SOCIAL SUPPORT AND THE QUALITY OF CHILD CARE PRACTICES AMONG PARTICIPANTS IN THE FAMILY CHILD CARE PARTNERSHIPS PROGRAM Except where reference is made to the work of others, the work described in this thesis is my own or was done in collaboration with my advisory committee. This thesis does not include proprietary or classified information. __________________________ Robin Irene Putnam Certificate of Approval: __________________________________ _________________________________ Jacquelyn Mize Ellen Abell, Chair Professor Associate Professor Human Development and Family Studies Human Development and Family Studies __________________________________ __________________________________ Scott Ketring Joe F. Pittman Associate Professor Interim Dean Human Development and Family Studies Graduate School SOCIAL SUPPORT AND THE QUALITY OF CHILD CARE PRACTICES AMONG PARTICIPANTS IN THE FAMILY CHILD CARE PARTNERSHIPS PROGRAM Robin Irene Putnam A Thesis Submitted to the Graduate Faculty of Auburn University in Partial Fulfillment of the Requirements for the Degree of Master of Science Auburn, Alabama December 15, 2006 iii SOCIAL SUPPORT AND THE QUALITY OF CHILD CARE PRACTICES AMONG PARTICIPANTS IN THE FAMILY CHILD CARE PARTNERSHIPS PROGRAM Robin Irene Putnam Permission is granted to Auburn University to make Copies of this thesis at its discretion, upon request of individuals or instructions and at their expense. The author reserves all publication rights. ________________________ Signature of Author ________________________ Date of Graduation iv SOCIAL SUPPORT AND THE QUALITY OF CHILD CARE PRACTICES AMONG PARTICIPANTS IN THE FAMILY CHILD CARE PARTNERSHIPS PROGRAM Robin Putnam Masters of Science, December 15, 2006 (B.A., Mars Hill College, 2003) 95 Typed Pages Directed by Ellen Abell The purpose of the study was to describe the social and professional support networks of family child care providers involved in a program designed to improve child care quality, and to examine whether the extent and quality of provider social support affect the quality of their care giving practices. Study participants were 109 family child care providers enrolled in the Family Child Care Partnerships (FCCP) program--an in- home, mentor-delivered, technical assistance and education program focusing on assisting providers to increase the quality of their caregiving practices and achieve national accreditation standards. Providers were randomly selected from two groups?those currently enrolled and those who became accredited during the period of their enrollment. MANOVAs were conducted to determine whether there were changes in social and professional support and three types of child care quality from Time 1 (at the time of enrollment) to Time 2 v (12 months or more subsequent to their enrollment), and whether changes in support were associated with changes in process or global quality.. Overall, the findings indicated that for both groups of providers, levels of professional involvement and global quality increased significantly. No significant increases in levels of social support or process quality were seen for either group. Increases in professional involvement were related significantly to increases in global quality. vi ACKNOWLEDGMENTS The author would like to thank Dr. Ellen Abell for her direction, encouragement, assistance and patience during this study. Her knowledge and guidance has been, and will continue to be, invaluable. Special thanks are extended to the other committee members, Dr. Jackie Mize for her help with statistics and Dr. Scott Ketring for taking part in this research. Thanks are also due to Dr. Ellaine Miller for all of her support and answering my many questions and to Amber Paulk for her constant encouragement and advice. My family has been supporting me during the course of this investigation. I certainly would not have made it this far in life without the trust and faith of all of these individuals. vii Style manual used: American Psychological Association Publication Manual Software used: Microsoft Word viii TABLE OF CONTENTS LIST OF TABLES???????????????????????????..ix LIST OF FIGURES??????????????????????????.?x I. INTRODUCTION??????????????????????...??1 II. LITERATURE REVIEW???????...????????????..?...8 Social Support and Quality Care???????..???????????...8 Training????????????????????????????.17 Family Child Care Partnerships???????????????????23 III. METHOD???????????????????????????..27 Participants???????????????????????????.27 Procedures???????????????????????????..28 Measures????????????????????????????29 Preliminary Analysis of Non-respondents???????????????.34 IV. RESULTS???????????????????????????.35 Provider Characteristics????????????????????.??36 Preliminary Analyses?????????????????????.??38 Social and Professional Support of Providers?????.???????.?..42 The Effects of Social Support on Caregiver Quality?????????.??44 Additional Analyses?????????????????????.??..49 V. DISCUSSION??????...???????????????????.54 Implications for Prior Research?????????????..???...??55 Implications for FCCP Training for Professional Support ??..??????.56 Limitations of the Study??????????????????????.60 Implications for Future Research??????????????????...62 Implications for Childcare Professionals??????????????...?64 Conclusions???????????????...???????????.65 REFERENCES?????????????????????...???.???..68 ix APPENDICES APPENDIX A--IRB-Approved Informed Consent Form????????????..74 APPENDIX B--Social Support Survey??????????????...????..76 APPENDIX C--Family Day Care Rating Scale????????????????.79 APPENDIX D--Caregiver Interaction Scale??????????????...??..83 ix LIST OF TABLES Table 1: Demographic Characteristics of All Study Participants by Group??...37 Table 2: T-tests for Key Variables??????????????????..39 Table 3: Correlations Among Study Variables??????????????41 x LIST OF FIGURES Figure 1 Estimated Marginal Means of Number of Associations???????43 Figure 2 Estimated Marginal Means of Professional Involvement Scale????43 Figure 3 Estimated Marginal Means of Total Caregiver Interaction Scale???..46 Figure 4 Estimated Marginal Means of Positive Relationship????????.46 Figure 5 Estimated Marginal Means of Family Daycare Rating Scale?????58 Figure 6 Estimated Marginal Means of FDCRS-PQ Subscale????????.52 1 I. INTRODUCTION Defining quality in child care has become an issue of debate in the professional literature. Most professionals agree however, that quality is comprised of many aspects and caregiver-child interactions are primary among them. Recently, the relationship between children and non-parental caregivers has come under increasing scrutiny. Child care advocates and researchers have argued for the importance of attachment security and other aspects of infant-caregiver relationships as a critical aspect of quality in infant and toddler programs (Elicker, 1997; Elicker, Fortner-Wood, & Noppe,1999; Raikes, 1996). Attachment security with caregivers has been found to be associated with peer competence, empathy, and achievement orientations (e.g., Howes, 1997; Howes, Matheson, & Hamilton, 1994). In addition, children with a more secure relationship with their teacher are more likely to have more opportunities to learn because of their willingness to explore their environment (Kruif et al., 2000), and are more likely to engage in higher-level or complex play with objects more frequently (Galinsky, Howes, Kontos, & Shinn, 1994). 2 Children?s social competence is strengthened by teachers who provide a warm, sensitive, and responsive model of interaction (Katz & McClellan, 1997). A teacher?s ability to be less directive, less harsh, and less detached can affect children?s sociability, because children with this type of teacher have more positive interactions, are more considerate, and are more sociable (Phillips, McCartney, & Scarr, 1987). Children with better teacher interactions also display higher levels of language development (Whitebook, Howes, & Philips, 1990) and show more competence in cognitive activities (Howes & Stewart, 1987). From these findings, primarily the result of research done in center-based child care settings, it has been assumed that these relationships generalize across other child care environments, including the family child care home. However, the family child care setting is quite different from center-based environments, in that it involves providing care to a small group of unrelated children by an adult working for pay out of her own home (Elicker, Fronter-Wood & Noppe, 1999). Consequently, on a daily basis, providers experience unique challenges, compared to center-based care. These include caring for a wide range of ages and developmental levels, working and living in the same space, often caring for their own children side by side with unrelated children, and administering and managing their own program. According to Trawick-Smith and Lambert (1995), family child care providers are among the loneliest and least appreciated of all professionals working with children. Parental demand for family child care services is very high, partly because the care offered is seen as flexible, personalized, low-cost, independent, and negotiated directly with parents (Taylor, Dunster, & Pollard, 1999). Families who use family child 3 care tend to emphasize the child-rearing philosophy of the provider and the homelike setting to justify their preference (Pence & Goelman, 1987; Britner & Phillips, 1995). Family child care providers are often seen as ?substitute mothers.? Parents will often say they want child care that treats their children as they would be treated by a mother at home. Many family child care providers feel that motherhood is an important job qualification and others even consider it the most important and only necessary requirement (Nelson, 1990). The requirements of operating a business, however, and the fact that the children whom providers care for are not their own, create some conflicts with the ideal of motherhood (Nelson, 1990). While some may see them as little more than baby sitters, family child care providers are professionals who operate their own business, seek new clientele, and establish relationships with the parents for whose children they care. Because the ideals of good business practice and motherhood do not always go hand in hand, family child care providers participate in a delicate balancing act of professionalism and nurturing. In line with good professional practice, family child care providers can establish a professional identity by seeking out and obtaining on-going training and by being actively involved in provider networks. Research shows that caregivers providing higher quality child care are better educated in child care related areas and take advantage of in- service and training opportunities (Howes, Smith, & Galinsky, 1995; Ghazvini & Mullis, 2002). Training is thought to lead to higher quality child care and to increase the likelihood that high quality providers will remain in the field (Mueller & Orimoto, 1995). Family child care training has been shown to improve overall quality of care and enhance 4 providers? commitment to quality care. Providers with more family child care training and more professional preparation are rated as more sensitive (Galinsky, Howes, Kontos, & Shinn, 1994). However, access to training opportunities for family child care providers is limited (Taylor, Dunster, & Pollard, 1999) because of the long hours they work each day. Family child care providers must start before the general work day and continue working in order for parents to get to and from work.. This long work day makes finding time for training difficult (DeBord, 1993). Family child care providers have varying needs when it comes to training. While most training programs focus on knowledge of child development and education, others believe that training programs should focus on providing networking support and reducing the isolation that many providers experience (Taylor, Dunster, & Pollard, 1999). In workshops and training sessions, providers have the opportunity to interact and connect with other family child care providers and child care professionals. Providers themselves deem these professional network opportunities essential to the long-term enhancement of quality care and working conditions (Taylor, et al., 1999). The relevance of such social support to the nurturing aspect of quality care giving is supported in the parenting literature. Studies suggest that supportive networks are associated with adequate child rearing (Belsky & Vondra, 1989; Bonds et. al., 2002; Cochran & Brassard, 1979; McLoyd, 1990; Powell, 1979). The extent of social support that parents receive has been linked to warmth, responsiveness, and role satisfaction as parents, while parental isolation is associated with impaired family functioning (Coletta, 1979; Crittenden, 1985; Pascoe, Loda, Jeffries, & Easp, 1981). Social support can act as a 5 buffer to stressors such as emotionally negative infants by decreasing depression and increasing responsiveness and sensitivity in mothers (Mertesacker, Haverkock, & Pauli- Pott, 2004). Given the primacy of the nurturing, care giving role in family child care, its idealized connection with motherhood, and its overlap into the personal life of the family child care provider, the importance of social support established by the parenting literature suggests that social support networks may also be important in regard to the quality of care provided in the family child care setting. The purpose of this study is to describe the social and professional support networks of family child care providers involved in a program designed to improve child care quality and to examine whether the extent and quality of provider social support affect the quality of their care giving practices. Licensed family child care providers in the state of Alabama who were participating or have participated in the Family Child Care Partnerships (FCCP) program were solicited as study participants. Key goals of the FCCP program are to increase the quality of care provided in the family child care home, with an eye toward accreditation, to promote professionalism in the child care community, and to increase provider knowledge of resources--especially social resources--that support the implementation of best practices and facilitate the establishment of provider associations. Participating providers receive a mentor who provides education and support. Providers attend group meetings where they receive instructional information, but also have the opportunity to build relationships with other providers in their area. Participation in local and national associations gives providers resources for support and extends their professional and social networks. The first research question guiding this study was whether participation in FCCP does, in fact, 6 increase providers? social and professional support networks. In order to assist providers in achieving higher levels of quality and accreditation status, FCCP evaluates the progress of its enrolled providers on a quarterly basis, using assessments of quality care giving that are common in the field. Researchers have examined three main types of quality in child care. Structural quality includes elements associated with the physical qualities of the facility, teacher-child ratios, teacher experience, and teacher training. Process quality includes those aspects of care that involve behaviors between the caregiver and the child, i.e., the interactions and interpersonal processes involved in care giving. Global quality measures consider both the structure and process of the care giving environment; however, global quality measures tend to place more emphasis on non-interactive aspects of quality. The processes that occur in the classroom and the quality of these processes influence children?s well-being and development outcomes (Peisner-Feinberg & Burchinal, 1997; Whitebook, Howes, &Phillips, 1989; Cryer, Tietze, Burchinal, Leal, & Placios, 1999). However, the relationship between quality and social support is rarely examined as a primary question, and the results of the few studies published are inconsistent. For example, a study examining many facets of family child care indicated that process quality is related to social support (e.g., Galinsky et al., 1994), while a similarly extensive study found that social support was only weakly related to global quality, and this relationship depended on how social support was measured (Weaver, 2001). 7 These mixed findings in the relationship between quality and social support, sparked two of the research questions that will be examined in this study. First, do the levels of social and professional support reported by family child care providers predict levels of process quality? And second, do the levels of a provider?s self-reported social and professional support predict levels of global quality? The quality of care that children receive is an important element in their early development. The findings from this study will be useful in providing information on the success of the FCCP program, which aims to improve the quality of care for children. Study results will inform program administrators and mentors about the effectiveness of their efforts to date and indicate program strengths and needs for improvement. Findings may also help to clarify the small literature in the family child care field on the potential benefits of social and professional networks for process quality indicators. Results may lead to recommendations for other programs or training efforts intended to increase quality care giving practices among family child care professionals. 8 II. LITERATURE REVIEW The purpose of this literature review is to examine the research on family child care providers? social support and training as they relate to quality. Distinctions in quality will be highlighted referring to the three types of quality--structural, process, and global-- researchers typically use to describe child care quality. First, the literature linking social support and quality care will be reviewed. Literature concerning the relationship between the social support experienced by child care providers and child care quality is limited. Consequently, studies about both family child care and center based child care providers are included in this review. Second, research linking training to quality will be reviewed. Also included here is a review of the literature concerning training for family child care providers and its effects on quality. Finally, the FCCP program goals and training procedures will be described. Social Support and Quality Care One of the most comprehensive studies examining family child care is by Kontos, Howes, Shinn, and Galinsky (1995). Several reports have been published from this study such as Galinsiky, Howes, Kontos, and Shin (1994). The study sample was gathered through multiple methods with a final sample size of 820 mothers of children in family child care or relative care from Texas, North Carolina, and California. Groups of low- income and minority children were over sampled to make possible for individual and comparative analysis. Mothers were then asked to refer their family child care provider. 9 Providers received one three-hour visit, usually during morning hours, when the target child (the child of the mother included in the study) was awake and engaged in typical activities. Observational information collected included adult-child ratios, the Care Giver Interaction Scale (Arnett, 1989), the Family Day Care Rating Scale (Harms & Clifford, 1989), attachment between caregiver and child, and four five-minute time sampling observations of the target child. A 12-page questionnaire was given to the provider at the end of the visit with a self-addressed envelope. The questionnaire assessed daily hassles, child rearing practices, job satisfaction, and job commitment. The Questionnaire on Social support (Crnic, Greenberg, Ragozin, Robinson, & Bashm, 1983) is a 16-item questionnaire, which assesses the availability and satisfaction of providers with their social support at three ecological levels: community, friendships, and intimate relationships. A measure for social integration into the family child care community was developed for this study, using the same format as the Questionnaire on Social Support (Crnic et al., 1983) with content specific to family child care and relative care providers. Results descriptive of the social support of family child care providers found that, as a group, providers in the study described themselves as only somewhat involved in their neighborhoods and with organized community groups; however, most (96%) felt that there was someone in their social network with whom they could share anger and happiness (Kontos et al., 1995). A composite variable for general social involvement was created from the individual items. The score could range from -1 to +1 and the providers had an average score on general social involvement of -.17. This score suggests that on average the family child care providers were slightly uninvolved. A composite score was also developed for satisfaction with general social support. Kontos and colleagues (1995) 10 found that the average for satisfaction was -.01, showing that providers were neither particularly satisfied nor dissatisfied with their social support. In regard to social integration into the family child care community, 25% of providers knew no other family child care providers and 42% had no contact with any other family child care providers during an average week. Over half of the providers had no contact with organized groups of family child care providers (Kontos, 1995). Providers who report more involvement with the family child care community (e.g., they belong to family child care associations, or participate in the Child and Adult Care Food Program) and are involved with other child care providers, are rated as more sensitive and responsive, and also more likely to be rated as offering good and adequate custodial care than those less involved (Galinsky et al., 1994). Kontos and Riessen (1993) explored child care quality from a socioecological perspective, asking how family day care providers? job satisfaction, commitment, and stress were predicted by providers? personal characteristics, child-rearing preferences, and characteristics of their program. The sample included 380 licensed family daycare providers in the state of North Dakota who returned mailed questionnaires. The majority of the providers were married, white, and belonged to a sponsored food program. Forty- six percent lived in metropolitan areas and 56% lived in rural areas. Researchers measured social support using the Questionnaire on Social support (Crnic et al. 1983). Other personal characteristics measured included age, experience, proportion of income from family day-care, education, training, child rearing preferences, and social support. Program characteristics measured included size, structure, materials/activities and fees. The outcome variables for this study were providers? reports 11 of job satisfaction, job stress, and job commitment. Results indicated that personal characteristics predicted job satisfaction and job stress. Perceived social support predicted all three aspects of a provider?s job attitudes-- satisfaction, commitment, and stress--such that increases in social support lead to more positive job attitudes. Providers who perceived more social support were more likely to be satisfied with their work, perceive less job stress, and be more committed to their jobs. Those with lower perceived social support reported less commitment and earned a lower portion of their household income from family child care. Authors suggest that the findings, over all, are consistent with the ecological framework?s suggestion that job satisfaction comes with a good fit between care givers? skills/needs and the demands of the work environment. The most important finding from this study with implications for the current study is that these data suggest that social support is an important characteristic of family child care providers relative to their job attitudes in satisfaction, commitment, and stress. However, the relationship drawn by the researchers between social support and child care quality is only inferred since they do not measure quality. By showing the relationship between social support and job satisfaction for family child care providers this study helps us see the importance of continuing research in the area of social support. In a study looking at a partnership between family child care programs and Early Head Start, Buell and Gamel-McCormick (2002) intensively interviewed four family child care providers in Delaware who had received weekly visits from an Early Care and Education Coordinator. Together with the Early Care and Education Coordinator, the providers developed a quality improvement plan, which addressed materials, training, 12 and knowledge needed to improve the quality of care they provided. The providers also received a stipend to purchase materials for the children in their care or to attend training and workshops in order to meet the goals of the quality improvement plan. The four family child care providers involved in the program were selected because they were considered leaders in the family child care community. Researchers gathered data through a semi-structured, open-ended interview describing their experiences with the program. The interview questions inquired about the benefits of participation in the program, as well as other issues, such as perceptions of care giving as a profession, unmet needs, and suggestions for how these needs could be met. Providers reported that their primary benefit from being in the program was the social support-- both instrumental and emotional-- that they received. Providers identified several types of instrumental support, such as assistance organizing all aspects of their program, curriculum development ideas, acquiring materials and equipment, and financial support for their own educational opportunities. Emotional benefits included building a relationship with their mentor and increasing their sense of self-esteem. One of the ways the providers felt the program could be more helpful would be to support the family child care provider networks and support groups that already exist in the community. The features of the program Buell and colleagues (2002) studied has many similarities to the program under consideration in the current study. The partnering of a mentor with the family child care provider as part of training is one of the most important similarities, and the social support the providers gained from this relationship has many possible implications to the current study. The finding that providers wanted more social 13 support and social networks shows the pertinence of further examining these relationships. In order to gain a better understanding of the factors related to care giving in family child care, Fischer and Eheart (1991) looked for separate and interactive effects of support networks, training, demographics, business practices, and job stability on family child care giving practices. A sample of providers in Kern County, California was stratified for urban and rural setting and randomly selected. The resulting sample consisted of 177 licensed and unlicensed providers. Telephone interviews collected data for predictor variables (support networks, training, demographics, business practices, and job stability). Support networks were measured using a coding system that gave one point for support services used such as child nutrition program, professional association, book loan, and county referral services. Training was divided into 8 levels, with level 1 indicating no training to level 8 indicating a 4-year degree related to child care and additional hours of training. From the telephone sample 90 participants were selected for home observations of which 36 participated. The observations averaged 3 hours to evaluate care giving practices using the Family Day Care Rating Scale (Harms & Clifford, 1988). The results concerning predictors of care giving practices show that training explains most of the variance in care giving practices, 52.5%. Support networks followed as the next significant predictor, explaining 12% of the variance. While this study does not measure process quality it did see significant change in global quality and reinforces the importance of support networks and training for family child care providers. 14 While Weaver (2001) explored many aspects of family child care providers, the questions concerning social support are the most pertinent for the current study. Weaver asked if and how licensed family child care providers? social support network resources are predictive of quality, if there are particular supportive resources that the majority of licensed providers find more helpful than others, and what the most important source of support available to licensed family child care providers is. All providers in the sample were licensed family child care providers in a particular county of Wisconsin and reflected the range of urban and rural communities of the county. A random sample of licensed providers were contacted, and a participation rate of 35% led to a sample of 65 currently practicing family child care providers (97% white) who have been providing care for at least one year and have a spouse or partner. Providers received one three-hour morning visit from an observer evaluate typical activities. The outcome variable for this study was global quality, as measured by the Family Day Care Rating Scale (Harms & Clifford, 1989). Predictor variables included demographics, adult attachment style, commitment to family child care, education and training; psychological resources consisted of well-being and depression. Social Support Network Resources were measured using an adaptation of the Sources of Help Questionnaire (Wan, Jaccard, & Ramey, 1996) where providers rated 15 predetermined sources of social support on a five-point scale. For example, ?please rate the helpfulness of your relatives as 1 (not helpful at all) to 5 (extremely helpful) or 0 (unavailable to you)?. Also, providers were asked to rate their agreement to a statement concerning the availability and willingness of others in their community to support them as a child care provider and to rate the support they received from their spouse or partner 15 on a five point scale (1 very unsupportive to 5 very supportive). In addition to the quantitative information taken, providers were asked an open ended question: ?Who is of most support to you in family child care and why?? Significant correlations were found between supportive social network resources and psychological well-being, depression and commitment to the profession, but not with quality of care. Provider rankings of quality of social support did not predict quality of care when examined in regression and multiple regression analyses. The total number of supportive resources available to licensed providers did predict quality of care; however; this relationship became non-significant when adjusting for family income.. When other variables were controlled for in a multiple regression analysis, psychological well-being was the strongest predictor of quality. Thus, the way social support was measured, whether as a global score or an indicator of the number of people identified as sources of support, affected the way it related to quality. Further insight into the possible relationships between social support and process quality as observed in center-based care was provided in an examination of predictors of quality care for young children between the ages of 15 months and three years (Ghazvini and Mullis, 2002). Researchers looked at process quality, structural quality, and global quality Thirteen of 30 child care centers randomly selected from the north Florida area participated. Social support was conceptualized as a caregiver characteristic and measured using the Questionnaire on Social Support, (Crnic et al., 1983). Other caregiver characteristics included perceived stress, and how caregivers described their job. Process quality was measured by the Child-Rearing Practices Report (Rickel & Biasatti, 1982) and the Caregiver Interaction Scale (Arnett, 1989). The components of structural quality 16 that were measured include hours of child contact, weekly salary, benefits, specialized training, use of planned activities, adult-child ratio, group size, use of child-designed space, and extent of housekeeping responsibilities. Global quality was measured using the Infant/Toddler Environmental Rating Scale (Harms et al., 1990). Structural quality was positively correlated with process quality, and global quality was positively correlated with both structural and process quality scores. Consistent with Kontos et al. (1995) caregivers in this sample reported only moderate levels of social involvement and satisfaction with their social support. Providers? satisfaction with social support was found to be negatively correlated with perceived stress. Findings from subsequent multiple regression analyses, which did not include social support, indicate that process quality was predicted by specialized training and low caregiver stress. Although this study does not show a relationship directly between care giver social support and quality, its findings about the relationship between satisfaction with social support and perceived stress and between perceived stress and process quality suggest asking whether social support may affect process quality by acting through its relationship with providers? stress (Ghazvini & Mullis, 2002). Several of these studies show that providers are not highly socially involved and many times isolated (Kontos, Howes, Shinn, and Galinsky, 1995; Galinsky, Howes, Kontos, and Shinn, 1994; Ghazvini and Mullis, 2002). Researchers have included measures of social involvement and social networks as part of their research, recognizing that social resources might play a role in providers? quality of care. However, rarely in the current literature is the extent or kind of social support available to providers examined a priori as a predictor of family child care quality. In addition, the concept of 17 quality has been variously measured in terms of process, structural, and global quality. There is some indication that process quality is related to social support (e.g., Galinsky, et al., 1994), but not to global or structural quality (Ghazvini & Mullis, 2002; Weaver, 2001). There is a gap in the family child care literature regarding the relationship between social support and quality; however, each study points to the possible importance of such a relationship and the need for further study. Training It is widely assumed that training can enhance the knowledge, skills, and practices of child care providers and lead to higher quality care. That is one of the goals of the Family Child Care Partnerships program, as it provides training for family childcare providers. Research that reports on the results for quality care giving of specific training programs for family child care providers, however, are limited. The first work to review comes again from the research activities of Kontos, Galinsky and their colleagues. Family child care providers who sought training had many of the same characteristics as the those who sought a professional network (Galinsky et al.,1994). Providers who sought training were among those who offered higher quality, warmer, and more attentive care. Providers with more family child care training were rated as more sensitive and less detached. Also, Galinsky et al. (1994) propose the conclusion that training providing constructive feedback helps improve practice, stimulates new ideas, and sparks renewed motivation. Training and its relationship to caregivers? childrearing attitudes and behaviors toward the children in their care were the areas of interest for Arnett (1989). He examined how differences in attitudes and behavior might exist between caregivers with different 18 levels of training. Data were collected from providers participating or intending to enroll in the Bermuda College Training Program, which is a two-year program including courses in communication, child development, child care business and preschool activities. The sample included 59 caregivers in 22 day-care centers on the island of Bermuda. All centers providing care on the island of Bermuda were recruited, and all but one participated. Caregivers were observed for two 45-minute periods on different days by different observers, who rated caregiver behaviors using the Caregiver Interaction Scale, developed by Arnett for this study. The measure contains 4 sub scales: Positive Interaction which rated warmth, enthusiasm, and developmental appropriateness of communication, Punitiveness which contained items concerning hostile, threatening, and harshly critical behavior, Permissiveness which contained items reflecting a lax approach to misbehavior, and Detachment which rated uninvolvement and disinterest in the children. Arnett also categorized provider training into four levels: level 1, no training; level 2, two courses of the program; level 3, all four courses of the program; and level 4, a 4-year degree in Early Childhood Education. Analyses of variance and co-variance indicated that caregivers with more extensive training (level 4) were rated higher on the Caregiver interaction Scale than caregivers at the other levels. Caregivers completing half (level 2) or all of the program (level 3) were less authoritarian in child rearing attitudes and more positive and less detached in their interaction with children than caregivers with no training. That there was no difference between levels two and three could be related to the fact that the first half of the program which providers in both levels 2 and 3 completed, focused on 19 communication and child development, issues directly relevant to interaction with children (as apposed to business practices, which was part of the training of the second half of the program). These finding show that level of training and possibly content of training have important influence on caregivers interaction quality with children. Arnett showed that care givers with training have higher quality, but which family child care providers seek out and benefit from training, and which drop out of training? These questions were addressed in an evaluation of the Family-to-Family training program (Kontos, Howes, & Galinsky, 1996). Each Family-to-Family program is developed by the community and family childcare providers to address local needs, by adapting existing curriculum to meet the prioritized standards and topics. The training involved instructional class time which totaled to about 15-25 hours, as well as home visits, the amount of which varied between sites. Study participants were recruited from San Fernando Valley, California, Dallas, Texas, and Charlotte, North Carolina as they enrolled in the Family-to-Family training program and numbered 130 providers. Another group of 112 regulated providers, not participating in a Family-to-Family program were identified through parent referrals and licensing lists and served as baseline data for typical child care quality. Pre- and post-training data were collected from the training group through interviews and a three-hour observation of quality of care. In each home, a target child was identified, usually the youngest child in care over 10 months. Time sampling observations were recorded for this child during the first hour of the visit. At the end of the visit a questionnaire was left with the provider. 20 Aspects of the family child care environment were measured as predictors of participation in training, including organization of the family child care home (e.g. organization of time, planning, and business and safety practices) and motivation to provide care. The outcome variable of quality was conceptualized in terms of process quality, structural quality and global quality. Process Quality was assessed using the Caregiver Interaction Scale (Arnett, 1989) and the Adult Involvement Scale (Howes & Stewert, 1987). Structural quality was measured with years of experience, training, and child to adult ratios. Global quality was assessed using the Family Day Care Rating Scale (Harmes & Clifford, 1989). After comparing providers who sought Family-to-Family training and typical regulated providers, no significant differences between the two groups in demographic characteristics were found. Providers who sought training were very similar to typical regulated providers, although the training group was more likely to see family child care as a stepping stone to other employment. Those that remained in training used more business practices than those who dropped out. Also, providers who completed training were more experienced in family child care than drop-outs. Post-training improvements included increases in business and safety practices and global quality. Process quality did not increase in post-tests for providers who participated in training. Findings suggested that classroom-based training can increase some aspects of quality practice, but appear to be less effective in achieving changes in process quality. Raikes, Raikes, and Wilcox (2005) examined subsidy receipt, regulation, and provider characteristics, including training, as they relate to quality in family child care homes. The stratified sample included 120 randomly selected family child care providers 21 from the state child care division files of, Kansas, Missouri, Nebraska, and Iowa. Phone interviews were conducted to collect general information about the provider and the type of child care provided. Observations provided data concerning global quality and caregiver sensitivity (process quality). State child care divisions provided additional data about the amount of regulation and subsidy dollars received by the provider and the number of children receiving subsidy. Data were also collected on education and training hours. Education and training are conceptualized as separate variables.. Global quality was measured using Family Day Care Rating Scale (Harms & Clifford, 1989). Caregiver sensitivity was measured using the Caregiver Interaction Scale (Arnett, 1989). Results indicated that providers who were more regulated had higher global quality and were more sensitive (i.e. had higher process quality). Providers who cared for a higher proportion of children receiving subsidies had lower global quality and process quality. Training hours did not predict process quality. Overall, higher education levels, more training, higher regulation, low subsidy density, and high quality tended to be related to each other. However, training hours did not predict process quality. Regulation moderates the association between education and process quality so that higher regulation combined with education resulted in higher process quality. Finally, Taylor, Dunster, and Polland (1999) examined issues that affect family child care providers as a specific population, hoping to provide better training for them. The study was designed to produce descriptions of current Canadian practice with respect to various forms of training for family child care providers, and deepen an understanding of what caregivers, parents, trainers, and other stakeholders see as the key training issues. 22 The population sampled included 298 Canadian?s involved in family child care who participated in focus groups. Written information from 258 organizations that offer training was also collected. Other focus groups included trainers, agency staff and parents who use family child care. Researchers coded each interview, looking for over arching themes, agreements, and disagreements within and between groups. The focus groups gave insight to the issues surrounding training for family child care providers. Caregivers frequently mentioned several barriers to training. These barriers include: time, financial costs, that training does not meet their needs, lack of recognition, negative public attitudes, isolation, limited awareness, and availability of training. On the positive side, another issue that frequently arose was social support. Caregivers pointed out that the social support they received from family, friends, other caregivers, and agencies and associations is what was most helpful to them when starting out. When asked what they felt they needed most, established, experienced caregivers mentioned networking with other caregivers in twice the number of focus groups compared to any other resource. Providers emphasized opportunities for networking as part of an ideal training program. Researchers concluded that effective training, to be useful, must link core content to context at every opportunity, and that meaningful training can not separate the process of imparting necessary information and skills from individual collective caregiver empowerment. Overall, findings suggest that family child care providers have different needs for training than the ideas, measures, and training currently recommended based on center-based child care practices. These needs include, for example, more flexible times for training, and dealing with the overlap between home and work. 23 The discrepancy between the training provided in the programs reviewed and what family child care providers need (Taylor et al., 1999) may account for the lack of change in process quality indicated by previously reviewed findings (Kontos, Howes, & Galinsky, 1996; Raikes, Raikes, & Wilcox, 2005). Only Arnett (1989) found his training regimen to be related to increases process quality. Taken together, these studies seem to suggest that level of training, content of training, and type of training have important influences on caregivers. It can be inferred from several of the studies (Kontos, Howes, & Galinsky 1996; Raikes, Raikes, & Wilcox, 2005; Taylor, Dunster, & Polland, 1999) that classroom-based or group-delivered training may not affect process quality. Even though process quality was not always affected, global and structural quality typically saw some increase, showing that training does benefit providers. But the question still remains: How can training increase process quality, that is impacting more than just what the provider knows, to include how she interats with the children in her care? Can training facilitate improved social networks and social support in ways that affect process quality? The current study will investigate these questions through an examination of a large, statewide training program specifically designed to improve the quality care practices of family child care providers through in-home mentoring, the use of targeted financial incentives, and the development of professional networks. The final sections of the review describe the Family Child Care Partnerships. Family Child Care Partnerships Family Child Care Partnerships (FCCP) is a statewide, in-home mentoring program funded by the Alabama Department of Human Resources, operated through the Department of Human Development and Family Studies at Auburn University. The 24 primary goal of the program is to increase the quality of care family child care providers offer and to assist them in attaining national accreditation standards. Participants in the program are licensed caregivers who provide in-home care to young children for a fee. Participation is voluntary and those involved receive weekly in-home training and technical assistance from trained program personnel (mentors). Training spans a range of topics including, for example, health, safety, arrangement of the space for child care, facilitation of children?s language, reasoning, and numeracy, activities for mixed-age groups, and business practices for home-based child care. Several methods are used to deliver this training. One of the primary methods mentors use is modeling activities and skills for providers. Group meetings and promotion of professional associations are other ways mentors seek to improve quality. FCCP mentors are trained to conduct individualized, in-home training on a weekly basis, for a period of time varying according to the individual needs of the providers. The average length of an individual mentoring visit is between 2 and 3 hours, but can range for 1 to 5 hours. Originally, 22 full-time mentors were assigned to work with a caseload of providers in their respective regions of the state. Over time, some mentors have left the program, with new mentors joining the corps of mentors. On average, mentors see 8 to 10 providers per week, but a mentor?s caseload can range from 5 to 15 providers. Over 500 providers have participated in FCCP since its inception in 2000, and currently 220 providers are enrolled and receiving mentoring services. The average provider, once enrolled in FCCP, participates for 22 months. Providers are recruited into the program through connections mentors made with other child care-related agencies 25 and groups. The length of time a provider remains in the program may vary. Providers are encouraged to participate as long as they continue to show some measurable progress. Providers receive visits from only one mentor, the mentor responsible for the region in which she lives. Some providers worked with more than one mentor if their original mentor had left the program. In summary, social support has not been consistently measured nor clearly defined in attempts to understand relationships in family child care. Also, training has not been shown to affect process quality with the exception of Arnett (1989). Therefore, it is important to understand more about the types of training that would lead to increased process quality. Further, taking into consideration the findings concerning the relationship between social support and child care quality, it remains unclear whether improved social networks would have any effect on caregiving quality of any type. Because FCCP is a long-term mentoring program, designed to improve both the quality of care and the quality of providers? social and professional networks, an opportunity to address these questions exists. The following research questions and hypotheses are proposed: RQ1: Does participation in the Family Child Care Partnerships mentoring program increase the perceived amount of social and professional support reported by family child care providers? H1: Extended participation (over 12 months) in FCCP will predict higher levels of provider-reported social and professional support than limited participation (less than 9 months). H2: The achievement of accreditation status among FCCP providers will predict 26 higher levels of social and professional support than non-achievement of accreditation status. RQ2: Do providers? levels of social and professional support affect family child care provider process quality? H3: Higher levels of reported social support will predict higher process quality. H4: Higher levels of reported professional support will predict higher process quality. RQ3: Do providers? levels of social and professional support affect family child care provider global quality? H5: Higher levels of reported social support will predict higher global quality. H6: Higher levels of reported professional support will predict higher global quality. 27 III. METHOD Participants Study subjects were licensed family child care providers from the state of Alabama currently, or at one time, enrolled in the Family Child Care Partnerships (FCCP) program. They were selected from a list of past and present FCCP participants provided by the Managing Director of FCCP. The primary selection criterion required that a provider have a Social Support Survey (taken upon enrollment into the program) on file. A secondary objective in the selection process was to identify providers who had dropped from the FCCP program after participating less than nine months. Anecdotal information gathered from FCCP staff suggested that providers who were unable to remain in the program for more than nine months experienced less connection to their mentors and did not attend the professional development opportunities provided for them (E. Abell, personal communication, October 2005). Three mutually exclusive groups of study subjects were identified: (1) 33 providers who dropped out before the nine-month mark; (2) 105 non-accredited, currently enrolled providers who, as of January 2006, had participated for at least 12 months; and (3) 26 accredited providers (i.e., providers who had achieved accreditation by the National Association of Family Child Care through the FCCP program). These groups are referred to, respectively, as: (1) Dropped; (2) Current; and (3) Accredited. 28 Procedures Data for the study were collected at two different time periods. Time 1 data were collected as part of standard FCCP program enrollment procedures. Newly enrolled providers entering the FCCP program were asked to complete a demographic data questionnaire and a social support interview with their mentor. In addition, within the first three months of a provider?s participation in the program, mentors provided evaluation data on the quality of providers? child care practices Thus, Time 1 data were already collected and entered for all study participants prior to undertaking the current study. For the Current and Accredited providers, the most recent 3-month evaluation was used to provide Time 2 indicators of caregiving quality. (It was decided that Time 2 quality evaluations would not be collected from the Dropped group.) Thus, Time 2 data were collected only on social support from providers in all three groups. Mentors collected these data from the Current and Accredited groups as part of their normal 3- month data collection procedures. Data from the Dropped group were collected through a mail survey. A packet containing a letter inviting participation in the study and two copies of an IRB-approved informed consent form (see Appendix A), the social support survey (see Appendix B), and a pre-addressed, postage-paid return envelope was mailed to the Dropped group. Payment of $10 was offered as compensation for those providers who elected to return a completed survey, as well as entry into a drawing for $50 of materials from the Lakeshore Learning Company. 29 Out of 34 Dropped providers, 2 sent back their surveys. Follow-up phone calls to solicit participation of those not returning surveys were made to 32 providers. Only 11 of completed phone calls were successfully made, as some phone numbers had been disconnected or were not answered. Of the 11 successfully contacted, no additional dropped providers agreed to participate. The resulting participation rate for the Dropped provider group was 6%. Members of the Current group were receiving mentoring visits at the time of data collection and had a Time 1 social support survey on file. Out of 105 contacted, 87 returned a Time 2 social support survey, for a participation rate of 83%. Members of the Accredited group achieved national accreditation status through their participation in FCCP and receive only occasional visits?usually quarterly--from a mentor. Out of 26 contacted, 22 returned a Time 2 social support survey, for a participation rate of 85%. As a result of the low participation rate and small number of returned surveys from the Dropped group, the research questions of this study were altered to focus on only those providers currently participating in FCCP and the possible differences between Accredited and Non-accredited groups on issues of quality and social support. Thus, data were examined for only two groups of providers, renamed as the Non- accredited group (N = 87) and Accredited group (N = 22). Measures Demographic data collected from the original FCCP enrollment surveys include age, gender, race, marital status, household and child care income, and rural/urban residential status. In addition, assessments of three types of child care quality were constructed with the existing data, and several indicators of social and professional 30 support were derived from the Social Support Survey. The quality assessments were announced visits conducted by the provider?s mentor. Mentors assess quality through observational measures once every quarter. Time 1 assessment was taken within the first three months of the provider?s enrollment into FCCP; Time 2 assessment was the most recent assessment on record for each provider at the time of the study. Assessments of Caregiving Quality Structural Quality. Studies have defined structural quality in a variety of ways. Most studies include teacher: child ratios, teacher experience, teacher training, and teacher education. For this study the elements of structural quality that were assessed include the type of family child care home operated by the provider?either group home (more than six children and an assistant) or family child care home (6 children or fewer and at least one caregiver)?the number of years of caregiving experience, and the provider?s level of educational achievement. This information was self-reported by providers on the demographic questionnaire that they completed at enrollment. Process Quality. The Caregiver Interaction Scale (CIS; Arnett, 1989) measures the quality of the interactions between the caregiver and the children in her care. The scale contains 26 items that assess the caregiver-child relationship on a four-point scale. It is an observational measure and the observer rates the caregiver on each item as (1) statement does not at all describe the provider, (2) statement describes the provider somewhat, (3) statement describes the provider quite a bit, or (4) statement describes the provider very much. Mentors were instructed to spend between 8 and 10 hours of observation with each provider before completing a CIS. 31 The CIS has four subscales: positive interaction, punitiveness, permissiveness and detachment. The positive interaction subscale includes items such as, ?listens attentively when children speak to her.? The punitiveness subscale includes indicators of behaviors associated with harsh punishment. An example of an item from this scale is ?speaks with irritation or hostility to the children.? The permissiveness subscale reflects behaviors associated with lax discipline, such as, ?Doesn?t reprimand children when they misbehave.? The detachment subscale reflects uninvolved or neglectful behaviors, such as, ?Doesn?t seem interested in the children?s activities.? Chronbach?s alphas were examined for each of the four scales (positive relationships ? = .90 for Time 1 and .90 for Time 2; permissive ? = .441 for Time 1 and .57 for Time 2; punitive ? = .55 for Time 1 and .70 for Time 2; detached ? = .44 for Time 1 and .46 for Time 2). Because of the generally low reliability for the subscales, the total CIS score was analyzed for reliability; the results show good reliability for the scale as a whole (? =.83 Time 1 and ? =.84 Time 2). For the purposes of this study, the Positive Relationship subscale and the Total CIS were used as indicators of process quality. Global Quality. Global quality was assessed using the nationally standardized Family Day Care Rating Scale (FDCRS; Harms & Clifford, 1989). The FDCRS is an observational measure containing 32 items, scored using a 7-poing-scale, and divided into six main sections (such as space and furnishings, basic care, learning activities, etc.). All items offer a description for scoring at the 1, 3, 5, and 7 anchors of the scale with 1 meaning inadequate, 3 meaning minimal, 5 meaning good, and 7 meaning excellent. Each anchor contains specific requirements for the score given. If a provider does not meet the requirements, she can not move on to the next highest score. Once all requirements for 32 the anchor score are met, the provider moves to the next anchor. An even number as a score on a single item indicates that the provider met all of the requirements for the previous anchor but not all of the requirements of the next anchor. Average scores ranging from 5 to 7 indicate high quality care, 3 to 4.9 indicate average quality care, and 1 to 2.9 indicate inadequate quality. For the purposes of this study, the single indicator of global quality is represented by the average of the total number of items scored. Mentors were trained to use the FDCRS by working through a video-based training provided by the FDCRS developers, reviewing scale items with a trained and experienced user of the scale, and practice observations in the field. Mentors were instructed to spend between 8 and 10 hours of observation with each provider before completing the FDCRS. Reliability data on mentor observations are not available; however, when the distributions of total FDCRS scores are examined across all the providers in each individual mentor?s caseload, there is substantial variance; this suggests that mentors are using the measure discriminately (Miller, 2005). Indicators of Social and Professional Support Indicators of social and professional support were taken from questions common to the Time 1 Perceived Stress and Social Support Interview and the Time 2 self-reported Social Support Survey. Providers answered five questions about the extent of their knowledge and use of other providers and provider networks as resources for themselves and their work. In addition, they completed a social network grid adapted from Bost (1995) containing a list of individuals the participant names as providing social support. Providers are asked to name ??all the individuals you know well or see on a daily or regular basis.? The grid includes the list of people in the social network, their relationship 33 to the provider, positive or negative influence, and amount of influence. (See the complete Social Support Survey in Appendix B). Social Support. The social network grid was reviewed and coded to identify the relationship of the supporter to the participant. Each individual indicated on the grid was categorized as being one of the following: a child care professional, a family member, a friend/acquaintance, a parent of an enrolled child, or an employee. The total number of persons identified within each category was summed, to represent the extent of different types of social support. However, preliminary analysis showed that these subcategories were not correlated with variables representing any of the three types of quality. The influence rating that the provider assigned each individual listed on the grid had a very low response rate (20 of the 109 participants). Thus, a single indicator, Total Number of Contacts, was constructed and used to represent the level of provider social support. Professional Support. Providers? answers to five questions from the Social Support Survey about the extent of their knowledge and use of other providers and provider networks were coded and summed to create a cumulative score on the Professional Involvement Scale (PIS). The questions asked were, (1)?How many other family child care providers do you know in your area/community?? (2)?How many of these providers do you feel you could call if you had a question or concern related to your work?? (3) In general, how often do you talk with another provider about your work?? (4)?How often would you say you take part in these meetings or activities?? (5)?Are you currently a member of any kind of child care provider organization?? Answers for the first four questions were coded as 0 through 4; the last question was coded as 0 equals no and 4 equals yes. A PIS score was computed by summing coded items. Higher total 34 scores represent more extensive knowledge of and involvement in professional networks. Reliability tests were performed on the PIS, resulting in a Chronbach?s alpha of .65. An additional question from the Social Support Survey asked providers to report the number of child care associations of which they are a member. This indicator of professional support is referred to as Number of Associations. Preliminary Analysis of Non-Respondents Chi-square and t-test analyses were conducted to examine whether differences existed at Time 1 on demographic indicators, social and professional support, and quality variables between the 34 members of the Dropped group and the combined members of the Current and Accredited groups at Time 1. The groups differed significantly on the social support indicator, Total Number of Contacts (t = 11.15). The combined members of the Current and Accredited groups had a mean Total Number of Contacts of 6.25, whereas the Dropped group had a mean of 2.33. There were no other significant differences between the groups. Chi-square and t-tests were also conducted to examine differences at Time 1 between all respondents and all non-respondents. Non-respondents differed significantly from respondents on Total Number of Contacts (t = 2.86). Respondents had a mean of 5.99 compared to the non-respondents, whose mean was 4.34. No other differences were found for any of the demographic indicators, social and professional support, and quality variables. 35 IV. RESULTS Frequencies describing study participants based on their information at the time of enrollment are presented, along with the results of chi-square analyses comparing the Accredited and Non-accredited groups on demographic data . T-tests examining mean differences between Accredited and Non-accredited groups were then performed for quality and support variables at Time 1 and Time 2. Following this, bivariate relationships among the key quality variables at Time 1 and at Time 2 are examined. As a result of the non-participation of providers who left the program prior to participating for nine months, the original study questions, which focused on the differences in quality and social support depending on participation, could not be addressed. Instead, the questions were revised to explore the differences between providers who had achieved accreditation status through program participation and those who had not yet done so. The results for the following questions are presented. 1. What are the differences between those providers who became accredited and those who did not in the amount of social and professional support they report, and do these differences change over time? 2. Does length of time in the program predict higher levels of provider- reported social and professional support from Time 1 to Time 2? 3. Do changes in provider levels of social and professional support from Time 1 to Time 2 affect process quality? 36 4. Do changes in provider levels of social and professional support from Time 1 to Time 2 affect global quality? Provider Characteristics Table 1 contains descriptive data on all of the providers participating in the study, as well as results of chi-square analyses comparing the demographics of the Accredited and Non-accredited groups. Just over half of all participating providers were African- American or another minority. Almost two-thirds of providers were over 40 and more than 80% reported being married at the time of enrolling in FCCP. Provider residence was divided nearly equally among those living in a rural area, a town or suburb, and a city. Half of the providers who reported their annual household income made over $30,000, while the rest made less than that. As a group, 71% reported their annual income from their child care business alone at $20,000 or less. However, over half of the Accredited group reported making over $20,000 in child care income yearly; in contrast, more than two-thirds of the providers in the Non-accredited group reported making less that $20,000 or less. This was the only demographic characteristic on which the Accredited and Non-accredited groups differed significantly. 37 Table 1 Demographic Characteristics by Group and Chi-Square Analysis Characteristic Non-accredited Accredited ? 2 All Ethnicity N=79 N=21 .309 N=100 White 46% 52% 47% Minorities 54% 48% 52% Age N=79 N=21 .810 N=100 Under 40 39% 29% 37% 41and over 61% 71% 63% Marital Status N=75 N=21 .033 N=96 Married 82% 81% 81% Single 1 % 19% 18% Geographic residence N=76 N=21 .782 N=97 Rural area 32% 43% 34% Town/suburb 33% 33% 33% City 30% 24 % 30% Total household income N=72 N=21 2.823 N=93 $30,000 or less 54% 33% 50% $30,001 and over 46% 67% 50% Child care income N=69 N=19 9.356** N=88 $20,000 or less 78% 42% 71% $20, 001 or over 22% 58% 29% Education N=75 N=21 3.927 N=96 High school or less 48% 24% 43% Some college, but no degree 39% 57% 43% Post-secondary degree 14% 19% 14% Child care type N=79 N=20 .028 N=99 Family home (up to 6) 62% 60% 62% Group family home (up to 12) 38% 40% 38% ** significant at the .01 level 38 Structural Quality Variables Two of the three variables used to indicate structural quality are described in Table 1--education and child care type. No significant differences were found between groups on these indicators (assessed only at Time 1). Only 14% of providers had achieved a post-secondary degree, while the remaining providers were split evenly between having some post-high school experience and having a high-school education or less. Over 60% of providers were licensed to serve up to six children, while the rest were licensed to operate family child care group homes, serving up to 12 children cared for by two adults. A t-test analysis found no significant difference between groups for the third indicator of structural quality, the number of years of paid family child care experience (see Table 2). Preliminary Analyses T-test analyses of mean differences between the Non-accredited and Accredited groups on key study variables were conducted (see Table 2). T-tests on social and professional support variables indicated that, at Time 1, the Accredited group had significantly higher means on the Number of Associations, Professional Involvement Scale, and Number of Contacts. At Time 2, the two professional support indicators remained significantly higher for the Accredited group, but the difference in social support was no longer significant. T-tests on quality indicators indicated that, at Time 1, the Accredited group had significantly higher means on the FDCRS, the total CIS, and the Positive Relationship subscale of the CIS. At Time 2, no significant differences were seen between groups on any of these quality indicators. Accredited providers had spent significantly more months in FCCP than had Non-accredited providers. 39 Table 2 T-tests Analyses between the Non-Accredited and Accredited Groups for Key Study Variables * significant at the .05 level **significant at the .01 level ***significant at the .001 level Variables Non-Accredited Accredited T Value N Mean SD N Mean SD T Number of Months in FCCP 87 35.87 18.08 22 59.73 11.06 -7.81*** Years in Child Care 80 10.79 9.49 21 7.60 6.81 1.45 Positive Relationship Time 1 87 32.33 5.35 21 35.14 4.35 -2.23* Positive Relationship Time 2 85 34.08 5.37 22 10.77 4.91 .86 Total CIS Time 1 87 8.201 7.52 22 9.95 7.64 -2.08* Total CIS Time 2 85 88.80 7.98 22 87.73 6.75 .58 FDCRS Time 1 86 3.90 1.05 22 4.82 1.40 -3.42*** FDCRS Time 2 87 5.08 .94 22 5.15 1.12 -.31 Number of Associations Time 1 79 .63 .91 20 1.55 1.50 -2.61* Number of Associations Time 2 87 1.28 1.01 22 2.09 1.11 -3.32** Professional Involvement Scale Time 1 86 2.39 1.36 22 3.11 .97 -2.88** Professional Involvement Scale Time 2 87 2.96 .98 22 3.36 .69 -2.45** Number of Contacts Time 1 80 5.97 3.43 20 7.50 3.02 -2.30** Number of Contacts Time 2 84 6.19 3.16 19 6.68 5.03 -.54 40 Correlations among key study variables are presented in Table 3. Of the structural quality variables, Education was positively correlated with the Number of Associations at Time 1, and Years of Experience was positively correlated with the Number of Contacts at Time 2. Structural quality was not associated with any of the process or global quality variables. The process quality variables--Positive Relations and CIS total--were positively correlated with each other and with the global quality indicator?FDCRS--at both Time 1 and Time 2. Correlational analyses of the social and professional support variables with the quality variables indicated that the social support indicator, Number of Contacts, was unrelated to any of the quality indicators at Time 1 or Time 2. Nor was it related to either of the professional support variables at either time. The professional support indicator, Professional Involvement Scale, was positively correlated with both indicators of process quality (Positive Relations and CIS total) at Time1 and with FDCRS at Time 2. The second professional support indicator, Number of Associations, was positively correlated with the Positive Relations subscale at Time 1, as well as with the Professional Involvement Scale at Time 1 and Time 2. 41 Tab l e 3 Co rr e l a t i o n s A m o ng S t udy V a ri a b l e s V a r i a b l e s 1 2 3 4 5 6 7 8 9 1. E duc a t i o n -- -- -- . 074 . 034 . 151 . 142 102 -. 080 2. T y pe o f Ch i l d C a r e . 048 -- -- -. 078 -. 026 -. 011 . 061 -. 039 . 108 3. Y e a r s o f E xpe ri e n c e -. 050 . 191 -- -. 124 . 010 -. 013 . 188 -. 034 . 207* 4 . P o s itiv e R e latio n s ( C I S ) .1 1 9 - . 1 3 1 .0 2 2 . 145 . 883** . 405** . 061 . 028 -. 096 5. CIS t o t a l . 136 -. 109 . 019 . 922** . 087 . 472** . 112 . 092 . 001 6. F D CR S . 067 . 043 -. 071 . 573** . 578** . 196* . 233* . 231* . 032 7. N u m b e r o f A s s o c i a t i o n s . 268* . 096 -. 047 . 207* . 141 . 148 . 309* . 526** . 143 8. P r o f e s s i o n a l I n v o l v e m e n t . 155 -. 014 -. 012 . 217* . 215* . 042 . 602** . 442** . 184 9. N u m b e r o f Co n t a c t s . 152 . 139 . 091 . 028 . 083 . 124 . 054 085 . 369* * C o r r e la tio n is s i g n if ic a n t at t h e 0 . 0 5 le v e l ( 2 - t a ile d ) , * * c o r r e latio n is s i g n if ic a n t a t t h e 0 . 0 1 le v e l ( 2 - t aile d ) . D i a g o n a l co n t a i n s co r r el a t i o n s b e t w een v a r i a b l e s at T i m e 1 an d Ti m e 2 . B e l o w d i a g o n al ar e co r r ela t i o n s a m o n g v a r i a b l e s at T i m e 1. A bov e t h e d i a g o n al ar e co r r el at i o n s am o n g v a r i a b l e s a t Ti m e 2 . 42 Social and Professional Support of Providers As shown in Table 2, there were no differences between the Accredited and Non- accredited groups on the amount of reported social support (Number of Contacts) at either Time 1 or Time 2. For indicators of professional support (Professional Involvement Scale and Number of Associations), the Accredited group means were significantly higher at both Time 1 and Time 2 than the mean scores for the Non-accredited group. To examine whether the amount of support reported by providers changed significantly over the course of their participation in FCCP, repeated measures MANOVAs were conducted for Time 1 and Time 2 levels of the social and professional network variables for all study participants. A significant increase from Time 1 to Time 2 occurred in the Number of Associations, F(1, 98) = 24.33, p = .000, and on the Professional Involvement Scale, F (1, 107) = 19.059, p = .000. No statistically significant change from Time 1 to Time 2 occurred for Number of Contacts, F(1, 95) = .192, p = .662. Thus, providers reported increased professional support, but no increases in social support, during the course of their enrollment in the program. To examine whether these increases differed for the Accredited versus Non- accredited groups (Status), repeated measures MANOVAs examined the possible interaction of group Status and change from Time 1 to Time 2. No significant interactions were found for Number of Associations by Status, F(1, 97) = .032, p = .859, or for Professional Involvement Scale by Status, F(1, 106) = 1.373, p = .224. Plots of the estimated marginal means for both groups indicate a similar rise from Time 1 to Time 2 in each professional support indicator (see figures 1 and 2). 43 12 Number of Associations 1.00 1.50 2.00 E s ti m a te d M a r g i n a l M e a n s status Non-accredited Accredited Estimated Marginal Means of Number of Associations 12 Professional Involvement Scale 2.20 2.40 2.60 2.80 3.00 3.20 3.40 E s ti m a te d M a r g i n al M e a n s status Non-accredited Accredited Estimated Marginal Means of Professional Involvement Scale Figure 1 Figure 2 Time Professional Involvement Scale Number of Associations Time 44 To determine whether the length of time spent in the program was associated with changes in professional support from Time 1 to Time 2, repeated measures MANOVAs were conducted for each professional support variable, with Number of Months in the program entered as a covariate. Results indicated that there were no main effects for Number of Months in the Program on Number of Associations, F(1, 97) = 1.054, p = .307, or on Professional Involvement Scale, F(1, 106) = .909, p = .342. In summary, these results indicate that social support did not significantly differ over time or between groups. However, there were significant increases over time for all providers on both professional support variables, regardless of group status or length of time in the program. Professional support reported by the Accredited group remained significantly higher at both Time 1 and Time 2 than that reported by the Non-accredited group. The Effects of Professional Support on Caregiver Quality To examine change in process quality from Time 1 to Time 2, repeated measures MANOVAs were performed on the two process quality indicators. Results indicated no statistically significant changes for either Total CIS, F(1, 106) = .434, p = .511, or the Positive Relationship subscale of CIS, F(1, 105) = 2.248, p = .137. However, because T- test results showed a significant difference between groups at Time 1 (see Table 2) for both of these measures, a repeated measures MANOVA including Status as a between- subjects factor was performed. Results showed a statistically significant interaction effect for CIS and Status, F(1, 105) = 3.971, p = .049, as well as for the Positive Relationship subscale and Status. F(1, 104) = 5.656, p = .019. An examination of figures 3 and 4 indicate that, although there is a significant interaction in the changes from Time 1 to 45 Time 2, for the Accredited group these changes are not in the hypothesized direction. . 46 12 Total CIS 87.00 88.00 89.00 90.00 91.00 E s t i m a t e d Mar g i n al Mean s status Non-accredited Accredited Estimated Marginal Means of Total CIS 12 Positive Relationship 32.00 33.00 34.00 35.00 E s t i m a t e d Mar g i n al Mean s status Non- Accredited Accredited Estimated Marginal Means of Postive Relationship Figure 3 Figure 4 Time Time Positive Relationship score CIS score 47 To examine change in global quality from Time 1 to Time 2, a repeated measures MANOVA was performed on the global quality indicator, FDCRS. Results indicated a statistically significant increase from Time 1 to Time 2, F (1, 107) = 58.914, p = .000. Next a repeated measures MANOVA including Status as a between-subjects factor was performed. Results indicated a statistically significant interaction effect for FDCRS and Status, F(1,106) = 7.178, p = .009. Figure 5 shows a plot of these results. 48 12 FDCRS 3.80 4.00 4.20 4.40 4.60 4.80 5.00 5.20 E s t i m a t e d M a r g i n al M eans status Non-accredited Accredited Estimated Marginal Means of FDCRS Scores Figure 5 FDCRS Time 49 To address whether changes in provider levels of professional support from Time 1 to Time 2 affected process quality, a separate two-factor within-subjects MANOVA for each process quality variable was performed. The results for the Total CIS scores repeated measures MANOVAs indicated no statistically significant interaction effects between Total CIS and Professional Involvement Scale, F(1, 105) = .000, p = .994, or between Total CIS and Number of Associations, F(1, 96) = .019, p = .891. In addition, no statistically significant interaction effects were found between Positive Relationship and Professional Involvement Scale, F(1, 104) = .390, p = .534, or between Positive Relationship and Number of Associations F(1, 105) = .289, p = .592. In order to examine global quality in relation to professional support, a two-factor within-subjects repeated measures MANOVAs examining FDCRS scores from Time 1 and Time 2 and professional support variables were performed. The results for both repeated measures MANOVAs indicated a statistically significant interaction effect for FDCRS scores and Professional Involvement Scale scores, F(1, 106) = 7.921, p = .006 and for FDCRS and Number of Associations F(1, 97) = 4.329, p = .040. Summarizing, global quality increased significantly from Time 1 to Time 2 for all study participants. The slope of increase in global quality for Non-accredited providers was significantly different than the increase for the Accredited group. Changes in levels of professional involvement were significantly associated with positive change in global quality. Process quality did not change significantly from Time 1 to Time 2 when examining all study participants. However, examining these results by group indicated that the slope of change unexpectedly declined for the Accredited group; in contrast change increased for the Non-Accredited group. Additional Analyses In order to follow up on the unexpected direction of change in the process quality 50 indicators, a sub-scale from the FDCRS was created to measure process quality in an alternate way. Four FDCRS items which specifically called on the observer to assess the nature and frequency of provider-child communication and interaction were selected to represent process quality: (1) Informal use of language with infants/toddlers, (2) Informal use of language with children 2 years and older, (3) Tone, and (4) Discipline. (To review the entire FDCRS scale, see Appendix C.) Reliability analyses for this measure, named, FDCRS-PQ, yielded Cronbach?s alphas of .92 for Time 1 and .91 for Time 2. Once reliability was established, differences between the groups were examined. Results of the Time 1 t-test indicated that the Accredited group began with a significantly higher mean that the Non-accredited group, t = -3.42, p = .001. Results of the Time 2 t-test indicated that the Accredited group remained significantly higher on this scale, t = -4.78, p =.000. Correlational analyses of the FDCRS-PQ found it to be positively associated at Time 1 with FDCRS (r = .80, p < .01), Total CIS (r = .56, p < .01) and Positive Relationship (r = .565, p < .01). At Time 2 the FDCRS-PQ was positively correlated with Education (r = .225, p < .05). The Time 1 and Time 2 FDCRS-PQ scores were correlated with each other at the .01 level, r = .30. FDCRS-PQ was not found to be significantly correlated with any of the social and professional support variables. To revisit the question concerning process quality and professional support, repeated measures MANOVAs were performed using the FDCRS-PQ. No differences over time were found, F(1, 103) = 2.812 p=.097. A between-subjects analysis of FDCRS- PQ scale examining Status (Accredited vs. Non-Accredited) with the Number of Months in the Program as a covariate also yielded no significant interactions (FDCRS-PQ 51 subscale and Status, F(1, 102) = .627 p=.430; FDCRS-PQ subscale and Number of months in the Program, F(1, 102) = 3.763, p = .055). For the plot of the separate lines by status see figure 6. 52 12 Process Quality FDCRS subscale 4.50 5.00 5.50 6.00 E s timated Margina l Means status Non-accredited Accredited Estimated Marginal Means of Process Quality FDCRS subscale Figure 6 Time FDCRS- Process Quality 53 Directly addressing the relationship between process quality and professional support, two-factor within-subjects repeated measures MANOVAs examining the FDCRS-PQ from Time 1 and Time 2 and professional support variables from Time 1 and Time 2 were performed. None of these interactions were significant (FDCRS-PQ and Professional Involvement Scale F(1, 102) =1.318, p = .254; FDCRS-PQ and Number of Associations, F(1,93) = 3.317, p = .072). Unlike the CIS and its Positive Relations subscale, the FDCRS-PQ showed increases in process quality in the expected direction for both groups of providers. 54 V. DISCUSSION Family child care is a child care environment that has unique challenges and advantages and is often the type of care parents seek out for their youngest children. Even so, family child care has received less attention than center-based care environments in research. The current study examined quality in family child care and the amount of social and professional support among providers participating in a quality enhancement training program (FCCP). Its aim was to address issues related to training and professionalism, two standing issues among family child care providers, by seeking to understand how professionally involved family child care providers participating in FCCP were over the course of their program involvement, and whether their involvement was related to the quality of care they provide. Overall, the findings indicated that for both groups of providers--those who, during their participation in FCCP later became accredited and those who did not?levels of professional involvement and global quality increased significantly. In contrast, no significant increases in levels of social support or process quality were seen for either group. Increases in professional involvement were significantly related to increases in global quality. Providers who achieved accreditation through participating in FCCP came into the program reporting significantly higher professional involvement than those who had not yet achieved accreditation. Although all providers became more professionally 55 involved than they were at enrollment, over time the accredited providers remained significantly more professionally involved than the non-accredited providers. In contrast, while accredited providers began FCCP at significantly higher levels of both global and process quality than the non-accredited providers, these differences faded over time, with the non-accredited providers appearing to ?catch up? to the accredited providers. An exception to this pattern was found for process quality when it was analyzed using a small subset of items from the Family Day Care Rating Scale; accredited providers maintained a higher level of process quality over time than non-accredited providers. Implications for Prior Research These findings are generally consistent with prior research that has found training and professional involvement to be related to quality caregiving practices. For example, Galinsky et al. (1994) found that providers were more likely to offer good or adequate custodial care (global quality) when they were more professionally involved, and caregivers who took advantage of in-service and training opportunities were found to provide higher quality child care (Fischer & Eheart, 1991; Howes et al., 1995; Ghazvini & Mullis, 2002). Current findings, however, do not support other links that Galinsky et al (1994) found, such as that between process quality and social support. In fact, it should be noted that social support, (measured in this study as the number of support figures identified), was not related to any of the three types of quality of concern in this study. Perhaps this is not a surprise, in light of Weaver?s (2001) findings that social support did not predict quality caregiving (in global terms), regardless of whether social support was measured as the number of supportive resources available to a provider or in terms of the perceived 56 quality of those resources. As was evident from the literature review, there are few studies to which these findings can be directly compared. In fact, the relationship between professional or social support and quality caregiving practices is rarely examined as a primary question. Furthermore, the concept of professional support has not been examined separately from social support. Doing so in the present study showed that professional support could be increased through participation in FCCP and that such changes could predict changes in providers? global quality, although not in their process quality. These findings support the views of family child care providers who have specifically suggested that opportunities for networking would be important to the quality of their caregiving and should be included in training programs (Taylor et al., 1999). Implications of FCCP Training for Professional Support With regard to how time spent in the FCCP program may have affected the amount of social and professional support providers reported, only professional support increased over time. That social support did not increase significantly over time suggests that the goals of FCCP are more professionally focused, encouraging providers to join local and national associations. Thus, the increase in professional support would be more expected than an increase in social support. However, the increase in professional support could be more closely linked to social support than it appears. Contact with other family child care providers was assessed as professional support for this study, but the social contact that occurred through increased activity in associations may have served personal needs for support as well as professional needs. 57 In considering what could have predicted the change in professional support over time, it is interesting to note that it was not the amount of time in the FCCP program that explained the increase in professional involvement. Rather it appears that providers were as likely to join a local or regional family child care association early in their participation in FCCP as they were to join later on. It may be that staying in the program facilitates providers becoming engaged, sooner or later, in a wider professional network than the one they had envisioned prior to participation in FCCP. It is also possible that the providers early on learn the value FCCP places on becoming involved and, as a result, increase and maintain their involvement with associations while enrolled in the program. In considering the differences between the accredited and non-accredited groups on levels of professional involvement, it is significant that both groups of providers reported becoming more professionally connected and increased their membership in professional associations. This is an explicit goal of the FCCP program to encourage providers to broaden their professional networks. However, it appears that how professionally involved a provider becomes is related to how involved she was before enrolling in FCCP. This suggests that motivation and personality play a role in the amount of professional involvement providers seek. Providers who became accredited while in FCCP were, for the most part, already involved in professional activities upon enrollment. It may be that these providers enrolled in FCCP with the specific motive of attaining nationally recognized credentials and viewed increasing their professional involvement as a means of supporting this goal. Meanwhile, non-accredited providers may have enrolled in FCCP for a variety of reasons that did not explicitly include the motivation to become accredited. FCCP offers 58 providers benefits other than accreditation support, such as a $500 equipment grant for materials and supplies targeting quality improvement needs, the ease of not having to seek required training hours in more distant locations, a mentor who is a source of encouragement and helps with activities, and easily accessible group meetings linking them to other family child care providers. Thus, many non-accredited providers who may not have seen accreditation as their primary goal nevertheless increased their professional involvement as a result of the activities encouraged by participation in the program. Another example of how motivation may underlie provider interest in being socially or professionally engaged becomes visible when considering those providers who did not respond or refused to participate in the study. These providers?primarily including those who dropped out of FCCP in the first 9 months--were found to differ from those who chose to participate by virtue of reporting less social support (fewer social contacts) than study participants. This seems to suggest that those willing to participate in a study such as this one, and in a program like FCCP, were simply more interested and more likely, in general, to choose to involve themselves in child care- related matters or with people, in general. Those who did not choose to be involved may have just had less desire to socialize with others. Implications of FCCP Training for Quality Caregiving In considering the differences between the accredited and non-accredited groups on quality, why would the wide differences seen between the groups on global quality indicators at enrollment not be maintained over time? One possibility is that, due to their high initial level of quality, the members of the accredited group had less to change, in contrast to the members of the non-accredited group who had more room to show 59 improvement on the FDCRS. The fact that the non-accredited group did reach quality levels that were, on average, not significantly different from the accredited group, suggests that program activities provided by FCCP were successful in improving quality care giving practices associated with global quality. In contrast to global quality, no significant change was observed for process quality over time or in connection with professional support. This was true whether process quality was examined using the Caregiver Interaction Scale (Arnett, 1989) or by the subset of FDCRS indicators. Thus, the program activities and mentoring that FCCP provides do not appear to impact process quality in a significant way, either directly or as a result of increased professional involvement. It may be that the professionally focused goals of FCCP and the many concrete changes that providers make to increase their overall quality take attention away from process quality. Considering process quality and its lack of increase in this study, why might it be hard to change? Process quality involves the interactions of the provider with the children in her care. These interactions reflect the behaviors and characteristics of the providers, in contrast to global quality, which largely includes factors such as the setup of furniture, health and safety practices, or daily eating routines. Global quality may change more easily or more quickly because the changes being made are more external. Process quality involves more personal changes. The provider is not simply being asked to change the position of a table or to check the smoke detector batteries, she is asked to change methods of discipline, and the way she responds to the children, not just what she is doing but the way she is doing it. These types of changes involve more ingrained behaviors and attitudes, and therefore more difficult and time consuming to change. 60 The lack of change in process quality suggests that FCCP consider the content of training provided by its mentors in regard to effecting significant provider progress in this area. Reviewing refocusing content to specifically address aspects of provider-child interaction may be necessary. Because process quality is so much more difficult to change due to issues of personality and style, mentoring processes and methods may need to be altered or redesigned to address in a non-threatening way these more sensitive interpersonal features of the caregiving environment. One suggestion would be to look to and possibly incorporate some of the methods using in training early childhood professionals in the academic environment, such as video-taping and self-critique. Limitations of the Study One of the major limitations of the current study is that it had no control group with which to compare the FCCP program participants. Thus, the question about whether participation in FCCP is actually the reason for the changes in professional support and quality cannot be answered directly. While a comparison group was sought from among providers who dropped out of FCCP before participating for 9 months, only two surveys sent by mail were returned. Contacting family childcare providers who had never participated in FCCP would have been ideal, but could not be done in the context of the current study. A future option that should be considered for gathering a comparable control group is to solicit the participation of family child care providers who are on a waiting list to become a part of FCCP. Their interest in FCCP could mean that they would be more willing to participate in a study seeking to provide information to improve the training that FCCP provides and possibly to agree to have their child care giving practices 61 observed and evaluated. On the down side, using providers who sought to participate in FCCP still would not yield a true control group, but it would at least allow comparison between providers receiving and not receiving regular mentor visits. The lack of a control group makes the findings of the current study specific to the family childcare providers who participate in FCCP. While the providers in the current study were randomly selected from among all FCCP providers, this group cannot be said to be representative of family child care providers in Alabama or beyond. Few data are available to describe the population of family child care providers in the state of Alabama or in general. Thus, the ability to generalize the current findings is necessarily limited. Another potential limitation is the relationship between the mentor and the provider, in that the mentor is also the observer for both the global and process quality measures. Mentors may let the relationship that they are building with a provider create a ?halo effect? in the scoring of the provider?s performance on any given day. Thus, the objectiveness of the mentors could be questioned. On the other hand, the frequency with which the mentor sees the provider during business hours could make her better able to assess the quality of care. Because the mentor comes so often to the home of the provider, she would have a better idea of what a typical day for this provider would be than possibly an observer coming on one day. Nevertheless, it could be argued the use of a trained independent observer would be a better method of data collection. Analyses of the data representing process quality brought up concerns about the CIS and its ability to accurately measure process quality, making it difficult to draw conclusions about one of the key study variables. Subsequent comparison with process 62 quality as measured by a subset of process-related FDCRS items indicates the possibility that a problem exists either in the CIS itself, the method of scoring, or in the training of the mentor-observers. With regard to scoring, a major difference between the CIS and the FDCRS is that the FDCRS provides concrete, descriptive anchors in a seven-point scoring system, which clearly outlines the elements necessary for each score. In contrast, the CIS simply provides four response options, ranging from ?not at all like the provider? to ?very much like the provider.? Furthermore, the training provided to prepare mentors to use the CIS is nothing like the training video and manual created for the FDCRS. Thus, there seems to be more subjectivity in the CIS than in the FDCRS, perhaps resulting in more measurement error. Preliminary reliability analyses suggested that error might exist as well, in that the reliability estimates for three of the four subscales were very low. Implications for Future Research In discussing possible avenues for future research, conceptualization and measurement issues surrounding both quality and social and professional support need to be addressed. The accuracy of the CIS in assessing process quality in family child care homes came under question when the direction of effects predicted was reversed. When the CIS has been used in previous research on process quality and its relationship to other aspects of child care--such as training--has been examined, few relationships have been found (e.g., Kontos et al., 1996; Raikes et al., 2005). In the literature reviewed for the current study, only one published article indicated that there was a relationship between family child care quality measured by the CIS and training, and this study was the one that introduced the CIS measure (Arnett, 1989). It may be that the trainings being evaluated did not change quality; however, it may also be that the CIS does not 63 adequately capture process quality in family child care. Comparison of the CIS with other more recently developed measures of process quality, such as the Observer Record of the Caregiving Environment (NICHD Early Child Care Research Network, 2005) could provide additional data about the reliability and validity of this assessment tool. An examination of the tools available to measure social support is also recommended. There is no standard measure of social support in the child care literature. Those few studies that conceive of social support as a variable to be included in the picture of child care quality, measure social support in a variety of ways. A related issue is the conceptualization of social support and/or support networks and whether or not these should include or be distinct from the notion of professional support and/or professional networks. Pertinent questions include: How should we conceptualize and measure social support versus professional support? Should we look at professional support availability differently from professional involvement? Future research should explore prior conceptualizations of and methods for the assessment of social and professional support and identify or develop a reliable and valid measure. A question for future research that comes out of the issue of obtaining provider participation in the current study is, Are providers who agree to participate in a training program like FCCP different from other licensed family child care providers who do not wish to participate in a training program? This question leads to the next, What can be done to improve the quality of care that children receive in family child care homes in which the provider does not seek training? An interesting comparison to consider would be the differences among providers who seek training, providers who excel in training, and those progressing more slowly in 64 training. How would their motivation differ in regard to high quality care and the changes necessary to incorporate best practices into their daily routines? Understanding these differences could shed light on how to reach each group of providers and help motivate them toward providing better quality child care. When considering what can influence global quality and the possible difficulties in affecting process quality, the question arises, what could positively influence process quality for family child care providers? This is a question that needs asking. How can the attitudes and ingrained practices of family child care providers be changed to improve process quality? An examination of what has been successful in changing attitudes and practices in other fields and whether these methods could translate into family child care would help answer these questions concerning the difficulty in changing process quality. The underlying attitudes associated with certain caregiving patterns would also be helpful in finding methods toward improvement of process qualtiy. Understanding what influences attitudes toward caregiving would give researchers a better foundation for designing programs to improve process quality. Implications for Childcare Professionals There are several implications of these findings for family child care providers and programming for family child care providers, particularly FCCP. Given that professional involvement did predict global quality, the emphasis that FCCP puts on becoming an active member of professional associations and making connections with other family child care providers and professionals is appropriate and conducive to improved quality. This is a positive aspect of the FCCP program. Linking family child care providers who tend to be isolated with others who experience the same difficulties 65 through local association meetings could be giving them a sense that they are not alone and that there are others they can call on for support. FCCP provides one such contact immediately, the mentor. The mentor is someone that the provider can talk to about questions concerning child care issues. Thinking about recommendations for the profession, one suggestion would be for licensing requirements to mandate that providers become members of local, regional, statewide, or national family child care associations. Licensing agents could offer information to family child care providers about opportunities to become professionally involved as a child care provider. Also, because participation in FCCP appeared to lead to providers seeking more professional support, long-term training programs that support professionalism should be available and recommended for family child care providers. Conclusions Professionalism in the field of early education has been a much-debated issue. How much should this field become a profession? The National Association for the Education for Young Children recently published an article (Freeman & Feeney, 2006) addressing what professionalism would look like in this field and asks for opinions on the matter. The characteristics that professionals should exhibit include autonomy (having child care professionals in positions to make policy recommendations), altruism (committed to the service the perform above profits), a service orientation (educate the public about the benefits of quality child care to society), indispensability, (recognized through training as the only resource to perform this service), and commitment to a code of ethics. ?Professional practice is not a precise set of easily learned behaviors applied by rote. Instead, professionals must have a firm foundation of the field?s principles and 66 recommended practices and rely on this knowledge and expertise to guide their decision making in the workplace? (Freeman & Feeney, 2006, pg. 11). As the findings of this study suggest, professional support predicts family child care global quality. This effect could be because the providers are learning from their involvement in professional activities about better practices in child care. One of the functions of child care associations is to hold meetings where they can give workshops on relevant child care. The publications that the national and local associations distribute keep providers up to date on current issues and practices in the field. Also, and most pertinent to the questions of the current study, are the contacts that providers make with other family child care providers in their area. The meetings give family child care providers the opportunity to talk the other providers they meet and discuss the issues they are having and learn how others have handled the same kinds of situations successfully or unsuccessfully. Bandura?s social cognitive theory (1986)--which outlines the interactions among behavior, cognitive and personal factors, and environmental events as a basis for explaining how individuals learn and perform new behaviors?can be seen to apply well to the current study?s findings. Learning can emerge from observation, such as with the mentor at the provider?s home demonstrating an activity. It can also come about indirectly from hearing about others? experiences. When providers have relationships with other providers, they not only have the opportunity to observe but also to listen to stories and learn from others by hearing about their experiences. They can then take this information just like the observation and generalize it to their own situations and expected consequences. 67 Family Child Care Partnerships has been successful as an training program designed to facilitate increases in professionalism and global quality. When revisiting the key goals of the FCCP program, increasing the quality of care provided in the family child care home, with an eye toward accreditation, promoting professionalism in the child care community, and increasing provider knowledge of resources--especially social resources--which support the implementation of best practices and facilitate the establishment of provider associations, the evidence of the current study indicates that FCCP is achieving its goals. Professionalism in early education and especially in family child care is an issue that will continue to undergo debate in the field. The quality of care that children receive is undisputedly an important element in their early development; however, there is still much research to be done on the issues raised in this study concerning the improvement of quality in family child care. 68 REFERENCES Arnett, J. (1989). Caregivers in day-care centers: does training matter? Journal of Applied Developmental Psychology, 10, 541-552. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, J.J.: Prentice-Hall. Belsky, J., & Vondra, J. (1989). Lessons from child abuse: the determinants of parenting. In Cicchetti & V. Carlson (Eds.), Child maltreatment: Theory and research on the causes and consequences of child abuse and neglect (pp. 153-202). New York: Cambridge University Press. Bonds, D. D., Gondoli, D. M., Sturge-Apple, M. L., & Salem, L. N.(2002) Parenting stress as a mediator of the relationship between parenting support and optimal parenting. Parenting Science and Practice, 2, 409-435. Britner, P. A. & Phillips, D. A. (1995). Predictors of parent and provider satisfaction with child day care dimensions: A comparison of center-based and family child day care. Child Welfare, 74, 1135-1169. Buell, M., Pfister, I. & Gamel-McCormick, M. (2002). Caring for the Caregiver: Early head start/family child care partnerships. Infant Mental Health Journal, 23, 213- 230. Crittenden, P. M. (1985). Social networks, quality of child rearing, and child development. Child Development, 56, 1299-1313. 69 Cochran, M. M., & Brassard, J. A. (1979). Child development and person social networks. Child Development, 50, 601-616. Colletta, N. D. (1979). Support systems after divorce: Incidence and impact. Journal of Marriage and the Family, 7, 837-846. Crnic, K., Greenberg, M.., Ragzin, A., Robinson, N., & Basham, N. (1983). Effects of stress and social supports in mothers in premature and full term infants. Child Development, 54, 209-217. Cryer, D., Tietze, W., Burchinal, M., Leal, T., & Palacios, J. (1999). Predicting process quality from structural quality in preschool programs: A cross-country comparison. Early Childhood Research Quarterly, 14, 339-361. DeBord, K. (1993). A little respect and eight more hours in the day: Family child care providers have special needs. Young Children, 48, 21-26. Elicker, J. (1997). Developing a relationship perspective in early childhood research. Early Education and Development, 8, 5-10. Elicker, J., Fortner-Wood, C., & Noppe, I. C. (1999). The context of infant attachment in family child care. Journal of Applied Developmental Psychology, 20, 319-336. Fischer, J. L, & Elheart, B. K. (1991). Family day care: A theoretical basis for improving quality. Early Childhood Research Quarterly, 6, 549-563. Freeman, N. K. & Feeney, S. (2006) The new face of early care and education: Who are we? Where are we going? Young Children, 61, 10-14. Galinskey, E., Howes, C., Kontos, S., & Shinn, M. (1994). The study of children in family child care and relative care. New York: Families and Work Institute. Ghazvini, A., & Mullis, R. L. (2002) Center-based care for young children: Examining 70 predictors of quality. Journal of Genetic Psychology, 163, 112-125. Harms, T., & Clifford, R. (1989). The family day care rating scale. New York: Teachers College Press. Howes, C. (1997). Teacher sensitivity, children?s attachment, and play with peers. Early Education and Development, 8, 41-49. Howes, C., Matheson, C. C., & Hamilton, C. E. (1994). Maternal, teacher, and child care history correlates of children?s relationships with peers. Child Development, 65, 264-273. Howes, C., Smith, E., & Galinsky, E. (1995). The Florida child care quality improvement study, interim report. New York: Families and Work Institute. Howes, C., & Stewart, P. (1987). Child?s play with adults, toys, and peers: An examination of family and child care influences. Developmental Psychology, 23, 423-430. Katz, L. G., & McClellan, D. E. (1997). Fostering children?s social competence: The teacher?s role. Washington, DC: National Association for the Education of Young Children. Kontos, S., Howes, C., & Galinsky, E. (1996). Does training make a difference to quality in family child care? Early Childhood Research Quarterly, 11, 427-445. Kontos, S., Howes, C., Shinn, M., & Galinsky, E. (1995). Quality in family child care and relative care. New York: Teachers College Press. Kontos, S & Riessen, J. (1993). Predictors of job satisfaction, job stress, and job commitment in family day care. Journal of Applied Developmental Psychology, 14, 427-441. 71 Kruif, R. E., Ridley, S. M., & Wakely, M. B. (2000). Classification of teachers? interaction behaviors in early childhood classrooms. Early Childhood Research Quarterly, 15, 247-268. McLoyd, V. C. (1990). The impact of economic hardship on Black families and children: Psychological distress, parenting and socioemotional development. Child Development, 61, 311-346. Mertesacker, B., Bade, U., Haverkock, A., & Pauli-Pott, U. (2004). Predicting maternal reactivity/sensitivity: The role of infant emotionality, maternal depressiveness/anxiety, and social support. Infant Mental Heal Journal, 25, 47- 61. Miller, E. B. (2005). Predictors of quality caregiving in the ?Family Child Care Partnerships? home visitation program. Unpublished doctoral dissertation, Auburn University. Mueller, C. W., & Orimoto, L. (1995). Factors related to the recruitment, training and retention of family child care providers. Child Welfare, 74, 1205-1238. NICHD Early Child Care Research Network (2005). A new guide for evaluating child care quality. In NICHD Early Child Care Research Network (Eds.), Child care and child development: Results from the NICHD study of early child care and youth development. New York: The Guilford Press. Nelson, M. (1990). Mothering others? children: The experiences of family day care providers. Signs: The Journal of Women in Culture and Society, 15, 586-605. Peisner-Feinberg, E., & Burchinal, M. (1995). Models of quality and children?s developmental outcomes. In S. Helburn (Ed.), Cost, quality and child outcomes in 72 child care centers: Technical report. Denver: University of Colorado, Department of Economic, Center for Research in Economic Social Policy. Pence, A., &Goelman, H. (1987). Who cares for the child in day care? An examination of caregivers from three types of care. Early Childhood Research Quarterly, 2, 315- 334. Phillips, D.A., McCartney, K., & Scarr, S. (1987). Child care quality and children?s social development. Developmental Psychology, 23, 537-543. Powell, D.R. (1997). Parents? contribution to the quality of child care arrangements. In C.J. Dunst & M. Wolery (Eds.), Advances in early education and day care; Family policy and practice in early child care (Vol. 9, pp. 133-155). Greenwich, CT: JAI Press. Taylor, A. R., Dunster, L., & Pollard, J. (1999). ?And this helps me how?: Family child care providers discuss training. Early Childhood Research Quarterly, 14, 285- 312. Trawick-Smith, J., & Lambert, L. (1995). The unique challenges of the family child care provider: Implications for professional development. Young Children, 50, 25-32. Pascoe, J. M., Loda, F. A., Jeffries, V., & Easp, J. A. (1981). The association between mothers? social support and provision of stimulation to their children. Developmental and Behavioral Pediatrics, 2, 15-19. Raikes, H., (1996). A secure base for babies: Applying attachment concepts to the infant care setting. Young Children, 51, 59-67. Raikes, A., Raikes, H., & Wilcox, B. (2005). Regulation, subsidy receipt and provider characteristics: What predicts quality in child care homes? Early Childhood 73 Research Quarterly, 20, 164-184. Rickel, A., & Biasatti, L. (1982). Modification of the Block childrearing practices report. Journal of Clinical Psychology, 38, 129-134. Wan, C. J., Jaccard, J., & Ramey, S. (1996). The relationship between social support and life satisfaction as a function of family structure. Journal of Marriage and the Family, 58, 502-545. Weaver, R.H. (2001). The habitat of licensed child care providers: The influences of caregiver, contextual, and group characteristics on quality of care. Unpublished doctoral dissertation, University of Wisconsin-Madison. Weaver, R.H. (2002). The roots of quality care: strengths of master providers. Young Children, 57, 16-22. Whitebook, M., Howes,C., & Phillips, D. (1989). Who cares? Child care teachers and the quality of care in America: Final report, National Child Care Staffing Study. Oakland, CA: Child Care Employee Project. 74 APPENDIX A IRB-Approved Informed Consent Form 75 INFORMED CONSENT FOR PARTICIPATING PROVIDERS IN THE FAMILY CHILD CARE PARTNERSHIPS SOCIAL SUPPORT STUDY You are invited to participate in a study designed to understand family child care provider?s social support networks and the quality of their interactions with the children in their care. This study is being conducted by Dr. Ellen Abell and Robin Putnam, a graduate student, both of whom are in the Department of Human Development and Family Studies and Auburn University. In this study we want to learn about who gives support to family child care providers to do their job. We want to inform organizations and training programs about how they can better support family child care providers. You were selected as a possible participant because of your previous involvement in the Family Child Care Partnerships program through Auburn University. If you decide to take part in the study, you will be agreeing to fill out the enclosed questionnaire. It contains questions about your general characteristics and your sources of social support and participation in family child care provider organizations. Participation is requested only once and the survey should take about 30 minutes to complete. When you return the survey to us in the pre-addressed, stamped envelope, we will reimburse you for your time in the amount of $10. We will also enter your name in a prize drawing, along with other providers who participate, to receive $50 in gift certificates to Lakeshore Learning (a company selling child development materials). As you fill out the survey, if there are questions that make you uncomfortable, leave them blank. Doing so will not affect your reimbursement or entry into the prize drawing. The information you provide will help us better understand the needs family child care providers have for support and educate other about your needs. The information you provide in connection with this study will be treated as private and kept confidential. Your name will not appear on your survey and will never be used in any publication or presentation that might be developed as a result of this study. The information collected from study participants, as a whole, may be published in Masters? thesis, in a professional journal, or presented at a professional meeting or training. Your decision whether or not to participate will not jeopardize your future relations with Auburn University or the Family Child Care Partnerships Program. The information you provide will not affect any benefits or services you receive with these or any other agency or program, now or in the future. If you have any questions, I invite you to contact Ellen Abell at (334)844-4151 (email: abellel@auburn.edu) or Robin Putnam at (979) 240-3464 (email: putnarl@auburn.edu). We will be happy to answer any questions you may have. For information regarding your rights as a research participant, you may contact the Auburn University Office of Human Subjects Research or the Institutional Review Board by phone at (334)844-5966 or email at hsubjec@auburn.edu or IRBChair@auburn.edu. HAVING READ THE INFORMATION PROVIDED, YOU MUST DECIDE WHETHER OR NOT TO PARTICIPATE IN THIS RESEARCH PROJECT. YOUR SIGNATURE INDICATES YOU WILLINGNESS TO PARTICIPATE. _______________________________ _______________________________ Participant?s signature Date Investigator obtaining consent Date _______________________________ _______________________________ Print Name Print Name _______________________________ Co-investigator?s signature Date 76 APPENDIX B Social Support Survey 77 The Child Care Profession and You Q-1 How many other family child care providers do you know in your area/community? (Please check one.) ____None ____1-2 ____3-5 ____6-10 ____over 10 Q-2 How many of these providers do you feel you could call if you had a question or concern related to your work? (Please check one.) ____None ____1-2 ____3-5 ____6-10 ____over 10 Q-3. In general, how often do you talk with another provider about your work? (Please check one.) ____Not at all ____Once in a while ____A couple of times a month ____1-3 times per week ____Every day Q-4 Do you know of any meetings or activities organized in your area for family child care providers? ____Yes ____No Q-5. How often are such meetings or activities offered in your area/community? (Please check one.) ____About once a week ____A couple times a month ____About once a month ____About once every 3 months ____I don?t know Q-6 How often do you attend these meetings or activities? (Please check one.) ____Don?t attend ____Attend once a year ____Attend a couple times a year ____Attend most of these when they are offered ____Attend everyone that is offered Q-7 Are you currently a member of any kind of child care provider organization or group? ____Yes ____No Q-8 If yes, please list the names of all of the child care organizations or groups to which you belong. Q-9. Now we would like to ask you some questions about the availability of people in your everyday life who support you or offer you support when you need or ask for it. Use the chart below to do three things. 78 Step 1. List all the individuals you know well or see on a daily or regular basis. Describe them in terms of their relationship to you. List everyone you can think of. And please print clearly. ? Step 2. Go through the list and put a ?P? or an ?N? to describe whether this person has a mostly positive (P) or negative (N) influence on you. Step 3. For each person, circle one number that best describes how much influence this person has on you. This person?s relationship to me is????????. Positive (P) OR Negative (N) No influence A little influence Some influence Quite a bit of influence A lot of influence A great deal of influence Example 1: my spouse P 1 2 3 4 5 6 Ex. 2: my child care assistant P 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 79 APPENDIX C Family Day Care Rating Scale 80 81 82 83 APPENDIX D Caregiver Interaction Scale 84 Caregiver Interaction Scale Mentor ID ________ Provider ID __________ Date of Completion______________ Item # Description Rating 1 = not at all 2 = somewhat 3 = quite a bit 4 = very much Notes 1. Speaks warmly to children 1 2 3 4 2. Seems critical of the children 1 2 3 4 3. Listens attentively when children speak to her 1 2 3 4 4. Places high value on obedience 1 2 3 4 5. Seems distant or detached from the children 1 2 3 4 6. Seems to enjoy the children 1 2 3 4 7. When the children misbehave, explains the reason for the rule they are breaking 1 2 3 4 8. Encourages the children to try new experiences 1 2 3 4 9. Does not try to exercise much control over the children 1 2 3 4 10. Sepaks with irritation or hostility to the children 1 2 3 4 11. Seems enthusiastic about the children?s activities and efforts 1 2 3 4 12. Threatens children in trying to control them 1 2 3 4 13. Spends considerable time in activities not involving interactions with the children 1 2 3 4 14. Pays positive attention to the children as individuals 1 2 3 4 15. Does not reprimand children when they misbehave 1 2 3 4 16. Talks to children on a level they can understand 1 2 3 4 17. Punishes the children without explanation 1 2 3 4 18. Exercises firmness when necessary 1 2 3 4 19. Encourages children to exhibit prosocial behavior, e.g. sharing 1 2 3 4 20. Finds fault easily with children 1 2 3 4 21. Does not seem interested in children?s activities 1 2 3 4 22. Seems to prohibit many of the things that children want to do 1 2 3 4 23. Does not supervise the children very closely 1 2 3 4 24. Expects the children to exercise self-control; e.g. to be undisruptive for group, teacher-led activities, to be able to stand in line calmly 1 2 3 4 25. When talking to children kneels, bends, or sits at their level to establish better eye contact 1 2 3 4 26. Seems unnecessarily harsh when scolding or prohibiting children 1 2 3 4 Rate the provider on how well each statement describes her. The statement describes the provider ?not at all?, etc.