a? . 61.; :51... .. 4 5-.....2 ’35.. 1 :15 ... guitar. 5 ‘ _ . .... Ex. affiniw. z anew... ,; u .363.- .31 . < is? .. s . I .t. .v ”5.5. am~ . flaunt .1 tun". .. . $5.33.. a; $3.6... 7.. . ‘13:... :S xflufiafi...i. .9. .5. b 1.: I).?..Zos its. 1.: 1...; . .3333?! I: 3.3 3...: Illasa im.t\‘. in... 5: : ..5ubx:.... 2:32.... xiii—m ...:......:.L¢. innit. 9.1.18, 3 a“; §01$3&3 23:5. .SiiIluu .1! g Eats! ‘151’1. r .6 .1}... 1...}: .111 3123) II- it‘ .9): xviivéol: In. an .‘X L‘n 0.3.1.113! I.lr|].>:¥ IlllllllllllllllllllIllHllllillllIll!HIIHIIHIHIIUHIWI 193 02048 8437 LIB RARY Michigan State University This is to certify that the dissertation entitled THE INFLUENCE OF A PARAPROF‘ESSIONAL, HOME VISITATION PARENT EDUCATION PROGRAM ON THE SOCIAL SUPPORT AND PARENTING BEHAVIORS OF LIMITED RESOURCE PARENTS presented by Dawn A. Contreras has been accepted towards fulfillment of the requirements for Ph.D. degree in Family and Child Ecology Date May 1. 2000 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE DEQQGZOO} 1;; l D U " : Q5 1192’ ”2307 APR 2 02910 11/00 mm.“ THE INFLUENCE OF A PARAPROFESSIONAL, HOME VISITATION PARENT EDUCATION PROGRAM ON THE SOCIAL SUPPORT AND PARENTING BEHAVIORS OF LIMITED RESOURCE PARENTS BY Dawn A. Contreras A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Family and Child Ecology 2000 Pit ITIII ex; as re p; (n ABSTRACT THE INFLUENCE OF A PARAPROFESSIONAL, HOME VISITATION PARENT EDUCATION PROGRAM ON THE SOCIAL SUPPORT AND PARENTING BEHAVIORS OF LIMITED RESOURCE PARENTS By Dawn A. Contreras Previous studies examining the effects of home visitation programs have found mixed results. The current study sought to expand upon previous research by examining the potential effects of home-based parent education on the quality and quantity of parental social support and parenting behaviors of limited resource parents. Relationships between social support and parenting behaviors were also tested. This study involved 63 parents in a treatment group and 60 parents in a comparison group. Results of ANCOVA analyses found significant differences at the .05 level of probability for the variables of satisfaction with social support and parenting behaviors. ANCOVA analyses did not find significant differences at the .05 level of probably for the variable of amount of social support. Pearson Product Moment Correlations were used to assess the relationships between social support and parenting behaviors. This analysis suggested that there might be a relationship between satisfaction with social support and parenting behaviors. Three out of four Pearson Product Moment correlations were significant at the .05 level of probability. Pearson Product Moment correlations did not show a relationship between amounts of social support and parenting behaviors. The implications of these findings are discussed. ACKNOWLEDGEMENTS I would like to thank my committee, Dr. June Youatt, Dr. Norma Bobbitt, Dr. Tom Luster and Dr. Ellen Whipple for their support, helpful feedback and encouragement throughout this process. I would particularly like to thank Dr. Youatt, who has served as an invaluable mentor and teacher throughout the years. I would also like to thank Michigan State University Extension and the Building Strong Families staff for partnering with me in this research. A special thanks to Dr. Karen Shirer for her support and Dr. Dawn Koger for her guidance and encouragement. Further, I would like to express appreciation and love to my family and friends. My husband, Joseph, Sr., gave me fresh inspiration and constant encouragement when the process seemed long. My father, Dr. George Harris and mother, Shirlene Harris, gave to me a love for Ieaming and a lifelong example of self-improvement. Other family members and friends were more than eager to help whenever possible. Finally, I would like to thank God for making all of this possible. TABLE OF CONTENTS CHAPTER 1 .......................................................................................................... 1 INTRODUCTION ................................................................................................ 1 Statement of the Problem ............................................................................... 6 Purpose of the Study ....................................................................................... 9 Significance of the Study ............................................................................... 11 Theoretical Framework ................................................................................. 12 Conceptual Map ............................................................................................ 17 Constructs and Definitions ............................................................................ 21 Research Questions ...................................................................................... 28 Assumptions .................................................................................................. 28 Hypotheses ................................................................................................... 30 Limitations ..................................................................................................... 31 Description of the Intervention ....................................................................... 34 CHAPTER II ........................................................................................................ 39 LITERATURE REVIEW ................................................................................... 39 Social Support Defined within the Context of Parenting Behaviors ............... 39 Providing Social Support Through Home-Based Parent Education .............. 55 Literature Review Conclusions ...................................................................... 72 CHAPTER III ....................................................................................................... 74 METHODS ....................................................................................................... 74 Research Design ........................................................................................... 74 Data Collection Procedures ........................................................................... 75 Research Sample .......................................................................................... 80 Instrumentation ........................................................................................... 101 Data Analysis .............................................................................................. 106 Ethical Considerations ................................................................................ 108 CHAPTER N .................................................................................................... 110 RESULTS ...................................................................................................... 1 10 Descriptive Analyses: .................................................................................. 1 10 Inferential Statistics: .................................................................................... 123 Main Hypothesis: Changes in satisfaction with social support following an intervention .................................................................................................. 123 Sub-hypothesis 1: Analysis of changes in overall amount of social support following an intervention .............................................................................. 129 Sub-hypothesis 2: Analysis of changes in parenting practices following an intervention .................................................................................................. 130 Sub-hypothesis 3: Analysis of the relationship between perceived satisfaction with social support and parenting behaviors. .............................................. 135 Sub-hypothesis 4: Analysis of the relationship between amount of social support and parenting behaviors. ................................................................ 137 CHAPTER V ..................................................................................................... 145 DISCUSSION OF RESULTS, CONCLUSIONS AND IMPLICATIONS FOR FUTURE RESEARCH ................................................................................... 145 BIBLIOGRAPHY ............................................................................................... 166 APPENDICES ................................................................................................... 175 APPENDIX A - INVITATION LETTER AND INTEREST FORM .................................. 176 APPENDIX B - FOLLOW-UP LETTER ............................................................... 179 APPENDIX C - FAMILY RECORD FORM ............................................................ 182 APPENDIX D - FAMILY SUPPORT FORM .......................................................... 184 APPENDIX E - HOME INVENTORY .................................................................. 186 APPENDIX F - PARENTING BEHAVIOR ASSESSMENT SCORING FORM ................ 188 APPENDIX G - UCRIHS APPROVAL LETTER ................................................... 190 APPENDIX H - PARENT CONSENT FORM - TREATMENT GROUP ........................ 192 APPENDIX I - PARENT CONSENT FORM - COMPARISON GROUP ....................... 194 vi TA; Ti—f TA T; TA TR TA TA TA TA T; T; T: Tr LIST OF TABLES TABLE 1: DEMOGRAPHIC CHARACTERISTICS OF ALL PARENTS ...................................... 84 TABLE 2: DEMOGRAPHIC CHARACTERISTICS OF ALL PARENTS ...................................... 86 TABLE 3: DEMOGRAPHIC CHARACTERISTICS OF PARENTS WHO COMPLETED THE STUDY .................................................................................................................................... 92 TABLE 4: DEMOGRAPHIC CHARACTERISTICS OF PARENTS WHO COMPLETED THE STUDY .................................................................................................................................... 93 TABLE 5: COMPARISON OF PRETEST SCORES FOR PERCEIVED SATISFACTION WITH SOCIAL SUPPORT ....................................................................................................... 99 TABLE 6: COMPARISON OF PRETEST SCORES FOR AMOUNT OF SOCIAL SUPPORT ....... 99 TABLE 7: COMPARISON OF PRETEST SCORES FOR PARENTING BEHAVIORS (PBA)... 100 TABLE 8: COMPARISON OF PRETEST SCORES FOR PARENTING BEHAVIORS (HOME) 100 TABLE 9: RATINGS OF SATISFACTION WITH SOCIAL SUPPORT FOR TREATMENT GROUP PARENTS .................................................................................................................. 112 TABLE 10: RATINGS OF SATISFACTION WITH SOCIAL SUPPORT FOR COMPARISON GROUP PARENTS ..................................................................................................... 113 TABLE 11: AMOUNT OF SOCIAL SUPPORT FOR TREATMENT GROUP PARENTS ............ 114 TABLE 12: AMOUNT OF SOCIAL SUPPORT FOR COMPARISON GROUP PARENTS ......... 117 TABLE 13: FREQUENCY OF PBA PARENTING BEHAVIORS (TREATMENT GROUP) ....... 119 TABLE 14: FREQUENCY OF PBA PARENTING BEHAVIORS (COMPARISON GROUP) 120 TABLE 15: FREQUENCY OF HOME PARENTING BEHAVIORS (TREATMENT GROUP)... 122 TABLE 16: FREQUENCY OF HOME PARENTING BEHAVIORS (COMPARISON GROUP). 123 vii T1 T1 Ti T; TA TA TABLE 17: EQUALITY OF GROUP MEANS FOR PERCEIVED SATISFACTION WITH SOCIAL SUPPORT .................................................................................................................. 125 TABLE 18: PAIRED T-TEST ANALYSIS OF PERCEIVED SATISFACTION WITH SPECIFIC SOURCES OF SOCIAL SUPPORT (TREATMENT GROUP) .......................................... 128 TABLE 19: EQUALITY OF GROUP MEANS FOR AMOUNT OF SOCIAL SUPPORT ............. 130 TABLE 20: EQUALITY OF GROUP MEANS FOR PARENTING BEHAVIORS (PBA) ........... 132 TABLE 21: EQUALITY OF GROUP MEANS FOR PARENTING BEHAVIORS (HOME) ...... 133 TABLE 22: PAIRED T-TEST ANALYSIS OF HOME SUB-SCALE (TREATMENT GROUP). 135 TABLE 23: CORRELATIONS OF SATISFACTION WITH SOCIAL SUPPORT AND PARENTING BEHAVIORS (PRE-INTERVENTION) ........................................................................... 136 TABLE 24: CORRELATIONS OF SATISFACTION WITH SOCIAL SUPPORT AND PARENTING BEHAVIORS (POST-INTERVENTION) ......................................................................... 137 TABLE 25: CORRELATIONS OF AMOUNT OF SOCIAL SUPPORT AND PARENTING BEHAVIORS (PRE-INTERVENTION) ........................................................................... 138 TABLE 26: CORRELATIONS OF AMOUNT OF SOCIAL SUPPORT AND PARENTING BEHAVIORS (POST-INTERVENTION) ......................................................................... 138 viii LIST OF FIGURES 121951351. DETERMINANTS OF PARENTING (BELSKY, 1984). ..................................... 13 Emma]. THE INFLUENCE PARENT EDUCATION ON PARENTING BEHAVIORS AND SOCIAL SUPPORT. ........................................................................................... 17 EIQUBEB. CONCEPTUAL MAP OF THE INFLUENCE OF PARENT EDUCATION ON PARENTING BEHAVIORS AND SOCIAL SUPPORT. ................................................. 18 EIQUBEA. OPERATIONAL MAP. ............................................................................... 21 13951355. PERSON-PROCESS-CONTEXT MODEL FOR HOME-VISITATION PROGRAM DELIVERY. ...................................................................................................... 59 EIGLLBEfi. RESEARCH DESIGN. ............................................................................... 75 CHAPTER 1 INTRODUCTION Optimally, parenting is a task that is done communally (Dam and Harding, 1999). Multiple sources of support give parents the assistance they need to carry out the job Of raising their Offspring. “Since the first parent-child dyad, adult caregivers have struggled with the demands presented by their children” (Daro and Harding, 1999, p.152). In response, parents have relied upon the support Of the human and material resources found within the Spheres Of their ecosystem. This has Often included informal sources of support, such as aid from a partner or spouse, kin, friends, neighbors, and CO-workers. These resources augmented the knowledge and Skills of parents by providing assistance. Social support is defined as the emotional, informational, and instrumental assistance that other people give to parents (Crockenberg, 1988). In recent times, however, a quickly changing society has stripped away many of the informal supports traditionally relied upon by parents to assist them in caregiving. Current societal conditions have left many parents bereft Of the support needed to meet the complex and taxing demands Of parenting. The reduction Of social support for parents is due to many factors, including greater mobility of families, smaller family sizes, and greater societal restrictions on informal exchanges (Carter, 1996; Crockenberg, 1988; Cochran, 1993; Webster-Stratton, 1997). Societal restrictions include such ecological factors as unsafe neighborhoods, and time constraints placed on dual-earner families (Cochran, 1993). Mobility has increased across all segments of our society. Economic and public policy changes have forced many people to seek employment that may cause them to commute for long distances or relocate. This phenomenon has Often served to isolate parents from extended family members and friends who have traditionally provided informal support. Changing household compositions have tended to produce smaller family Sizes. This Often equates to fewer people in the family to assist in the caregiving role. The Single parent household is a more common occurrence than ever before. In 1995 the percentage of Children in Michigan living in a Single parent household was twenty-eight percent. This was a higher rate than the national average Of twenty-six percent of Children living in a Single-parent household (Kids Count Data Book, 1998). Although single parent households have always been a relatively common family form, the causes Of solitary parenting have changed (Smith, Cudaback, Goddard, and Myers-Walls, 1994). In the past, Single parent families were typically formed as a result Of a parent becoming a widow or widower. Now solitary parenting is more often attributed tO divorce or never being married. As a result of these changes, “single parents may be more isolated, and perhaps more disillusioned than the single parents Of the past” (Smith, Cudaback, Goddard, and Myers-Walls, 1994, p. 7). Of particular concern are low-income Single parents, many Of whom are mothers. “Socioeconomically disadvantaged mothers are more likely to report depression, isolation from family members and {)4 Vi a: he I. K neighbors and less support available from informal or formal networks” (Webster- Stratton, 1997, p. 157). Time restraints placed on parents due to employment Obligations is a third variable that can leave families isolated and unable to develop and utilize informal social support networks. Often neighbors dO not know each other. This diminishes the possibility of informal resource exchange between caregivers. Another key deterrent to a strong informal support base for parents is a society filled with Violence and crime. Parents cannot seek assistance from an environment that is unsafe. “Increasing violence and accompanying fears for one’s safety and that of one’s children lead many people to isolate themselves and become less involved in their communities, which in turn results in deteriorating social networks, lack Of social support for individuals and families, and greater isolation” (Webster-Stratton, 1997, p. 157). During this period of diminished informal social support some parents have turned to more formal sources Of support for assistance with parenting (Carter, 1996; Crockenberg, 1988). However, the formal method of traditional classroom instruction has not always been proven to be an effective means of changing parenting behavior (Webster-Stratton, 1997). Changes in parenting behavior typically occur when the information iS personalized and individual support is Offered (Kagan and Weissbourd, 1994). Just giving information to parents is not enough to cause changes in parenting practices. “Research suggests that programs which Simply provide information about parenting and child development, without interaction with families and provision of other nc he services and supports, are unlikely to change parenting practices and lead to enhanced Child development” (Weiss and Rittenburg, 1991, p. 5). In other words, parent education that is prescriptive and standard for all situations may not be received well by the parent. Instead parents appear to have greater increases in knowledge, confidence and Skills when the treatment combines flexibility with structure and appropriate dosage. These problems are compounded when working with audiences who have not been successful utilizing formal support systems in the past. Audiences who have been unsuccessful working with formal support systems in other endeavors may be either reluctant or unable to access these same systems with parenting concerns. Such may be the case with limited resource parents. Limited resource parents are defined as those parents who have low educational attainment and few monetary resources. “Past failures in school are a major barrier to participation in adult education programs. Individuals who View themselves as academic failures as children Often continue to believe they cannot succeed as adult students. Many have great difficulty Changing this perception. Unlike middle-Class participants who have experienced success and satisfaction in education, limited resource adults typically avoid participation in educational programs because it represents another potential opportunity for failure" (Pearson, 1995, p. 1). In addition to low educational attainment, limited resource parents also have few monetary resources. Lack of resources may prohibit potential learners from accessing formal sources of parent education because Of an inability to acquire transportation, child care, or program fees. Social support is an important element in parenting because it helps to boost parents’ emotional and physical ability to provide an Optimal caregiving environment for their children. Research has found that higher levels of social support can enhance parental coping skills, decrease parents’ psychological or physical pathology and provide relief from daily burdens that might otherwise accumulate to incapacitate the parent, or press him or her into inappropriate or even abusive parenting behaviors (Cochran and Niego, 1995; Cohen and Wills, 1988; Thompson, 1995). With an inability to access either formal or informal sources of parent education and support, limited resource parents are at a higher risk for parental dysfunction and Child maltreatment (Burke, Chandy, Dannerbeck and Wilson- Watt, 1998; Garbarino and Whittaker, 1983; Thompson, 1995). “Indeed studies have shown that economic deprivation combined with a lack Of social support creates an especially dangerous situation for children in terms of child abuse” (Webster-Stratton, 1997, p. 157). Home-visitation parent education programs have risen up in response to the lack Of support for limited resource parents of young children (Daro and Harding, 1999; Gomby, Culross, and Behrman, 1999). Conceptually, the home- visitation program is a treatment modality that can provide support to parents in a unique manner. By bringing supports to the parent’s environment, “home Visitors can see the environments in which families live, gain a better understanding of the families’ needs, and therefore tailor services to meet those needs. The relationships forged between home Visitors and parents can break through loneliness and isolation and serve as the first step in linking families to their communities” (Gomby, Culross, and Behrman, 1999, p. 5). In theory, home-visitation parenting programs appear to be a treatment modality that would effectively support parents Of young children in their caregiving responsibilities. TO date, empirical evidence has not been able to verify this proposition for home-based parent education programs. The results of studies assessing the effectiveness of home visitation programs have been conflictual. Studies have not been able to answer the critical research questions that are needed to determine if home Visiting is an effective service delivery method. There are still many questions regarding the optimal conditions for home-Visitation programs, including program protocol, such as the number, intensity, frequency and content of visits; staffing experience, supervision and training; recipient Characteristics and evaluation procedures (Gomby, Culross, and Behrman, 1999). Statement of the Problem Supporting parents in their role Of raising children has long been @Wrtant responsibility for our society (Carter, 1996). Most social support for parents traditionally came from informal means, such as family, friends, community organizations and church groups (Carter, 1996). These support groups provided parents with advice, physical assistance, and perspective. Wthepastthreedecades parents have found traditional forms of social support to be greatly diminished. Increasingly, parents have had to 6 SE Ll“ Ki' 8C en a?) me search for alternative forms of social support to assist them in the role Of raising their children. Mmfiethggsfgtproviding parent educationbave—beeefound to be ineffective for some parents, particularly limited resource parents of young children (Webster-Stratton, 1997). Traditional forms of parent education have tended to be formal in£style and behavioralistic in nature, focusing primarily on skill acquisition (First and Way, 1995). “Forlqw-ingme f 'ies articularly, parent management training programs n S more broadl on building M ‘,__-__/ ' community networks and parent support. As the number Of people whom low- income parents feel they can rely on for informal assistance increases and as they feel more satisfied with their social support, the more likely they are to be nurturing and positive in their parenting and the less likely they are to report problematic behavior, compared with low-income mothers who feel isolated and dissatisfied with their social support” (Webster-Stratton, 1997, p. 158). In order to provide support to limited resource parents, it has been posited ———_._.._ ——-—-—_ ,.,____ ._ [13:1 paraprofessional, home-visitation delivery of parent education can increase parental skills and act as a bridge to other kinds Of community assistance (Olds, Kitzman, Cole, and Robinson, 1997). (Hewever, not enough is known about the ‘ ”PM ”fig/F— actual effects $199999 parent educatiOn thrOUghthis delivery method; In fact, M -2___‘ empirical studies have shown mixed results for changes in parenting behaviors __'7 .p-......---— and minimal data has been presented on the effectiveness Of this delivery method to change parental social support networks. Studies Showing evidence in support of Changes in parenting behaviors have associated this delivery method with increased parent-child interaction (Dam and Harding, 1999), maternal parenting efficacy (Duggon, McFarlane, Windham, Rohde, Salkever, Fuddy, Rosenberg, Buchbinder and Sia, 1999; Luster, Perlstadt, McKinney, Sims and Juang, 1996); matemal-child attachment (Jacobson and Frye, 1991), child IQ (Blair, Ramey and Harden, 1995; Liaw, Meisels and Brooks-Gunn, 1995), maternal involvement (Boger, Richter, Humetz and Haas, 1986), better prenatal care and fewer pre—tenTI births (Rogers, Peoples-Sheps and Suchindran, 1996). Other studies have not found evidence to associate home-visitation parent education with changes in parenting outcomes (Barth, 1991; Choi, Berger and Flunn, 1997; Siegel, Bauman, Schaefer, Saunders and Ingram, 1980; Silver, lreys, Bauman and Stein, 1997). These mixed results create the need for additional researchassessing the effects E W of home-visitation programs on parenting behaVIOFS‘F ”The need for researchon the effects Of home-visitation programs on social support is evide? Currently, @MQELWWU aboutjheeffects ofafihom‘e- visitation parent education program on parental changes in social support network (Cochran and Niego, 1995; Webster-Stratton, 1997). “In the research literature, little reference is made to therapeutic strategies used to decrease families’ isolation Or promote their involvement or to build a sense of community, either within the group or the larger community. Indeed compared with our well- developed research methods of measuring changes in parenting behavior, new? \V studies report measuring outcomes having to do with social networks, parents' ~— Sense of support or their involvement in their Children’s education or school” (Webster-Stratton, 1997, p. 158). In keepingwith this_deficit_9_t§mpiri_car 993553 recent study of home-visitation programs found that social support changes were .- l’SeldOrhn—inCIUOTSII a: an investigative variable (Gomby, Culross, 8. Behrrnan, 1999)./ Purpose Of the Study The purpose of this research was to examine the effects Of a paraprofessional, home-Visitation parent education program on the social support and parenting behaviors of limited resource parents. Specifically, this research assessed whether or not social support levels increased and perceptions of positive parenting behaviors improved as a result Of participation in a home- based, parent education program. Analysis was conducted toldetermine what Changes occurred in the entire treatment group, as well as subgroups of the sample. This research also explored correlations between social support and parenting behaviors. The overall design Of the Building Strong Families (BSF) program was thought to contribute to the results of this investigation. Specifically, the program model was predicted to influence Changes in parenting behaviors and social support because of its paraprofessional, home-based design. First Of all, it was anticipated that the use of a paraprofessional instructor would help the parent form a positive relationship between the parent and instructor (Gomby, Culross and Behrrnan, 1999). It has been suggested that a relationship builds between the paraprofessional home visitor and parent because the home visitor is viewed as having a similar background as the parent and having faced similar challenges (Gomby, Culross and Behrrnan, 1999). The closeness between the parent and instructor promotes a sense Of trust and encourages the parent to disclose his or her needs and desires. The instructor is then able to present information that is customized to the specific issues within the family. This helps the parent accept the information and seek to incorporate it into his or her lifestyle. Secondly, the paraprofessional, home-based design was anticipated to contribute to the results of this study because it is believed to help the instructor serve as a role model for the parent (Gomby, Culross and Behrman, 1999). AS the home visitor role models the information that is taught in the curriculum, the parent is able to View the Skills being demonstrated. Parents are more comfortable practicing and Ieaming skills that they have seen demonstrated for them. A third way in which this paraprofessional, home-based design was anticipated to effect the results of the study was by the instructor serving as a form of social support for the parent until additional linkages could be built in his or her support network. The instructor provided emotional and informational support to the parent while parenting behaviors and social support levels were being increased. A final element speculated to contribute to the effectiveness Of the program was the content of the BSF program. The BSF materials are designed to meet the specific needs of a limited resource adult audience. They are written at a third-grade reading level and are graphically illustrated in four-color pictures. Also, they are written to build upon the strengths of the parent and allow the parent to contribute to the discussion. These curriculum attributes are thought to 10 build the parent’s confidence in his or her parenting ideas, while still gaining additional information. Further, the materials are specifically written to provide information on positive parenting behaviors and building social support networks. The flipcharts and videotapes give information and demonstrate how to build skills in these two areas. Experiential activities help the parent practice skills in building parenting behaviors and social support networks. For example, one unit of the BSF curriculum asks the parent to set a goal and work toward accomplishing the goal by contacting local resources to assist him or her. The activity suggests that local resources may include a neighbor, friend, or formal support services, such as a lawyer, doctor or agency person. Research has shown that a theoretically based program design, that incorporates relationship-building and education into home visits, is effective in creating change in parents (Olds, Henderson, Jr., Kitzman, Eckenrode, Cole, and Tatelbaum, 1999; Olds, Kitzman, Cole and Robinson, 1997). Based on this empirical evidence, the overall design Of the BSF program was posited to contribute to the effectiveness of the intervention to create Changes in parenting behaviors and social support in limited resource parents. Significance of the Study The findings Of this study _will be helpfliljnmrthejmjhiunderstanding of the influencegfw paraprofessional, home-visitation parent education programs on (I, the social support levels and parenting behaviorsofiflimited resource parents.__ “-mmr—t—uh—é. W guide the focus and direction_gfparent support - H—-——--u-.-.._. programs for limited resource parEIts. Currently administrators and staff in ___——”"/-~/ "w — ‘ "’ -' '——‘--""'""m '— _- _ __t_._ _ _. ___ ll Mpolrcy and programmrng systems are strugngecrsrons regardrng how ,_.,__..-——_. —-—_~.___,__'—————' _._._.—-' best to provide parenting information to parents. Important decisjgnsregarding staffing, program protocol and content can effect the budget and administration Of parent education programs. Paraprofessionals are less expensive to hire, but are they effective in helping parents change behaviors? Also, if paraprofessionals are used, what expenses need to be invested in training and supervision of staff? And finally, are home visits an effective method of increasing support for limited resource parents? [this paraprofessional home-Visitation program is found to be effective in increasing social support levels and positive parenting behaviors in low-income parents, then program administrators should consideriroadening their parent education program to include social support strengthening units. Program \ . ——-‘ planners may want to add such elements as training in interpersonal communication skills, problem-solving, and understanding ways to build support so that both within family support and external sources of support may be enhanced in the family unit (Webster-Stratton, 1997). ,tJl/timately, increasing K socialsupport levels in families would help buffer the negative influences Of poverty anidTOth-er life stressors on parenting Skills, decrease feelings of isolation and depression for parents and promote a positive environment for children to be raised. Theoretical Framework The foundation Of the theoretical framework for this study was Belsky’s Determinants Of Parenting (1984). This model originated from the literature on child maltreatment and posits that parenting behavior is determined by three 12 primary factors: parent characteristics; child characteristics and contextual sources of support, including marital relations, work and social support network (Belsky, 1984). This model is shown on Figure one. A Process model Of the Determinants Of Parenting > Social . support Mantal network Relatrons / Parent’s Parenting Child Developmental F» Characteristics -> Behaviors "' Characteristics Hrstory and Personality /‘ \ I Child \ Work Development Legend: Bi-directional Relationship: —> Unidirectional Relationship: H figural. Determinants of Parenting (Belsky, 1984). One determinant Of parenting behavior is parent characteristics. Parent characteristics is defined as personality. According to Belsky (1990) personality is influenced by the parent’s temperament, emotional states and maturity level. These variables work together to play a strong role in the parent’s ability to respond to his or her child. Studies have shown that parents who are psychologically healthy, with high levels of ego development and feelings of self efficacy, tend to be positively associated with parenting behaviors that are more warm, sensitive and responsive to their infants (Belsky, 1990.) Conversely, l3 parents who are depressed, with low levels of self efficacy and control, tend to display parenting behaviors that are less affectionate, responsive and more controlling with their young children. Personality, according to Belsky (1984), is also effected by the parent’s developmental history. Studies have shown that parents who were classified as securely attached in childhood tend to provide more emotional support and assistance when interacting with their young children (Belsky, 1990). They are also more likely to rear infants who develop secure attachments to them. Parents with insecure attachment histories may be at risk Of repeating the poor parent and Child interaction they experienced in their family Of origin. The domain Of parent personality is also indirectly effected by the variables of work, marital relations and social support network. For example, a parent who feels supported by a partner, coworker or friend feels better about him or herself and is able to invest more emotional and physically energy into positive parenting. A second determinant of parenting behavior is child Characteristics. Child Characteristics are hypothesized to have direct effects on both parenting behaviors and Child development. This variable includes such elements as the child’s age, stage of development and temperament. Difficult child temperament attributes, such as negative mood, high activity level and inclination to disobey, tend to be associated with more negative and controlling behavior in parents (Belsky, 1990). I4 The third determinant of parenting behavior is contextual sources Of support. This determinant is portrayed in Belsky's model as the concepts of marital relations, work and social support network. All three of these sources of support effect parenting both directly and indirectly through the variable Of parent's characteristics and personality. Marital relations are posited to be the strongest contextual variable effecting parental functioning (Belsky, 1990). Marital relations can be a resiliency factor that protects high-risk families from poor parent and child interaction. Research suggests that a supportive relationship with a partner can help parents break the cycle of intergenerational transmission of negative, rejecting and insensitive maternal care. This is particularly true for parents who are at risk Of poor parent and child interaction because of seriously discordant families Of origin, teenage status or abusive histories (Belsky, 1990). If the relationship with a spouse is the primary support system for parents, it is likely that the personal social support network between parents and their friends, kin, neighbors, and associates is the next most important system of support (Belsky and Vondra, 1989; Belsky, 1984). The personal social support network is comprised of all the people outside the immediate family, who the parent knows and thinks is important to him or her. In many households, the social support network may serve as the principle system of support. This is especially true for single parent households or families where the spousal relationship is in turmoil. Studies have shown that caregivers with more social support tend to provide more sensitive and positive care to their Children (Belsky, 15 1990.) “Social support can serve as a buffer against the stress of difficult child Characteristics or environmental challenges. Conversely, low levels of social support have been associated with inconsistent parenting, increased punitiveness, decreased use of reasoning as a discipline strategy and fewer parental rewards” (Whipple and Webster-Stratton, 1991, p. 88). This study sought to add to the field Of knowledge about parenting determinants by exploring three unresolved aspects. One area this study examined was the effects of a home-Visitation parent education program on the parent’s social support network. Specifically, this aspect Of the study assessed the effects of home-visitation parent education on the parent's quality and quantity of social support. This relationship is graphically portrayed on Figure two as a bi-directional line between home-visitation parent education and social support network. The second unresolved area this study examined was the effectiveness of using home-visitation parent education as a form of social support to help parents improve parenting skills. This area of study looked at the direct effects of home-Visitation parent education on the parenting behaviors of limited resource parents. This is shown through the unidirectional arrow on Figure two originating at the home Visitation parent education and going to the variable of parenting behaviors. The third area this study investigated was the relationship between social support network and parenting behaviors Of limited resource parents. This aspect Of study examined relationships among the quality and quantity of social 16 support and the parenting behaviors of limited resource women. Specifically, this study looked at the effects of social support on parenting behaviors. The thick arrow between social support and parenting behaviors on Figure two depicts the posited direct effects Of social support on parenting behaviors. The thick arrow from social support through parent's personality to parenting behaviors graphically Shows the indirect effects between social support and parenting behaviors. Home- S°°ia' visitation support Parent network . j Education Marital Relations ,z\. I Developmental I.) Parent’s _> Parenting {_ ChIId History Personality Characteristics \l 1 Legend: w rk h'l Bi-directional Relationship: <—> O gezgmpment Unidirectional Relationship:——-> Research Emphasis Areas: Bolded and enlarged arrows figural. The Influence Parent Education on Parenting Behaviors and Social Support. Conceptual Map The conceptual map for this study reflected an ecological approach to studying the influence Of home-visitation parent education on parenting behaviors. This approach examined the input, transformation and output Of 17 parents in their role Of raising Children. A model Of the posited evolution within parenting behaviors is depicted in Figure three. The element Of “time one” on the model illustrates the input of energy-matter and information that parents received. This included such elements as parent characteristics, social support network and Child characteristics. Salient parent Characteristics for this research study were income level and educational level. These variables have been found to have both a direct effect on parenting behaviors, as well as an indirect effect through the variable of social support network (Crockenberg, 1988). TIME ONE - Input Parent Characteristics: *Income *Education A \4 Parenting Behaviors I.» TIME TWO - Output Social support network Paraprofessional, Home-visitation Parent Education 4 * Parenting Behaviors (Post) (Pre) Program \ 4 I /‘ x. X . . | Parent SOCIa' Ch'Id _ . | Characteristics: Network Charactenstrcs: “Education- (Bil'th t0 3 years) ' Parenting I Knowledge Legend: Bi-directional Relationship: £———+ Single Direction Relationship: ——> Transforrnation-Direct Effect: ——> Transformation-Indirect Effec : — — I} EigIILeji. Conceptual Map of the Influence Of Parent Education on Parenting Behaviors and Social Support. 18 Social support network is an input variable that also has both a direct effect on parenting behaviors and an indirect effect through parent Characteristics. Social support network members can directly provide input to parenting behaviors by giving the parent information on parenting issues or by providing physical assistance for parenting tasks. Indirect input would be provided when the social support network member provides emotional support to the parent that helps him or her feel better about him or herself. This increased sense of emotional well being is then transferred to improved parenting behaviors. The final input variable in this conceptual map was Child characteristics. This study specifically identified the child Characteristic of age as a salient variable in parenting behaviors. The focus of this research was on parenting children who are birth to age 3. Child Characteristics were posited to have a bi- directional influence on parenting behaviors. Transformation for the parenting behaviors Of parents was hypothesized to begin as the parent started the home-visitation parent education program. Paraprofessionals were used to deliver the information because it was anticipated that high-risk parents would be better able to form a trusting relationship with the paraprofessional. A trusting relationship between the paraprofessional and parent was believed to be a key component in helping limited resource parents transform current parenting behaviors. “By making efforts to maintain a consistently supportive relationship, the home Visitor Shows the parent that positive caring relationships are possible. The parent begins to 19 see herself as someone who deserves support and attention, and by extension, sees her child as deserving the same” (Olds, Kitzman, Cole and Robinson, 1997, p. 20). The output for formal support is reflected in “time two" Of the model (Figure three). As transformation took place parents were posited to improve their parenting behaviors. This was hypothesized to be a direct effect of the home- visitation parent education program. Parenting behaviors were also hypothesized to Change through indirect effects from parent Characteristics and social support network. The home-visitation parent education program was posited to increase education in parent characteristics and overall amounts Of social support network. These increases in social support network and education were hypothesized to have an indirect effect on parenting behaviors. Operational Map An operational map was designed to show how the constructs and variables were assessed in this study. Specifically, this study looked at how a paraprofessional home-Visitation program influenced the social support network and parenting behaviors Of limited income parents Of young children. The operational map is diagrammed in Figure four. 20 CONTROL INDEPENDENT DEPENDENT VARIABLES VARIABLE VARIABLES Parent ..............> Characteristics 5 Social support : network (+) Income (+) ' (Control Variable) - l Paraprofessional v (Egngtargfcgzgible) Home-Visitation Style _> Parenting Parent Education Beh(a+1;lors Child ' Characteristics Age: 0-3 yrs. (?) (Control Variable) Legend: Direct Effect: ——> Indirect Effect: - .. .> Research Emphasis Areas: Bold Lettering (+) Positive Relationship (?) Unknown Relationship Eigurgfi. Operational Map. Constructs and Definitions This section provides conceptual and operational definitions for the key concepts included in this research. | I IIII'II'E IEI l' W Parent education was conceptually defined as “programs, services, and resources Offered to parents and caregivers that are designed to support and empower them or increase their capacity and confidence in raising healthy children” (Smith, 1999, p. 149). 21 eratignaLQeflnjtign, Parent education was operationally defined in this study as the BSF program. BSF is a parenting program developed by Michigan State University Extension. It is targeted at low-income, low-literacy parents of Children from birth to age 3. The eight-to-twelve week program is taught by trained paraprofessionals. Using multicultural, cartoon-style flipcharts, videos and experiential activities, the four lesson modules give parents information about child development, positive discipline, parent-child interaction and goal- setting. D | l I I . | | , S . | | Conceptual Definition, Social support has been conceptually defined as "the emotional, instrumental, or informational help that other people provide an individual" (Crockenberg, 1988 p. 141). Qpecatignalflefinitign, Social support was operationally defined in this Deal, 1994). This instrument is a self-report questionnaire. It measures quantity and perceived satisfaction with support from a variety of sources, inciuding support from a spouse/partner, kin,informal sourcesand formal sources. a —.—..._ Wu-ufi-H‘ _ ,___.. .——__. ——._. w- u-..— ...__ Specifically, this study examined the parent’s level of social support. Level of social support was defined in this study as the parent’s self-reported rate of BPBIIIYEDSSWPIEQFIEI support. The parent’s rate offiqnantitybf social“ support was calculated as the number of social support items on the Family Support Scale that the parent reported was available to him or her prior to the intervention and after the intervention. The parent’s quality of social support was 22 determined by assessing the mean ratings of parental satisfaction with the various sources of social support listed on the Family Support Scale prior to the intervention and after the intervention. [1 IIII'II'E I'Bl' Who, Parenting behaviors have been conceptually defined as actions used by a parent or caregiver to promote optimal development in a Child (Carter, 1996). Q l' l D n 'I' In this research study parenting behavior was operationally defined as the scores achieved on two instruments. The first instrument selected to measure parenting behaviors was the Home Observation for Measurements of the Environment (Caldwell and Bradley, 1984). There are 45 items in the inventory within Six subscales. The subscales include emotional and verbal responsivity of the parent, acceptance of the Child's behavior, organization of the physical environment, provision of appropriate play materials, parents' involvement with the child and opportunities for variety in daily stimulation. Each item on the inventory is scored with a yes (indicating that the item is present) or no (indicating that the item is not present). The number of “yes” answers is totaled to give a .score. The second assessment tool selected to measure parenting behavior was the BSF: Parenting Behavior Assessment. Items from the BSF: Parenting Behavior Assessment were read to the participants and shown on a flipchart. Participants ranked the frequency of performing each parenting item by placing a 23 matching card for each behavior in an envelope that indicated the participant’s perceived level of implementation. This procedure was completed both as a pretest and as a posttest. H . .l‘. . .__ 1 1.. :1" _ --“.. - . .- .._ 1 W Limited resource parents have been conceptually defined as anyone who carries the responsibility for raising a child and has below average access to resources, particularly financial resources (Carter, 1996). QperatignaLDeflnjtign, The operational definition of limited resource parents in this study was parents of 0-3 year old Children who are at or below 185% of the poverty level. C III! 'II . Ql'll’! anggptuaLQefinjtign, Child’s age has been conceptually defined as the age of a Child, as recorded in months. QgenatignaLDefinitm The variable of child’s age was operationally defined in this study as the number of months a parent indicated that his or her Child has been alive. The child’s age was recorded by the parent on the Family Record Form upon enrolling in the BSF program. chELKexletms angggtuaLQeflnjtign, Home Visitation has been conceptually defined as a method of delivering information or services where visitors are sent into families homes to provide information, health care or psychological or other support services (Gomby, Larson, Lewitt & Behnnan, 1993). 24 Qnemtignalfleflnitm This study operationally defined home visitation as the number of times a paraprofessional entered a parent’s home to deliver parenting information, as recorded on the parent’s Family Record Form. E 'l C 'I' CommunLQeflnjtign, Family composition was conceptually defined as a term to characterize the members within a household. Qperatignalflefinitign, In this study, family composition was operationally defined as the category a parent chose on the Family Record Form when asked to tell if his or her family is a “single parent household”, “two-parent household”, “extended family” or “other.” EmniCiiY ancgptualfleflnjfign, Ethnicity has been conceptually defined as a category identifying people according to a race or large groups of people classed according to common traits or customs. QgemtignaLDgflnitign. Ethnicity was operationally defined in this study as the ethnic group that the parent used to identify him or herself on the Family Record Form. The options included “Caucasian”, “African-American”, “Hispanic”, “Asian” and “Multi—Cultural.” W W Residence was conceptually defined as the geographical location where one resides. 25 QuentignaLQefirmign, In this study residence was operationally defined as the location where the parent indicated he or she lived at the time of the study. Categories in this study included “rural” and “urban”. Education ngepjuaLDeflnitign, Education has been conceptually defined as the number of years a person has completed schooling at a formal institution. W This study’s operational definition of education was the last grade completed by the parent, as reported on the Family Record Form. IQtaLMODtblILIDmme angepjualflefinjtign, Total monthly income was conceptually defined as the amount of money a parent earns in the period of one month. eratignaLQeflnjtign. Total monthly income was operationally defined as the amount of money a parent said he or she earned in one month’s time as indicated by the category selected on the Family Record Form. The categories included “$800 or less per month”, $801-$1,000 per month”, $1001-$1,200 per month” or “1,201 or more per month”. Eaflisiinam angegtuaLQeflnjfign, Participant was conceptually defined as a person who becomes involved in an activity. Qmmtignalfleflnitign. In this study participant was operationally defined as a parent who enrolled in the BSF parent education program. 26 NumheLQLthIdLen ConceptuaLDefinjtign, Number of Children was conceptually defined as the total number of children for whom the parent gives care. Queraflgnalfleflnjflgn, Number of children was operationally defined as the total number Of children the parent listed under the “family members” section of the Family Record Form. E 'II' S! E 'l' [ESEl-I I IS ancegtuaLDefinitiQn, Treatment group was conceptually defined as a sample of people who participate in a treatment or intervention as a part of a research project. eratignaLDgflnjtign, Treatment group was operationally defined in this study as parents who completed the BSF program in the six counties participating in the study and during the time frame of the research. Qanaflsnnfimun angeptuaLQeflnjtign, Comparison group was conceptually defined as a sample of people who do not participate in a treatment or intervention being tested in a research project. Qperatignalflefinjtign, In this study comparison group was operationally defined as parents in the Six participating counties who were eligible to participate in the BSF program, but were not currently involved in it. Many of these parents were already on waiting lists to become involved in the Building Strong Families program during the time frame of the research project. 27 Research Questions This study asked the following research questions: 1. DO levels of perceived satisfaction with social support Change for those parents who participate in a home Visitation parent education program, as compared to parents who do not participate in a home based parent education program? Does the amount of social support change for those parents who participate in a home visitation parent education program, as compared to parents who do not participate in a home based parent education program? Do parenting behaviors change for those parents who participate in a home visitation parent education program, as compared to parents who do not participate in a home based parent education program? IS there a relationship between perceived satisfaction with social support and parenting behaviors for limited resource parents of young children? Is there a relationship between the quantity of social support and parenting behaviors for limited resource parents of young children? Assumptions This research was based on several assumptions about social support, home-visitation parent education programs and parenting behaviors. The primary assumptions included in this study are listed below: W: Social support is a determinant of parenting behavior (Belsky, 1990). 28 mm: The amount of social support a parent has and his/her degree of satisfaction with the support are modifiable (Crockenberg, 1988). The plasticity of parental social support allows its parameters to change as influences Change. Assumptignfl: Home-visitation parent education programs can be a form of social support (Carter, 1996; Dunst, Trivette and Deal, 1988). Assumptjgnfl; Paraprofessional, home-visitation parent education can influence parenting behaviors (Hiatt, Sampson and Baird, 1997). Assumptignfi: Paraprofessional, home-visitation parent education can serve as a gateway to other forms of social support (Olds, Kitzman, Cole and Robinson, 1 997). Assumpjignfi: Self-report of social support is an effective method of measurement. Individual perceptions of Offered support determine the degree of helpfulness of the support to the recipient. The recipient will find the support to be helpful only when the assistance ( 1) is requested by the recipient, (2) matches the parent’s perceived needs and (3) does not make the recipient feels overly obligated and unable to reciprocate (Dunst, Trivette, and Deal, 1994; Fisher, Nadler and Whitcher-Alagna, 1982). This type of information must be acquired through self-report. Assumptjgmil; Supportive parenting behaviors are defined in American culture as responsive, sensitive and age-appropriate care for children. 29 Sensitive, responsive and age-appropriate care fosters self-regard, self- protection and self-control. These are traits valued in the American culture (Gearity, 1996, Jennings, Stagg and Connors, 1991). _ W: There is a need to develop theory about the relationship between social support and parenting behaviors in limited resource parents. There is plenty of evidence that social support, in general, facilitates positive parenting behaviors. What remains to be explored are the specific components of the mega-construct known as social support and how the components effect individual sub-populations, particularly high-risk families (Jennings, Stagg and Connors, 1991 ). Hypotheses The research hypotheses in this study were based on the literature review. This research tested five hypotheses about the influence of home-visitation parent education programs on social support and parenting behaviors of limited resource parents. I I . I I II . Limited resource parents involved in a paraprofessional, home-visitation parent education program will Show an increase in levels of perceived satisfaction with social support as compared to limited resource parents in a comparison group. firmness: 1. Limited resource parents involved in a paraprofessional, home-Visitation parent education program will show an increase in overall amounts of social support as compared to limited resource parents in a comparison group. 30 2. Limited resource parents involved in a paraprofessional, home-Visitation parent education program will Show an increase in positive parenting behaviors as compared to limited resource parents in a comparison group. 3. There is a positive relationship between perceived satisfaction with social support and supportive parenting behaviors in limited resource parents. 4. There is a positive relationship between overall amounts of social support and supportive parenting behaviors in limited resource parents. Limitations While much effort was made to reduce the limitations of this study, some limitations still exist and need to be acknowledged. One limitation of this study was in the research design. Much importance has been placed on the need to evaluate home-visitation programs using an evaluation design that has a randomly selected control and treatment group (Gomby, 1999; Gomby, Culross and Behrman, 1999). In an evaluation design that has random assignment every participant has an equal chance of being in the treatment or the control group. Such a design was not possible for this study because of ethical considerations. In this study program recipients needed the parenting class, and it would be unethical to deny the parent a needed and desired service to his or her family. The fact that the comparison group was not randomly drawn from the same population as the treatment group led to the limitation that the comparison group parents might be different from the parents in the treatment group. Many parents Sign Up for home-visitation parent education programs and never complete them. One analysis of several home-visitation programs found that 31 between 20% to 67% of the families enrolled in the programs left the program before the scheduled number of visits (Gomby, Culross and Behrman, 1999). In fact, high attrition rates is a consistent problem with home-Visitation programs, with many home-visitation programs having annual dropout rates of 50% or higher (Baker, Piotrkowski and Brooks-Gunn, 1999; Gomby, 1999; Gomby, Culross and Behrman, 1999; St. Pierre and Layzer, 1999). Knowing that many of the parents on the waiting list may never complete the entire program, comparing the parents on a waiting list with the parents who complete the program may not be comparing equivalent groups of parents. In cases where the groups were not equivalent, statistical analyses were conducted to compensate for the differences. Another limitation to using parents on waiting lists as the comparison group was that often waiting lists were very short or nonexistent in these programs. It was difficult to recruit enough families to participate in the control group during the same amount of time as instructors were able to recruit families for the treatment group. Consequently, the time frame for the control group recruitment was much longer than the period of time for the treatment group recruitment. While the time frame for recruiting the treatment group was only four months, control families were recruited over a ten-month period. A fourth limitation of this study was the instrument used to measure the construct of social support. The social support instrument was Chosen because it allowed the researcher to measure both quantity and perceived satisfaction with social support. Social support items were rated by the parent on a scale from 32 (5)~ex1[emg|y_he|p.tul to (1 )—ng_t_at_a|Lh_elpM. lf social support items were not available to the parent it was rated as a (0). A limitation of the survey was that when a ”not available“ item was coded as a zero, it lowered the mean score for satisfaction. The quandary is that sometimes less overall support is not as stressful as more support that is perceived as unsatisfactory. A second limitation on the social support instrument is the way in which it measures amount Of social support. The instrument uses pre-determined I I sources of social support to assess changes. Parents are asked to indicate if the specific sources of social support on the scale are available to him or her. It is possible that the social supports listed on the instrument are not the ones being utilized by the parent. The small sample size used in this study was another limitation. Although power analysis showed that a sample size of at least 50 participants per group would Show results at an alpha level of .05 and a moderate effect level of .80, smaller effects would not show significance (Pecora, Fraser, Nelson, McCrosky and Meezan, 1995). This limitation prevented the investigator from analyzing differences in group means for certain subgroups of the sample. A Sixth limitation to this study was the probability that some of the parents in the treatment group were court-mandated to attend the program, whereas all of the parents in the comparison group had enrolled in the program on a voluntary basis. Since the BSF program does not collect data at entry on whether or not the parent is participating on a voluntary basis, this study is not able to separate court-mandated parents from the other parents in the sample. 33 Generalizability of the findings was another possible limitation to this study. Although the sample was taken from a mixture of parents from rural, suburban and urban areas, it is possible that parents who become involved in the BSF parenting program at Michigan State University Extension are different from parents who participate in other home-Visitation parent education programs. Until other home-Visitation parent education programs confirm the results, findings in this study are only generalizable to those parents who participate in the Michigan State University Extension BSF program. A final limitation to this study was its lack of ability to measure long-term changes in parenting behavior and social support. Parental changes were measured on the last day of the program intervention. In order to document long-term Change, a longitudinal study would need to be conducted with parents who participate in this intervention. Description of the Intervention The intervention used in this study was participation in the BSF parent education program. BSF is a paraprofessional, home-Visitation parent education program. The target audience is limited resource parents of Children, ages birth to three years. Michigan State University Extension in East Lansing, MI developed the BSF program in 1989. Based on evidence that limited resource families did not have access to parent education information and support, the program was created to respond to an identified deficit of parenting support for this population. Specifically, the BSF program was developed to provide limited resource parents 34 with the knowledge, skills and support necessary to help their children reach their potential. The core elements of the program are weekly parent education classes delivered through home visits. Home visits were selected as the main information distribution modality because of the administrating organization’s previous success with this delivery method for limited resource parents of young children. For thirty years, Michigan State University Extension has used home- visitation as a means of delivering food and nutrition information to this population. The benefits found in Michigan State University’s home-based food and nutrition program are ones that were also important to the parent education program. Home visitation facilitated access to people who were geographically or psychologically isolated and not able to attend classes held outside of the home. It also allowed instructors to view parents in transactions with the resources in the local environment and customize services to meet the specific needs of the program recipients. And finally, home-visitation helped parents develop a one-tO-one relationship that promoted trust between instructors and parents and encouraged parents to extend this trust to other service providers in the community. Ultimately, this benefit was theorized to reduce isolation for parents, build social support networks and increase parents’ resources. Each BSF program is staffed by a program supervisor and instructors. The program supervisors have a minimum of a bachelor’s degree and are experienced in human service delivery. Responsibilities of the program supervisor include day-to—day administration of the program, training and 35 assessment of the instructors and fiscal management. They coordinate weekly program meetings to allow instructors to share ideas and provide support for each other. Program supervisors receive a minimum of twelve days of training each year, including biannual management meetings. Instructors for the BSF program are paraprofessional staff who hold a high school diploma or a GED. Responsibilities of the instructors include conducting weekly home Visits with parents, data collection for use in program assessment and record keeping. Each instructor receives a minimum of twelve days of training each year, including training in child development, positive discipline, parent-child interaction, parental goal-setting and problem-solving, home- visitation delivery Skills, diversity training, adult Ieaming education, accessing community resources, building family strengths and resilience, time management, dealing with difficult people or overcoming barriers. All instructors also attend a biannual meeting to receive staff development and program updates. Developers of the BSF program chose to use paraprofessionals as parent education instructors because of the theorized benefits of peer educators. By using paraprofessional staff as service providers it is posited that the program is better able to respect the values and needs of the audience. The paraprofessionals are indigenous to the community being served. Parent education instructors are parents from the same neighborhoods in which they are teaching and come from similar backgrounds to the target audience. The similarities in backgrounds help the paraprofessionals present information in a 36 manner that is nonjudgmental and Offer resources that are acceptable to the program recipients. It also helps the paraprofessionals quickly form a trusting relationship with the parents and serve as positive role models. All Michigan State University affiliated programs have access to a state team of trainers. At the time of this study the BSF program had on staff two state program leaders and a specialist in parenting. The purpose of the state staff is to provide staff development opportunities for local program staff, assist in staff hiring and evaluation, conduct statewide program oversight and evaluation and develop curriculum materials. The curriculum used by the BSF program has four units. The units cover the topics of child development, positive discipline, parent-child interaction and parental problem solving and goal setting. The child development, positive discipline and parent-Child interaction units are each separated into age segments. These units are divided into the following ages: birth-3 months, 4-6 months, 7-12 months, 12-24 months and 24-36 months. Program protocol suggests that instructors cover all of the information for the current age of the child and all of the information for the age that the Child will be reaching next. The information is presented through multi-cultural, four-color flipcharts. The flipcharts are augmented by Videotapes that role model the content being taught through the flipcharts. Parents are encouraged to practice the skills taught in the lessons through experiential activities. Lesson review sheets also help the parent remember the material after the lesson has finished. 37 Lessons last approximately one hour per Visit and are presented weekly. Program protocol suggests that the program is a minimum of eight weeks in length. The mean number of visits per family statewide is 12 Visits, with the range being a minimum of 8 weeks to a maximum of 27 weeks. Instructors are not required to deny a parent additional Classes if there is material that the parent still desires to leam. Program protocol encourages the instructor to customize the lessons for the parent. Program recipients are asked to express their parenting questions and needs to the instructor. The instructor then is able to present the lessons or information in the order that is most helpful to the parent. Full-time instructors typically present the entire BSF curriculum to a minimum of 45 families in a year. They carry a caseload of 10 to 12 families at a time. Program recipients are recruited through community agencies, such as the Family Independence Agency and the local health department. In order to become involved in the program the parent must have a child who is in the age range of birth to three years of age. Many programs also have a recommended income guideline for participation in the program. The income criterion is often 185% of the poverty rate. Participation in the program is voluntary, unless the parent is court mandated to take the classes. 38 CHAPTER II LITERATURE REVIEW This study focused on the parenting determinant of social support because social support is a construct that is considered to have a degree of plasticity. Unlike other influences, social support is assumed to be relatively changeable. If families do not have enough social support, it may be possible to increase that support in order to improve parenting. It is this possibility for change, along with the potential impact of that change on parents and children, that has focused the attention of service providers on social support (Crockenberg, 1988, p. 142). Social Support Defined within the Context of Parenting Behaviors Social support has been defined as "the emotional, instrumentame informational help that other people provide an individual” (Crockenbei‘g, (1988.3. 141 ). The first element mentioned in Crockenberg’s definition of social support is “rat-.— yemotional supporg Emotional support typically is provided in the context of intimate relationships and friendships. ”It refers to the expressions of empathy and encouragement that convey to parents that they are understood and capable of working through difficulties in order to do a good job in that role” (Crockenberg, 1988, p. 141). Emotional support can "enhance one’s confidence in the interpersonal resources that are available for coping as well as provide outlets for the release of tension and anxiety" (Thompson, 1995, p. 57). Instrumental support is a second element within the social support construct. Instrumental support involves assisting the parent with concrete help so that the number of tasks or responsibilities a parent must perform is reduced. Typically assistance comes in the form of household and child care tasks . - --.--._..____._.._-_'..I .- -. 39 (Crockenberg, 1988). (”Instrumental aid may decrease the amount of stress in the parent’s life by directly resolving instrumental problems or providing the recipient with increased time (Cohen and Wills, 1985). The degree of support instrumental aid provides to the parent depends on the interpretation of the support. When assistance is (I) requested by the parent, (2) match the parent’s perceived need and (3) does not make the recipient feel overly obligated and unable to reciprocate, instrumental aid has been found to increase maternal well-being and parental functioning (Dunst, Trivette, and Deal, 1994; Fisher, Nadler and Whitcher-Alagna, 1982). Informational support is a third element of the social support construct. :lnformational support refers to advise or information concerning child care or parenting” (Crockenberg, 1988, p. 141). It provides parents with help in defining, understanding, and coping with problematic events, Informational support assists parents by helping them reappraise a stressor as benign or suggesting appropriate coping responses to counter a perceived lack of control (Cohen and Wills, 1985). An example of informational support is suggesting appropriate techniques for dealing with a child care problem or helping parents realize that the behavior is normal for the age of the child and undue worry is not necessary. || S'IS IEIIIE I'BI' "T he complex and multidimensional social support networks of parents affect parental behavior both directly and indirectly because of network associates who provide support, impose stress, Offer information, give advice, and otherwise influence adults as workers, community members, and parents” 40 (Thompson, 1995, p. 41). Crockenberg (1988) has posited that social support may influence quality of parenting through modalities involving both direct and indirect effects. Social support may influence quality of parenting as a reducer of stress events (direct effects) or as a buffer and generator or activator of active coping (buffer effects). 191th? direct e‘ffectsflcdelsocial support serves as a proactive method of coping with life events. The direct effects model hypothesizes that social support serves as an important component to overall feelings of well-being, regardless of the stress levels. Social support is posited to relate to overall well-being by providing positive affect and a sense of predictability and stability in one’s life Situation. It is also thought that embeddedness in a social support network may help a parent avoid negative experiences that might otherwise increase the occurrence of psychological or physical pathology (Cohen and Wills, 1985). One way in which the direct effects form of social support functions is by the performance of behaviors which help to eliminate events that may be perceived as stressful. An example of this would be the receipt of help with baby-Sitting so the parent does not become physically drained of energy. A second way in which direct effects functions is by providing support so that the parents does not perceive the event as a stressor. The buffering effectsrnedei posits that support is related to well-being when the parent is experiencing a stress reaction. This model suggests that social support protects or “buffers” the stress reaction so that the parent is less likely to encounter the potentially pathogenic influence of stressful events. Social 41 support is thought to Intervene at two different points in the causal link between stress and illness in this model. First, social support may come between the stressful events and the experienced stress reaction, helping the parent to reduce or eliminate the stress response. Once the parent is experiencing a stress response, instrumental support may reduce the stress, thereby eliminating the stress reaction. “Baby-sitting, childrearing advice, and financial assistance simply provide relief from daily burdens that might otherwise accumulate to incapacitate the parent, or press her or him into inappropriate or even abusive behavior patterns” Cochran and Niego, 1995, p. 403). The support may also come in the form of noninstrumental support. For example, emotional encouragement may build the parent's psychological state so the parent can cope better with the stress. This could eliminate a stress response so that the parent is not adversely affected. Secondly, the buffer may come between the stress reaction and the onset Of illness. In this case, social support can help the parent learn appropriate coping responses, so that the parent is not negatively affected (Cohen and WillS, 1985). This type of support may be either in the form of emotional support, through expressions of empathy, or informational support, such as advice or instruction, WW Emotional support and informational support are likely to be responsive to a wide range of stressful events. In contrast, instrumental support and social companionship functions are assumed to be effective when the resources they provide are Closely linked to the specific need elicited by a stressful event (Cohen 42 and Wills, 1985). Cohen and Wills, 1985 suggest that social support, regardless of the content type, will only have stress-reducing effects when four conditions are present. The four conditions include (1) the stressor is one that is socially acceptable and does not result in feelings of guilt and Shame; (2) discussion of the stressor will not be detrimental to one’s relationship with a comparison other; (3) the support is provided by people who are perceived as providers of accurate information and (4) the support group communicates a relatively calm reaction to the potential stressor. The four conditions mentioned by Cohen and Wills, 1985 have several implications for using highly trained, paraprofessionals to break isolation in high risk families. Paraprofessionals are usually peer instructors who are often perceived as friends. They can gain credibility with high risk families because they are viewed as having experienced many of the same problems. Strong training for paraprofessionals will help them react calmly to the family’s stressors and provide information that the family will receive as accurate (Powell, 1993; Wasik, 1993). S I S l i E | l E l' . IheMamaLBelatignsmg. In Belsky’s ”Determinants of Parental Functioning” model, it is hypothesized that the marital relationship is the first- order support system, with inherent potential for exerting the most positive or negative effect on parenting functioning (Belsky, 1984). "Spousal support of both the emotional (e.g., love, intimacy) and instrumental (e.g., child care tasks) variety is associated with enhanced parental perfonnance—of mothers and fathers alike” (Belsky, 1990, p. 887). Several studies have since moved toward 43 substantiating Belsky's hypothesis. Evidence suggests that spousal support is a key factor in determining parental functioning (Cooley and Unger, 1991; Crockenberg, S, 1987; Lewitt, Weber and Clark, 1986). These studies seem to show that spousal support has direct, indirect and moderating effects on the quality of parenting. A spouse might directly effect the quality of parenting by giving appropriate advice or concrete help to the other spouse regarding care of the Child (Simons and Johnson, 1996). This type of direct support helps the parent provide higher quality care. Another type of spousal support may have an indirect effect on the quality of parenting. Support may take place via a third variable, such as psychological parental stress (Belsky and Vondra, 1989). When the spouse gives emotional support, such as expressions of warmth and caring, the other spouse’s emotional state may be so effected that quality of parenting raises as a result (Simons and Johnson,1996) Finally, spousal support may serve as a mediating variable. For example, a spouse may buffer the perceived stress of an event by providing emotional or informational support. These supportive behaviors may increase the other spouse's emotional state so that quality of parenting does not suffer (Simons and Johnson,1996) W If the relationship with a spouse is the primary support system for parents, it is likely that the personal social support network between parents and their friends, kin, neighbors, and associates is the next most Important system of support (Belsky and Vondra, 1989;‘_Belsky, 1984). The . personai social support network is comprised of all the people outside the immediate family, who the parent knows and thinks is important to him or her. In many households, the social support network may serve as the principle system of support. This is especially true in the case of single pane—n: hcuSelPIds or families where the spousal relationship is initunntE/Supportive social support networks can enhance the parent’s self esteem, increase problem solving abilities, build skill level and reduce stress levels so that general well-being and parental functioning are enhanced (Cole, Kitzman, Olds and Sidora, 1998; Crockenberg, 1988; Cochran and Brassard, 1979; Dunst, Trivette, and Deal, 1994). A study of 55 married women found that women who reported high levels of social support during the prenatal assessment subsequently reported higher levels of self-confidence in the parenting role and less depression 3 months after the delivery. Specifically, women who had a larger personal social support network on whom they could rely on for a variety of social provisions were found to have more confidence in their abilities to perform well as mother. This confidence was an effective deterrent to depression (Cutrona and Troutman, 1986) A study using parents of preschool children also found a positive relationship between social support and maternal well-being. This study involved 69 white, middle-class mothers and their preschool age children. It was conducted to see if social support predicted maternal behaviors. The results found that larger and better quality social support networks were associated with 45 less perceived parenting task difficulty, resulting in parental functioning that was more warm and nurturing than parents who did not have strong social support (Melson, Ludd, and Hsu, 1993). A third study looked at a slightly different sample and used an operational definition of social support, as “natural mentors." This study of 127 young African American mothers analyzed the influence of natural mentors on maternal well- being. The study found that satisfaction with social support and intangible support were negatively related to depression, but instrumental aid was found to be positively related to levels of depression. Those mothers who utilized more intangible support and were more satisfied with this support were less depressed. In this study, the mentors seemed to enhance the mother‘s capacity to benefit from their social support networks. Conversely, the mothers who used their mentors primarily for instrumental support or were less satisfied with their social support were found to be more depressed (Rhodes, Ebert, and Fischer, 1992). The use of instrumental support may have been associated with increased depression because it made the parent feel incompetent in her role as a parent. Similar results regarding the use of instrumental support byyoung panents were found in another study. In this study, instrumental aid was provided to teen ~F—_ __._..a-- parents by their maternal grandmothers. The aid was primarily hyging for the parent and child care assistance. The study found that when the teen lived with “man“ her mother for an extended period there was a less stimulating environment, 46 lower maternal responsiveness toward the Child and decreased child outcomes (Cooley and Unger, 1991). Although the association between instrumental aid and maternal well- being has mixed results in the previous studies, the use of informational and emotional support appears to be consistently associated with positive maternal effects. Continued support for this hypothesis is found in the Jacobson and Frye study. In 1991, Jacobson and Frye used 46 low-income mothers to evaluate the influence of maternal social support on the development of attachment. The participants were divided into an experimental group and a control group. The experimental group was assigned a home Visitor. The home visitor provided the participant with empathy and information on pregnancy, early child care and development. The home Visits began monthly after recruitment, increased in intervals toward the time of delivery, continued weekly until the child was 2-3 months in age and than decreased to monthly until the child was one year of age. Infants of mothers in the experimental groups were rated more securely attached than the control group. The findings provide experimental evidence regarding the importance Of social support on infant attachment. The results were irrespective of the child’s temperament. The authors suggest that the women in this study responded positively to the study treatment because ”they were particularly vulnerable to the lack of social support given their isolation and the number of stressful events occurring in their lives . . . Social support appeared to be necessary in promoting secure attachment at all levels of ego functioning in this sample. Nonetheless, it may be less critical for women with 47 levels of ego functioning higher than those found here” (Jacobson and Frye, 1991, p. 580). A similar study was conducted by Boger, Richter, Kernetz and Haas in 1986. In this study, 48 first-time mothers were randomly assigned to treatment and control status prior to entering a perinatal positive parenting program. Twelve treatment and twelve control mothers were placed in subgroups by maternal age, age 23 and younger and age 27 and older. The treatment group received an initial contact in the hospital room by a parent volunteer. This was followed by the parent volunteer giving telephone and home-visitation support during the initial postpartum period. The third component consisted of peer support groups, facilitated by the parent volunteers. Support was primarily emotional and informational. The results of the study found that the treatment parents had significantly better matemaI-involvement and child nurturing environments than the control parents at 15 months postpartum. Also, treatment mothers initiated interactions with their infants more often than did the control mothers. .,'. .....' l_ ..:u':- ‘..H .. i.” ”L. The importance of studying social support systems in the context of population subgroups has been stressed by several researchers (Bo, 1994; Dilworth-Anderson and Marshall, 1996; Rohrle and Sommer, 1994; and Vaux, 1988). Comparative research is important because 1) our society is complex and multi-cultural, 2) attitudes, beliefs, and values from subcultures form the macrosystem that filters down into the behaviors carried out in the microsystem of individuals, 3) social support that is not equitably distributed across the 48 population may be shaped by public policy, 4) ecological niches of our society’s subgroups manifest distinct patterns of stressors and support availability, and 5) differences in social support patterns may result in epidemiological variations between sub-populations. Key research samples used for comparative social support research include population subsets that vary in ethnicity, gender, social class and martial status (Bo, 1994; Cochran and Niego, 1995; Dilworth-Anderson and Marshall, 1996 and Vaux, 1988). WW Researchers have found that social support variations among ethnic groups may be due to past sociohistorical experiences (Dilworth-Anderson and Marshall, 1996). “The specific history of a group helped shape the cultural context in which social support is given” (Dilworth-Anderson and Marshall, 1996, p. 76). For example, Dilworth-Anderson and Marshall (1996) suggest that residual influences from the slave community have made it so a social support system that values survival is still evident today within the African-American subculture. Support comes from within the subgroup to protect the itself, instead of outside the subgroup. Kin are called upon to provide various support, including both emotional support and instrumental support, such as helping with household tasks or parenting. Parenting is considered a community activity, which includes the use of “othermothers” to assist the biological mothers (Dilworth-Anderson and Marshall, 1996). The social support systems of Hispanic families in the United States may be strongly influenced by their migratory history. Many families migrate into the United States from other counties, including Mexico, Puerto Rico, Cuba, and 49 Central American counties. A migratory lifestyle necessitates that social support come from kin. “These family-centered cultures include a social support system that reflects close and distant kin and the godparents of children in the family” (Dilworth-Anderson and Marshall, 1996, p. 70). Dilworth-Anderson and Marshall (1996) suggest that the family unit in the Hispanic family is responsible for complete support of the family, including emotional, instrumental and material assistance. Social supportin the Asian American family often is characterized by ancient and traditional values for harmony and shame (Dilworth-Anderson and Marshall, 1996). Harmony and shame help preserve a social support system that is family centered and promotes a common sense of giving and receiving help. Social support that promotes unity is valued. A similar value for unity and harmony is also found in the Native American family (Dilworth-Anderson and Marshall, 1996). A cultural history in discrimination has fostered the need for unity within the subgroup. The extended family is relied on for all types of support. Wins, Social class can be defined as the family's combination of income, educational level of parents and the status and complexity of the occupations in which the parents engage (Cochran and Niego, 1995). One comparative study of social class and social support involving 1,000 people found that educational level was a strong predictor of social support (Fischer, 1982). “Others things being equal, the more educational credentials respondents had, the more socially active they were, the larger their networks, 50 the more companionship they reported, the more intimate their relations and the wider the geographic range of their ties. In general, education by itself meant broader, deeper, and richer networks (Fischer, 1982, p. 252). Fisher also found that household income was a salient predictor of social support. Even with education held constant, people with more income were found to have more non-kin in their social support networks and were more likely to indicate that they had adequate levels of companionship and practical support then were respondents with fewer monetary resources (Fischer 1982). The effects of poverty and social support levels on parenting are notable. “Poor parents are unable to purchase services (eg. child care) that reduce stress on child rearing. Perceiving a lack of social support may intensify feelings of hopelessness, which in turn may influence the way poor parents interact with their children. Poor parents who reported few sources of assistance to draw upon in a crisis were especially likely to report that they yelled at or slapped their children 'very often.’ Economic deprivation combined with a lack of social support creates a dangerous situation for children” (Smith, 1999, p. 408). _ WWWnother important set of subgroups for social support comparative research are households who vary by marital status. According to Vaux, 1988, married couples tend to report higher levels of social support than divorced, widowed, or separated households. Married respondents report having more confidants, larger networks and higher appraisal levels for their social support. Divorced couples report the lowest levels of social support out of all household types (Vaux, 1988). 51 Household composition is a salient variable in epidemiology studies. Married men and women report being less depressed. This state is probably due to the higher levels of social support in married couples (Vaux, 1988). Wm; A final set of population subgroups for social support comparative research is males and females. Studies have found differences in social support systems based on gender to be inconsistent and mixed (Vaux, 1988). There is some evidence to suggest that females may be advantaged in the area of emotional support. This is particularly true in the area of informal support, such as peers. It has been suggested that, in order to determine the true effects of social support systems on gender subgroups, further analysis needs to be done (Vaux, 1988). Ironically, females tend to be more vulnerable to higher levels of distress and depression than males (Vaux, 1988). This is curious when coupled with the suggested finding that females may have more emotional support. Such variables as societal status, work load and role strain may be contributing factors to higher levels of distress in females. WED. The research has found that structural factors operating through the constructs of ethnicity, culture, class, gender and household composition place constraints on the capacity of parents to build networks that adequately support them in their role of child rearing (Cochran and Niego, 1995). These constraints yield a smaller pool of eligibles from which parents are able to build and maintain personal support networks. “African-American parents, nonethnic White parents, parents with relatively little 52 education, and parents living in cultures shaped by beliefs that lead to narrow definitions of the women’s role, all have smaller pools of potential network counterparts. Constraints accumulate for single parents, who often have less access to relatives, further education, jobs paying a decent salary, and housing in neighborhoods that are supportive of neighboring activities” (Cochran and Niego, 1995, p. 411). [:01 o Imnggao 0A.. ovoli-‘o A; o.’.:.. 1.1.... The literature makes a strong statement that some type of social support is important for positive parental function. So, does the antithesis of this hypothesis also hold true? Do diminished levels of social support promote parental dysfunction or maltreatment? The answer to that question appears to be true. Studies support the hypothesis that diminished social support resources tends to promote parental dysfunction or maltreatment of children. (Burke, Chandy, Dannerbeck and Wilson-Watt, 1998; Garbarino and Crouter, 1978; Garbarino, 1977a; Garbarino, 1977b; Polansky, Gaudin, Jr., Ammons and Davis, 1985; Smith,1999; Thompson, 1995). “The maladaptive behavior of abusing parents is, in part, the result of the absence of stress- and anxiety-reducing mechanisms provided by strong, supportive social networks. Less support creates more anxiety and leads to fewer adaptive responses to stressful situations” (Smith, 1999, p. 407). Child maltreatment research has suggested that there are several conditions that combine to create an environment for child maltreatment. The conditions are (1) a cultural context that condones domestic violence in general 53 and violence toward children in particular, (2) families that experience social or economic stress, (3) the parenting “style” of the child’s caregivers that is accepting of child maltreatment and (4) child characteristics, which act as a stimulus. The factor that allows these forces to result in abuse is “isolation.” (Garbarino and Sherman, 1980; Garbarino1977a; Garbarino1977b; Smith, 1999; Thompson, 1995) Isolation reduces the social supports that are available to the family. It reduces the family’s ability to seek assistance from others. It also limits the child’s contacts for help. One study linked maltreatment to the overall balance of stresses and supports in the neighborhood context of families. The likelihood of child maltreatment was hypothesized to vary in direct relation to the availability, adequacy, and use mode of the family’s supportive resources in the community. Twenty neighborhoods and 93 census tracts were involved in the study. The study results showed the strongest predictors of child maltreatment were linked to the social supports available in the area (Garbarino and Crouter, 1978). Polansky, Gaudin, Jr., Ammons and Davis, 1985 conducted a study with 154 neglectful mothers and 154 non-neglectful mothers to examine differences in their social support networks. They wanted to see if the neighborhoods of the neglectful mothers were less caring and inferior in their general quality of life. They also wanted to see if the neglectful parents have less access to helping networks and were more socially isolated. In this study the neglectful mothers reported less social support from informal sources. They felt they had fewer people to ask for practical or emotional support. They reported higher levels of 54 loneliness then the non-neglectful mothers. They also described their neighborhoods as less friendly and less helpful than the mothers who were non- neglectful. “Neglecting and maltreating mothers, although in much greater need of support than most others parents, are likely to avoid potential sources of help or act in ways that discouraged others from offering help . . .. Individuals in the community are often less responsive to offer help to parents experiencing chronic stress, especially when the conditions are stigmatizing. Parents who are too emotionally fragmented to be effective with their children may be shunned as undeserving by the community. For example, parents whose child rearing falls markedly below accepted norms, as evidenced by neglect, are likely to be distanced by their neighbors and others in the community” (Smith, 1999, p. 408). It would appear that the children of parents who have a history of poor parenting and are isolated by their community are in a very dangerous situation. Provldlng Soclal Support Through Home-Based Parent Educatlon Because this study investigates social support as provided by a paraprofessional, home visitation, parent education program, relevant research will be presented on the effectiveness of home-based parent education. This section of the literature review provides an examination of the strengths and limitations of paraprofessional home-visitation programs designed to support pregnant women and parents of young children. Paraprofessionals have been defined as “individuals who have not received baccalaureate, post-baccalaureate or professional training, but have practical experience in the community and are familiar with local resources” (Hiatt, Sampson and Baird, 1997, p. 78). 55 A plethora of programs using the home-visitation style of service delivery to pregnant women and parents of young children has evolved throughout this century (Gomby, Culross, and Behrman, 1999; Olds and Korfmacher, 1997). There are many benefits to the use of home-visitation as a service delivery model. It provides the home visitor an opportunity to see the program recipient in active transactions with people and resources in his or her environment. lnfon'nation gleaned from viewing such transactions allows the visitor to customize the trajectory of the service to best meet the needs of the parent. This method of program customization and delivery is an especially important mode of resource dissemination for people who are psychologically or geographically isolated and might not obtain services or resources in any other way. Once allowed into the home, visitors can serve as a bridge to other community resources and extra-familial social interactions (Gomby, Culross, and Behrman, 1999; Gomby, Larson, Lewitt and Behrman, 1993). Although the benefits of home-visitation as a form of social support seem intuitive, they have not always been supported by evidence in the empirical literature. The research that has been conducted has provided mixed results (Gomby, Culross, and Behrman, 1999). Studies showing support of the paraprofessional home-visitation model have associated paraprofessional social support with increased parent-child interaction (Daro and Harding, 1999), maternal parenting efficacy (Duggon, McFarlane, Windham, Rohde, Salkever, Fuddy, Rosenberg, Buchbinder and Sia, 1999); matemal-child attachment (Jacobson and Frye, 1991), child IQ (Blair, Ramey and Harden, 1995; Liaw, 56 Meisels and Brooks-Gunn, 1995), maternal involvement (Boger, Richter, Humetz and Haas, 1986), better prenatal care and fewer pre-term births (Rogers, Peoples-Sheps and Suchindran, 1996). Other studies have not found a strong association between paraprofessional home-visitation support model and parenting outcomes (Barth, 1991; Choi, Berger and Flunn, 1997; Luster, Perlstadt, McKinney, Sims and Juang, 1996; Siegel, Bauman, Schaefer, Saunders and Ingram, 1980; Silver, lreys, Bauman and Stein, 1997) One possible reason for this lack of empirical evidence is that home visiting is a method of delivery service that is not homogenous in structure or quality from program to program (Baker, Piotrkowski, and Brooks-Gunn, 1999; Gomby, Culross, and Behrman, 1999). The literature can not make generalizable statements about the effectiveness of a service delivery model that has diverse, often-inadequate designs, implementation styles and evaluation methods. Home-visitation models vary in the experience and background of their staff (professionals, paraprofessionals, volunteers, college students), staff training, program protocol, goals, content, rate and intensity of delivery, clients and rigor of evaluation regimes (Gomby, Culross, and Behrman, 1999; Gomby, Larson, Lewitt, and Behrman, 1993). E D'Ell "II' S'IS l One possible element contributing to the mixed evaluation results found in home-visitation programs might be the program design. An important characteristic of home-visitation programs is the conceptual design of the program (Gomby, Larson, Lewitt and Behrman, 1993; Gray and Wandersman, 1980). A limitation of some home-visitation programs is that they are not based 57 on a theoretically sound conceptual design. An effective, theoretically sound conceptual design has evolved from trials involving home-visitation programs delivered by professionals (Olds, Henderson, Jr., Kitzman, Eckenrode, Cole, and Tatelbaum, 1999; Olds, Kitzman, Cole and Robinson, 1997). Recently this model was adapted and proposed as a standard for home-visitation programs delivered by paraprofessionals (Hiatt, Sampson and Baird, 1997). Clinical trials on the paraprofessional home-visitation model are currently being conducted in Denver, Colorado (Olds, Henderson, Jr., Kitzman, Eckenrode, Cole, and Tatelbaum, 1999; Olds, Henderson, Jr., Kitzman, Eckenrode, Cole and Tarelbaum, 1998; Olds, Kitzman, Cole and Robinson, 1997). The model is based on elements from human ecological theory, mattachmfienttheory and self-efficacy theory; Human ecological theory is reflected ‘95..--“ in the elements of person, process and context, as shown in Figure five (Olds ’~-/ Kitzman, Cole and Robinson, 1997). The element of “process” in this delivery model encompasses the (emotional, infOrmational and instrumental support that a paraprofessional can Ie—qun—O give to the parent during the home visits. It includes the relationship that forms _;g ____—_H . during the program between the home visitor and the parent, the education that is transmitted, and the goal setting and problem solving processes that occur “Mn—— “‘5‘...— .— withir/i‘ the parent and between the parent and others in his or her social network. .--.. “an, The element of “person” includes both the parent and the child. It focuses ,/ ./'J on the transformations in parents' behavioral and psychological characteristics and outcomes from the child that occur as a result of the home visit. Changes 58 from the parent's adaptive behavior can resonate outward to more external environments of the parent and child. -_- -_—....__.__.._ The third and final element from the “Person-Process-Context" model is — -—-_.-.-_—...-.- the context. In order to facilitate parental mediation of environmental contexts, ”-___,___ the home-visitation program should include partner or kin supports in the program, as much as possible (Olds, Kitzman, Cole and Robinson, 1997). The program should also facilitate the extension of parents’ social networks, by creating connections with informal and formal resources. Context Culture SchoolNVork Community Services r ' Family/Friends. f Child Low Birth Weight Program Process Parent Pre-Term Delivery Relationship Psychological Maltreatment Educatlon <—-> Resources H Injuries Goal Setting . Developmental & Problem Solvrng Adaptive Behavioral Behavior Processes EMILE—5- Person-Process-Context Model for Home-visitation Program Delivery. Note: Program influences on pregnancy outcomes, child health and development, and maternal life-course (Olds, Kitzman, Cole, and Robinson 1997) 59 One of the rationales for testing the efficacy of paraprofessional home- visitation in the Denver trial grew from the hypothesis that the power of the intervention might be enhanced if the visitors were viewed as having attributes similar to the families they serve and if the visitors were viewed as having overcome challenging situations in their own lives (Olds, Kitzman, Cole and Robinson, 1997 p. 18). A close therapeutic alliance between the parent and paraprofessional home visitor helps the parent Ieam to trust supportive relationships, review his or her own childrearing history and role models a positive relationship that may be replicated between the parent and child (Gomby, Culross, and Behrman, 1999). “By making efforts to maintain a consistently supportive relationship, the home visitor shows the parent that positive caring relationships are possible. The parent begins to see herself as someone who deserves support and attention, and by extension, sees her child as deserving the same” (Olds, Kitzman, Cole and Robinson, 1997, p. 20). AdherenceJoneEmramnesign Paraprofessional home-visitation programs not only need to have a theoretically sound conceptual design, they also need staff who adhere closely to the program design. Evaluation results of paraprofessional home-visitation programs may be diminished because staff members do not adhere closely to the established protocol of the program (Baker, Piotrkowski, and Brooks-Gunn, 1999; Duggan, McFarlane, Windham, Rohde, Salkever, Fuddy, Rosenberg, Buchbinger and Sia, 1999; Gomby, Culross and Behrman, 1999; Wagner and Clayton, 1999). One analysis of several home-visitation programs found that “programs’ curricula are not always delivered with fidelity to the models” (Gomby, 60 Culross and Behrman, 1999, p. 17). Moreover, the researchers found that many home-visitation programs encouraged home visitors to deviate from the established protocol so that the program could be customized to the individual needs of the client. Such deviations make it difficult to evaluate the program because program recipients may not have received a uniform intervention. It has been posited that “the individualization of service content and delivery inherent in home visiting programs may make it hard to see differences across a whole group, because, in fact, the group is not getting the same treatment” (Gomby, 1999, p. 42). Another variable that may contribute to implementation challenges and effect evaluation results is staffing. Home visiting programs are still wrestling with questions regarding the amount of experience, training and supervision that staff need in order to implement an effective home visiting program. Very little research has been conducted comparing the effectiveness of paraprofessional versus professional instructors in home-based parent education programs. This lack of empirical evidence has led “most researchers to believe that it is not possible at this time to conclude that individuals from a particular professional or educational discipline are better home visitors than others, but it seems likely that extremely well-trained visitors are needed to serve families that face multiple complex issues. . . . No matter what their skill level, close supervision is needed to help home visitors deal with the emotional stresses of the job and maintain 61 objectivity, prevent drift from program protocols, and provide an opportunity for reflection and professional growth” (Gomby, Culross and Behrman, 1999, p. 18). E B . . I Studies show that paraprofessional home-visitation programs are most effective with subgroups of parents (Gomby, Culross and Behrman, 1999). Paraprofessional home-visitation programs appear to be most effective with those parents who have some risk factors, such as being young, socially isolated, economically disadvantaged, inexperienced with parenting or limited in education (Brooks-Gunn, Gross, Kraemer, Spiker and Shapiro, 1992; Jacobson and Frye 1991; Olds, Henderson, Jr., Kitzman, Eckenrode, Cole, and Tatelbaum, 1999; Olds and Kitzman, 1993). “This might be because home visiting services help place a supportive floor underneath the neediest families, or because those families feel the strongest need and motivation to change. Or perhaps the area in which home visiting programs' effects can be observed most easily is among the group that is the neediest because that group has the most room for improvement” (Gomby, Culross, and Behrman, 1999, p. 21). Paraprofessional home-visitation models also seem to be effective with recipients who have requested services (Gomby, Larson, Lewitt and Behrman, 1993; Osofsky, Culp and Ware, 1988). It has been less effective with parents who feel they either do not want or need the services after they become involved in the program. “Intervention participants may not passively receive what is given or served, rather, they may actively interact in a variety of ways and at different levels with the intervention staff and activities. In addition, participants' 62 level of activity affects their development and intervention outcomes” (Liaw, Meisels and Brooks-Gunn, 1995, p. 408). A study of 130 adolescent mothers compared the outcomes of program participants who were actively involved in a program with those who did not become actively involved with it (Osofsky, Culp and Ware, 1988). The purpose of the study was to examine the effects of social support on parenting behaviors and child outcomes.Analysis was conducted by dividing the participants into two groups. One group included the participants that were activity involved in the program by keeping appointments and following through on program goals and activities (takers). The second group consisted of those participants who were not activity involved in the program (non-takers). Analysis was conducted on the “takers” and “nontakers”. At thirteen months of age significant differences were found on the Bayley Mental Scale, feeding interactions and on maternal play. The “takers” scored higher than the “nontakers” on each of these variables. At thirty-six months highly significant differences were found on five out of the six areas in the HOME scale. A large study of families of low birth weight and premature infants also focused on the parents’ level of participation. In this study 985 families of low birth weight and premature infants participated in a clinical trial to look at the effects of early intervention on intellectual development of low birth weight, premature infants (Blair, Ramey and Harden, 1995; Brooks-Gunn, Gross, Kraemer, Spiker and Shapiro, 1992; Liaw, Meisels and Brooks-Gunn, 1995). Infants were randomly assigned to an intervention group (n=377) and a control 63 group (n=608). The control group received periodic assessments and referrals. The intervention group received the assessments and referrals, plus a home- visitation program and the opportunity to participate in a child development center and parent support group meetings. Analysis was conducted on the parents’ level of involvement (high, medium, low) and child IQ scores (Blair, Ramey, and Harden, 1995). Research results showed that at 24 and 36 months, higher mental development index and IQ were associated with higher levels of participation in the program. The parents’ level of active participation continued to make more significant contributions to both child IQ and the HOME scores at age 3 ( Liaw, Meisels, and Brooks-Gunn, 1995). Although paraprofessional home-visitation programs have been found to be effective with high-risk families who need and want prevention services, they have not shown as much effectiveness with families who are already involved in multiple and complex high-risk behaviors and are in need of professional intervention services (Barth, 1991; Choi, Berger and Flynn, 1997; Olds and Kitzman, 1993). A study of 154 families involved with Child Protective Services or the Pennanency Planning Unit of the Department of Social Services illustrates the limitations of using paraprofessionals to meet the meets of families experiencing severe problems. The purpose of the study was to determine the influence of volunteer parent-aid home visits on the ability of children to remain with their parents instead of being placed in foster care (Choi, Berger and Flynn, 1997). Parents involved in the study were asked to choose if they would like to have the services of a parent-aid (n=83) or choose to not receive this service (n=71). Parent-aids made weekly (or more frequent) home visits to the parents for a mean of 18 months. No curriculum was used during the home visits. The purpose of the visits was to provide emotional and instrumental support. Results of the study indicated that the parents who received home visits did not have significantly different outcomes than the control group. Multivariate analysis showed that the most powerful predictors of program outcomes were the variables of substance abuse, followed by the amount of contact with case planners. Another recent study looked at the use of the paraprofessional home- visitation model with multi-problem families. This study of 191 women looked at the effects of social support on child abuse rates (Barth, 1991). All of the women in the study were at risk of child abuse. Participants were randomly assigned to a treatment or control group. The control group received an assessment and referrals to community service. The treatment group received “CPEP” services, consisting of twice monthly home visits by a paraprofessional for six months. The home visits focused on the identification and completion of goals. It also included instrumental assistance, such as help with transportation or child care. Results of the study showed no advantages on self-report measures for the treatment or on follow-up reports of child abuse. “Paraprofessional services like CPEP appear not to be designed for unilateral intervention on behalf of highly distressed families. Paraprofessionals may be overrnatched by the 65 multiple problems of families. The results support the provision of CEPEP types of services for families that have not already begun destructive patterns of interaction with their children.” (Barth, 1991, p. 372). The literature has provided much empirical evidence supporting the hypothesis that paraprofessional home-visitation programs do not benefit all families equally. As identified earlier, home visiting programs appear to benefit subgroups of families with varying characteristics. “Existing research does not enable conclusions about which families are best suited to which home visiting models or which are best suited to home visiting versus some other service- delivery strategy. Research does however, clearly suggest that programs will not produce benefits across the whole population of families with young children. (Gomby, Culross, and Behrman, 1999, p. 21). E D I' || | 'l [E D I' A meta-analysis from thirty-nine studies was conducted to determine the effectiveness of home-based programs (Hattie, Sharpley and Rogers, 1984). The results of the study showed that one of the key variables in determining the effectiveness for either paraprofessionals or professionals was the duration of the program, number of sessions and the total hours of contact. The longer, more frequent and more intense the sessions, the greater the program effect. More recent studies have added support to the proposition that program length, duration and intensity are positively related to program effectiveness (Baker, Piotrkowski and Brooks-Gunn, 1999; Duggan, McFarlane, Windham, Rohde, Salkever, Fuddy, Rosenberg, Buchbinder and Sia, 1999; Gomby, 1999; Gomby, 66 Culross, and Behrman, 1999; Gomby, Larson, Lewitt, and Behrman, 1993; National Committee to Prevent Child Abuse, 1992; Wagner and Clayton, 1999). The influence of intensity, length and duration of home-visitation services is illustrated in a study of 6,514 pregnant teens (Rogers, Peoples-Sheps and Suchindran, 1996). The study was conducted to evaluate the impact of a paraprofessional home-visitation program on prenatal care use, lower birth weight and preterm births. The treatment group (n=1,901) was pregnant teens who agreed to participate in the social support program. A comparison group was pregnant teens in neighboring counties (n=4,613) who did not have access to the program. The treatment group received home visits monthly throughout the pregnancy, at the hospital during delivery and then regularly during the first year. Each visit was structured with educational lessons and assistance with community services. Results of this study showed that a higher number of teens in the treatment group initiated prenatal care early in the pregnancy, had adequate prenatal care and fewer pre-terrn births. The study did not find an effect on rates of low birth weight. The authors suggest that the home visits were not intense enough to create the behavioral changes needed to affect birth weight. Yet, the length and frequency of the visits prior to the delivery were enough to impact the outcome variables of prenatal care and term of delivery. These findings seem to imply that all three variables of frequency, duration and intensity need to be at 67 adequate levels to have an influence on different program outcomes. This phenomenon is known as the dosage effect. There is a lack of research determining the minimum number of home visits needed to effect change in parents of young children. “Although no studies have been conducted to demonstrate the minimum number of home visits necessary before change can occur, it seems intuitively reasonable that some threshold number of visits must be crossed before change can occur and that too few visits will hamper the formation of the relationship between home visitor and parent and result in spotty coverage of the program’s curriculum. While precise minimum threshold is unknown, some researchers have speculated that four visits or three to six months of services may be required before change can occur" (Gomby, Culross, and Behrman, 1999, p. 16). Other studies have shown that a minimum of six to eighteen months of high intensity services are needed to change attitudes and strengthen parenting (National Research Council, 1993; Whipple, 1999; Whipple and Wilson, 1996). It appears that short-term, low- intensity programs are not sufficient by themselves to change long-term parental behaviors. Two recent studies demonstrated the need for adequate dosage in services designed to create changes in parental behavior (Whipple, 1999; Whipple and Wilson, 1996). These studies found that services were more effective when the parents receive four contacts per week over a period of 15 to 24 months. These rates of intensity and duration were found to contribute to 68 changes in parents’ level of depression, stress, parenting ability and conflict resolution skills (Whipple 1999; Whipple and Wilson, 1996). E 'l I | | I B I I. The issue of program involvement is a prevailing challenge within home- visitation programs. ~ It has been suggested that recipients of home-visitation program receive an average of about half the number of visits recommended in the program protocol (Gomby, Culross, and Behrman, 1999). “The low levels of involvement may reflect a lack of interest on the part of families, the chaotic nature of some families’ lives or their inability to juggle time commitments between their home visiting programs and their responsibilities to work, extended family, and children. Whatever the cause, once an appointment is missed, home visitors with tight caseloads may find that they are unable to reschedule a visit until the next regular appointment time rolls around, with the consequence that families receive less intensive services than planned” (Gomby, Culross, and Behrman, 1999, p. 15). Ultimately, low program involvement may mean fewer benefits to the family. W During the last decade a few paraprofessional home-visitation programs have begun to use true experimental designs for their research studies. These rigorous evaluation studies have not consistently shown evidence supporting the use of a paraprofessional home-visitation program. One possible reason for this lack of empirical support is a masking of the treatment effect, by using control groups that were actually receiving some type of treatment (Silver, lreys, Bauman, and Stein, 1997). 69 A study of 365 low-income mothers of five to eight year old children illustrates the problem of a control group receiving treatment (Silver, lreys, Bauman, and Stein, 1997). The purpose of this study was to show the effects of a social support program on maternal variables of psychiatric well-being. The mothers were randomly assigned to a treatment group (n=174) and a control group (n=169). The treatment group was expected to receive at least six home visits from a paraprofessional during a 12 month period and at least bi-weekly telephone calls. However, few of the mothers actually received all six home visits. The mean number was three visits. The purpose of the visits was to share information on child development and accessing community resources and provide emotional support. No specific curriculum was used. The control group received assessments at program entry, six months, twelve months and eighteen months into the program. The results of the study found no significant differences between the groups on a psychiatric symptom index or its sub-scales. Further analysis showed that the experimental group had more baseline symptoms. The treatment group’s progress was significantly greater, relative to their initial status. Overall, the program effects were minimal. The strengths of this study were the large sample sizes, the random selection and the ecological orientation to the program. The limitation was that the visits were so minimal in number that a relationship between the visitor and parent may not have had time to form. In actuality, the control group may have received a comparable number of visits through the assessment schedule and 70 have received similar benefits to the treatment group. “In examining mothers’ evaluative comments in the final research interview (six months postterrnination of the intervention), we noticed that many control group mothers, as well as those in the experimental group, said they felt that participating in the research itself (as distinct from the intervention) had been helpful because it allowed them to express their feelings about their children’s illnesses and showed that someone cared about their experiences and opinions. Their statements suggested that being a part of the study itself could have had some unintended beneficial effects that may have masked some of the intervention’s impact. Thus, participating in the research interview may have mimicked a low-dose support intervention” (Silver, lreys, Bauman, and Stein, 1997, p. 262). A similar masking of treatment effect may have also happened in another study. The purpose of this study was to examine the effects of a fairly intensive family support program on the quality of care teen parents provide for their children and look at mediating factors for parental care (Luster, Perlstadt, McKinney, Sims and Juang, 1996). Eighty teen parents were randomly assigned to a home visited group (n=43) and a standard program comparison group (n=40). All of the teen parents were interviewed during the prenatal period, at six months postpartum and twelve months postpartum. The comparison group received the assessment and assistance and information, primarily over the telephone and through the mail. The home visited group received the assessments and weekly home visits by a paraprofessional family advocate for 71 over a year. The home visits provided the parents with emotional support, access to community resources, information about child development and care, and goal-setting information. In order to conduct the home visits, family advocates received three months of training and ongoing supervision. The results of the study showed a significant main effect on the HOME score for the home visited group and a significant main effect on ethnicity. The African American teen parents in the home visited group scored higher than the comparison group. There was also a significant main effect on two of the subscales. The home visited group had higher scores on emotional and verbal responsibility and maternal involvement. Significant effects were not found on maternal psychological well-being. One reason why the program effects for the home visited group may not have been large in magnitude is because the comparison group received a lot of assistance. Similar to the previous study, participants in both groups expressed satisfaction in their program and may have gained from the services. Literature Review Concluslons This review of literature summarized the current research base for the variable of social support as a global construct impacting parenting behaviors and paraprofessional home-visitation as a specific form of social support. Although a substantial amount of research has provided evidence in support of an overall positive relationship between social support and supportive parenting behaviors, there is much more research needed to determine the type of social support needed to benefit diverse family situations. In this case, the research 72 remains unclear regarding what form of social support facilitates optimal parenting behaviors in limited resource parents of young children. 73 CHAPTER III METHODS This chapter describes the methodology of this study. Six sections are included in this chapter: (1) Research Design, (2) Research Collection Procedures, (3) Research Sample, (4) Instrumentation, (5) Data Analysis, and (6) Ethical Considerations. Research Deslgn This research project used a two-group pretest, posttest design. The two groups included a treatment and comparison group, each consisting of a nonrandom, convenience sample. Ethical considerations prohibited this study from using random selection. Some parents involved in the treatment being utilized in this study had a court-mandated need for the intervention. These circumstances dictated that parents complete the treatment program as soon as it was logistically possible. Participants in both the treatment group and the comparison group completed a battery of pretests to determine initial scores on parenting behaviors and social support levels. Treatment group parents then participated in the BSF program for a period of 8 to 14 weeks. The mean number of sessions was 10.25. Comparison group parents did not participate in any type of intervention during this period. At the completion of the BSF program participants in both the treatment group and comparison group completed a battery of posttests to determine changes in parenting behaviors and social support levels. The research design is diagrammed in Figure six. 74 2-Group Protest, Posttest Research Design 93le EBEIESI INIEBMENIIQN EQSILESI Treatment X X X Comparison X X Eigurefi. Research Design. Data Collection Procedures Counties were invited to participate in the study in January 1999. Letters of invitation were sent to twenty-three BSF home-visitation programs in Michigan. A copy of the letter of invitation is in Appendix A. Six programs agreed to participate in the study by sending a letter of acceptance to the investigator. A follow-up letter was sent to all programs that indicated an interest in becoming involved in the study. The follow-up letter contained information regarding the intent of the study and details of the procedure. It also invited them to a training meeting. A copy of the follow-up letter is included in Appendix B. The investigator met with the staff of the participating programs at a training meeting on January 14, 1999. During this meeting the study protocol was explained to the program staff and all paraprofessionals were instructed on the use of the instruments. The program staff and paraprofessionals were not apprised of the specific hypotheses being investigated in this study. This information was omitted to help reduce evaluator bias during data collection. Data collection took place between February 1, 1999 and January 30, 2000. 75 Initial data collection occurred on the first visit with the family. During this visit the parenting instructor introduced the project to the parent and received permission to collect data. After permission had been received and a signed consent form had been obtained, each parent then was given an identification number. The identification number was used as the only identifier on all data collection instruments, so that the data was kept confidential. Parents in the treatment group completed the intake form, Family Social Support Survey and Parenting Behavior Assessment at the initial visit, prior to the delivery of the program. All of these instruments were completed through instructor interviews. The intake form asked the parent to report on a variety of variables related to the parent's demographics. This form contained such items as parents’ and children’s gender and ages, family composition, ethnicity, residence, household income and education. This form took approximately 5 to 10 minutes to complete. When completing the Family S9931 Support Survey ttbejnétmgtgruwguld, read the parent the instructions at the top of the form and then ask the parent to indicate the level of the helpfulness of each social support item on the form. The instructor circled the parent’s indicated level of helpfulness on the form. If a social support item was not available to the parent, the instructor would simply circle the letters NA to indicate that response. Parenting instructors were allowed to clarify any social support item by giving examples of people or services that would fall into a category, however they were not permitted to give comments 76 that may influence the parent’s responses. The Family Social Support Survey took approximately 10 minutes for parents to complete. To complete the Parenting Behavior Assessment parents were asked to think of one of their children and tell the instructor the age of that child. The child’s age was recorded at the top of the Parenting Behavior Assessment form. The parent was then asked to respond to items on the assessment with just that child in mind. This helped the parent respond to parenting behavior items without having to think about behaviors for children of different ages and temperaments. Next the parent was given five envelopes with each envelope representing one of the following responses: “like me—always”, “like me-often”, “like me- sometimes”, “like me-seldom”, and “like me-never.” During the assessment parents were given 32 cards with one parenting behavior printed on each card. Parents placed each of the cards in the envelope that most closely represented the parents' perception of the frequency at which he or she performs each parenting behavior. After the parent had placed all 32 cards in one of the five envelopes, the tops of the envelopes were folded over and given to the instructor. The instructor took the envelopes back to the office to score the activity. All the behavior cards placed in the envelope labeled “never“ were coded as a one. The behaviors found in the envelope labeled “seldom” were coded as a two. The scores of three and four were given to the behaviors in the envelopes labeled “sometimes” and “often”, respectively. Scores of five were given to all the behaviors in the envelope labeled “always.” The Parenting 77 Behavior Assessment took approximately 20 minutes to complete. The entire pretesting process was typically completed in 40 minutes. During the initial visit the instructors were also asked to observe or ask questions about forty-five parenting behaviors listed on the HOME Inventory. The majority of the parenting behaviors on the HOME Inventory could be observed during the visit. A few of the items needed to be asked of the parent. Examples of the items that needed to be asked of the parent included, “Is the child taken regularly to the doctor's office or clinic?” and “Is the child taken to the grocery store at least once a week?" The types of behaviors the instructor was asked to observe fell into the categories of emotional and verbal responsively, acceptance of the child’s behavior, organization of the environment, provision of play materials, parental involvement with the child and opportunities for variety. The instructors’ observations were recorded on the HOME Inventory form immediately after the visit. After the completion of the pretests, instructors began the BSF program with the parents in the treatment group. During weekly home visits, instructors would provide information and support to the parent in the areas of child development, positive discipline, parent-child interaction and parental goal- setting and problem-solving. The program would last a mean number of 10.25 weeks. However the actual number of weeks would vary from parent to parent. The range for program length had a minimum of 8 weeks and a maximum of 14 weeks. 78 At the conclusion of the program parents were asked to complete the Family Social Support Survey and the Parenting Behavior Assessment a second time, using the same protocol as was used on the initial visit. Instructors were also asked to observe the parent’s behavior and ask the questions needed to complete the HOME Inventory. Immediately after the final visit instructors would record their observations for the HOME Inventory. The comparison group for this research project was parents in the six participating counties who were eligible to participate in the BSF program, but were not currently involved in it. Many of these parents were already on waiting lists to become involved in the BSF program during the time frame of the research project. BSF instructors collected the data from the comparison group parents. The initial data collection served as an introduction to the program, allowing the BSF instructor to become acquainted with the parent. The protocol for data collection was similar to the treatment group. At the initial visit the instructor obtained a signed consent form, assigned the parent an identification number and collected the pretest data for the intake form, Family Social Support Survey and the Parenting Behavior Assessment. Instructors would also make the observations and ask the questions necessary to complete the HOME instrument. Approximately 8 to 12 weeks later, the BSF instructor collected the data a second time. Immediately following the second data collection, the instructor began program delivery. 79 All parents involved in the research project received a $10.00 gift certificate to a local department store. The intended sample size consisted of 60 families in both the treatment and the comparison group. Each group was over- sampled to help counteract the effect of program attrition. The actual number of parents recruited for the treatment group was 72 parents and the number of parents recruited for the comparison group was 66 parents. Power analysis recommends a sample size of at least 50 for an alpha of .05 and a moderate effect size of .8 (Pecora, Fraser, Nelson, McCrosky and Meezan, 1995). Research Sample The unit of analysis for this study was parents of young children. The population for the treatment group was all of the parents participating in the Michigan State University BSF program. BSF is a home-visitation, parent education program that is coordinated by Michigan State University Extension. Michigan State University is a land grant university in East Lansing, MI. In January 1999 there were 23 BSF programs that were coordinated by Michigan State University Extension. All the programs in Michigan were invited to participate in this research study. Six of the programs elected to participate in the research. The participating counties were Clinton County, Ingham County, Jackson County, Huron County, Sanilac County and VanBuren County. These counties are located in a mixture of rural and urban areas in Michigan. Between February 1, 1999 and May 1, 1999, the six participating programs invited all new enrollees in the BSF program to become part of this research study. Ninety-five parents of young children were invited to participate 8O in the study. Seventy-one of the parents agreed to become involved. This is 75% of the sample (n=71). I O I - I. ”no.1 o. o ‘; Drum .0. ._|.o: o . .0 s: A]. -.l ,0. ‘0 'I I: i: ‘-. I 0‘01 .10 I o s‘ A I0 l: I30 o 21023 Comparisons were made between the Building Strong Family enrollees who agreed to participate in the research and the enrollees who declined to participate. Independent t-tests were conducted for the ratio demographic variables. Specifically, independent t-tests were conducted for the variables of education, mother’s age, father’s age and number of children in the family. Independent t-tests determined that equal variances were assumed for the variables of participant’s education (t=.047, p=.963), mother’s age (t=-1.257, p=.221) and number of children (t=.324, p=.747). One ratio variable was significantly different at the .05 level of probability. Statistically significant differences were found in the variable of father's age (t=-2.958, p=.005). In general, male BSF enrollees who declined participation in the research were older. The paternal age for those enrollees who did not choose to participate in the research was 31.70, whereas the male BSF enrollees who agreed to become involved in the study were a mean age of 25.3 years. Chi-square computations were conducted to determine group equivalency for the categorical variables. This included the variables of ethnicity, gender, household composition, gender and income. Results from the Pearson Chi Square showed that the groups were equivalent for the variables of ethnicity (p=.222), residence (p=.698), and gender (p=.716). Statistically significant differences were found for the variables of household composition (p=.031) and 81 income (p=.017). BSF enrollees who declined to participate in the research tended to be from single parent households (57.1%) and reported earning $800.00 or less a month (68.4%). In comparison, BSF enrollees who agreed to participate in the research had 22.6% of families who were from single parent households and 40% of families reported earning $800.00 or less a month. WW Analysis was conducted on key demographic variables for BSF enrollees who agreed to participate in the research project. Means and standard deviations were computed for all the interval and ratio variables. Table 1 shows the descriptive statistics for maternal and paternal age, education and number of children. The maternal age of parents in the treatment group ranged from 15 years to 38 years, with a mean age of 22.7 years. The standard deviation for maternal age was 4.9 years. Twenty-six percent of the mothers in the treatment group were under the age of twenty (n=16). The other 74% of mothers were between 20 and 38 years of age (n=47). Fathers in the treatment group were slightly older in age, in comparison to mothers. The mean age of fathers was 25.3 years, with a range from 17 to 48 years. The standard deviation for paternal age was 5.9 years. Nine percent of fathers were under 20 years of age (n=1). The other 91% of the males were aged 20 years or older (n=7). The mean educational attainment for the parents in the treatment group was 11.6 years of education (SD=1.95). The last grade completed by parents 82 ranged from 8th grade to college graduate. Thirty-nine percent of the parents did not have a high school diploma (n=28). The other 61% had a 12th grade education or higher (n=43). The mean number of children within families was 1.7 (SD=1.1). Number of children ranged from one child to six children in the family. The largest percentage of parents had one child (62%, n=44). 83 Table 1: WWW VARIABLE RESEARCH COMPARISON T VALUE PARTICIPANTS GROUP AND P PARTICIPANTS Total Number 71 69 Maternal Age: Mean 22.7 years 25.4 years Standard 4.9 years 5.6 years -3.27 (.001) Deviation 15 years 16 years Minimum 38 years 40 years Maximum Paternal Age: Mean 25.3 years 29 years Standard 5.9 years 8.4 years -3.17 (.002) Deviation 17 years 20 years Minimum 48 years 43 years Maximum Last Grade Completed: Mean 11.6 years 11.7 years -.45 (.656) Standard 1.95 years 1.4 years Deviation 8 years 9 years Minimum 20 years 18 years Maximum Number of Children: Mean 1.7 children 2.5 children -3.60 Standard 1 .1 1 (.0005) Deviation 1 child 1 child Minimum 6 children 5 children Maximum Percentages and frequencies were computed for all the categorical variables. Categorical variables for this study included gender, residence, household composition, income and ethnicity. These descriptive statistics are reported on Table 2. 84 A majority of the participants in the treatment group were female (89%, n=63). Eleven percent of the parents in the treatment group were males. The frequency for males was 8 parents. The geographic location of participants reflected a mixture of rural and urban residents. Fifty-four percent of parents in the treatment group lived in a rural area (n=38). Forty-six percent of the treatment group parents lived in an urban area in Michigan (n=33). The dominant types of household composition were two-parent households (62%, n=44) and single-parent households (22%, n=16). Other parents in the treatment group indicated that they lived in extended family households (16%, n=1 1 ). The majority of the treatment group parents earned less than $1 ,000 a month (69%, n=49). The modal income level was $800.00 or less a month (40%, n=28). Twenty-nine percent of parents indicated that they earned between $801 and $1,000 a month (n=21). Ten percent earned between $1,001 and $1,200 (n=7). The 21% of participants earned over $1,200 a month (n=15). Treatment group parents identified themselves as from a variety of ethnic groups. Seventy-three percent of the participants were Caucasian (n=52). Eleven percent of the participants were African-American (n=8), seven percent were Hispanic (n=5), seven percent were Asian (n=5) and the final two percent of participants identified themselves as “multi-cultural"(n=1). 85 Table 2: Qemootanhrcfiharactensnmtlslflarents RESEARCH COMPARISON (P) VARIABLE PARTICIPANTS GROUP PEARSON PARTICIPANTS CHI % N % N SQUARE Gender: Female 89% 63 61% 42 .0005 Male 11% 8 39% 27 Residence: Rural 54% 38 55% 38 .523 Urban 46% 33 45% 31 Household composition: 22% 16 31% 21 Single Parent 62% 44 63% 43 .255 Two-Parent 16% 1 1 6% 5 Extended Family Monthly Income: $800 or less 40% 28 48% 33 $801-$1,000 29% 21 31 % 21 .078 $1,001-1,200 10% 7 15% 10 1,201 or more 21% 15 6% 5 Ethnicity: Caucasian 73% 52 73% 50 African-American 11% 8 15% 10 .039 Hispanic 7% 5 12% 9 Asian 7% 5 0% 0 Multi-Cultural 2% 1 0% 0 C . G D | . , Sixty-nine parents agreed to become part of the comparison group for this research study. Analysis was conducted on key demographic variables for parents in the comparison group (See Tables 1 and 2). Table 1 shows the means and standard deviations of the comparison group for all interval and ratio demographic variables. The mean maternal age of comparison group participants was 25.4 years (SD=5.6). The range of ages for mothers was 16 86 years to 40 years. Five percent of the female participants in the comparison group were under the age of 20 (n=3). The rest of the female participants in the comparison group were over 20 years of age (n=39). The mean age of fathers in the comparison group was 29 years (SD=8.4). The range of ages for males was 20 years to 43 years of age. There were not any male participants in the comparison group who were teenage parents. All of the males in this group were over 20 years of age (n=27). The mean educational level of comparison group participants was 11.7 years of education (SD=1.4). The range of education for the comparison group participants was a minimum of 9 years of education to a maximum of 18 years of education. Twenty-six percent of the comparison group participants did not have a high school diploma (n=18). The other 74% of the participants in the comparison group had a high school diploma or higher (n=51). The mean number of children in the homes of comparison group parents was 2.5 (SD=1). The range of number of children in the homes of comparison group participants varied from one child to five children. About half of the comparison group participants had one or two children in the home (54%, n=37). The other comparison group participants had three or four children (46%, n=32). Table 2 shows the comparison group’s percentages and frequencies for all the categorical demographic variables. The comparison group parents had a mixture of males and females in the sample. Sixty-one percent of the comparison group participants were female (n=42). Thirty-nine percent were 87 males (n=27). This is different from the treatment group sample where 89% of the participants were females. A slight majority of the comparison group parents lived in rural areas (55%, =38). Forty-five percent of the parents lived in urban areas (n=31). This mix of rural and urban parents is reflective of the geographic balance of the counties that participated in the study. Similar to the treatment group parents, the majority of parents lived in two- parent families (63%, n=43). Other household composition styles represented in the sample included single parent homes (31%. n=21) and extended family homes (6%. n=5). Seventy-nine percent of the comparison group parents indicated that they earned less than $1 .000 per month (n=54). This was a slightly higher percentage of parents than the treatment group. Sixty-nine percent of the treatment group parents earned less than $1,000 per month (n=49). The other twenty-one percent of comparison group parents earned $1,001 or more a month (n=15). Seventy-three percent of the parents in the comparison group identified themselves as Caucasian (n=50). Fifteen percent of the comparison group parents identified themselves as African-American (n=10) and 12% of the comparison group parents Identified themselves as Hispanic (n=9). None of the comparison group parents identified themselves as Asian or “multi-cultural”. 88 O O I O .I'.'IA.. 0 L3 .;|'.L. «0| 10 sources of social support were included. Pearson Product Moment correlations were re-calculated to see if parents with higher levels of social support would show a relationship with parenting behaviors. Again, the investigator found no relationship between amount of 138 social support and parenting behaviors. Calculations for pre-intervention scores showed a correlation of .20 and a probability of .057. Post-intervention scores showed a correlation of -.089 and a probability of .227. These analyses appear to show that the results of this study were not due to an adequacy factor. Clearly more analysis needs to be conducted to determine if a relationship exists between amount of social support and parenting behaviors and if so, what factors contribute to the relationship. I I E l . t S . I S I Since significant findings were found in this study for the dependent variables of satisfaction with social support and parenting behaviors, the investigator felt it was prudent to assess intragroup variations among subgroups of the sample. The empirical literature has found that subgroups of parents may have differing levels of social support because of environmental-societal conditions. Findings have suggested that structural factors operating through the constructs of ethnicity, culture, class, gender and household composition place constraints on the capacity of parents to build networks that adequately support them in their role of child rearing (Cochran and Niego, 1995). These constraints may yield a smaller pool of eligibles from which parents are able to build and maintain personal support networks. “African-American parents, nonethnic White parents, parents with relatively little education, and parents living in cultures shaped by beliefs that lead to narrow definitions of the women’s role, all have smaller pools of potential network counterparts. Constraints accumulate for single parents, who often have less access to relatives, further education, jobs 139 paying a decent salary, and housing in neighborhoods that are supportive of neighboring activities” (Cochran and Niego, 1995, p. 411). Independent t-tests were run to determine if there were within group differences among the sample for the variable of perceived satisfaction with social support. Analyses were conducted on pre-intervention and post- intervention scores. Key demographics variables were grouped for the analysis. Ethnicity was grouped as 1=Caucasian and 2=minority. The category of “number of children” was grouped as 1=one child and 2=two or more children. Education was grouped as 1=Iess than a high diploma and 2=eamed a high school diploma or higher. Monthly income was grouped as 1=$800 to $1,000 and 2: $1,001 and higher. The variable of household composition was grouped as 1=single parent home and 2=two-parent home. Paternal and maternal ages were grouped as 1=oz "F .EoEom ”N .moEzoEow um deco ”v .m>m>>_< ”m ”28¢. .8385 220 >E 20: 56cm mac .8: _ .>_o._mm 90.93 9 Bio >E 39:85 _ .9239. fig; EEo >E 3:286 _ .moto on 5:; Eco >E tokEoo _ .8205 ome EEo >E «2 _ .oEmee .5395 2:0 >E some“ 29 _ .EEo >E 53> moEmm Ea _ oE 9 53 .o 3:38 ome 9 Eco >E omeSooco _ .moEz 028m vcm «£35 £8 Lon— >>o=m _ .905 new 33620 5:69 .8 25.0 no .688 or: 9 Eco >E 0x9 _ .mo>mcmnm_E on coca EEo >E xcmaw _ .EEo >E omfia _ .EEo >E 53> oE: 9.6:on >o.Eo _ .550 no: :0 3:5 on 9 Eco >E ommSooco _ .26 >5 9 can. _ 58 as 52995 220 >5 9 8:9 _ .memxm you 9 8:20 m 220 >E 02m _ .>m_a o>=om E 25 >E 5:5 uo>_o>c_ 8m _ .EEEE m mm: 220 >E can; E_mo «om _ 6:2qu xma 2 Eco >E 39:85 _ .oSomm ucm 3mm 6.8 EEo >E 2o; _ .283 550 950.6 micron—Eco on 9 Eco >E 9m: _ .220 >E 5:5 mxoon «m x02 _ Sea 9 220 >E .2 moEmm a: ome _ .xofi m_ on cogs 256 no L208 9: 9 Eco >E 0x9 _ .temEE upon 2 Eco >E ommSREo _ .560 E8 2 oE: EEo >E ozm _ $5.20 .650 5:5 Ea 2 Eco >E 39:85 _ .EEo >E E2 m=E= «mm _ .53, $3 9 Eco >E ..2 355 $305 _ .coaaac meE 26: .mucmc ..o xoo_ 355 26: Scam EEo >E 9 £9 _ .65 53> .3ch on 9 220 >E ommasooco _ flan. a clean .lnomolIo “mom a c. 0. 3.9.3. ed. 6.20 to and. ”..an32 EquEan. "E3". mctoom ._o_>ucem mczconam 189 APPENDIX G UCRIHS Approval Letter 190 mgr RESEARCH AND - GRADUATE STUDIES diversity Committee as Research Involving Ream Subjects (UCRIHS! W Star Universly 46 War We Eat new. “idiom 488244046 517155-218) FAX- Slim-2976 greasy-nausea» MICHIGAN STATE 0 N I v E R s I r Y January 4, I999 TO: Dr. Dawn Contreras 103 Human Ecology RE: [RB # 94564 CATEGORY: l-C TITLE: THE INFLUENCE OF A PARAPROFESSIONAL, HOME VISITATION PARENT EDUCATION PROGRAM ON THE PARENTING BEHAVIORS AND SOCIAL SUPPORT OF LIMITED RESOURCE PARENTS ANNUAL APROVAL DATE: February 27, 1998 REVISION REQUESTED: December 18, 1998 REVISION APPROVAL DATE: January 4, 1999 The University Committee on Research Involving Human Subjects' (UCRIHS) review of this project is complete and I am pleased to advise that the rights and welfare of the human subjects appear to be adequately protected and methods to obtain informed consent are appropriate. Therefore, the UCRIHS APPROVED THIS PROJECI‘S REVISION. RENEWALS: UCRIHS approval is valid for one calendar year, beginning with the approval date shown above. Projects continuing beyond one year must be renewed with the green renewal form. A maximum of four such expedited renewals possible. Investigators wishing to continue a project beyond that time need to submit it again for a complete review.- REVISIONS: UCRIHS must review any changes in procedures involving human subjects, prior to initiation of the change. If this is done at the time of renewal, please use the‘green renewal form. To revise an approved protocol at any other time during the year, send your written request to the UCRIHS Chair, requesting revised approval and referencing the project‘s IRBII and title. Include in your request a description of the change and any revised instruments, consent forms or advertisements that are applicable. PROBLEMS/CHANGES: Should either of the following arise during the course of the work, notify UCRIHS promptly: 1) problems (unexpected side effects, complaints, etc.) involving human subjects or 2) changes in the research environment or new information indicating greater risk to the human subjects than existed when the protocol was previously reviewed and approved. If we can he of firrther assistance, please contact us at $17 355-2180 or via email. UCRIHS@pilot.msu. cdu. Sincerely, Ewawaaée David E. Wright, PhD. UCRIHS Chair new: db ‘91 CC: APPENDIX H Parent Consent Form — Treatment Group 192 Michigan State University Extension Children, Youth and Family Programs Parent Consent Form Building Strong Families Participant ‘ Name of Parent Participant Number We are conducting an evaluation to better serve parents of young children throughout Michigan. Your participation in this project is voluntary and services are still available should you choose not to participate. You have the right to refuse to answer any question or even refuse to participate at any time, without penalty. ‘ ‘ Information collected will be kept confidential to the maximmn extent of the law. Your name will not be put on any of the evaluation forms. The forms will be kept separate from anything that would allow someone to link responses to your name. Also, no individual responses will be used when reporting the results of this evaluation. You may receive a copy of the evaluation results by calling Dawn Contreras at 517-355-6586. The evaluation will take about 20 to 30 minutes of your time. It involves completing a parenting-skills assessment at the beginning of the Building Strong Families program. You will also be asked to complete the same forms at the end of the program. If at any time you have questions or concerns about this evaluation, you may call Dawn Contreras at 517-355- 6586 or Dr. Wright at 517-355-2180. I agree to voluntarily participate in this research project: Signature of Parent Date Sharcd\Dawn\Evaluation Manual\Consent Forrnwpd 193 APPENDIX I Parent Consent Form - Comparlson Group 194 Michigan State University Extension Children, Youth and Family Programs Parent Consent Form - #2 Name of Parent Participant Number We are conducting an evaluation to better serve parents of young children throughout Michigan. Your participation in this project is voluntary and services are still available should you choose not to participate. You have the right to refuse to answer any question or even refuse to participate at any time, without penalty. Information collected will be kept confidential to the maximum extent of the law. Your name will not be put on any of the evaluation forms. The forms will be kept separate fi'om anything that would allow someone to link responses to your name. Also, no individual responses will be used when reporting the results of this evaluation. You may receive a copy of the evaluation results by calling Dawn Contreras at 517-355-6586. The evaluation will take about 20 to 30 minutes of your time. The evaluation will take place while you are waiting to become involved with the Building Strong Families program. It involves completing an initial parenting-skills assessment and a social support inventory today. Then, in the next 1-2 months, you will be asked to complete the same forms a second time. To thank you for your participation in this evaluation, you will receive a $10.00 gift certificate after the final forms have been completed. If at any time you have questions or concerns about this evaluation, you may call Dawn Contreras at 517-355-6586 or Dr. Wright at 517-355-2180. I agree to voluntarily participate in this research project: Signature of Parent Date Sharcd\Dawn\Evaluation ManuahConsent Formwpd 1 95 I‘IICHIGAN STATE UNIV. LIBRARIES IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIHIIIIIIIIIIII 31293020488437