The Impacts of Non-Parental Care on Child Development

2. Annotated Bibliography

2.1.1 Non-Parental Caregiving

Clarke-Stewart, K. Alison; Allhusen, Virginia D.; Clements, Darlene C. In M.C. Bornstein (ed.) Handbook of Parenting: Volume 3. Status and social conditions of parenting. (1995). Mahwah, New Jersey: Lawrence Erlbaum Associates. p. 151-176

2.1.1.1 Keywords:

"ecology" of centre-based and family daycare, theories of non-parental care, family roles of non-parental caregivers, attachment figures, United States.

Background: This is a chapter in a handbook on parenting in which the authors review the theories and studies that are concerned with the effects of non-parental care on young children's social and cognitive development. Three specific theories are identified and discussed. Attachment theory argues that an infant's relationship with an adult attachment figure is of vital importance for his or her subsequent psychological development. Indeed, it is on the basis of the quality of the attachment relationship formed during their first year of life that "infants construct an internal working model of the attachment figure." Such a model consists of "a set of expectations concerning that person's availability and a complementary view of the self as worthy or unworthy of such care." However, at a time when most infants are cared for by several adults during their first year of life, the view of mother as the primary attachment figure may not be completely valid. In fact, it is possible that an infant builds a distinct working model for each caregiver he or she is exposed to. Nevertheless, it seems some developmental psychologists (Belsky, 1988, for example) continue to argue that the potential for a child's emotional adjustment is enhanced when he or she is separated from his or her mother during the first year of life. The authors contend that the negative effects of non-parental care on infants are not as dramatic as attachment theorists suggest. They substantiate this claim by referring to the fact that throughout the history of mankind, the majority of children in the world have received care from multiple caregivers and continue to do so.

The second set of theories discussed in this chapter is the socio-biological theories, which have attracted much attention in recent years. These theories are relevant to the issue of non-parental caregiving because they are based on the premise that people demonstrate favouritism and protective behaviour towards other who are genetically similar. Hence, biological relatives are most likely to provide the best quality care for their children. The authors argue that it does not automatically follow from this line of argument that parents or other relatives are equipped to offer the best care. "Professional" caregivers, for example, because they have less invested in the child's future, may actually be more objective about the child's behaviour. These tendencies may balance or outweigh the advantages of parental or familial child care.

The third set of theories discussed in the chapter focus on the stimulation of children's cognitive and social development. Such theories were particularly popular in the United States in the 1960s. Moreover, they were the basis upon which the pre-school intervention movement, which included Project Head Start, was developed. Essentially, cognitive and social development theorists believe that the environment has a determining effect on child development. Therefore, if children receive non-parental care in "a materially and verbally stimulating daycare environment with a moderate number of children," they will have higher cognitive development and experience an easier transition into elementary school. These theories also maintain that interaction with peers beginning at an early age fosters social competence, thus favouring child care arrangements outside the family.

2.1.1.2 Keywords:

"ecology" of centre-based and family daycare, theories of non-parental care, family daycare home, roles of non-parental caregivers, attachment figures, United States.

Background: This is a chapter in a handbook on parenting in which the role of the non-familial care provider as an important figure in the social networks of children is discussed. Three roles of non-parental caregivers are identified. According to the authors the caregiver is likely to occupy a place in the child's hierarchy of attachment figures. Also discussed are differences between caregivers and mothers.

Teacher roles: In daycare centres, caregivers tend to think of themselves as teachers and to use more academic teaching methods. In contrast, teaching in family daycare is more informal. Children learn through free exploration and in the context of "real-life" tasks and situations. Social rules are an important element of what caregivers in homes and in centres may teach children.

Disciplinarians: This role may be especially important in a setting populated by a large group of children who are close in age. Praise is one method through which caregivers try to discipline or manage children.

Nurturers and Attachment Figures: Daycare providers are natural candidates for attachment figures in the hierarchy of children's attachment as they spend significant portions of each day involved in caregiving and interacting with children. A growing body of research suggests that not only do children in daycare form attachment relationships with their caregivers, but that this has positive implications for children's development. Children who are rated as securely attached with their caregivers are more competent in their interactions with peers and adults.

Reported in the chapter is one study in which children who were insecure with both their mothers and their caregivers were rated lower in social competence than children who had at least one secure relationship. This suggests that a secure attachment with a daycare provider can compensate for an insecure relationship with the mother. Studies have consistently shown, however, that although children form attachments to their child care providers, they prefer their mothers to these other caregivers.

Caregivers versus Mothers: In terms of disciplinary styles, caregivers in daycare centres and nursery schools have been observed to be less authoritarian, less critical, and more likely to help and respond to children's initiation of play than mothers. Children also see their mother's role as being distinct from their caregiver's role: Mothers are perceived to be more involved in the children's physical care, whereas pre-school teachers are seen more as providing play and stimulation. Compared to teachers, home care providers interact more with each child individually, especially when there are only one or two children in the care arrangement, and they may be more positive and sensitive in their approach to children. They also do more supervisory disciplining. Compared to mothers, family daycare providers are more emotionally distant, engaging in less positive physical contact as well as being less playful and stimulating with the child.

2.1.1.3 Keywords:

"ecology" of centre-based and family daycare, theories of non-parental care, family daycare home, attachment figures, United States.

Care in child's own home: The authors identify this type of care arrangement as quite distinct from other more formal kinds of daycare in that it allows the child "to remain in a familiar, secure place." It is also seen as most likely to give the parents an opportunity to monitor the behaviour of the caregiver. The caregiver in turn is able to provide each child with individual attention. According to the authors, if the in-home caregiver is related to the child, this is "the most economical and stable of all daycare arrangements." If on the other hand, the caregiver is not a relative, this form of care is the least stable. Yet, if the caregiver is trained in child development, for example, a professional nanny, "in-home care is the most expensive kind of care."

Daycare home and family daycare home: Both of these terms refer to a care arrangement in which "a woman cares for a small group of children in her own home." The care provider "may or may not be related to the child; she/he may or may not be trained in early childhood education." Furthermore, "the number of children in the daycare home may range from 1 to 6 (family daycare home) or 6 to 12 (group daycare home)," and the basis of this arrangement "varies from an informal agreement about shared caregiving between friends to a formal, supervised network of licensed daycare homes." The authors examine a number of features of this type of care arrangement. Firstly, "daycare homes are usually located near the child's home, and therefore, they are easy to get to and are in a familiar neighbourhood where people are likely to share the parent's values and circumstances." Secondly, "the mother may have more control over what happens to her child in a daycare home than she would in a daycare centre because she can give instructions to a daycare home provider that she would not be able to give a daycare centre teacher." Thirdly, a daycare home provider is usually "flexible about taking children of different ages" and this form of care "offers the child an opportunity to interact with a handful of children of different ages rather than a large group of age-mates." Fourthly, the daycare home "provides new experiences for the child while at the same time providing continuity with the kind of family care the child is used to - in a home setting with a 'mother figure'." And finally, daycare homes are "unlikely to offer organized educational games or structured activities." Rather, the children spend most of their day in free play. In short, the main goal of most family daycare providers is to "provide a warm "homelike" atmosphere for the children."

Daycare centre: According to the authors, this is the most visible and easily identified child care arrangement; "the one most people think of when they speak of 'daycare'." The authors suggest "a centre may provide care for fewer than 15 children or more than 300" and "usually divides children into classes or groups according to their age." Furthermore, centres commonly have "some staff with training in child development and they offer children the chance to play with peers, often with educational materials." Compared with other daycare arrangements, centres are also considered "relatively stable and publicly accountable."

2.1.2 The NICHD Study of Early Child Care

The NICHD Early Child Care Research Network. http://www.nih.gov/nichd/html/publications.hmtl. Downloaded on October 14, 1998.

Keywords: National Institute of Child Health and Human Development, child care experiences, developmental outcomes, child care settings, United States.

Background: In 1991, the National Institute of Child Health and Human Development (NICHD), National Institute of Health, and U.S. Department of Health and Human Services, set out to develop a comprehensive, longitudinal study about the unique contribution that child care characteristics and experiences make to children's development - above and beyond the contributions made by the family and child characteristics.

Methods: A total of 1,364 children and their families from diverse economic and ethnic backgrounds in 10 locations across the United States were enrolled in the study. Children in the sample were being cared for in a wide variety of child care settings: care by fathers, other relatives, in-home caregivers, child care home providers, and centre-based care.

Independent variables: Many characteristics of the children (such as gender and temperament) and their environment were measured. Child care characteristics measured included the age of entry into care, quantity of care, stability of care, quality of care, and type of care; other aspects of child care, such as the provider's formal education, specialized training, child care experience, and beliefs about child rearing. The adult to child ratio, group size, physical environment and other safety and health characteristics were also measured.

Dependent variables: These included children's cognitive and language development, children's relationship with their mothers, and their self-control, compliance and problem behaviours, as well as peer relations and physical health. Children's cognitive development and school readiness were measured using standardized tests; language development was assessed through standardized tests and maternal reports.

Results: Close to half of the infants were cared for by a relative when they first entered care, but there was an acute shift towards reliance on child care centres and family daycare homes during the course of, as well as after, the first year of life. Economic factors accounted primarily for both the amount of non-parental care, the age of entry into care, and type and quality of care infants received. Family characteristics, including the family income and the mother's education, were strong predictors of children's outcomes — for both those children in nearly exclusive maternal care, and those children in extensive non-parental child care. These findings suggest that the influence of families on children's development is not significantly reduced or changed by extensive non-parental care. Of the family characteristics, sensitivity of the mother was the strongest predictor of children's behaviour. Child care quality was the most consistent predictor of children's behaviour. Children in care receiving more sensitive and responsive attention had fewer caregiver-reported problems at age two and three.

The higher the quality of child care (more positive language stimulation and interaction between the child and provider), the greater the child's language abilities at 15, 24, and 36 months, the better the child's cognitive development at age two, and the more school readiness the child showed at age three. Among children in care for more than 10 hours per week, those in centre care, and to a lesser extent, those in child care homes, performed better on cognitive and language measures than children in other types of care (including exclusive care by the mother), when the quality of the caregiver-child interaction was taken into account.

Centre care is associated with better cognitive and language outcomes and a higher level of school readiness, as compared to outcomes in other settings with comparable quality of care. Group care resulted in fewer reports of problem behaviour at age three. Instability of care, as measured by the number of entries into new care arrangements, led to higher probability of insecure attachment in infancy if mothers were not providing sensitive and responsive care.

2.1.3 The Importance of Quality Child Care

Kohen, Dafna; Hertzman, Clyde. In J. Douglas Willms (ed.) Vulnerable Children (in press).

Keywords: Characteristics of quality child care, child care arrangements, disadvantaged children, children's linguistic, cognitive, and social outcomes, NLSCY data, provincial differences, regulated care, Canada.

Background: The term "daycare" is used in a general sense to cover various types of non-parental care. The authors identify the following types of non-parental care: Care in a regulated daycare centre; care in the home of a care provider which is regulated (regulated family daycare); care in the home of a care provider which is not regulated (unregulated family daycare); care in the child's home by a paid sitter or nanny; and other informal arrangements, such as care by a grand-parent or other relative.

Use of Child Care Arrangements: The sample consisted of children aged four and five years old who are cared for by someone other than their parents or guardians on a regular basis. Forty percent of Canadian children aged 4 and 5 spend part of their week in some type of care arrangement so that their parents can study or work outside the home. The majority of these children are cared for outside their home in an unregulated daycare (14.5%) or in a regulated daycare (12.2%). Unregulated care in the home by a non-relative (e.g., a baby-sitter or nanny) accounts for only 6.1% of these preschoolers (aged four and five) care arrangements, while 7.3% are cared for by a relative.

2.2 Quality of Care and Child Development

2.2.1 Using the NLSCY to Study the Effects of Child Care on Child Development, Final Report

Cleveland, Gordon; Douglas Hyatt. (1997). University of Toronto at Scarborough.

Keywords: process quality, NLSCY, quality indicators, survey research, "production process," Canada.

Background: The report commissioned by ARB examines the most relevant ways in which to use the NLSCY to study the effects of NPC on child development. It outlines a theoretical background to conceptualize the issue and discusses the main issues for measurement. It concludes with a section on specific recommendations for future cycles of the survey.

It suggests an input-output model through which to consider the impacts of NPC (inputs) on child development (outputs). During the "production process" inputs are transformed into outputs. When the authors use the concept of child care as an input they are referring to non-parental child care, whether parents are working or studying or doing neither. Decisions about child care inputs are made by governments, parents, and the growing child in the context of decisions about resources and investments (based on a model of family decision-making derived from Leiborwitz, 1974 and amended by Haveman and Wolfe, 1995. See article for full reference).

Independent variables: The authors list the following elements of NPC as having the potential to influence child outcomes: type of child care used; history of types of child care used; amount of use of non-parental care; structural measures of the quality of child care used; non-structural (process) measures of the quality of child care used; auspice of care used (i.e., non-profit, commercial, public); type of care or quality of care used at crucial ages or transition periods for the child; number of different non-parental caregivers in early years; consistency of expectations among different caregivers in early years; consistency of expectations and attitudes among parents and caregivers in early years; and interactions of any of the above.

According to the authors the quality of non-parental child care is of crucial importance. However, it is insufficient to use structural measures of quality as a proxy for process quality in child care (there is reasonable evidence that the links between process quality and any other indicator of quality are not sufficiently well established). The standard "process" indicators in the U.S. for measuring centre-based child care are the Early Childhood Environment Rating Scale (ECERS) for children over 30 months and the Infant-Toddler Environment Rating Scale (ITERS) for children 30 months and under (see Harms and Clifford, 1980, 1990 for both). For family home child care, the Family Daycare Rating Scale (FDCRS; Harms and Clifford, 1989) is identified. The H.O.M.E. scale, the Belsky and Walker Spot Observation Checklist, and the Observational Record of the Caregiving Environment (ORCE) are the indexes which might be particularly relevant for the NLSCY, because they apparently are able to rate the quality of child care using a scale which is common to several different types of child care (e.g., both centre-based and home-based types of care).

Recommendations for the NLSCY: The authors conclude that it would be desirable to have on-site observations of child care quality occur several times over the pre-school life of each child. They recommend that this is the best way to assess "process" quality in current child care arrangements. If on-site observations are not possible then the authors recommend that parental (i.e. subjective) evaluation of child care quality should become the first priority amongst questions asked. The authors also recommend that additional data on child care inputs should be collected via additional questions in the parent questionnaire.

2.2.2 Non-Parental Caregiving

2.2.2.1 Keywords:

"ecology" of centre-based and family daycare, theories of non-parental care, literature review, family daycare home, roles of non-parental caregivers, attachment figures, United States.

Background: This is a chapter in a handbook on parenting. The authors argue global indices of quality are not helpful for uncovering connections between children's development and specific kinds of daycare experiences. More useful are those studies in which researchers examine the predictability of separate components of the daycare environment for separate developmental outcomes. The authors pay special attention to studying the caregiver "herself" - her behaviour or the factors that influence her behaviour.

Caregiver's behaviours: Studies clearly indicate that the caregiver's behaviour towards the child is a central mediator of the child's experience in non-parental care and, consequently, affects the child's development.

Caregivers' characteristics: Certain characteristics of the caregiver may each play a role in the way s/he interacts with the child. These characteristics include:

Education and training: Many researchers have demonstrated a link between the level of education and/or training that a caregiver has received and his or her behaviour with the children in his or her care.
Experience: Studies have shown that with fewer than 2 or 3 years of experience in child care there is a tendency for daycare providers to be passive and not initiate educational activities.
Stability and consistency: Children who experience many changes in their child care arrangement (either because of caregiver turnover or because the parents change the arrangement) have been shown to perform poorly on intelligence tests, and to be less competent in their play with adults and peers. However, there may be a ceiling on the positive effects of caregiver stability: Beyond 3 or 4 years, there is no evidence that stability improves the quality of care. Caregiver stability is important not only because it is facilitates the formation of close relationships between children and their daycare providers, but also because stability is an indicator of good working conditions, adequate wages, and high staff morale.
Degree of commitment to, or emotional investment in, the child: This is a factor to consider although there is no data to support the view that non-familial caregivers are less emotionally committed to the children under their care.
Gender: Does the sex of the caregiver make a difference in children's experiences in child care? Very little research is available to help answer this question, primarily because the overwhelming majority of child care providers are women. Research shows that both mothers and fathers tend to interact more with same-sex than with opposite-sex infants. Is this also true for non-parental caregivers? Studies that compare male and female caregivers are needed to answer the question of whether the caregiver's sex and the child's sex influence the quality of care that non-parental caregivers provide.

2.2.2.2 Keywords:

"ecology" of centre-based and family daycare, theories of non-parental care, literature review, family daycare home, roles of non-parental caregivers, attachment figures, United States.

Background: In this chapter on non-parental caregiving the factors that influence caregiver behaviour are discussed.

Factors: Structural aspects of the caregiving environment and the inclusion of an educational curriculum in the program have been repeatedly shown to be important indicators of daycare quality. These include:

Adult-Child Ratio and Group Size: In studies where the adult-child ratio is reduced, for example from 1:4 to 1:12, children have been found to experience less contact with the caregiver. This includes having "fewer of their questions answered, to engage in shorter conversations, and to be subject to more prohibitions."
There is a suggestion that children in classes with a heterogeneous age mix behave more competently than those in homogeneous groups. The authors contend that in mixed-age pre-school classes, "children have been observed to exhibit fewer dominance activities (hitting, kicking, demanding objects), more language (asking questions, conversing, imitating), more co-operation (offering objects), more altruism, and to increase in persistence, flexibility, intelligence, and positive response to a stranger."
Physical Environment: Several studies have revealed that children's behaviour and development are linked to aspects of the physical environment of the daycare setting. These may include "division of the classroom into interest areas and availability of varied, age-appropriate, and educational toys, materials, and equipment." According to the authors, however, simply adding "novel" materials to pre-school classrooms does not necessarily lead to cognitive or other gains.
Educational Program: Studies of daycare have also shown that children in more educationally oriented daycare programs differ in developmental outcomes from those in less educationally oriented programs.

2.2.3 Assessing Quality in Child Care Setting

Doherty, Gillian. Spring (1998). Ottawa: Canadian Child Care Federation. http://www.cfc-efc.ca/docs/00001225.htm. Downloaded on October 15, 1998.

Keywords: child care experience, non-parental care situation, "critical periods" of development, external variables, quality and forms of stimulation, neurology, parental characteristics, Canada.

Child care quality has traditionally been defined in terms of program inputs; for example, a relatively small number of children per adult, or a relatively high number of caregivers with training in childhood education. While these inputs do enhance the probability of creating "warm, supportive and stimulating interactions between child care providers and children," according to the author, they do not guarantee a quality experience for children. An alternative approach for evaluating quality according to the author is examining observational and neuro-biological research data in order to identify the type of experiences that are beneficial or potentially harmful for children. Discussed are the ideas of Begley (1996) and Kuhl (1997), two leading neurologists who begin their research with the premise that humans are born at a much earlier stage of individual development than the young of all other mammalian species. Of particular relevance to the issue of child development is their thesis that prior to the age of six a child needs specific "inputs" at "critical periods." If these "inputs" are not received, or are not received at the appropriate moment, development will not occur.

2.3 Effects of Non-Parental Care on Infants

2.3.1 Child Care in the First Year of Life

The NICHD Early Child Care Research Network. Merrill-Palmer Quarterly. (1997). Vol. 43, no. 3. p. 340-360.

Keywords: non-maternal child care, infants, first year, descriptive statistics, child care histories, multiplicity and stability of care, United States.

Background: This article provides information on child care during the first 12 months of life, specifically, the initiation and amount of infant child care, child care history patterns, and the type, multiplicity, and stability of care used by parents of infants.

Methods: The data reported here were collected as part of the NICHD Study of Early Child Care
- a prospective, multi-site study of the short- and long-term effects of infant care, conducted in the 1990s, based on a sample of families in which over half of the mothers indicated at birth that they planned to return to work full-time. Mothers were contacted in the hospital at the time of birth and recruited into the study when their infants were 1 month of age. A total of 1,364 families with healthy newborns were enrolled in the study. Data for the analyses presented in this article focus on the 1,089 children whose "non-maternal care" began before their first birthday. Mothers and their children were visited in their homes when the children were 1 month old. Mothers were administered a standardized interview containing questions about the household composition, child care, the family's health, and the education, employment history, and status of the family members, as well as a set of questionnaires. Families were subsequently telephoned when the infant was 3, 5, 9, and 12 months of age to update information about child care and family characteristics.

Variables used to profile child care:

Use of child care: In interviews mothers were asked to indicate whether "anyone other than yourself is caring for [CHILD'S NAME] on a regular basis." Use of child care could be for any reason and for any number of hours as long as the infant was in the arrangement on a regular basis; occasional care was excluded. Care provided by fathers and partners was included if it was regularly scheduled when the mother was not present. Care provided by the mother herself was included as a child care arrangement when the mother provided care for the infant at her workplace; this included care, for example, by mothers who cared for their own infants while they worked as child care providers.
Age at entry into care: The child's age (in months) at the time s/he first entered non-maternal care. Age of entry is reported for three groups: infants entering any child care, infants entering 10 or more hours of care, and infants entering 30 or more hours.
Amount of care: At each assessment period, mothers reported the number of hours per week that their infant was cared for on a regular basis by someone other than themselves or by themselves while they were working.
Type of care: Children's primary child care arrangements were classified into one of 10 categories: (1) mother at work, (2) father/partner, (3) grandparent in child's home, (4) grandparent in grandparent's home, (5) other relative in child's home, (6) other relative in the relative's home, (7) non-relative in the child's home, (8) non-relative in a child care home, (9) child care centre, and (10) other (e.g. sibling care).
History of care: Information provided by mothers on all starts and stops of specific types of child care arrangements since the previous interview. Children were divided into six unique history patterns: (1) exclusive maternal care through the first year of life, (2) some child care from the father or mother's partner, but no other non-maternal arrangement, (3) one continuous non-parental care arrangement, (4) one non-parental arrangement and then a return to exclusive maternal care, (5) more than one continuous child care arrangement, and (6) more than one arrangement with at least one termination.
Multiplicity and stability of care: At each assessment, mothers were asked to report the number of arrangements (up to three) begun for the child and the number of arrangements (up to three) stopped since the previous interview. Two measures were created from these data (these reflect non-parental arrangements only): (1) multiplicity of care arrangements - the total number of arrangements used at entry into child care, and (2) stability of care - the total number of arrangements (simultaneous and sequential) used over the course of the first year of life.

Results: Excluding father/partner care, 72% of the infants experienced some non-maternal child care with an average age at entry of 3.31 months. At the time of first entry into non-maternal care, infants were in care on average for 29 hours per week. Excluding father care, the average hours of child care per week was 28.

For infants who entered into 10 or more hours of care per week, the average number of hours was 33, and for infants who entered into 30 or more hours per week, the number of average hours was 42. At entry, the most common forms of care were father/partner care (25%), care in a child care home by a non-relative (24%), and care by relatives, including grandparents, in either the child's or the relative's home (23%). Children in a child care centre (12%) or in-home care by a non-relative (12%) were less common. By 12 months of age, care by fathers/partners and grandparents had decreased and care in centres and in child care home had increased. Of those infants who experienced some non-parental care, 38% were enrolled in just one arrangement and, of these children, 32% had one arrangement and then returned to exclusive parental care. Of the infants who experienced more than one non-parental arrangement, 93% experienced at least one termination of an arrangement. Over the course of the first year, infants who experienced non-parental care were in more than two such arrangements, on average; over a third (36.5%) had at least three different non-parental arrangements.

2.3.2 Non-Parental Caregiving

Keywords: "ecology" of centre-based and family daycare, theories of non-parental care, literature review, family daycare home, roles of non-parental caregivers, attachment figures, United States.

Background: This chapter in a parenting handbook highlights general conclusions reached by the authors on the developmental outcomes for infants with and without non-parental child care experiences.

Developmental outcomes: Researchers continue to come to grips with the question of whether or not early, extensive non-parental care leads to emotional insecurity and social maladjustment. In this chapter the authors express their skepticism about the argument that quality of the mother-child relationship is less secure for children of employed mothers than for children of non-employed mothers. Much of the evidence is based on experiments using the "Strange Situation." Yet, such an instrument seems highly inappropriate for assessing the attachment relationships of daycare children as these children have become accustomed to separations from their mothers. In other words, we do not know if "the avoidant behaviour observed in some of these children is truly reflective of insecurity or if it represents an adaptation by daycare infants toward greater independence and ease with infant-mother separation." According to the authors, an alternative method, which involves rating children's behaviour with their mothers at home after observing a substantial period of unstructured, natural interaction has not revealed significant differences between children with extensive, early daycare experience and those without. The authors argue that the quality of the specific caregiving arrangements experienced by the individual child as well as other influential environmental factors in the child's ecology must be taken into account.

2.3.3 Early Child Care and Self-Control, Compliance, and Problem Behavior at Twenty Four and Thirty Six Months, The NICHD Early Child Care Research Network

Child Development. (1998). Vol. 69, no. 4. p. 1145-1170

Keywords: interviews, observations of child care setting, quality-of-care measures, "selection effects," non-compliance, problem-type behaviour, developmental transitions, group-type care, stability of care, "sleeper effects," United States.

Background: Early child care experience continues to be linked with increased levels of non-compliance, aggression, and problem behaviour. Although evidence suggests quality of care must also be considered, it has thus far not been studied. The current investigation was designed to examine the multiple features of care, including quality. The authors studied whether or not child care experiences in the first years of life (e.g., quantity, quality, stability) predict self-control, compliance, and problem behaviour at 24 and 36 months of age and whether child and family factors moderate these effects.

Methods: More than 1,000 children with diverse child care and family experiences were repeatedly monitored in this study. Child care data were obtained from mothers' reports during face-to-face interviews at 1, 6, 15, 24, and 36 months, as well as from phone calls every 3 months beginning at 3 months of age. Observations of the non-maternal care site and interviews with caregivers were conducted when the child was 6, 15, 24, and 36 months of age. At 24 and 36 months of age, child behaviour was observed both in a natural setting and in a laboratory.

Independent variables: (a) family and child variables including gender, child temperament, positive qualities of maternal behaviour and infant-mother attachment security ("Strange Situation" measured at 15 months); and (b) child care variables, including (1) age of entry, (2) quantity of care, (3) stability of care, (4) group-type, (5) quality of care. Child's age in weeks, when reported non-maternal child care experiences totaled at least 10 hours per week; Sum of all hours in all non-maternal arrangements, scored as mean hours per week during the period 4-15 months (first year), 16-24 months (second year), and 25-36 months (third year); Number of different child care arrangements the child experienced as reported by the mother during the same three age periods described above; Child's primary non-maternal care arrangement (where the child spent the most time, or the more formal, institutional arrangement if time across settings was equal) at each of four ages (6, 15, 24, and 36 months). Children were considered to be in group-type care if there were at least three other (non-sibling) children in addition to the study child in the child care arrangement; An instrument using observational data designed specifically for this study but based on the ORCE (see NICHD Early Child Care Network, 1996) was used.

Dependent variables: Child functioning assessment (at 24 and 36 months) obtained from (1) mother and caregiver reports of child's problem behaviour and social competence; (2) laboratory assessments; and (3) child care observations. (1) The Child Behaviour Checklist 2/3 was used to assess behaviour problems. The Adaptive Social Behaviour Inventory was used to measure social competence and disruptive behaviour. (2) Instruments used in the laboratory included The Cleanup Task, Compliance with Bayley Test Examiner, Resistance to Temptation, and the Three Boxes Interaction Procedure. (3) Child functioning was measured using the modified ORCE.

Results: The full set of child care predictors never accounted for more than 3% of the variance in the measures of self-control, compliance, and behaviour problems. In fact, they accounted for even less when family processes reflecting attachment security and sensitive mothering were controlled. It remains possible that larger effects may emerge if specific variables were examined (e.g., impoverished children, single-parent homes, depressed mothers). Among child care variables, quality of care remained the most consistent predictor at both 2 and 3 years of age. But, even though quality of care proved to be the most consistent of the child care variables in predicting children's compliance and problem behaviour, the amount of explained variance was modest, and standardized regression coefficients never exceeded 0.16 in the cumulative-experience analyses or 0.13 in the lagged-and-concurrent-analyses. Moreover, all but two of the six significant effects of cumulative quality of care were reduced to non-significance once family variables were controlled.

Although 2-year-olds who spent more time in non-maternal care were reported by their mothers to be less co-operative, and by their caregivers to exhibit more behaviour problems (after controlling for selection effects and child characteristics), by the time they were 3 years of age, no significant effects for amount of child care experience could be detected. Authors argue it is plausible that the effects could re-emerge when children are challenged by another developmental transition, such as the entrance to school. It is however, also possible that as child development is a very complex process, it is not explained sufficiently by the study variables.

The composite measure of sensitive mothering² in this study proved to be relatively strong in predicting child outcomes. Significant standardized regression coefficients for this predictor ranged from an absolute value of 0.20 to an absolute value of 0.36 in the cumulative-experience analyses. Consistent with evidence from other studies, these results indicate that more responsive and positive mothering in the first 2 years predicted fewer mother-reported problems and more positive interaction between mother and child in the laboratory at 24 months.

2.3.4 Environmental Changes and Children's Competencies

Kohen, Dafna; Hertzman, Clyde; Wiens, Michele. Applied Research Branch, Human Resources Development Canada. Working Paper W-98-25E, 1998.

Keywords: NLSCY, child care, environmental changes (residential moves, changes in schools, child care arrangements), child outcomes.

Background: Environmental changes are common in the lives of Canadian children. Authors in this study examine the impact of residential moves, and changes in child care arrangements and schools on the outcomes of children aged 0-11 (thus extending the applicability of this entry beyond simply an examination of infant outcomes). This summary will focus only on the outcomes related to changes in child care arrangements.

Independent variables: Family socio-demographic characteristics, Mental Health of the Person Most Knowledgeable (PMK) about the Child.

Dependent variables: Infants/Toddlers (0-3 years) — Motor Social Development (15 questions measuring dimensions of motor, social, and cognitive development); temperament (Infant Characteristics Questionnaire Revised). Preschoolers (4-5 years) — verbal competence (PPVT-R); behavioural problems (43 item checklist). School ages (6-11 years) — behavioural problems, math achievement (Mathematics Computation Test of the Standardized Achievement Tests, Second Edition (CAT/2); grade repetition.

Results: In the 12 months before the survey 23% of infants and toddlers had experienced a change in care arrangements, most often because the program had terminated (33%) or because the PMK was dissatisfied with the care arrangement (20%). Mothers of children who had experienced changes in care arrangements were more likely to have reported poorer mental health and the children were more likely to be male. These children were also more likely to be characterized as having a difficult temperament. For preschool age children, 25% of the 41% who were in care had experienced a change in care arrangements over the past year. The reasons for the change were similar to those seen in the younger children. The mothers of these children were more likely to have higher educational levels, lower household incomes, and poorer mental health. Even after controlling for socio-demographic characteristics, children who had experienced a change had lower PPVT scores than those who did not. Twenty two percent of school age children had experienced a change in care arrangements, most often because of termination of program availability (31%) or a family move (16%). These mothers were more apt to be single parents, have higher educational attainments, and poorer mental health and their children were more likely to be male. Children in this age group who had experienced a child care change had 30% higher odds of having high behavioural problem scores than those who had not. These results substantiate the importance of stability in child care.

2.4 Effects of Non-Parental Care on Preschoolers

2.4.1 Effects of Age of Entry, Daycare Quality, and Family Characteristics on Preschool Behavior

Hausfather, Albert; Toharia, Angeles; LaRoche Catherine; Engelsmann, Frank. Journal of Child Psychology and Psychiatry. (1997). Vol. 38, no. 4. p. 441-48

Keywords: Child daycare, daycare quality, child behaviour, outcomes, high-quality care, "interactive effects," age of entry, Canada (Montreal).

Background: In 1990, Howes, et al. discovered that there were important interactive effects between age of entry and daycare quality: Children who entered low quality daycare as infants were the least compliant and least likely to self-regulate. Yet, these and other authors have never included daycare centres of excellent quality in their research and therefore, have been unable to evaluate the differential effects of extremes in quality. This article tests the hypothesis that the quality of the daycare program is the main determinant of risk or benefit of attending non-parental care from an early age.

Methods: A cross-sectional design was developed, which included an assessment of the present daycare centre quality as well as a detailed history of the child's daycare experiences from birth. Only centre-based group daycare was studied because it is the child care environment most suitable for a standard evaluation. All 66 licensed daycare centres receiving infants for at least the past 5 years in the Montreal area were initially contacted. Of the 66 licensed daycare centres receiving infants, 24 agreed to participate. In addition a questionnaire was given to parents.

Independent variables: (a) family characteristics; (b) age of entry into child care; and (c) quality of child care. Quality of care was evaluated by two independent raters using two direct observation assessment scales: the Early Childhood Environment Rating Scale (ECERS, Harms and Clifford, 1980), and the Early Childhood Classroom Observation Scale (ECCSO, Bredekamp, 1987). The purpose of the second scale was to enrich the assessment of some specific dimensions of quality care such as the relationship between the child and caregivers and nature of the program, and to assess further the validity of the ECERS.

Life events inventory: A life events inventory was completed by the parents (see the Life Events Questionnaire adapted from the questionnaire by Coddington, 1972). Parenting stress (see the Parenting Stress Index, PSI-short form; Abidin, 1990). Family Evaluation (Parental feelings, perceptions and the family mode of functioning with regard to two basic dimensions; its adaptability and its cohesion, were measured using the Family Adaptability and Cohesion Evaluation Scale, FACES-II, abbreviated form; Olson, Russell & Sprenkle, 1989).

Dependent variables: Pre-school behaviour: Daycare teachers, unaware of the hypotheses of the research evaluated the 155 4-5 year-old children attending the Montreal daycare centres on the Social Competence Scale (SCS, Kohn and Rosman, 1973) and the Pre-school Behaviour Checklist (PBC, McGuire and Richman, 1986).

Results: The positive or negative effects of age of entry to daycare on the child's later group behaviour are significantly mediated by the daycare quality. In other words, an early age of entry into group daycare should not, in itself, be considered a risk factor. Early entry into group daycare may have detrimental or beneficial effects, depending on the centre quality conditions and the family background of the child. A longer stay in low quality daycare contributes significantly to the child's display of anger and defiance in the group setting. In high-quality care, early and longer attendance, combined with favourable family circumstances are predictors of the child's level of interest and participation in the group setting. The possible negative effects of an early group experience (angry behaviour in children) and of family stress may be sufficiently buffered when excellent-quality daycare is provided to the child.

2.4.2 Effects of Daycare on the Development of Cognitive Abilities in 8-Year-Olds: A Longitudinal Study

Broberg, Anders G; Hwang, C. Philip; Lamb, Michael E; Wessels, Holger. Developmental Psychology. (1997). Vol. 33, no. 1. p. 62-69

Keywords: child characteristics, cognitive abilities, quality of home care, quality of alternative care, "sleeper effects," Sweden (Goteborg).

Background: The impact of child care on child development is difficult to quantify when the samples are not randomly selected and thus pre-existing differences between children may be present in different child care trajectories. This study was designed to avoid this problem by recruiting children between 12 and 24 months of age before they were launched on different child care trajectories and by following them over a period of 7 years. The study was conducted in Goteborg, Sweden and 146 children (72 girls) were enrolled when they averaged 16 months of age. None of the children had experienced regular out-of-home care at the time of enrolment, but within 3 months, 54 entered centre care and 33 entered family daycare. According to the authors, evidence of the effects of daycare quality on children's cognitive abilities (except Clarke-Stewart's 1991 study of the links between verbal abilities and quality) is quite inconclusive.

Methods: The authors recruited 146 children from the waiting lists for public child care facilities in all areas of Goteborg, in 1982 and 1983. Children were included in the study if they were (a) between 12 and 24 months old, (b) firstborn, or at least not living with a sibling under 12 years of age, (c) living with both parents (married or unmarried), and (d) not attending regular daycare. Because of a shortage of places, not all children were assigned to public child care facilities. Fifty-four of the children were enrolled in public daycare centres (centre daycare or CDC group), 33 entered family daycare (FDC group) setting (8 public and 25 private), and 59 children remained in the exclusive care of their parents (home care or HC group). The authors assessed the effects of early child care arrangements with measures of the type, amount, and quality of out-of-home care. After the initial assessment, children were followed for seven years through four additional phases of data collection that took place when they averaged 28, 40, 80, and 101 months of age.

Independent variables: (a) child characteristics, including gender, temperament, number of siblings, and prior cognitive abilities (hypothesized to be the most powerful predictors of later cognitive skills); (b) family background variables including, socio-economic status, quality of the home environment, and the extent of paternal involvement; and (c) quality of non-parental care.

(a) Temperament was measured as inhibition. In Phase III, when children averaged 40 months of age, mother's ratings of eight items from Block and Block's (1980) California Child Q-Sort and observers' ratings of the children's lack of sociability with a strange adult and lack of involvement in a peer-play situation were combined into an index of inhibition. Inhibition was also measured by pre-school teacher's ratings in Phases III and IV on seven items from Baumrind's (1968) Pre-school Behavior Q-Sort. In order to measure the number of siblings, in phases II, III, IV, and V, the authors coded whether siblings had been born since their last visit. (b) Paternal involvement in all phases was assessed using ratings by the parents on the Parental Responsibility Questionnaire (PRQ; Lamb, et al., 1988). Child's time in daycare was indexed by the total number of months (whether full time or part time) he or she had spent in (1) centre-based daycare; (2) family-based daycare; and (3) exclusive home care up to and including Phase III. Quality of home environment was measured using age-appropriate versions of Caldwell's HOME Inventory (Caldwell, 1970; Caldwell & Bradley, 1984). During the individual phases, full or sub-scales of the infant, the pre-school, or the full elementary school version of HOME were used. The Belsky and Walker (1980) Spot Observation Checklist was also used in the first three phases to measure the quality of home care. (c) Quality of non-parental care: Both structural and dynamic measures were used. Structural aspects were assessed with information obtained in a short interview with caregivers regarding the number of children in the group, the number of caregivers, the average number of hours the target child spent in out-of-home care each day, the "longest" regular day the child spent in daycare, and the total number of caregivers who had been working in the group within the last 12 months. This information was then combined into an index and a summary measure for the early childhood period was constructed by averaging scores from the assessments at 16, 28, and 40 months. To tap dynamic aspects of the alternative care setting, the Belsky and Walker (1980) Spot Observation Checklist was also used in the first three phases of data collection. Scores from the first three assessments were then summed for analytic purposes.

Dependent variables: (a) Verbal abilities and (b) mathematical abilities. In phases II, III, and IV, verbal ability was assessed using objective tests. In Phases II and III, the Language sub-scale of the Griffiths Developmental Studies (Griffiths, 1954, 1970) was used. Two verbal sub-scales (reading and vocabulary) of a Standardized School Readiness Test (Ljungblad, 1967/1989) were administered individually to children prior to their entry into elementary school. When the children were in second grade, a standardized test of reading abilities was used. To measure mathematical abilities, the numerical sub-scales of the Standardized School Readiness Test (Ljungblad, 1967/1989) were administered individually to the children prior to their entry into elementary school. The sub-scale scores were summed to yield total mathematical ability scores. When the children were in second grade, a standardized test of mathematical ability was used (Ljung & Pettersson, 1990).

Results: To identify those variables that explained measurable amounts of variance in cognitive ability, a stepwise regression analysis was conducted. The tested verbal and mathematical abilities at 101 months of age were used as the dependent variables. Findings indicated: Earlier measures of cognitive ability were the best predictors of cognitive performance at 8 years of age. The type of early child care experienced proved to be predictive as well. Daycare had positive "sleeper effects" on cognitive development. Children who had spent more months in centre-based daycare before they were 40 months old obtained higher scores on tests of cognitive ability than did other children. For children who had spent 3 or more pre-school years in out-of-home care, quality of non-parental care was also predictive. Dynamic measures of quality (quality of adult-child interaction) predicted verbal abilities, whereas structural measures of quality (child-staff ratio, group size, age, and range) predicted mathematical ability. Family socio-economic status did not predict performance. Part of the reason for this may be that socio-economic status is of less predictive importance in Sweden than in other countries.

2.4.3 Daycare Participation as Protective Factor in the Cognitive Development of Low-Income Children

Caughy, Margaret O'Brien; DiPietro, Janet A; Strobino, Donna M. Child Development. (1994). Vol. 65. p. 457- 471

Keywords: protective factors, disadvantaged children, low-income children, daycare participation, reduced risk, United States.

Background: Existing knowledge about the impact of daycare on the cognitive functioning of disadvantaged children has come primarily from studies of centre-based programs, designed specifically to foster cognitive development. These centre-based programs, made up of special enrichment curricula, do not represent the child care experience of most disadvantaged children. The goal of the authors here is to study the impact that enrolment in ordinary/routine child care during the first 3 years of life has on a child's cognitive development. The authors test a model proposed by Dasai, et al. (1989) in which the impact of daycare participation varies with family income.

Methods: The authors conduct a secondary analysis of The National Longitudinal Survey of Youth (NLSY), a national American data set that includes a national probability sample of over 12,000 young men and women between the ages of 14 and 21. The original purpose of the NLSY was to analyze the labour market participation of this sample of young adults. In 1982, however, the NLSY received additional support to allow for data collection on the children of women in the NLSY sample, and in 1986, the NLSY completed in-home assessments of almost 5,000 children to measure social, behavioural, and cognitive functioning. Data were collected from a variety of sources, including maternal reports, interviewer observation, and direct assessments of the child using quality control procedures. The NLSY continues to evaluate the children every 2 years, and in 1990 the total number of children assessed exceeded 8,000. The sample used in this article consists of 867 5- and 6-year-old children who completed the 1986 assessment wave of the NLSY.

Dependent variables: (a) Daycare participation and (b) quality of the home environment. Information about daycare participation was derived from interviews with mothers. As a part of the NLSY cohort, mothers were interviewed yearly beginning in 1979. Information routinely collected as part of the interview included household composition, income, education, and employment status. In the annual 1986 interview, mothers were also questioned regarding daycare participation. The daycare participation variable was constructed from three binary variables about whether or not the child was enrolled in daycare during the first 3 years of life. These variables were used to create two additional variables reflecting the pattern of daycare use: the total number of years during the first 3 years of life that the child participated in daycare, and the year daycare began. In addition to this information, the predominant daycare arrangement was reported for each of the first 3 years of life and was categorized as (1) in the child's own home, (2) in another home, or (3) daycare centre/school. Quality was assessed using a shortened version of the HOME scale (Bradley & Caldwell, 1979). The internal consistency of the HOME-SF was assessed and the Cronbach's alpha was 0.67 for 5-year-olds and 0.70 for 6-year-olds.

Dependent variables: Academic Readiness. Academic readiness was measured by the 1970 version of the Peabody Individual Achievement Test (PIAT). Three of the PIAT sub-tests (mathematics, reading recognition, and reading comprehension) were administered.

Results: The first set of analyses tested the effect of daycare participation and its interaction with family income on cognitive outcome using an analysis of variance (ANOVA). In the second set of analyses, multiple linear regression was used to investigate the relation between patterns of daycare participation and outcome, controlling for confounding variables.

Children from impoverished environments who started attending daycare before their first birthday had higher reading recognition scores than children from comparable home environments who did not attend daycare at all.

Daycare participation during the first 3 years of life is positively related to the subsequent development of mathematics and reading skills for children from impoverished environments. This relation is strongest for reading skills if participation commences before the child's second birthday. For mathematics skills, it appears that centre-based care, in particular, exerts a protective effect over the developmental liabilities of an impoverished environment.

Some adverse effects of daycare attendance were also detected. For children from homes at the upper end of the HOME-SF scale, initiation of daycare before the second birthday was negatively associated with reading recognition performance, and participation in daycare that was centre-based or in the child's own home was negatively associated with mathematics performance.

Interestingly, the correlation between the HOME-SF score and family income was modest, at best. Family income can influence the quality of the home environment directly by influencing the resources available to the family for the purchase of items that are assessed as part of the HOME-SF. However the HOME-SF also measures other aspects of the home environment, for example, maternal responsiveness and affection toward the child.

Limitations: The daycare variables used in this study were limited to maternal recall of participation during the first 3 years for those children who were 5-6 years old in 1986. In addition, the NLSY did not include information regarding the number of hours per week spent in daycare or the staffing patterns and qualifications of the daycare providers.

2.4.4 Child-Care History and Kindergarten Adjustment

Bates, John E.; Marvinney, Denny; Kelly, Timothy; Dodge, Kenneth A.; Bennett, David S.; Pettit, Gregory S. Developmental Psychology. (1994). Vol. 30, no. 5. p. 690-700

Keywords: extensiveness of non-parental care, child characteristics, child adjustment problems, behavioral development, child care histories, United States (Nashville, Tennessee; Knoxville, Tennessee; Bloomington, Indiana).

Background: This three-site longitudinal study, conducted over a two-year period examines the controversial issue of the timing and the extensiveness of non-parental care as it attempts to understand the origins of aggressive behaviour in children and kindergarten adjustment. The authors argue that the developmental stage of the child is a key factor if he/she experiences daycare. According to the authors, what are needed are studies that distinguish the timing of care as opposed to the total amount of care received up to the time that adjustment outcomes were measured. Since the authors were able to consider the full array of daycare histories, rather than only children who started a particular arrangement and continued for some specified time, they were able to disentangle the effects of timing and total amount of daycare.

Methods: 589 families of 5-year-old children who were about the enter kindergarten in Nashville, Tennessee (n = 204), Knoxville, Tennessee (n = 204) and Bloomington, Indiana (n = 181) were recruited. Of the families approached, 75% agreed to participate. In the summer before kindergarten or within the first weeks of school, parents were separately given questionnaires and were interviewed for 1.5 hours at home. Although mothers' interview data was central to analyses, in 388 of the cases, fathers' interview data were available for comparison. Children were given a social-cognitive interview during this same period. Children were followed up after they had been in kindergarten 3 months or more and assessed with teacher, peer, and observer measures of social adjustment in school. The 588 children followed (one family dropped out after initial assessment) were in 145 classrooms in 14 different schools.

Independent variables: Daycare use: Daycare was measured retrospectively. While this method merits some caution, measurement was antecedent to the child's experience in kindergarten, so it cannot be asserted that kindergarten adjustment biased parents' recollections of daycare. Furthermore, on the basis of Vandell and Corasaniti's (1990) validity studies, it appears that mother's retrospective reports of roughly defined amounts of daycare do tend to be accurate. Yarrow, Campbell, and Burton (1970) had similar findings.

Dependent variables: Social-cognitive measures: Negative adjustment (based on teacher, peer, and observer composites); teacher-reported aggression (from the Teacher's Report Form; Achenbach and Edelbrock, 1986), plus reactive and proactive aggression from the Teacher Checklist (Coie and Dodge, 1988); peer-rated aggression (see Coie, Dodge, and Coppotelli, 1982); observer-rated aggression; and positive adjustment (teacher positive peer relations, peer popular/skilled).

Results: After controlling for SES, family stress, family structure, and marital quality, children who spent more time in child care during their first 5 years scored lower on a composite measure of positive adjustment (i.e., peer popularity or teacher-rated peer competence) and higher on a composite measure of maladjustment (i.e., teacher-rated behaviour problems, peer-rated aggression, peer dislike, or observed aggression) than children with less child care experience.

Even though child care experience in the first year of life did not predict behavioural maladjustment when child care experience in the second through fifth year was controlled, analyses did reveal that extensive care in the first year, coupled with extensive care thereafter, was associated with increased problematic functioning in kindergarten.

Limitations: Authors were not in a position to evaluate the role played by quality of care and assumed that the average level of care was below the high-quality centres reported in most studies.

2.4.5 Vulnerable Children: The Importance of Quality Child Care

Kohen, Dafna; Hertzman, Clyde. In J. Douglas Willms (ed.), Vulnerable Children (in press).

Background: This paper examines whether children's competencies (social and cognitive skills) differ across types of care arrangements, before and after adjusting for family structure and parents' socioeconomic status.

Methods: Three sets of secondary analyses are conducted on data from the NLSCY, a nationally representative longitudinal study of children and youth in Canada. The sample for the first two analyses was limited to children aged 4 and 5 since this was the group whose developmental assessments were available. The third analysis, concerned with the feasibility of providing regulated child care to all families with incomes below $35,000 for their children from birth to age 5.

Independent variables: Type(s) of care arrangements. This variable was coded as either regulated daycare, which is usually given in a regulated daycare centre, but may also include regulated family daycare; unregulated care in the home by a sitter or nanny; unregulated care outside the home, such as family daycare; and care by a relative, either inside or outside the home.

Dependent variables: Three measures of competency were used: Prosocial behaviour, the presence of a behaviour disorder, and PPVT-R scores (Pre-school vocabulary).

Results: In the case of children from low-income families, those who participate in care arrangements outside the home, either regulated or unregulated, have superior vocabulary skills to those who are cared for at home by a relative or those who use no care arrangements at all. The differences are substantial: for families with incomes of $15,000, the difference between children who participate in child care and those who do not is about 4 points on the PPVT-R test. Less than one-half of Canadian 4-year old children from low-income families attend a regulated daycare or pre-kindergarten program despite the potential benefits to development.

2.4.6 Non-Parental Caregiving

Clarke-Stewart, K. Alison; Allhusen, Virginia D.; Clements, Darlene C. In M.C. Bornstein (ed.) Handbook of Parenting: Volume 3. Status and Social Conditions of Parenting. (1995). Mahwah, New Jersey: Lawrence Erlbaum Associates. p. 151-176

Background: Presented are conclusions derived from studies on the developmental outcomes for preschoolers with and without non-parental care experiences.

Summary: With respect to cognitive development, there is a substantial body of research suggesting that the children who attend relatively high-quality daycare centres or early childhood programs perform better on cognitive tests than children from comparable family backgrounds who do not. This outcome is usually not observed for children in daycare homes or with in-home caregivers. In fact, on various measures of intellectual development, children in family daycare or in-home care perform at levels similar to children at home with their mothers.

Since differences between children in daycare centres and daycare homes are less when the daycare homes are of high quality, it would appear that the quality of care is more important than the type of care. For example, in one study reported by the authors in this chapter, although the language competence of children in unlicensed daycare homes was found to be inferior to that of children in centres, the language competence of children in regulated homes was equivalent (Goelman and Pence, 1987). In another study, when care in daycare homes was enriched by the experimental addition of an educational curriculum, the intellectual performance of the children was observed to improve to the level of children in daycare centres (Goodman & Andrews, 1981).

Children who attend daycare programs have also been shown to be different from children without non-parental care experience in their social behaviour. Compared with children without non-parental care experience, children with such experience are more "self-confident, outgoing, assertive, verbally expressive, self-sufficient, and comfortable, and less distressed, timid, and fearful in new situations." The authors argue they are also more independent from their mothers and exhibit more social skills. On the negative side, along with independence there may exist a lack of politeness and compliance with mothers' and caregivers' requests. The authors point out that one of the reasons for the different outcomes for cognitive and social development may be that few child care educational programs emphasize teaching children social rules. The authors define social rules as "effective ways of solving social problems." An alternative, though highly suspect argument is that "these aggressive children are acting out the emotional maladjustment they have suffered as a consequence of having formed an insecure attachment to their mothers earlier on."

2.4.7 Relations Between Preschool Children's Child-Care Experiences and Concurrent Development: The Cost, Quality, and Outcomes Study

Peisner-Feinberge, Ellen S.; Burchinal, Margaret R. Merrill-Palmer Quarterly. (1997). Vol. 43, no. 3. p. 451-477

Keywords: child care centres, quality of care assessments, children's cognitive and socio-emotional outcomes, children with at risk backgrounds, United States.

Background: This study was designed to test whether child and family characteristics moderate the relation between child care quality and pre-school children's outcomes with a large sample of community child care centres and children from diverse family backgrounds.

Methods: This article is based on data gathered during the first year (1993) of the Cost, Quality, and Outcomes Project (CQOP), a study which was designed to examine the relations among child care costs, quality, and longitudinal outcomes for children in full-time care in community child care centres. Sampling for the CQOP was conducted in four stages. First, child care centres and then classrooms within centres were selected. Next, a sub-sample of pre-school classrooms was chosen. And finally, children and families were selected from this sub-sample. These procedures yielded 757 (52% boys) children averaging 4.3 years of age. Data collection took place in two phases, with child care quality measured first, and child outcome data gathered from children, teachers, parents second.

Independent variables: (a) child and family characteristics, and (b) child care classroom quality. Demographic information was obtained from parent surveys. Each classroom was observed for approximately 3 to 4 hours in a single visit and four observational measures were used: (1) classroom environment was measured using the Early Childhood Environment Rating Scale; (2) teacher sensitivity was measured with the Caregiver Interaction Scale; (3) child-centredness was assessed by the UCLA Early Childhood Observation Form; and (4) teacher responsiveness was measured with the Adult Involvement Scale.

Dependent variables : Child's cognitive and socio-emotional developmental status. Data were gathered from (a) individual assessments and (b) teacher ratings. Receptive language comprehension/ability was measured using the Peabody Picture Vocabulary Test-Revised (PPVT-R; Dunn & Dunn, 1981); pre-academic skills were assessed using the Woodcock-Johnson Tests of Achievement-Revised (WJ-R; Woodcock & Johnson, 1990); and children's self-perceptions of competence and attitudes toward child care were measured by the Attitudes/Perceptions of Competence (Stipek, 1993). The Classroom Behaviour Inventory (CBI; Schaefer, Edgerton, & Aaronson, 1978), and Student-Teacher Relationship Scale (STRS; Pianta, 1992) were also used.

Results: After adjusting for child and family characteristics, the two child care quality variables were both jointly and individually related to better PPVT-R language scores.

The teacher's rating of closeness with the child was not significantly related to receptive language, whereas the observed quality index was positively associated with better receptive language skills as measured by PPVT-R. Neither child care measure was significantly related to pre-math skills. Only closer teacher-child relationships were associated with higher ratings of cognitive/attention skills. Fewer behaviour problems were associated with closer teacher-child relationships and, modestly associated with lower observed quality. Whereas better quality care had a positive influence on cognitive and socio-emotional outcomes for children, in several instances higher quality care had an even stronger positive influence for children from less advantaged backgrounds. There was no evidence that children from more advantaged backgrounds are buffered from the potentially harmful effects of poor quality care by the influences of the family. Quality of care affects the developmental outcomes of children from all backgrounds. Although the authors found significant relations between child care quality and all the child outcomes studied after controlling for background factors, the associations were modest. The strongest finding was in the area of children's language development.

2.5 Effects of After-School Care on Child Development

2.5.1 Patterns of After-School Care in Middle Childhood: Risk Factors and Developmental Outcomes

Pettit, Gregory S.; Laird, Robert S; Bates, John E.; Dodge Kenneth A. Merrill-Palmer Quarterly. (1997). Vol. 43, no.3. p. 515-38

Keywords: after-school care (ASC), self-care, types of care, care-outcome linkage, social competence, behavioural and academic adjustment, adult-supervised activity-oriented (A-O) care, United States.

Background: Most studies that have compared children's developmental outcomes on the basis of after-school care (ASC) arrangements have construed ASC exclusively as a single type. It seems likely however, that most school-aged children are involved in a variety of care arrangements throughout a typical week. This study seeks to capture more fully the range of ASC experiences common among school-aged children by evaluating longitudinally the impact of patterns of ASC on children's behaviour problems, social skills, and academic performance. Patterns here refer to (a) the extent of usual weekly involvement in different ASC arrangements, (b) combinations of arrangements occurring with some regularity, and (c) overall number of ASC arrangements used.

Methods: This study was completed as part of the ongoing Child Development Project, a multi-site longitudinal study of children's social adaptation. Participating families were recruited in two cohorts (1987 and 1988) from each of the three sites: Nashville and Knoxville, Tennessee, and Bloomington, Indiana. The sample consisted of 585 families at the first assessment prior to kindergarten. Follow-up assessments of the children were conducted in kindergarten and every grade thereafter through Grade 6; follow-up family assessments (via parent questionnaires) were conducted in the summer following kindergarten and in all subsequent summers.

Independent variables: (a) children's involvement in differing types of ASC in Grades 1, 3, and 5, (b) family SES, (c) child sex, and (d) kindergarten adjustment. ASC data were derived from guided recall in-home interviews conducted with mothers in the summer between the children's fifth-and sixth-grade school years. Out-of-school care arrangements were classified into eleven types.

Dependent variables: (a) behavioural problems, (b) social competence in peer relations, and (c) academic performance in Grade 6 (measured by grade point average and by standardized test scores).

Results: Mothers reported that their children spent an average of 8.7 hours (SD = 8.2) per week in non-parental care in first grade, 9.8 hours (SD = 8.7) in third grade, and 10.6 hours (SD = 8.5) in fifth grade. The overall complexity of non-parental care increased across grades, from a mean of 1.3 (SD = .09) arrangements in Grade 1, to 1.6 (SD = 1.0) in Grade 3, to 1.8 (SD = 1.0) in Grade 5 (all t test comparisons p< .001). Amount of self-care, as a factor by itself, and in interaction with SES, was associated with later adjustment problems. Lower SES children experiencing higher amounts of self-care in first grade were rated by teachers (Teacher Report Form; Achenbach & Edelbrock, 1986) as having more externalizing problems in Grade 6 than lower SES children not experiencing self-care. However, daycare and sitter/relative care appeared to buffer the effects of low SES in that lower SES children experiencing these types of care showed better adjustment than lower SES children not involved in these types of care.

2.5.2 Children of the National Longitudinal Survey of Youth: Choices in After-School Care and Child Development

Vandell, Deborah Lowe; Ramanan, Janaki. Developmental Psychology. (1991). Vol. 27, no. 4. p. 637-643

Keywords: latchkey arrangements, adult supervised after-school care, single mothers after-school care, "at risk" children, National Longitudinal Survey of Youth, behavioural problems, cognitive assessments, United States.

Background: This study sought to address four questions: (a) Are there differences in demographic and family characteristics associated with children's after-school care arrangements? (b) Are there differences in child development associated with after-school care? (c) Are associations between after-school care and child development maintained after controlling for differences in family characteristics? (d) Do different types of after-school arrangements differentially affect children from low-income, single-parent, or ethnic-racial minority households?

Methods: These questions were examined by using a national data set (U.S.) called the National Longitudinal Survey of Youth (NLSY). The NLSY was begun in 1979 as a survey of youth aged from 14 to 19 years. The original sample was selected with an over-representation of Blacks, Hispanics, and low-income Whites. In 1986, the survey was broadened to incorporate data from 4,953 children of the original female NLSY respondents. Children were directly assessed, and mothers were asked about their employment and child care used, in addition to other demographic information. This study focuses on the subset of third-, fourth-, and fifth-grade children within the larger NLSY data set. A total of 390 children (199 girls) were examined. At the time of the assessments (1986) 47.3% of the children lived in households whose incomes fell below the poverty line. Fifty-two percent of the children were members of single-parent households and most were born to adolescent mothers.

Independent variables: (a) type of after-school arrangement, (b) quality of the home environment. Three types of after-school arrangements were designated: Latchkey care - if the child cared for himself or herself or was cared for by a sibling less than 15 years old; other-adult care - if the child's after-school care was provided by the father, grandmother, other relative, baby sitter, or daycare centre; mother care - if the mother cared for the child after school. The shortened form of Caldwell and Bradley's HOME scale was used.

Dependent variables: (a) social, (b) emotional and, (c) cognitive development. A revised form of the Behaviour Problems index (Peterson & Zill, 1986) was completed by the child's mother. Children completed two scales from Harter's (1984) Self Perception Profile of Children. Three sub-scales from the Peabody Individual Achievement Test (PIAT) were used to measure children's mathematics, reading comprehension, and reading recognition. The Peabody Picture Vocabulary Test (PPVT) was used as an indicator of the children's verbal functioning. The Digit Span Sub-scale of the Wechsler Adult Intelligence Scale-Revised (WISC-R) was used as an indicator of children's short-term memory and attentiveness.

Results: Families using mother care after school had about 70% of the income of those families using latchkey or other-adult care. Families using mother care after school as opposed to latchkey or other-adult care also had poorer home environments as measured by the HOME-SF scale. Differences in emotional support, but not cognitive stimulation, were the source of the overall differences on the HOME-SF scale. There were no significant differences in the emotional support provided by families using latchkey or other-adult care. Children who returned home to their mothers after school were typically rated as having more antisocial behaviours, more peer conflicts, and lower PPVT test scores than did children in other types of after-school care. The subsequent interaction analyses that controlled for differences in family income and emotional support revealed that these effects were restricted to single-parent families. These observations parallel other recent studies that have found negative developmental outcomes associated with exclusive care by single mothers.

Limitations: The sample was relatively small with limited statistical power, especially for detecting interactions between type of after-school care and family functioning. Data were based solely on mother's report.

2.6 Child Care Selection

2.6.1 Early Child-Care Selection: Variation by Geographic Location, Maternal Characteristics, and Family Structure

Fuller, Bruce; Keiley, Margaret K; Singer Judith D; Wolf, Anne. Developmental Psychology. (1998). Vol. 34, no. 5. p. 1129-1144

2.6.1.1 Keywords:

Initial non-parental care choices, child care selection process, National Longitudinal Survey of Youth (NLSY), United States.

Background: Little knowledge is available about which families are most likely to select non-parental care and at what age in the child's life. This paucity of research makes it difficult to determine: When young children from different types of families enter non-parental care; whether access to early child care is equally distributed across diverse families; whether early child care selection puts only certain youngsters on a developmental trajectory that eventually includes exposure to centre-based pre-school; and whether the apparent effects of formal programs, such as enrichment programs, stem from the pre-school "treatment" per se or from prior selection effects rooted in the family's attributes and practices.

Research designs to date assume that home effects are direct, rather than recognizing that a parent's management of his or her child's time outside the home may yield additional indirect effects. The child care selection process over time, commencing with the child's care following birth, needs to be adequately modeled. Three explanations have been advanced for why families differ in their propensity to select non-parental or centre-based child care.

Household economy and social class: One initial American article on selection of infant care shows that high-income and impoverished families display the greatest probabilities of selecting non-parental providers (NICHD). Some argue that working-class households are least able to use centres in the U.S., in as much as they are ineligible for subsidies and cannot afford private fees. Yet, low-income families in the U.S. (with incomes under $15,000 annually) are only somewhat more likely to use kin members and less likely to select centre-based care, compared with affluent households earning more than $50,000 (Hofferth, Brayfield, Deich, & Holcomb, 1991). Thus, in terms of type of arrangements, social class may not play a big role.

Family demographics and social structure: Better-educated parents are consistently found to be most likely to place their children in non-parental care. In the U.S., at least, many studies have found racial and ethnic differences in the use of child care, although the reasons for these differentials remain unclear. Variability across ethnic communities in the organized supply of centres or family daycare homes may further contribute to differences in the use of non-parental care. Parents, as they have more children, are less likely to select non-parental providers or they shift to less expensive forms of care.

Organisational-level factors: Variation in the supply of non-parental care across states (provinces) and localities may also influence the individual family's propensity to use care. Early studies used family-level data to draw inferences about supply, access, and the distribution of quality of non-parental care providers across local communities. The effect of mothers' previous employment has been found to vary depending on the child's age and on the number of other children in the home. This illustrates that the effects of any economic factor (in this case previous employment) on child care selection must be understood in the context of the family's social structure, not solely on parent's attributes.

2.6.1.2 Keywords:

Initial non-parental care choices, child care selection process, National Child Care Survey (NCCS), discrete-time survival analysis, child care histories, United States.

Background: This article provides initial evidence as to which families are more likely to use non-parental forms of child care (babysitters, family daycare homes, kin members, and centres) and at what age initial entry into these settings occurs. The authors begin be reviewing what is known about the child care selection process. They argue that the evidence continues to focus on later selection of centre-based programs for children 3-4 years of age and relies on traditional statistical methods - linear and logistic regression analysis. Such methods, it is argued, are ill suited for analyzing the unfolding pattern of child care selection over time. As an alternative, discrete-time survival analysis is used to model the event histories of a national probability sample of 2,614 children under 6 to determine whether and if so when, they first entered non-parental care. Commonly used by bio-statisticians studying human lifetimes, survival analysis can be used to study how long it takes for any event to occur, even when the event is within an individual's (or his or her parent's) control.

Methods: In early 1990, 4392 parents with at least one child under age 13 participated in the NCCS, a telephone survey focusing on child care issues. This article focuses on the youngest child in the sub-sample of 2,614 households with a child age 6 or under who had not yet entered kindergarten, in which a mother was present, and for whom the parent provided information on whether and, if so, when, the child first entered non-parental care.

Independent variables: (a) geographic location; (b) mother's characteristics - education, age at first birth, race/ethnicity, working before birth; and (c) family structure - single-parent family, number of siblings.

Dependent variable: entry (time) into non-parental child care (in months).

Results: Few differences were found in terms of urbanicity but large differences in terms of region. Unequal distribution of child care supply was observed across regions and communities. Nearly 1/4 of all children were placed in their first non-parental child care setting in their first 5 months of life. As children grow older, the probability of placement declined. Many women not known to be working during pregnancy placed their children into non-parental care. A positive relationship between mother's education and child care use was found. The older a woman was when she began her family, the less likely she was to place her children in care.

The effect of maternal employment during pregnancy was profound but varied both by the child's age and by the number of other children in the home. Among mothers who choose not to place the first-born child in care, the effect of previous employment diminished over time, disappearing entirely by the time the child turned 3. But if the target child was third born or more, the probability of placement was much lower, even if the mother was working before the child was born. In other words, the effects of family size are not linear, but are observed most strongly after a third child is born. But maternal employment during pregnancy also had effects independent of the child's or family's specific demographic characteristics (e.g., single-parent status, maternal education, ethnicity, or mother's age at first birth). Controlling for the effects of geographic location, maternal demographics, and ethnicity, the odds of entry into care for children from single-parent families were nearly twice as high (1.9) as those for children from two-parent families. Overall, as others studies have found, mothers who are single, who worked during pregnancy, and who have no other children, are most likely to use non-parental care.

2.6.2 Measuring Non-Parental Care in the NLSCY: Content and Process Issues

Norris, Christina; Brink, Satya; Mosher, Patricia. Applied Research Branch, Strategic Policy, Human Resources Development Canada. Technical Paper T-00-1E. (August 1999).

Keywords: Non-Parental care, child care, child outcomes, characteristics of child care, patterns of non-parental care in Canada.

Background: In 1998, the Human Resources Investment Branch (HRIB) of HRDC released a series of reports profiling the characteristics and work environment of non-parental care providers working in the regulated and unregulated care sectors. The purpose of this technical paper is to use the work undertaken by HRIB on the child care sector to examine the content and process issues related to the measurement of non-parental care in the NLSCY. The paper also examines the pattern of non-parental care use in Canada, and the family and socio-demographic characteristics associated with use of specific forms of care.

Findings: In 1994, 33.2% of children between the ages of 0 and 11 years were in some form of non-parental care and there appears to be an emerging profile of children's care providers. Those who provide care in their own homes are typically older, married, and have children of their own. Cost of child care is influenced by both the age of the child and the type of care provided and care providers working in the child's home generally receive the lowest incomes.

As a family's socio-economic status (SES) increases so does their reliance on, usually unregulated, non-parental care (families of lower SES seem to rely mostly on daycare centres). Mothers under 40 are also more likely to employ some form of non-parental care.

Types of non-parental care have been shown to vary across the regions of Canada. Family child care and daycare centres are least used in the Atlantic Provinces and are most used in Quebec. Relative care is most popular in both the Atlantic Region and British Columbia.

Non-Parental care use is nearly equivalent between one- and two-parent families at 39.1% and 40.0%, respectively. However, two-parent families are more inclined to use unregulated and in-home care by a non-relative, whereas one-parent families tend to use daycare centres.

Families living in rural areas use non-parental care slightly less (34.9%) than families that live in the city (39.7% - 42.2%).

The structural aspects of non-parental care that should be taken into consideration for measurement include: group size, the quality of the physical setting, the training the caregiver has received, and the caregiver-to-child ratio. The process measures involve quantifying the interaction between the child and the caregiver as well as the actual care given. The affective aspects, the developmental "appropriateness" of the child care experience, and the caregiver's sensitivity and responsiveness should also be taken into consideration when measuring the quality of non-parental care.

²The sensitive mothering variable is a combination of two measures. The mothers' behaviours in a videotaped interaction between her and her child at home or in the lab, and the mother's behaviour as assessed by HOME (Home Observation for Measurement of the Environment) at 6, 15 and 36 months.

Contents

The Impacts of Non-Parental Care on Child Development - August 1999