support: Theory, Research, and Intervention, 3rd ed. New York: Praeger Publishers. 190 Vaux, A. (1992). Assessment of social support. In H.O.F. Veiel & U. Baumann (Eds.), The Meaning and Measurement of Social Support, (pp. 193-216). New York, NY: Hemisphere Publishing Corporation. Walsh, F. (2016). Strengthening family resilience, 3rd ed. New York, NY: Guilford Publications. Weinman, J., Wright, S. & Johnston, M. (1995). Measures in Health Psychology: A User’s Portfolio. Windsor: Nfer-Nelson. Weiss, R. S. (Ed.) (1974). The provisions of social relationships. In Z. Rubin (Ed.), Doing Unto Others (pp. 17-26). Englewood Cliffs, NJ: Prentice Hall. Whenan, R., Oxlad, M., & Lushington, K. (2009). Factors associated with foster carer well- being, satisfaction and intention to continue providing out-of-home care. Children and Youth Services Review, 31, 752–760. https://doi.org/10.1016/j.childyouth.2009.02.001 Wilke, D. J., Rakes, S., & Randolph, K. (2019). Predictors of early departure among recently hired child welfare workers. Social Work, 64, 188–197. Retrieved from https://watermark.silverchair.com/swz020.pdf?token=AQECAHi208BE49Ooan9kkhW_Erc y7Dm3ZL_9Cf3qfKAc485ysgAAAkEwggI9BgkqhkiG9w0BBwagggIuMIICKgIBADCCA iMGCSqGSIb3DQEHATAeBglghkgBZQMEAS4wEQQMKQOPj0050gc_mxklAgEQgIIB 9B4Eri_CZ_110qeXxWE6BVGYJpkzunz-Pz8wjH1GiF9bY_4d Wilson, K., Sinclair, I., & Gibbs, I. (2000). The trouble with foster care: The impact of stressful “events” on foster carers. British Journal of Social Work, 30, 193–209. https://doi.org/10.1093/bjsw/30.2.193 Zimet, G. D., Dahlem, N. W., Zimet, S. G., & Farley, G. K. (1988). The Multidimensional Scale of Perceived Social Support. Journal of Personality Assessment, 52(1), 30–41. https://doi.org/10.1207/s15327752jpa5201_2 Zinn, A. (2009). Foster family characteristics, kinship, and permanence. Social Service Review, 83, 185–219. Retrieved from http://www.journals.uchicago.edu/t-and-c 191