Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Measuring the Outcomes of Day Care Jeffrey R. Savers, Rochelle Beck, and Joan Bissell Day care in the United States comprises a very hetero- geneous collection of "programs"--some of them public, some private, some institutional, some informal. It can be a large, smoothly run, full-day nursery school in a splendid facility with highly professional staff; it can also be Mrs. Jones taking care of Mrs. Smith's kids in the Jones' family playroom. Its goals and functions are as varied as its sponsors and practitioners, and its clients range from infants to school-age children, from the poorest to the wealthiest families. This heterogene- ity poses major challenges for outcome measurement. It requires an arsenal of measures appropriate to different goals, different settings, and different client populations. Problems of measurement are exacerbated by the fact that day care is highly politicized, in the broadest sense of the term. Day care has many ~constituencies"; many groups with divergent interests have different perceptions of its proper goals and functions. A few examples illustrate the range: Day care can be seen as a service to children, intended to equip them with School readiness" skills or to support the development of their social skills and emotional strengths. It can be seen as a service to parents, designed to free them for work or other pursuits. It can be seen as a family support service, intended to strengthen families by allowing them to increase earnings while still meeting their child- rearing responsibilities. It can be seen as a societal tool, designed to increase employment and upward mobility, augment the tax base, and reduce the welfare rolls. It can be seen as a vehicle for delivering services such as 109
110 health care, nutrition, parent education, family counsel- ing, and the like to low-income families. While these views are not necessarily incompatible, each points to a different kind of emphasis in outcome research. Day care also has organized opponents, who see it as undermining the family and who see government support for day care as unwarranted intrusion into family rights and responsibil- ities. This negative view, too, has implications regard- ing unintended outcomes, which should be considered in evaluating day care programs. Issues of measurement for day care demonstrations are rendered even more complex by the fact that demonstrations can address two distinct types of questions, which might be termed "program" questions and "policy" questions. Program questions have to with the best ways of operating day care programs, e.g., the most effective methods of recruiting and training staff, the most effective "curricula" or activities to use with children, or the most effective means of eliciting parent involvement. Policy questions have to do with the proper roles of the various levels of government and with the most effective means of achieving governmental goals. Examples include: Should the federal government encourage out-of-home care for young children, through subsidies or other incentives, or should it subsidize parents who stay home and care for their own children? Which day care subsidy mechanisms (e.g., vouchers, direct purchase of care, income disregards) maximize parental choice? Which maximize quality of care? Which maximize cost-effectiveness? What is the most appropriate and/or effective division of labor among the federal government, states, and localities in regulating and monitoring the quality of care? Should health and social services be delivered through "client-oriented" day care or through more l Income disregard is a system under which mothers receiving Aid to Families with Dependent Children (AFDC) under Title IV-A of the Social Security Act are allowed to earn income above the maximum levels normally permitted for those receiving such aid, provided that the surplus is spent for child care. The system is designed to prevent the cost of child care from becoming a barrier to prevent welfare mothers from entering the labor force.
111 specialized service agencies, such as health clinics? Clearly, demonstrations addressed to these different levels of question require different outcome measures. For example, different measures are needed to assess an exemplary day care program designed to demonstrate innovative techniques for educating children and a model information and referral system or a voucher experiment designed to demonstrate ways to stimulate private initiative and maximize parental choice. Given the multiplicity of goals of day care programs, of day care constituencies, and of demonstrations in day care, it seems obvious that outcome research must itself be multifaceted. No single study could address all of the (quite legitimate) concerns sketched above. Thus day care confronts researchers with the need to stake out their turf clearly--to start with an explicit framework of values, goals, concepts, and questions and to recogniz that alternative frameworks exist and to take that broader context into account. While any one study must be limited in focus, it is important for researchers, and those who interpret research, not to draw erroneous policy conclu- sions from data that address only one domain of concern. For example, it might be the case, as some have argued, that the most cost-effective way to enhance children's cognitive development is through the education of parents and/or in-home intervention with very young children. But such a claim (if true) would not necessarily argue against support of developmental day care, since the latter potentially provides other benefits to families. (Other examples are cited below, particularly in connec- tion with our discussion of the effects of day care on employment and family income.) The admittedly ambitious aim of this paper is to map the broad terrain of outcome measurement for day care from a bird's-eye view. We begin with a brief overview of day care in the United States and a discussion of the concerns of the many constituencies of day care--children, parents, providers, researchers, and policy makers. The body of the paper provides a taxonomy of potential out- comes addressed to these multiple concerns, surveys the current status of measurement with respect to each class of outcomes, and suggests needed additions and improve- ments. The final section summarizes the paper's main conclusions about the state of outcome measurement for day care and its recommendations for the future. e
112 BACKGROUND Day Care in the United States: An Overview The use of day care in the United States, although not entirely work related, is intimately linked to the labor force participation of women, which has increased dramat- ically in recent decades. In 1950 only one fifth of all mothers with children under 18 were employed; by 1978 the proportion had increased to more than one half. The largest percentage increase occurred among women with children under six, whose labor force participation nearly tripled (from 14 percent to 40 percent) during this period (Congressional Budget Office, 1978:44). Labor force participation of mothers is highest among women who head single-parent families. Among two-parent families, the labor force participation of mothers is greatest when the father's income is low (Johnson and Hayghe, 1977). The above data can be and have been interpreted as evidence for an abiding and increasing demand for child care. There is, however, heated controversy over the proper public response to this apparent increase in need. Some commentators, often remarking on the activist family policies of the governments of other industrial- ized nations, have argued for increased public subsidy of child care. Others have argued that the increase in demand has been exaggerated and that private market mechanisms are adequate to cope with it. Still others have decried the labor force trends as indicators of the decline of the family, construing the demand for increased subsidy as an invitation to increased government encroach- ment on family rights. It is not our purpose to take sides on these issues but to describe the day care "market" as it currently exists, the role of the govern ment within that market, and certain new developments that seem likely to raise salient program and policy issues in the 1980s. Subsequent sections of the paper outline concomitant issues of outcome measurement. We use the term "day care" broadly to mean care provided on a regular basis by persons other than immediate family members (parents, live-in grandparents, older siblings). However, our primary focus is on paid care by nonrelatives, provided in the child's home or elsewhere. Full- or part-time day care is an experience shared by large numbers of American children. According to a -
113 national consumer survey published in 1975 (UNCO, Inc., 1975, Vol. II:6-8 to 6-11), more than 5 million children age 13 or younger were at that time cared for essentially full time by someone other than a parent, i.e., for 30 or more hours per week. Another 6 million children receiving care from persons other than their parents for periods between 10 and 30 hours per week. Somewhat more than half the children in full-time care were supervised by someone other than a relative, and about two thirds were in care outside their own homes--statistics that also imply, of course, that in-home care and care by relatives accounted for a large portion of full-time nonparental care. Most out-of-home care is "family day care," provided in the care giver's home to small numbers of children (six or fewer, including the care giver's own children, by federal regulatory definition.) According to the consumer survey, approximately 1.3 million family day care homes serve 2.4 million children full time, 2.8 million children from 10 to 29 hours per week, and much larger numbers on an occasional basis. Only about 900,000 children received care in centers during 1976-1977, according to a national telephone survey of more than 3,000 centers, roughly one of every six in the country (Coelen et al., 1978). (This survey employed a relatively strict definition of the term "day care center" and excluded mixed care arrangements, in which children are in nursery school for part of the day and in family day care for the rest of the day.) Children of different ages are distributed unequally across types of care. Preschoolers (ages three through five) are the predominant age group among children in full-time care by nonrelatives. School-age children predominate among those in part-time care. Among children who receive care in their own homes, more than half are of school age, and almost 30 percent are children under three (infants and toddlers); only 20 percent are in the preschool age range. Conversely, as suggested by the consumer survey and confirmed in the recent national telephone survey of day care centers (Coelen et al., 1978), most of the children served in centers (70 percent) are preschoolers; the remainder is divided equally between younger and older children. In family day care, children under three are the largest group served. There is some controversy over the proper interpreta- tion of those distributional facts. Surveys of parental
114 preferences (e.g., Hill, 1977, 1978; Steinberg and Green, 1979; Rowe et. al., 1972; Fosburg and Hawkins, 1981) indicate that parents are relatively reluctant to place very young children in day care and, when they do so, prefer to use in-home care or home-like, family day care, often within close proximity to the home. As children approach preschool age, parents are more willing to use out-of-home care and are more likely to turn to day care centers as sources of group educational and social experiences that may help prepare children for school. When children reach school age, the school itself provides group experiences; parents again turn to informal in-home or family day care arrangements to provide supervision during after-school hours. On one hand, the existing pattern of care can be seen as a reflection of parental preferences, thereby reflecting--for those who believe that parents know what is best for their children--the interests of children. On the other hand, parents cannot choose forms of care that are unavailable or beyond their means, and they are unlikely to state preferences for forms of care about which they know little. Thus the widespread preference for and use of informal, small-scale arrangements may in part reflect a lack of awareness and/or access to other forms of care, particularly formal, enter-based care, especially for children of school age and those under three. Whatever the reasons for the distributional facts, the facts themselves represent important realities with which demonstration projects and outcome measurement must reckon. Outcome measurement in the domain of child development has, for valid historical and theoretical reasons, focused on the effects of center care; particular attention has been given to its effects on infants and toddlers. Though this research has yielded relatively clear and valuable insights, it has concentrated on the least-used form of care and on an age group that is underrepresented in that form of care. (Fewer than 40,000 children under two are in center care, most of them in the Southwest, as reported by Ruopp et al., 1979.) Until recently, research has neglected the informal care arrangements that affect most children under three and many older children as well. Although the role of government in child care is a bitterly debated topic, massive involvement of government at all levels, especially the federal level, is already a reality--with which outcome measurement must deal if it is to be relevant to policy. In fiscal 1977 estimated , _ _ _
115 federal and state expenditures for day care and other early childhood programs exceeded $2.7 billion. This sum includes expenditures on services other than day care as conventionally defined, e.g., Head Start and the Department of Agriculture's Child Care Food Program, which provides food subsidies to child care facilities serving children from low-income families. However, its largest component is the $809 million spent on day care through grants to states under Title XX of the Social Security Act, followed by $500 million in tax revenues foregone under the child care tax credit. It also includes an estimated $500 million in state and local matching funds (Congressional Budget Office, 1978). Although federal funds are used to purchase care in all types of facilities, those monies targeted for the poor are disproportionately allocated to centers. About 70 percent of Title XX funds are spent in centers, 17 percent in family day care homes, and 14 percent on in-home care (U.S. Department of Health, Education, and Welfare, 1978). Approximately 200,000 children from low-income families receive center care that is wholly or partially subsidized. However, low-income parents who receive subsidies through the AFDC income-disregard mechanism disproportionally choose family day care over center care. Closely linked to governmental funding of child care is governmental regulation. States and a few localities maintain licensing codes, which set standards that child care facilities must meet in order to be allowed to operate. These codes affect virtually all centers and, in some areas, family day care homes as well. Most family day care, however, is unlicensed--90 percent, according to one survey (Westinghouse Learning Corp. and Westat Research, Inc., 1971)--and licensing requirements for family day care, where they exist, tend to be enforced erratically if at all. In-home nonparental care is not subject to licensing. In addition, the federal government maintains purchasing standards, which specify the types of facilities in which federal dollars may be spent. These standards, the Federal Interagency Day Care Requirements, established in 1968 and currently being revised (see the Federal Register, March 19, 1980) are stricter than the licensing-standards of most states, especially with respect to required ratios of numbers of staff to children and have therefore been controversial because of their potential cost implications. In practice, federal purchasing standards, like state licensing requirements,
116 have affected centers more than family day care homes and are likely to continue to do so. Also, the federal standards primarily affect the care purchased by the states and their local delegate agencies using Title XX money; care purchased by the poor under the income- disregard mechanism and care purchased by middle- and upper-class families using the tax credit are effectively unregulated by the federal government. Thus, whether one views federal and state regulations as necessary and benevolent attempts to set a floor under the quality of care or an unwarranted intrusion of government in the child care market, some form of regulation is a reality for most centers but for only a fraction of family day care homes. In sum, government at all levels is heavily involved in child care. Government purchases or underwrites care for large numbers of children, primarily for the poor but also for the more advantaged classes (through tax credits). The principal policy tools used by government for influencing the type and quality of care received by children have been funding strategies and regulations. This public presence in the day care market has not been guided by a coherent national child care policy. It has arisen in part as a by-product of other policies designed to support low-income families or to induce low-income single parents to work, thus reducing welfare expendi- tures, and in part as an effort to provide tax relief to the middle class. Federal support for child care to low-income families coexists with other federal policies, such as Aid to Families with Dependent Children, which subsidizes parents to care for their children at home. In this regard, American day care policy stands in sharp contrast to the policies of other industrial nations in Europe. The European countries make a much larger relative investment in children and families, and they have relatively clear-cut policies designed to encourage either parental care or parental employment, depending on their respective labor markets (Kamerman and Kahn, 1978). There have been repeated calls by prestigious groups in the United States for a national child care policy (Keniston and the Carnegie Council on Children, 1977; National Research Council, 1976). It is impossible to say whether the 1980s will see a serious attempt to establish such a policy. However, with or without such an attempt, it is clear that debates about funding and regulation will continue, and that policy researchers will be called on to produce data relevant to those debates.
117 An overview of the American day care scene would not be complete without some discussion of relatively new developments that may pose new policy questions for the 1980s. Some states, such as California, have begun to experiment with funding mechanisms such as vouchers, designed to increase parental discretion in the purchase of care and to capitalize on the responsiveness of private providers. A related development is the growth of infor- mation and referral services, some of them publicly subsi- dized; these services are designed to facilitate the match between parental needs and existing child care resources. The increased labor force participation of women also has led to the beginnings of new demands on unions and employ- ers to include child care in employee benefit packages. New experiments with union- or industry-supported child care may be in the offing. The financing of day care is likely to become an increasingly salient issue, as the field becomes increasingly professionalized and as day care workers--among the nation's lowest paid--seek recognition and increased compensation for their services. Informal, low-cost care by friends and relatives may absorb less of the latent demand than it has in the past as women who heretofore provided such services enter the labor force. These developments, and others as yet unforeseen, are sure to create needs for new forms of evaluation design and new outcome measures. Who Cares About Day Care- and What Do They Care About? As suggested earlier, day care has many constitu- encies--groups sharing common interests and perceptions of the aims and functions of day care. Some of these groups, such as children and parents, are beneficiaries of day care; others, such as researchers and policy makers, are gatekeepers r who control public information and decision making. Some are providers and the persons who train them. The interests and perceptions of the various groups are not mutually exclusive; they overlap and intersect at many points. Moreover, the views of the various groups need not necessarily receive equal weight in the choice or the development of outcome measures; a case could be made, for example, that the needs of chil- dren and families are paramount. Nevertheless it is useful to enumerate the constituencies and identify the outcome measures most salient for each, to provide a
118 comprehensive framework within which existing measures can be located and evaluated. First among the constituencies are children. They are not a political constituency in the usual sense; they do not speak for themselves, individually or collectively, but rely on adult advocates to express their needs and defend their rights. Yet they nonetheless have needs and interests that may differ from those of every other group, perhaps even their parents. They need physical activities and educational experiences to stimulate their develop- ment. They need to interact with adults and other children in order to begin to learn about themselves-- what they enjoy, what they do well, what they want to be--and about how to form relationships with others. Some adult advocates feel that they need to begin to develop a sense of their cultural as well as personal identities. Moreover, given that many children spend 8 to 10 hours of their 12-hour waking day in care, the quality of life available to them while in care is a prime concern in itself, regardless of its developmental effects. Children need a safe and pleasant physical environment, appealing and nutritious food, and, in some cases, special services such as diagnostic screening and health care, which may be available only through day care. The need to measure development comprehensively--not to rely on traditional measures of cognitive skill or ability that have been used in evaluating other programs for children--has been widely recognized but only partially met by day care researchers. The need to measure immediate quality of life has barely been acknowledged as such, although relevant aspects of the environment have been studied. Second are parents, who may have several purposes for using day care in addition to providing the child with a pleasant and stimulating environment: to enable a second or single parent to enter the labor market, to learn about child rearing, to feel less isolated, to help get through temporary crises. Availability of day care might permit some mothers to participate in vocational education, thus improving their marketable skills. Availability of day care might permit single mothers to work, and fewer might apply for welfare as a result. Parents might feel more confident about their abilities to raise their families, as a consequence both of their improved economic situation and of the help and advice given by the day care providers. The consequent reductions in stress might even result in fewer single parents being
119 institutionalized or referred for psychiatric care. Obviously, a wide range of measures is needed to address these questions and to capture the equally wide range of outcomes that parents might expect from day care. Third are families. As a unit the family has somewhat different needs from its individual members. For example, a family may need a day care program to help it maintain a viable income, to help it stay intact during a troubled time, to help its members interact more positively, or to prevent negative interactions such as spouse or child abuse. Does the availability of day care decrease the need for foster care or institutional placements? Would fewer families disintegrate under economic pressures or in times of illness or crisis if they had day care arrangements to relieve some of the daily burdens of child- rearing? Are families more nurturing if they have outside sources of respite or advice? Is the incidence of spouse or child abuse reduced as a result? Again, a range of measures that goes beyond that normally associated with day care is indicated. Fourth are communities. While we do not usually think of communities per se as having an interest in day care, there may be legitimate outcomes worth measuring from the perspective of the community. For example, a community with insufficient or low-quality day care programs may have higher welfare expenditures or it may discourage families with two wage earners from living in it. A community offering high-quality care, by contrast, may be able to attract businesses and families and thereby increase the tax base. Aside from the various economic implications, the availability of day care services may in part characterize a community as hospitable or not, intimate or not, accessible or not, a good place for families or not--a characterization that may itself affect families living in that community, their interactions, and their expectations for their children. Fifth are care givers. Employees in the growing day care industry comprise a wide variety of people: from neighbors with no training or professional experience, who baby-sit for several children, to highly trained professionals in large day care centers, with theories, equipment, and routines for handling groups of children. What is measured, how it is measured, and the interpreta- tion of the results directly affect livelihood, reputa- tion, self-image, and future income. While the employees often share with the children and parents a concern for many of the outcomes deemed important in child care (such
120 as the quality of daily interaction and the happiness of the child), their stake as providers of this service colors their outlook and their need for information. Their working conditions, wages, job stability, profes- sional recognition, and professional growth are concerns that others may not share. And evaluation, especially if it involves direct observation of their interactions with children, may be threatening to them and may therefore require extensive consultation, explanation, and justification. Sixth is the research community, particularly (1) experts on child development and the family and (2) policy researchers and program evaluators. Though these two groups may overlap in membership and outlook, they are distinct in objective: The first seeks basic and applied knowledge about children's psychological growth and family functioning, and the second studies the effects of programs and public policies. Often the research conducted by both groups focuses on the concerns of parents, providers, and communities, but this is not always the case. Because researchers view day care from particular theoretical perspectives, through the lenses of particular research techniques, the measures they choose sometimes communicate important information to other researchers but are incomprehensible or irrelevant to parents and care givers. For example, many parents and community advocates were genuinely surprised when the most widely publicized early evaluations of Head Start looked primarily at intellectual development and used measures narrowly applicable to school success. Many had never seen that outcome as the overriding purpose of the program, and they wondered why reports concerning such important outcomes such as the delivery of health, social, and nutrition services did not receive equal attention or why some of the dynamics of introducing a community- controlled program into poor and minority neighborhoods could not be measured. Day care research has suffered from a similar narrowness of focus. Finally, there are policy makers and government program managers at the local, state, and federal levels. Respon- sible for decisions about the allocation of resources and the administration of programs, for the creation and implementation of laws and regulations, for assessing the needs of children and families, and for setting priorities and creating programs to address those needs, policy makers often are interested in outcomes that have to do with the functioning of the service delivery system.
121 While the effects of programs on individual children and families are important, in many cases the outcomes most relevant to policy makers address issues of access, equity, and efficiency that transcend the concerns of individual children, parents, or providers: How many children have been served? At what unit cost? From what revenues? Is the delivery system working efficiently? Is it freeing or draining local funds for other needed public services? Have families increased and hence contributed additional federal, local taxes? What is the total day care capacity in the community? What ways are there to measure unmet needs reliably? What is the nature of the day care market? Is it mixed enough (i.e., public vs. private, school vs. nonschool, family vs. center, subsidized vs. fee) to allow for real choice? Is parents' knowledge about day care options sufficient or increasing, so they can make informed choices among services? Do different types of care (e.g., family day care homes, small group homes, larger day care centers) work together in an integrated manner, by sharing resources, making referrals, and so on? Have parents who are active in the decision making of their day care facilities become more active in other community institutions or political processes? Have licensing or other regulatory mechanisms improved the quality of care? Limited its availability? Changed the nature of the people or organizations that enter the field? Is day care an efficient vehicle for the delivery of other services to children, such as health care? What agency or agencies, at what level or levels of government, should be responsible for day care policy? The answers to many of these and other important questions may involve simple accounting procedures and may not fit into any theoretical framework. Yet these atheoretical indicators may have as much value in assessing day care programs as research findings arising from traditional experimental designs and measures. Questions such as the ones above apply in different ways and to different degrees at the different levels of government; hence, different "systemic" outcome measures are salient for different policy audiences. Local policy makers operate within a framework of laws and regulations established at higher levels. They tend to be concerned with issues of compliance and with the detailed fit of available services to local needs. State policy makers and program managers tend to be concerned with issues of equity, access, and cost of services across localities their earnings state, and
122 within the state as well as with trade-offs among human services within constraints of the state budget and of discretionary federal programs, such as Title XX. They are also concerned with issues of accountability and monitoring and with the administrative machinery needed to carry out these functions. Federal policy makers are concerned with issues of uniformity of basic levels of service across states, leadership, fostering research and innovation, other goals important to the nation as a whole (such as a literate citizenry, a productive work force, and low unemployment and tax rates), and equality of opportunity for women and minorities. Thus, policy makers at the national level often frame the purposes of programs and outcomes to be measured in terms that are somewhat removed from the concerns of individual children, parents, or local communities. Sometimes they formulate outcomes in terms of the rhetoric that led to successful legislation of the program or that pertains to other salient political goals. For example, day care often "sells" as a device for reducing welfare rolls because many middle-class voters view the reduction of welfare expenditures as a legitimate goal but are reluctant to support underwriting additional social services. From this perspective, the reduction of welfare rolls or overall taxpayer cost savings (the cost of welfare compared with the cost of child care, training, and job placement) become salient outcomes; direct measures of services received do not suffice. Existing research has made only sporadic use of a theoretical indicators of the functioning of the child day care system as a whole and has not systematically addressed the concerns of policy makers and program managers at the various levels of government. A TAXONOMY OF DAY CARE OUTCOMES As the foregoing discussion suggests, the researcher who sets out to evaluate a demonstration program in day care is confronted with a confusing welter of potential program and policy objectives. Maximizing achievement of some objectives may not be fully compatible with maximiz- ing the achievement of others. Inevitably the researcher must choose to stress certain outcomes, or trade-offs among outcomes, while downplaying or ignoring others. What is crucial is that this choice be informed and deliberate--that it be based on an appreciation of the
123 goals of a particular demonstration and or the information needs of the intended audiences of the evaluation and not on expediency, convention, or failure to consider import- ant outcome domains. To facilitate informed choice, this section sets forth a taxonomy of potential outcomes, based on consideration of the interests of the many constituencies of day care. The taxonomy is intended to be systematic and complete in identifying broad categories of outcomes to be considered. Within these broad categories, numerous specific outcomes are listed; however, at this more specific level the taxonomy is intended to be suggestive rather than exhaus- tive. Many of the outcomes discussed are not usually thought of as such; they include, for example, measures of service delivery and of the quality of the physical and social environment provided to children. In most existing research, such variables, if considered at all, are treated as "independent"; dependent measures (outcomes) in most studies are measures of developmental change in children. We argue, however, that traditional conceptions of outcomes, derived primarily from develop- mental psychology, must be broadened to take account of the diverse purposes of day care demonstrations and the concerns of its constituencies. The section also surveys some of the major types of measures used in existing studies and comments on their adequacy; gaps in measurement are identified wherever they exist. Again, the intent is not to review every measure ever used in day care research but rather to identify broad areas of strength and weakness in current measurement. Although substantive findings are mentioned, the discussion is not a comprehensive review of the literature nor does it comment systematically on the quality of research designs or the soundness of substantive conclusions. The taxonomy of measures represents a widening circle, beginning with children and the effects of day care on their daily experiences and development, then spiraling outward to encompass providers, parents and families, the community, and ultimately the entire child care and social service delivery systems. Children's Experiences in Day Care Day care is a physical and social environment in which children spend a substantial portion of their waking
124 hours. Therefore, it does not seem unreasonable to begin thinking about the effects of day care on children by asking what kinds of experiences various day care settings provide and whether these experiences are intrinsically good or bad. Curiously, this approach has been little used in day care studies in developmental psychology and evaluation research (with exceptions to be described shortly), although it is used routinely by parents in deciding whether and where to place their children in care. Researchers have tended to view the experiences of children as means, not ends--the ends being various forms of developmental change, such as enhanced cognitive or social skills. This view is also implicit in many discussions of day care policy. For example, part of the justification for federal support of care for children from low-income families lies in the presumed educational and socializing effects of the preschool group experience. Individual developmental change is perfectly appropri- ate to use as one standard in assessing the benefits of day care for children; however, there are serious technical and philosophical reasons for objecting to exclusive use of this standard. Such a narrow focus of evaluation places the burden of proof of merit on measures and modes of analysis that, given the current state of the art, inherently limit the ability of a program to demonstrate its worth. Moreover, exclusive focus on individual change ignores the goals and practices of many day care programs and providers, and it implies value judgments that are open to question, particularly with respect to the justification for public subsidy for the care of the children of the poor. The emphasis on developmental outcome measures reflects an assumption, ubiquitous but often tacit in debates among policy makers and researchers, that early childhood pro- grams are justified primarily by future gains to the child and/or to society, such as enhanced educational achieve- ment, enhanced employability and income, and reduced delinquency and dependence on welfare. Rarely in such discussions are programs justified by immediate benefits to the child--the child's opportunity to spend several years of his or her life in a good environment, both in day care itself and in the family. Public expenditures on children are viewed as investments in the future, not as purchases of goods and services to be consumed in the present. This sort of thinking seems so natural that it is hardly ever questioned in some circles, but it is by no means the only way to think about programs for children.
125 Historically, such single-minded preoccupation with individual development would have seemed aberrant. In the early 1900s child care programs, such as the day nurseries of settlement houses in low-income areas in cities serving minority populations (considered then, too, to be "culturally deprived"), did not measure success in terms of psychological growth. Instead, the number of baths and delousings per child per week, the number of shoes cleaned, the number of garments disinfected, and the number of slum babies saved from filth and degradation were the outcomes valued by administrators, philanthro- pists, and policy makers. The measures were accounts of these direct services in columns in ledgers capturing the theme of the Progressive Era's discovery of the management and professionalization of human services. To cite a more contemporary contrast, no one would dream of justifying public programs for the elderly primarily in terms of their future contribution to society. Advocates for such programs base their claims on the humanitarian premise that society has an obligation to provide a decent life _ . _ . . . _ _ . . tor those tor whom it has assumed some degree ot financial responsibility. Surely, similar reasoning could be applied to children, particularly the children of the poor. We do not wish to overstate the case or pose a false dichotomy. Obviously, parents and providers care about children's futures, and money spent on young children may well be a wise long-term social investment. Moreover, there is evidence that various indices of the quality of the day care environment are linked to indices of individ- ual growth. For example, in the New York Infant Day Care Study (Golden et al., 1978), two-year-olds who experienced a high degree of cognitive stimulation in the day care environment performed better than other children on measures of language comprehension and social competence at age three. In the National Day Care Study (Ruopp et al., 1979), preschool children in classrooms with high levels of cooperation and engagement in activities involving reflection and innovation on the part of the child also performed well on standardized tests of cognitive development. Nevertheless, potential future gains do not obviate the need for evaluations to give equal weight to the present--to the child's immediate needs and experience. To address issues of the "quality of lifer within day care itself requires at least three classes of measures: (1) measures of the quality of the physical environment; _ _ , ~
126 (2) measures of the quantity and quality of ~supplement- ary" services to children, such as nutrition and health care (including screening and diagnostic services); and, most importantly, (3) measures of the quality of inter- action among children and between adults and children in the day care setting. (Measures of the quality of life in the family are discussed later.) The Physical Environment Evaluations of child care facilities often include descriptions of the physical environment. Objective descriptors abound: square footage of indoor and outdoor space, inventories of equipment and materials, counts and checklists of health- and safety-related features such as numbers of toilets and fire exits, protection around electrical outlets, sanitary features of kitchen facil- ities, etc. Many of these physical characteristics are covered in state and local licensing codes and health, fire, and safety regulations. Thus, minimal character- istics necessary for safety and sanitation are fairly well established (by common sense and the practical experience of providers and relevant monitoring agencies). To go beyond the basics to subtler descriptors of environmental quality is more difficult. Crowding or its absence, lighting, color, noise level, the accessibility of materials as opposed to their sheer physical presence, the layout of space as opposed to its sheer size, the presence or absence of private places, and countless other physical characteristics of child care settings can potentially affect children's behavior within those settings. Two recent review papers on the effects of the physical environment in day care (Prescott and David, 1976; Krovant et al., 1976) cite a number of relevant studies on the behavioral effects of square footage available per child and a few studies of acoustics, play equipment, and other features of the environment. However, both papers are striking in the contrast they present between the poverty of systematic empirical knowledge and the wealth of opinion about the impact of the environment on children. A few studies attempt to define global features of the environment, such as "softness" or "inclusion-seclusion potential," or to examine the physical environment as part of a broader "closed vs. open" atmosphere (e.g., Prescott et al., 1967, 1972, 1975; Prescott and Milich,
127 1974). However, most studies focus on one objective feature of the environment--usually square footage per child--and attempt to relate it to one or more behavioral variables, especially the amount of social interaction (positive and negative) and the frequency of aggressive or destructive acts (see Prescott and David, 1976, for references). Existing research thus provides little or no basis for understanding how features of the environment (e.g., the amount of space and its arrangement) interact. In short, the physical environment is the subject of detailed prescriptions by regulatory agencies and advocacy groups, but these prescriptions are based at best on the practical experience of providers and at worst on specula- tion. There is a lack, not of potential measures, but of well-founded knowledge about which measures to use and how to combine specific indicators so as to form more general and meaningful variables characterizing the physical environment. Supplementary Services Day care facilities, especially centers serving children from low-income families, often provide "supplementary" services such as nutrition, health, and dental care. Such services are supplementary only in the sense that they go beyond supervision of children during the working day. For children who may not receive them elsewhere, these services may be fundamental to the child's well-being. (Day care facilities also frequently offer services such as parent counseling, which can potentially affect parent-child relations and family functioning; these services are discussed later.) Health and related services pose problems of measure- ment that are analogous to those posed by the physical environment. For example, it is a fairly straightforward (though perhaps burdensome) matter to keep records of screenings and immunizations. It is not at all straight- forward, however, to determine whether these services actually improve children's health. (See Levine and Palfrey in this volume for a thorough discussion of the difficulties involved.) In addition, because these services are not universal in day care and because different facilities arrange for them in different ways (some by direct provision and some by referrals), service delivery measures themselves can be ambiguous and difficult to standardize. For example,
128 it is obviously inappropriate to equate "referrals" that require working parents to take the initiative in securing services for their children with referrals in which the day care facility center makes contact with the providing agency, arranges transportation, and does any necessary follow-up. Without careful attention to the differences in the ways in which services are made ~available" to children, measures are likely to be perfunctory and unrevealing. Interaction with Care Givers and Peers The study of children's behavior in group settings, including but not limited to day care, has an extensive history in developmental psychology. Until recently, most studies were theoretically motivated, designed to identify consistent dimensions of behavior and sometimes to relate them to characteristics of the setting or the supervising adult (for example, see Baumrind and Black, 1967, Becker and Krug, 1964, Kohn and Rosman, 1972, Peterson, 1961, Schaefer, 1961). More recent studies of children's behavior in natural settings have examined the effects of day care, usually in comparison to home rearing--although comparisons of different day care settings have gained increasing attention in the past few years. Useful reviews that cover these studies, among others, are provided by Belsky and Steinberg (1978), Belsky et al. (1981), Etaugh (1980), Hoffman (1974), Meyer (1976), and Riccicuti (1976). Many of the day-care-related studies have used frequency counts or ratings of behavior to draw inferences about traits of individual children, which might potentially be influenced by the day care environ- ment; thus, the proper place to consider the outcome measures used in these studies is the section below that deals with measures of developmental change. Other observation studies, however, have provided data that can be interpreted as characterizing the social climate of the environment to which the child is exposed; measures used in the latter studies are discussed below. (We recognize that the distinction between studies of the social environment and studies of social development is often difficult to draw and that many studies of the environment are motivated by its personal effects on development. Nevertheless, we maintain the distinction because we believe it is important philosophically.)
129 A number of studies have examined the behavior of care givers and children of various ages in different settings--centers, family day care, and the child's own home. There also exist comparative studies of day care centers with different configurations of staff and children (e.g., staff-child ratios, age mixes, group sizes, levels of staff training) and different physical resources (e.g., space and equipment). One recent study has compared different types of family day care-- unlicensed, licensed, and "sponsored" (the latter term referring to homes that are part of larger child care systems). In almost all of these studies, measures are ratings or frequency counts based on natural observations. A partial list of variables examined in these studies includes the following: 1. Care giver nurturance, responsiveness, and care giver-child contact (see Cochran, 1977; Heinicke et al., 1973; Rubenstein et al., 1977; Ruopp et al., 1979). 2. Care giver restrictiveness, "management" and behavior, and emphasis on rules and routines (see Cochran, 1977; Fosburg and Hawkins, 1981; Prescott et al., 1967; Ruopp et al., 1979; Stallings and Porter, 1980). 3. Cognitive and verbal stimulation and teaching on the part of the care giver (see Carew, 1979; Cochran, 1977; Fiene, 1973; Fosburg and Hawkins, 1981; Hawkins et al., 1979). 4. Provision of opportunities for children to initiate activities (see Heinicke et al., 1973). 5. Involvement in activities on the part of children (see Golden et al., 1978; Prescott, 1973; Ruopp et al., 1979). 6. Conflict, aggressiveness, and destructiveness on the part of children (see Carew, 1979; Butt and Vaizey, 1966; Loo, 1972; Shapiro, 1975). 7. Isolation, inactivity, and aimless wandering by children (see Carew, 1979; Fosburg and Hawkins, 1981; Rubenstein and Howe s, 1979; Ruopp et al., 1979; Shapiro, 1975). Be Overt distress (crying) among infants (e.g., Rubenstein and Howe s, 1979; Ruopp et al., 1979). These studies, varying widely in scope and emphasis, suggest that naturalistic observations of children and care givers can potentially be used to capture important elements of quality in child care and to discriminate among different types of day care environments. However,
130 observational studies raise important practical and methodological issues that have received insufficient attention from many researchers using observational techniques in day care settings. Some studies have used time-sampled counts of relatively fine-grained, objec- tively defined behaviors. This approach provides a record that is both detailed and faithful to the temporal prevalence of events. However, it requires well-trained observers and is expensive and time-consuming. And it is subject to the criticism that there is no necessary relationship between the psychological significance of an event and its frequency or duration. Studies based on global ratings of the classroom environment by observers in effect filter the flow of events through the eye--that is, the value system and the implicit or explicit psycho- logical theory--of the beholder. This approach thus has the potential advantage of weighting events according to their significance and the potential disadvantage of greater observer bias than the event-record approach. There has been little or no comparative study of pictures of the same day care settings painted by the two different methods. (In one case, in which the same children were studied by different researchers using the two different methods, rather different pictures, especially of the children's "aggressiveness," emerged (compare Lay and Meyer, 1973, with Schwarz et al., 1974). Moreover, while observer bias and interobserver agreement have received attention from researchers, other serious sources of bias have received much less attention. It has long been known that instability of behavior is a threat to the reliability of behavioral measures (Medley and Mitzel, 1963). Mathematical techniques exist for assessing the distortions introduced by observers, fluctuation of behavior over time, and other sources (Cronbach et al., 1972). However, only a few recent studies have put them to use. Thus, if observational measures are to fulfill their promise, a great deal must be learned about the properties of alternative recording strategies and possible trade-offs between expense and objectivity. Developmental Change As indicated earlier, most research on the effects of day care arising within the disciplines of developmental psychology and early childhood education has focused on changes in children's social and cognitive development.
131 For expository convenience we will distinguish between developmental and educational lines of research on day care, although the boundary between the two is fuzzy. The former is concerned primarily with children's socio- emotional development and interpersonal skills. The latter is concerned primarily with cognitive ability and achievement and, to a lesser extent, with practical, self-care skills. The two lines of research correspond to two different views of day care: as a socializing environment and as a mini-school. Though most contempo- rary day care, at least in centers, incorporates elements of both views and though some research projects attempt to assess outcomes in both domains, the distinction is worth maintaining because the two emphases have different implications for the choice of outcome measures. Until recently, research on developmental change in day care focused on comparisons between children reared for substantial periods in group care environments and children reared at home or children reared in group environments for considerably shorter periods. This research was intended to measure the outcomes of group care per se, not of particular kinds of group care, although it in fact concentrates on a relatively narrow and not particularly typical range of group care environ- ments, as several reviewers (Belsky and Steinberg, 1978; Etaugh, 1980) have noted. In the past few years, researchers have begun to compare different types of day care, such as in-home care, center care, and family day care (e.g., Golden et al., 1978; Clarke-Stewart, 1979, 1980). __ types of care; for example, Ruopp et al. variations within center care that were associated with different staffing and grouping patterns. Research on children in group settings began as early as the 1930s, when the first studies of the effects of nursery school entered the literature of child develop- ment. Some of these studies provided our earliest demonstrations that preschool education can boost the scores of disadvantaged children on standardized tests of ability and achievement. For example, one study (Barrett and Koch, 1930) found that orphanage children gained 20 points on the Merrill-Palmer Test after six months' exposure to nursery school. However, most of the early nursery schools were not hothouses for cognitive develop- ment. Primarily serving middle- and upper-class children, they were devoted to developing the "whole child." Play, arts and crafts, and a general emphasis on human relations Others have examined variations within particular (1979) examined
132 were prominent in their curricula. Correspondingly, early studies generally looked for social benefits in the form of increased participation, cooperation, impulse control, and communicative skills on the part of preschool children (e.g., Parten, 1932). While these studies succeeded in documenting interesting aspects of children's growth and behavior in group settings, they were on the whole method- ologically naive by modern standards and were inconclusive in their attempts to demonstrate that nursery school conveys special benefits in contrast to home rearing. Another line of research, beginning a decade or so after the nursery school studies and extending into the 1960s, sheds much-needed light on the dark underside of child care: the care of infants and young children in institutions. Rene Spitz's influential essays documented appalling rates of apathy and morbidity among infants in institutions where care was inadequate and inconsistent (Spitz, 1945). Subsequent studies found retardation of the onset of vocalization, motor skills such as crawling and creeping, visually guided reaching and grasping, smiling, and other forms of responsiveness to the physical and social environment (e.g., Provence and Lipton, 1962; White, 1969, Dennis, 1941; Dennis and Najarian, 1957; Paraskevopoulos and Hunt, 1971). (That these early deficits would have enduring consequences was often implicitly assumed and therefore rarely investigated directly). These observations were interpreted as evidence of the devastating effects of early maternal deprivation (Bowlby, 1969), though later work called this interpretation into question, suggesting that general physical and social stimulation--not specifically maternal interaction--is what the worst institutional settings lack (e.g., Gouin-Decarie, 1965). Modern day care in the United States may bear consider able resemblance to nursery school (in fact the two may be indistinguishable, save for the length of the session), but only in a minority of cases does it bear much resem- blance to the bleak, underfunded, understaffed institu- tions observed by Spitz and others. Nevertheless, studies of institutionalized children, like studies of nursery school, continue to exert an influence on our thinking (e.g., Fraiberg, 1977). Public debate about the merits and perils of day care as an environment for children continues to center on issues raised by these studies. Critics allege that day care weakens the bond between mother and child, robbing the child of the security and emotional attachment necessary for healthy development. -
133 Such criticisms mean that demonstrations in day care, more than demonstration programs in other child-related areas (e.g., in early education or health care), must not only prove their positive benefits but must also prove that they do not actually harm children. Consequently, the outcomes measured in early studies of nursery sabools and of institutional care remain relevant today, as do some of the actual measures used in those early studies. Many contemporary studies of the impact of day care can trace their intellectual descent from the early studies of institutionalization through the theoretical formulations of Bowlby (1969), with a link by marriage to experimental techniques for assessing the degree and quality of mother- child bonding. Most prominent among these are variants on the Ainsworth "strange situation n techniques (Ainsworth and Wittig, 1969), in which the child is separated from its mother and introduced to a strange adult. The amount of exploratory behavior shown in the mother's presence; the amount of distress, approach, and avoidance shown in response to the stranger; and the amount of proximity- seeking shown on the mother's return are used as indices of the security of the child's attachment to the mother. By comparing the behavior of home-reared children with that of children who have spent a substantial proportion of their early years in day care, studies using the "strange situation" and adaptations thereof have directly addressed the issue of whether exposure to day care weakens the mother-child bond. Thoughtful, comprehensive discussions of these studies appear in Belsky and Steinberg (1978) and Belsky et al. (1981). These reviewers point out that the "strange .. . . .. . ~ . ~ ~ ~ _ ~ ~ _~ , ~ _~_ w~ I. , situation" is designee tor use wltn collared oecw=en 1 ~ and 18 months of age and that it entails a complex coding system (Ainsworth et al. 1978). Used appropriately it is reliable, valid, and predictive of later social develop- ment (Sroufe, 1979). However, many day care researchers have used invalidated variants of the "strange situation," often with children two years old or older. For example, several studies have examined the child's reactions to separation and reunion with the mother during dropoff and pickup at day care, recording frequencies of distress, clinging, avoidance of the mother, exploratory behavior, and the like. While separation and reunion may be important to study, they cannot be assumed to measure attachment in the same way as does the "strange situation"--particularly when children are considerably older than one year, when overt attachment is most salient.
134 Subject to the above caveat, Belsky and his colleagues report that most studies find no deleterious effects of day care on attachment. Reports of such effects (e.g., Blehar, 1974; Ricciuti, 1974; Cochran, 1977) are difficult to replicate or open to alternative explanations. How- ever, one recent study using the "strange situation" in the manner prescribed by Ainsworth (Vaughn et al., 1980) suggests that there may be damage to the attachment relationship for certain highly vulnerable infants placed in day care before one year of age. The nursery school studies also have descendants, though the line of descent is less clear and direct than is the case for studies of attachment. Research on child care turned toward heavy emphasis on cognitive skills during the 1960s. Influential basic research studies and syntheses pointed to the malleability of intelligence (Hunt, 1961; Bloom, 1964). Numerous preschool education programs sprang up, many of them affiliated with univer- sities, most directed at compensating for presumed environmental deficits experienced by the children of the poor. Positive results from many of these programs, based primarily on improvements in children's scores on standardized tests of ability and achievement, soon appeared in the literature of developmental psychology and early childhood education (see Weikart, in this volume.) And, of course, Head Start was established, soon to be followed by the widely publicized Westinghouse-Ohio evaluation, based almost exclusively on standardized tests (Westinghouse Learning Corporation and Ohio University, 1969). Research on day care, as opposed to compensatory education, began to appear somewhat later, as new policy issues came to the fore under pressure from the women's movement, advocacy groups, and labor market trends. What is striking is that many studies on the impact of group care continued to include standardized measures of cognitive ability and achievement, such as the Bayley Scales, the Stanford-Binet Test, the Preschool Inventory Test, and the Peabody tests, in their outcome measurement batteries. The general finding of this large body of work is that day care has no effect, positive or negative, on the scores of children from relatively advantaged backgrounds, and for children from low-income families day care seems to forestall the decline in test scores that usually occurs with age (Belsky and Steinberg, 1978). Fewer studies have examined indices of cognitive and linguistic development other than general scores on standardized tests. The list of more specific abilities
135 studied, however, is long and rather impressive-- problem-solving, abstraction, and playfulness, measured through ratings of natural behavior (Macrae and Herbert-Jackson, 1976; Schwarz et al. 1974); concept formation, memory and recognition vocabulary, measured with standardized tests (Kagan et al., 1976); age of onset of speech and complexity and maturity of speech patterns (Fiene, 1973; Cochran, 1977); and for infants a variety of motor skills. The various studies do not lend them- selves to easy summary; no overwhelming positive or negative effects of day care have emerged. What is more important here is the sheer variety of outcomes and outcome measures, and the fact that no consensus has emerged as to what should be measured and how. In the domain of social development, the picture is even more complex. A wide variety of individual traits and social skills have been assessed by means of global ratings or systematic frequency counts based on naturally occurring behaviors and by means of tests, administered both verbally and as structured problem situations to which the child must respond. Variables assessed on the basis of natural observations include dependency, nurturance, sociability toward peers, attitudes toward the care giver, cooperation with peers and adults, hostility, aggressiveness, general activity level, assertiveness, conformity, and exploratory behavior; tests and structured situations have been used to assess curiosity, the capacity to adopt the perspective of others (social role-taking), the capacity to give assistance, relationship to parents, sex typing, impulse control, and cooperation. Examples of studies employing natural and/or structured measures of social development include Clarke-Stewart (1979, 1980), Caldwell et al. (1970), Schwarz et al. (1974), Lay and Meyer (1973), Macrae and Herbert-Jackson (1976), Moore (1975), Doyle (1975), and Lippman and Grote (1974). Again, no simple summary of findings is possible; what is important for our purposes is the wide range of outcomes for which measurement has been attempted and the lack of convergence on a particular set of outcomes or measures. How good are the various development and educational measures that have been discussed, and how useful for evaluating programs and shaping policy? To answer these questions measure by measure would require a long dissertation indeed, but a number of general comments can be made.
136 , _ , Contrary to what one occasionally hears, there is no lack of candidate outcome measures for a wide variety of cognitive and social skills. However, as a reader of an early draft of this paper put it, there is good reason to question whether any of the candidates merit election. ~c is striking tnat a relatively small set of (inter- correlated) measures of general cognitive skills are used in study after study, while anarchy reigns in the measure- ment of social development and more differentiated cognitive skills. The attraction of standardized cognitive measures such as IQ appears to derive from their relatively high reliabilities (in the traditional psychometric sense) and their predictive validity against a criterion of success in school as well as from the historical influences of Head Start and its precursors. However, despite their widespread use, there is equally widespread dissatisfac- tion with those measures, even among many who use them. There are many reasons for dissatisfaction: Poor and minority children score less well on the tests than other children, leading to charges of cultural bias. The tests are generally designed to be insensitive to specific learning experiences, making them questionable as outcome measures for intervention programs of any kind. The most widely used tests do not attempt to measure creativity, persistence, flexibility, and resourcefulness in attacking problems or a host of other aspects of cognitive skill and style that may ultimately indicate much about a child's potential as a learner or future ability to use what is learned. Unfortunately, instruments designed to measure the latter aspects of cognitive development, though influential in basic research, have on the whole not demonstrated the reliabilities and predictive validities of the general ability measures, nor have they achieved public acceptance and widespread use in evalua- tion as measures of intellectual potential. There is a serious question in the psychometric literature an to how measurable these traits are and how separable from general intellectual ability. Similarly in the area of social skills, a bewildering variety of potential measures exists (see compendia by Johnson and Bommarito, 1971; Walker, 1973). Used primarily by highly trained researchers in academic settings, these measures have nevertheless not been impressive on psychometric grounds, especially when used by researchers other than their developers and especially when used in field settings. Although a few brave souls
137 have stepped forward to suggest a definitive instrument battery for measuring "social competence" as an outcome of early childhood programs (Zigler and Trickett, 1978), no single instrument, let alone battery, has commanded widespread acceptance. It is not for lack of effort in the basic research community that measures of cognitive style and socio- emotional development lag behind standardized tests of general cognitive and linguistic skill on psychometric grounds. When years of effort fail to produce a desired result, it is worth asking whether the enterprise is misconceived. Trait measures are inherently individual- istic. They focus on characteristics of the child, not on the social matrix within which those characteristics are nurtured. However, there is massive evidence through- out the literature of child development (summarized most pointedly by Bronfenbrenner, 1979) that situational and cultural contexts profoundly affect young children's behavior. Thus, while the search for better trait measures should and will continue, perhaps researchers should also begin to devote equal effort to finding better ways of characterizing child/environment systems. Many day care programs have begun to try to produce and sustain change in children by changing their home environ- ments through services to families. The individualistic focus of trait measures diverts attention from the impact of day care on family functioning and family-community relations. Such impacts are not only "goods" in them- selves but may redound to the benefit of the child. For example, if day care relieves economic pressure and consequent interpersonal stress within the family, the child can potentially benefit from the improved home environment. Similarly, when day care acts as a vehicle for connecting families with community services (e.g., health care or food programs), the child is again likely to be a beneficiary. In the same vein, while there is evidence for longitudinal stability of some social traits, such as aggressiveness and dependency, from the elementary school years on, evidence for stability of all but a few traits from the infant, toddler, and preschool years is slight (Kagan and Moss, 1962). There is evidence for stability of certain broad features of temperament from infancy on, chiefly in the work of Chess, Thomas, and their colleagues (e.g., Thomas et al., 1969). However, it is questionable whether temperamental differences are very susceptible to environment and whether temperament, as distinguished
138 from social behavior and social skills, is what day care providers try to influence. It is also difficult to know how to search for longitudinal stability in social traits. We cannot simply assume that a given pattern of behavior in adults or older children derives developmentally from a superficially similar pattern in younger children. We need a theoretical framework to tell us which behavior patterns should be associated over time. Thus, to demand that any program for young children prove its worth by demonstrating that it produces enduring change in social behavior is to demand a great deal. We have much to learn about the time course of effects of early childhood programs. The typical finding for measures of cognitive gain is an appreciable effect by the end of the program, gradually diminishing as the child progresses in school. On the other hand, there is exciting recent evidence of long-term "sleepers effects, which manifest themselves in the late elementary school years or beyond. Recent work suggests that these long- term effects may be due partly to socialization--to changes in work habits, motivation, and the like (see Weikart, in this volume.) There is also some evidence for sleeper effects in social development, though not for effects specifically attributable to intervention programs (Kagan and Moss, 1962). Until we understand the temporal structure of intervention effects in the social domain, evaluations of the effects of day care and intervention programs in this domain will remain hit-or-miss. Finally, almost without exception the variables, measures, and study designs arising from developmental and educational research reflect scant attention to the information needs of policy makers. Much of the research is, properly, motivated by theoretical concerns. Even where the concerns are practical, they tend to be narrow. Virtually all of the measures used within developmental psychology and early childhood education address pieces of a single concern of policy makers, i.e., the quality of programs, in particular their benefits for children. This focus is obviously appropriate to the fields in question--but if the policy maker's broad concerns for access, equity, and efficiency as well as quality are to be addressed, disciplinary boundaries will have to be broken and new, integrative efforts at measurement must be undertaken. Some tentative steps already taken in this direction are discussed in the later section on the effects of alternative day care policies on the child care delivery system.
139 Outcomes for Care Givers We do not often think of outcomes for providers of a service°-and outcomes for care givers are, admittedly, secondary relative to outcomes for children and families. However, day care is somewhat unusual in this regard. Such outcomes are of interest to federal policy makers, as evidenced by the facts that (1) a provision for employ- ment of low-income mothers was formerly included in federal day care purchasing standards and (2) the govern ment has supported a credentialing organization, the Child Development Associate (CDA) consortium, which sets standards and implements procedures by which day care workers can receive a formal, transferable certificate of competence for on-the-job experience and training. Potential outcomes for care givers thus include changes in income, working conditions, job satisfaction, and professional growth as a consequence of experience and training. To our knowledge the effects of the CDA program (on care givers or on children) have not been evaluated. More generally, while the low wages of care givers have been documented and their behavior has been studied insofar as it affects children, attitudes and behaviors have not been assessed from the perspective of care givers themselves to any significant extent. A great deal remains to be learned about what care givers like and do not like about their jobs, what kinds of training and other assistance they find useful, what causes "burnout," and what kinds of care-giving arrangements are best for them. Outcomes for Families and Communities An earlier section suggested several ways in which day - care programs might benefit parents and families, with concomitant benefits for the child. This section takes up the theme of effects on parents, families, and the wider community directly. In this domain, unlike that of developmental effects, there is not a wide range of measures from previous research to consider. Basically there have been four types of relevant research: (1) studies of effects of day care on parent-child interaction and family functioning; (2) studies of parental prefer- ences and satisfaction; (3) studies of parent-provider relationships; and, at a different level, (4) studies of the effects of the availability of child care on parental
140 employment and income. Only the fourth type of research, on economic impact, has been at all extensive. Parent-Child Interaction and Family Functioning A few studies have examined effects of day care on mother-child interaction in laboratory settings (Ramey and Mills, 1975; Falender and Heber, 1976). Some results suggest that low-income mothers of children in day care interact more and give more positive and less negative feedback in a training task than mothers of home-reared children, but these results are subject to various inter- pretations, and later work has found no differences in parent-child interaction (Ramey et al., 1979; Farran and Ramey, 1980). Several studies have used Caldwell's Index of Home Stimulation (a set of descriptors of the home that have been shown to correlate with intellectual achieve- ment) with mixed results as to whether day care produces any differences, positive or negative (Fowler and Khan, 1974, 1975; Ramey and Mills, 1975). Other measures used in studies of family impact have been Schaeffer and Aaronson's Maternal Inventory, designed to measure the mother's attitude toward and interest in her child (Fowler and Kahn, 1974, 1975); parents' self-reports of knowledge about childrearing (Steinberg and Green, 1979); parents' attitudes toward children's rights (Ramey et al., 1981); mothers' self-reports of marital satisfaction (Meyers, 1973; Steinberg and Green, 1978); and family structure and functioning as measured by the St. Paul's Profile (Golden et al., 1978). Several comments can be made about those studies as a group. First, and most important, there are few such studies (and several of these cited are either unpublished or still in progress). The potential effects of day care on the child's experience in the family and on the climate, cohesiveness, and strength of the family itself are neglected areas of research that deserve much greater attention. We lack a systematic theoretical treatment of the aspects of family functioning that might be affected by day care. Second, the above listing does not differentiate day care programs that have active parent counseling or education components from those that do not. Presumably, effects on mother-child interaction ought to be greatest in programs that teach mothers how to care for their children; similarly, marital satisfaction and family
141 functioning ought to profit most from programs with family counseling components. There is a clear need to tailor outcome measures to the goals and practices of programs. Also, there is a substantial literature on the effects of parent education and counseling (e.g., Brim, 1959; Good son and Hess, 1978). Obviously, studies of day care programs with such components would do well to borrow from that literature; the few existing studies have begun to do so. Again, lacking systematic thought on day care as it affects the family, we are in need of more work to determine under what circumstances such supplementary services are useful and supportive to families and under what circumstances they are superfluous or intrusive. Third, despite evidence of a growing concern with day care's effects on the family, a potential larger step has not been taken. There do not yet appear to be studies that relate the availability of child care at the community level to social indicators bearing on the health of families within communities--e.g., rates of divorce and desertion, child abuse, out-of-home placements of children, etc. Parent Preferences and Parent Satisfaction A number of studies have used interviews and question- naires to assess parent preferences for different types of day care arrangements as well as parents' satisfaction with the child care facilities they use (e.g., Hill, 1977, 1978; Steinberg and Green, 1979; Rowe et al., 1972; Fosburg and Hawkins, 1981). In general, these studies show that parents are reasonably well satisfied with the care they use and that they prefer care close to home, usually family day care or in-home care for infants and toddlers and center care for preschoolers, because of its presumed educational benefits. However, their options in choosing particular day care facilities are set largely on the basis of cost, location, and schedule. Some ambiguities of interpretation surrounding the preference data were alluded to earlier. Concerning satisfaction data, existing results are largely tainted by the fact that most such information has been collected from parents whose children are currently in day care; presumably those who were seriously dissatisfied would have withdrawn their children from day care and themselves from the researchers' samples. In general, this is an area of investigation that seems ripe for imaginative
142 instrumentation and research design, aimed at providing qualitative insight into the reasons for parental preferences and satisfaction or dissatisfaction. Existing preference data tell us little more than the facts the market itself reveals. Parent-Care Giver Relations Relations between parents and the people that serve their children are a topic of intense political interest. Federal day care purchasing standards require that parents be represented on the boards of centers serving large numbers of subsidized children and that parent participa- tion in general be encouraged. However, as a reviewer of the literature on parent involvement--and a strong advocate for it--notes, "the assumption that some form of parent participation would increase parental proximity to surrogate child care establishments, and that increased proximity would improve the quality of child care or ease the child's transition from home to institution, has never been directly assessed" (rein, 1976). In the years since that review, several projects have examined parent-care giver relations. One study focused on the frequency and nature of contacts and on center policies that facilitate or discourage contact (Powell, 1977, 1978). Measures were based on self-reports in interviews and questionnaires. Another study focused on the conflicting expectations of both black and white parents and staff in educationally oriented day care centers operated by the California public schools (Joffe, 1977). A third study examined parents' preferences for different kinds of participation and found that they preferred active roles in the classroom (e.g., as volunteer aides) as well as participation in educational programs and social events (Travers et al., 1977; Layzer, 1980). Parents in federally funded centers were far more likely than those in parent-fee centers to want to participate in center governance through membership on governing boards. A great deal remains to be learned about relations between parents and providers, especially about how these relations affect children. Continuities and discontinuities related to social class and ethnic culture are important to investigate in this regard.
143 Effects on Employment and Income The impact of the availability and/or subsidy of day care on families' employment patterns and wages is an outcome important to policy makers, because federal and state day care policies have often been determined in relation to broader economic policies and, more specif- ically, in relation to welfare-reform objectives. Several studies have examined the effect of child care on women's labor force participation (Conly, 1975; Ditmore and Prosser, 1973; Jusenius and Shortlidge, 1975; Kurz et al., 1975; Shaw, 1974). The principal conclusion drawn from these studies and related data has been that the provision of day care does not significantly affect the labor force participation of women. The principal deter- minants are rather the availability of suitable jobs and the existence of other barriers to employment of low- income women. The studies suggest that readily accessible child care services induce no more than 10 percent of nonworking, low-income mothers to enter the work force; the figure is higher if families having children under three are excluded. Although the availability of day care may not induce many women to enter the labor force, there is some evidence that those mothers who are enabled to work by the availability of day care enjoy economic benefits, such as higher income (Peters, 1973), more education (tally, 1973), Ramey et al., 1981), and enhanced work skills (Ramey et al., 1981). Not all studies find such benefits, however. The New York Infant Care Study (Golden et al., 1978) found few effects on income. Perhaps more important for policy purposes, the cost- effectiveness of day care seems particularly questionable for parents at the low end of the income scale. Several studies have shown that it would be less costly for the government to provide income maintenance than to provide day care in "developmental," center-based programs while requiring low-income mothers to work (Rivlin, 1973; Woolsey, 1977). In our view the studies cited above highlight the needs, cited at the beginning of this paper, to frame the goals and value assumptions of outcome measurement clearly and to draw policy conclusions with great care, consider- ing not only findings in a single outcome domain but taking account of other domains as well. Consider first the finding concerning the relatively modest impact of day care on women's labor force participation. It is not
144 at all clear that increasing such participation is a generally agreed upon policy goal. (From the point of view of those who fear ill effects of day care on children, it is clearly to be avoided.) If the goal of day care policy is to ensure that children will be adequately cared for when their mothers choose to work and are able to find employment, it is irrelevant whether the availability of day care affects women's propensity to enter the labor force. Only if day care is seen primarily as a device for reducing the welfare rolls is its modest effect on labor force participation important And even then, the 10 percent increase need not be viewed as small; in the absence of prior data, it merely estab- lishes an empirical benchmark--a realistic expectation-- against which alternative policies can be evaluated. Consider next the studies that suggest that income maintenance may be more cost-effective than day care. Such studies weigh immediate economic benefits (employ- ment, income, reductions in public expenditure, tax contributions) against the costs of care. However, consideration of nonmonetary costs and benefits and/or long-term benefits--psychological effects on the child and parent, possible effects on family conflict and dissolution, possible long-term improvements in the mother's earning power, and possible increases in the likeliood that the child will become economically self-sufficient--might shift the balance of interpreta- tion. Also, cost-benefit studies of subsidized day care have for the most part been based on the assumption that developmental day care centers will be the orincinal delivery mechanism. strafe that informal day care arrangements such as neighborhood family day care are currently used by working mothers with far greater frequency than are formal centers. The fundamental question regarding what kind of day care should be eligible for subsidy must be examined as a policy issue in connection with any thorough assess- ment of the cost and benefits of providing day care to low-income families as a concomitant to employment. The above remarks relate to a broader issue: the need to improve the theoretical framework for linking day care policies to family economic functioning. While some research exists attempting to model the complex inter- relationships involved in these areas, further development is necessary. Without such development, predictions and interpretations of the effects of day care on employment behavior and economic outcomes are frequently based on , - , ~ However, utilization studies demon -
145 assumptions that do not reflect the range of alternative policies available to decision makers or the complexities of actual relationships in these areas. Conclusion In conclusion it is important to stress that whole areas of investigation are missing from the list of four topics considered above. To the degree that day care provides parents with counseling and advocacy, intended to make them more effective in dealing with social service agencies, schools, and other community institutions, potential effects on relations between parents and such institutions are critical outcomes to examine. These outcomes have not been investigated in connection with day care, though they have been examined in connection with comprehensive family service programs attached to Head Start. The difficult issues that arise in assessing these programs are discussed by Hewett and Deloria, in this volume. Even more broadly, the effects of a healthy day care market on the economic and social well-being of a community have, to our knowledge, never been studied. To do so would require careful selection of communities and ingenious use of a wide range of social indicators, not only the labor force participation of mothers but also the influx of industry, in-migration and out-migration of families, perhaps even real estate values. This kind of investigation would clearly bear little resemblance to a laboratory study; it would attempt to explicate complex community dynamics, blending qualitative and quantitative information with the ultimate aim of logical coherence and empirical plausibility, rather than statistical generalizability in the strict sense. Effects of Alternative Day Care Policies on the Child Care Service Delivery System This section returns to the issue with which our taxonomy of day care outcomes began: service delivery. We began by considering the services received by the individual child and argued that they could legitimately be considered outcomes of care for some purposes. We end by considering the effects that alternative governmental policies might have on the aggregate profile of day care services in a community, state, or the nation.
146 Policy makers have a number of broad concerns with respect to any social service program: (a) access to services on the part of the target population; (b) equity of service delivery across ethnic groups, residents of different geographic regions, and any other subgroups of the population with a legitimate claim to services; (c) quality and effectiveness of programs in achieving service objectives; and (d) efficiency of delivery--cost effectiveness and administrative burden relative to other means of delivering the same service or to alternative uses of funds. Most of this paper has addressed issues of quality and effectiveness, as has most research on day care. However, issues of access, equity, and efficiency will loom large in the 1980s as increased demands are made on both public and private systems for providing child care. In fact, as argued by Kennedy and McDaniels in this volume, policy makers, particularly at the federal level, often leave determinations of quality and effectiveness to consumers, providers, professionals, and local administrators, on the theory that these issues are best addressed in the context of local needs and interests. As indicated in our introduction, policy makers have but a few primary tools for influencing the profile of care--funding mechanisms, regulations, and administrative guidelines and practices. They need to know how the profile of services and associated costs will be affected by policy choices between emphasis on categorical block- grant funding, as under Title XX, and mechanisms such as income disregard, the tax credit, and vouchers, which give greater scope to parental initiative, or between strict and lenient regulatory policies. They also need to know how public services interact with preexisting private mechanisms of service delivery. For example, does public subsidy of day care reduce dependence on the extended family as a source of child care? (Answering this basic service delivery question is a prerequisite to answering questions about whether day care strengthens or weakens families, destroys or extends natural support mechanisms.) Policy makers also have secondary tools such as the ability to provide training and information and referral services, which can potentially enhance the operation of the system as a whole. They need to know how these services affect the supply of qualified care givers and the supply and demand for different types of care. As suggested earlier, many of the policy maker's concerns are systemic, in that he or she cares not about
147 a program in isolation but about a program in the contex of other programs or service delivery mechanisms--both those that actually exist and those that might be funded These systemic service delivery questions can be addressed through a wide variety of atheoretical indicators--quantitative measures deriving from no particular conceptual framework. Examples include the numbers of clients served, types of services provided, costs of services, capacity and utilization rates, licensure rates for different types of facilities, cost variations across types of facilities, employment characteristics of the day care industry, etc. Researchers have not in general made much use of such measures, and as a consequence research has been less informative to policy than it might be. Precisely because these measures are atheoretical, they tend to fall outside the concerns of the specific disciplines and to seem intellectually uninteresting. What has been missed is the potential intellectual excitement of constructing a picture of an extremely complex phenomenon--the day care market--within which these measures would take on meaning. CONCLUS IONS AND RECOMMENDATIONS First, the outcomes of day care are difficult to measure because different constituencies have different, often conflicting views about what day care is intended to accomplish, what is important to know, and what measures are appropriate to use. This multiplicity of views can lead to confusion, distortion, misuse, and unfulfilled expectations between the public and the policy makers regarding the results of research; hence, it may undermine the credibility of research itself. Further- more, to the extent that the concerns of some constitu- encies are excluded from research altogether r important questions may go unanswered, and the utility of research may be further reduced. Second, day care research in developmental psychology has for the most part sought to compare the long-term developmental effects resulting from group care as opposed to home rearing. Much attention has focused on attachment to the mother and on cognitive skills as measured by standardized tests. While many other aspects of social behavior and development have been examined in individual studies, no consensus on critical variables or
148 measures has emerged. Moreover, little is actually known about the duration of effects, their generality across situations and other important properties. Measures of the child's immediate experience--his or her interactions with care givers and peers--have received insufficient attention as indices of the quality of care. Measures of impact of enhanced nutrition, health care, and other "supplementary" services have been neglected to an even greater degree. Only limited attempts have been made to differentiate the effects of center care from those of family day care; in-home care by persons other than the parent--relatives or paid helpers--has hardly been studied at all. Third, except in the areas of income and employment, effects of day care on parents, families, and communities have received insufficient attention. Little is known about effects on family functioning, family-community relations, or the social climate of the community as a whole. For example, virtually nothing is known about the impact of day care on child abuse, marital stability, out-of-home placements, etc. Fourth, perhaps because day care research has focused more on program than on policy issues, aggregate atheoretical indicators of service delivery have rarely been treated as outcomes. Matters of great concern to policy makers--numbers of children served, availability of services to members of ethnic minority groups, the handicapped, residents of rural as well as urban areas, etc.--have not aroused the intellectual interest of researchers. In only a few studies have questions of cost been juxtaposed with questions of quality. Thus the policy maker's legitimate concerns with access, equity, cost-effectiveness, and efficient management have largely been bypassed in favor of researchers' preoccupation with selected aspects of the quality of care, thwarting the application of simple, useful measures to larger questions and widening the gulf between research and policy. Systemic effects, such as the impact of alternative funding mechanisms and regulations on the availability of different types of care, or the impact of publicly subsidized care on informal service delivery systems, such as the extended family, have not been adequately studied. Finally, a fundamental conclusion not stressed earlier that underlies many of the more specific conclusions documented in the foregoing pages is that existing day care research lacks an overarching theoretical framework.
149 Day care has been studied in the absence of an integrative theory that deals with it as a developmental environment, part of an interlocking web of social institutions, and a social service regulated by government policies. Until these issues are confronted, basic research and technical tinkering to improve outcome measures will probably have limited payoffs. Future efforts in science and technology are more likely to lead to fundamental improvements in outcome measurement if systematic efforts are made to address the enumerated shortcomings of existing approaches, which we enumerate below. Processes must be established to ensure that the views of all relevant constituencies are reflected in choosing outcomes. While not every constituency has a stake in every evaluation, those who do must be represented if a particular evaluation is to be credible. Researchers must learn to ask, systematically and routinely: What outcomes are important? For whom? What measures will communicate most meaningfully? To whom? What actions or decisions will be based on the resulting information? A process for ensuring that these questions are asked and that the answers generated inform the measurement process may avert or reduce the inappropriateness, misuse, lack of credibility, and limited usefulness that have afflicted many past studies of day care. In the domain of developmental effects, dependence on a handful of standardized measures of cognitive skills and an unstandardized grab bag of measures of social behavior must be reduced. To do so will ultimately require nothing less than a broader, deeper, and more systematic concept of development. To fully understand the developmental effects of day care, we need an empirical taxonomy of skills and dispositions and a clear understanding of how cognitive and socioemotional characteristics interrelate. We need to understand which aspects of behavior are situationally controlled and which can be expected to generalize across situations. We need to understand the time course of development of various characteristics, so that we know when to expect short-term but transient effects, when to expect longitudinal stability, and when to expect sleeper effects. All this, of course, is a tall order, tantamount to saying that we need a more mature science of develop- mental psychology. But until we begin to achieve these goals, we must be extremely cautious about evaluating programs in terms of their developmental effects as currently measurable. Thus this recommendation is only
150 in part a predictable plea of the researcher for more research; it is also an exhortation to modesty and, implicitly, a suggestion that basic researchers may profit from attention to applied problems. A serious attempt to understand how different day care environments foster or impede development can only lead to a more thorough understanding of development itself. The immediate experiences of children in day care must be seen as ends in themselves, not merely means--as immediate outcomes, which may also be processes through which long-term developmental outcomes are achieved. Quality of care from the point of view of the child depends not only on lonq-term developmental effects but, · . _ . . . . . . . . . In the rlrst Instance, on the physical ana numan envlron- ment in which the child spends a substantial portion of his or her time. To treat children's experiences as outcomes requires that we develop descriptors of the physical environment that capture what is important for children--elusive qualities such as privacy, accessibil- ity, stimulation, and, of course, safety. More importantly, we need richer yet more practical ways to describe human interaction in the day care setting through refinements and extensions of existing observation systems and rating scales. Once again, an interplay of basic and applied research is indicated. Specific effort must be devoted to inventing measures of children's experiences and development that capture the distinctive advantages and disadvantages of the different environments in which they receive nonparental care, i.e., centers, family day care homes, the homes of relatives, and their own homes. Very different claims are made by proponents of the various types of care; for example, centers are said to provide group experiences that prepare the child for school, while family day care is said to provide a home-like environment and in-home care the security of the home itself. Efforts at measurement have begun to address those claims, but until this work comes to fruition, the claims will remain in the realm of rhetoric, and parental choices will be made and policy debates conducted on essentially ideological grounds. Attention must be paid to the host of existing atheoretical indicators, such as counts of children served and descriptors of services delivered. Such measures have immediate practical utility as management devices and tools of accountability. In addition, they have enormous potential value in assessing the systemic
151 outcomes of day care and (intended and unintended) systemic effects of day care policies, e.g., shifts in the distribution of types ~ ~ ~ ~ and regulatory policies. of care in response to funding What is needed is not so much development of new measures, but development of a better conceptual framework for interpreting existing measures. Many atheoretical indicators have little meaning in themselves; most are influenced by many factors and therefore are ambiguous when viewed in isolation. Yet when interpreted in contexts provided by well-chosen questions, well-conceived studies, and other measures, they can yield invaluable insights. Finally, several of the above recommendations imply what is perhaps the most fundamental recommendation of all: We must somehow grope our way toward an approach to day care that is less fragmented than that afforded by existing disciplines, that will allow us to comprehend day care as a whole rather than as a collection of disconnected elements or relationships. Perhaps the most promising conceptual framework currently available for integrating the disparate levels of description that have so far been applied to day care is the ~ecological" framework proposed by Bronfenbrenner (1979). That framework has already been applied to day care by Bronfenbrenner and others (Belsky et al., 1981). We are aware that pleas for holistic social science can become hollow cliches and that science often proceeds by analysis and dissection. Yet ultimately science puts its intellectual pieces back together in a new and more meaningful way. Perhaps through interdisciplinary borrowing and sustained attention to day care itself, as a developmental environment, an adjunct support for families, a social service, and a policy tool, we can begin to reassemble the jigsaw-puzzle picture left us by existing studies in psychology, sociology, economics, and policy research. REFERENCES Ainsworth, M. D. S., Blehar, M., Waters, E., and Wall, S. (1978) Patterns of Attachment. Hillsdale, N.J.: Lawrence Erlbaum. Ainsworth, M. D. S., and Wittig, B. A. (1969) Attachment and exploratory behavior of one-year-olds in a strange situation. In B. M. Foss, ea., Determinants of Infant Behavior. Volume 4. London: Methuen.
152 Barrett, H. S., and Koch, H. C. (1930) The effect of nursery school training upon the mental test performance of a group of orphanage children. The Pedagogical Seminary and Journal of Genetic Psychology 37:102-122. Baumrind, D., and Black, A. E. (1967) Socialization practices associated with dimensions of competence in preschool boys and girls. Child Development 38:291-327. Becker, W. C., and Krug, R. S. (1964) A circumplex model for social behavior in children. Child Development 35:371-396. Belsky, J., and Steinberg, L. D. (1978) The effects of day care: a critical review Child Development 49:929-949. Belsky, J., Steinberg, L. D., and Walker, A. (1981) The ecology of day care. In M. Lamb, ea., Childrearing in Non-Traditional Families. Hilledale, N.J.: Lawrence Erlbaum. Blehar, M. (1974) Bloom, B. (1964) Bowlby, J (1969) e . Anxious attachment and defensive reactions associated with day care. Child Development 45:683-692. Stability and Change in Human Characteristics. New York: Sons, Inc. Attachment and Loss. Volume New York: Basic Books. John Wiley & Attachment. Brim, O. G. (1959) Education for Child Rearing. New York: Russell Sage Foundation. Bronfenbrenner, U. (1979) The Ecology of Human Development. Cambridge, Mass.: Harvard University Press. Caldwell, B. M., Wright, C. M., Honig, A. S., and Tannenbaum, J. (1970) Infant care and attachment. American Journal of OrthopsychiatrY 40:397-412. Carew, J. (1979) Observation Study of Care Givers and Children in Day Care Homes: Preliminary Results from Home Observations. Paper presented at the biennial meeting of the Society for Research in Child Development, San Francisco, April.
153 Clarke-Stewart, K. A. (1979) Assessing Social Development. Paper presented at the biennial meeting of the Society for Research in Child Development, San Francisco, March. Clarke-Stewart, K. A. (1980) Complementary strategies for studying day care and social development. In S. Kilmer, ea., Advances in Early Education and Day Care. Greenwich, Conn.: JAI Press Inc. Cochran, M. (1977) A comparison of group day and family child-rearing patterns in Sweden. Child Development 48:702-707. Coelen, C., Glantz, F., and Calore, D. (1978) Day Care Centers in the U.S.: A National Profile 1976-1977. Volume III of the final report of the National Day Care Study. Cambridge, Mass.: Abt Books. Congressional Budget Office (1978) Children and Preschool: Options for Federal Support. Washington, D.C.: U.S. Government Printing Office. Conly, S. R. (1975) Subsidized Day Care and the Employment of Lower Income Mothers: A Case Study. Ph.D. dissertation, Department of Economics, University of South Carolina. Cronbach, L. H., Gleser, G. C., Nanda, H., and Rajaratnam, N. (1972) The Dependability of Behavioral Measures: Theory of Generalizability for Scores and Profiles. New York: John Wiley & Sons, Inc. Dennis, W. (1941) Infant development under conditions of restricted practice and of minimum social stimulation. Genetic Psychological Monographs 23:143-184. Dennis, W., and Najarian, P. (1957) Infant development under environmental handicap. Psychological Monographs 71:(Whole No. 436). Ditmore, J., and Prosser, W. R. (1973) A Study of Day Care's Effect on the Labor Force Participation of Low-Income Mothers. Washington, D.C.: Office of Economic Opportunity.
154 Doyle, A. B. (1975) Infant development in day care. Developmental Psychology 11:655-656. Etaugh, C. (1980) Effects of nonmaternal care on children. American Psychologist 35:309-319. Falender, C., and Heber, R. (1976) Mother-Child Interaction and Participation in a Longitudinal Program. Unpublished paper. Rehabilitation Research and Training Center, Madison, Wisconsin. Farran, D., and Ramey, C. (1980) Social class differences in dyadic involvement during infancy. Child Development 51:254-257. Fein, G. (1976) Infant day care and the family: regulatory strategies to ensure parent participation. Unpublished paper prepared for the Assistant Secretary for Planning and Evaluation, U.S. Department of Health, Education, and Welfare. Fiene, R. J. (1973) The differential structural characteristics of sentences formed by preschool children in family and group care centers. Unpublished paper. Department of Psychology, State University of New York at Stony Brook, August. Fosburg, S., and Hawkins, P. (1981) Final Report of the National Family Day Care Home Study. Volume I: Summary. Cambridge, Mass.: Abt Books. Fowler, W., and Khan, N. (1974) The Development of a Prototype Infant and Child Day Care Center in Metropolitan Toronto. Ontario Institute for Studies in Education. Year III Progress Report. Unpublished. Fowler, W., and Kahn, N. (1975) The Development of a Prototype Infant and Child Day Care Center in Metropolitan Toronto. Ontario Institute for Studies in Education. Year IV Progress Report. Unpublished. Fraiberg, S. (1977) Every Child's Birthright: In Defense of Mothering. New York: Basic Books.
155 Golden, M., Rosenbluth, L., Grossi, M., Policare, H., Freeman, H., and Brownlee, E. (1978) The New York Infant Day Care Study. New York: Medical and Health Research Association of New York City. Goodson, B. D., and Hess, R. D. (1978) The effects of parent training programs on child performance and parent behavior. In B. Long-Term Gains from Early Brown, ea., Found: _ Education. Boulder, Colo.: Westview Press. Gouin-Decarie, T. (1965) Intelligence and Affectivity in Early Childhood. New York: International Universities Press. Hawkins, P., Wilcox, M., Gillis, G., Porter, A., and Carew, J. (1979) Observation Study of Care Givers and Children in Day Care Programs. Paper presented at the biennial meeting of the Society for Research in Child Development, San Francisco, Spring. Heinicke, C., Friedman, D., Prescott, E., Puncel, C., and Sale, J. (1973) The organization of day care: considerations relating to the mental health of child and family. American Journal of Orthopsychiatry 43:8-22. Hill, C. R. (1977) The child care market: a review of the evidence and implications for federal policy. Unpublished paper prepared for the Assistant Secretary for Planning and Evaluation, U.S. Department of Health, Education, and Welfare. Hill, C. R. (1978) Private demand for child care: implications forpublic policy. E 2:523-545. Hoffman, L. W. (1974) Effects of maternal employment on the child. Developmental Psychology 10:204-208. Hunt, J. McV. (1961) Intelligence and Experience. New York: Ronald Press. Hutt, C., and Vaizey, M. J. (1966) Differential effects of group density on social behavior. Nature 209:1371-1372.
156 Joffe, C. E. (1977) Friendly Intruders: Childcare Professionals and Family Life. Berkeley, Calif.: University of California Press. Johnson, B. L., and Hayghe, H. (1977) Labor force participation of married women, March 1976. Monthly Labor Review 100:32-36. Johnson, O. G., and Bommarito, J. W. (1971) Tests and Measurements in Child Development: A Handbook. San Francisco, Calif.: Jossey-Bass, Inc., Publishers. Jusenius, C. L., and Shortlidge, R. L. (1975) Dual Careers: A Longitudinal Study of Labor Market Experience of Women, Columbus, Ohio: Ohio State University, Center for Human Resource Research. Kamerman, S., and Kahn, A. (1978) Family Policy: Government and Families in Fourteen Countries. University Press. Kagan, J., and Moss, H. (1962) Birth to Maturity. Sons, Inc. New York: Columbia New York: John Wiley & Kagan, J., Kearsley, R. B., and Zelazo, P. R. (1976) The Effects of Infant Day-Care on Psychological Development. In The Effects of Early Experience on Child Development. Symposium presented at the meeting of the American Association for the Advancement of Science, Boston, February. Keniston, K., and the Carnegie Council on Children. (1977) All Our Children: The American Family Under Pressure. New York: Harcourt Brace Jovanovich, Inc. Kobn, M. L., and Rosman, B. L. (1972) A social competence scale and symptom checklist for the preschool child: factor dimensions. Their cross-instrument generality and longitudinal persistence. Developmental Psychology 6:430-444. Kruvant, C., Redish, G., Dodge, D. T., Hurt, N. J., Passantino, R. J., and Sheehan, R. (1976) The effects on children of the organization and the design of the day care physical environment--appropriateness of the federal interagency day care requirements. Unpublished paper prepared for the Assistant Secretary for
157 Planning and Evaluation, U.S. Department of Health, Education, and Welfare. Kurz, M., Robins, R., and Spiegelman, R. (1975) A Study of the Demand for Child Care by Working Mothers. Menlo Park, Calif.: Stanford Research Institute. Lally, R. (1973) The Family Development Research Program. Progress Report. Unpublished paper. Syracuse University. Lay, M., and Meyer, W. (1973) Teacher/child behaviors in an open environment day care program. Unpublished paper. Syracuse University Children's Center. Layzer, J. (1980) Interviews with parents. In Technical Appendices to the National Day Care Study. Volume IV-B:185-216 of the final report of the National Day Care Study. Cambridge, Mass.: Abt Associates, Inc. Lippman, M. A., and Grate, B. H. (1974) Socioemotional Effects of Day Care. Final project report. Grant No. OCD-CB-219, Office of Child Development. Loo, C. (1972) The effects of spatial density on the social behavior of children. Journal of Applied Social Psychology 2:372-382. Macrae, J. W., and Herbert-Jackson, E. (1976) Are behavioral effects of infant day care programs specific? Developmental Psychology 12:269-270. Medley, D. M., and Mitzel, H. E. (1963) Measuring classroom behavior by systematic observation. In N. L. Gage, ea., Handbook of Research on Teachinq. Chicago, Ill.: Rand McNally. Meyer, W. (1976) Staffing Characteristics and Child Outcomes. Unpublished paper prepared for the Assistant Secretary for Planning and Evaluation, U.S. Department of Health, Education, and Welfare. Meyers, L. (1973) The relationship between substitute child care, maternal employment and female marital satisfaction. In D. Peters, ea., A Summary of the Pennsylvania Day Care Study. University Park, Pa.: The Pennsylvania State University.
158 Moore, T. (1975) Exclusive early mothering and its alternatives: the outcome to adolescence. Scandinavian Journal of Psychology 16:255-272. National Research Council (1976) Toward A National Policy for Children and . Families. Advisory Committee on Child Development. Washington, D.C.: National Academy of Sciences. Paraskevopoulos, J., and Hunt, J. McV. (1971) Object construction and imitation under differing conditions of rearing. Journal of Genetic Psychology 119:301-321. Parten, M. B. (1932) Social participation among preschool children. Journal of Abnormal and Social Psychology 27:243-269. Peters, D. (1973) A Summary of the Pennsylvania Day Care Study. University Park, Pa.: The Pennsylvania State University. Peterson, D. R. (1961) Behavior problems in middle childhood. Journal of Consulting Psychology 25:205-209. Powell, D. R. tl977) The Interface Be Programs: A Study of Parent-Caregiver Relationships. Detroit, Mich.: Merrill-Palmer Institute. Powell, D. R. (1978) The interpersonal relationship between parents and care givers in day care settings. American Journal of Orthopsychiatry 48:680-689. Prescott E. (1973) A Comparison of Three Types of Day Care and Nursery School/Home Care. Paper presented at the biennial meeting of the Society for Research in Child Development, Philadelphia, Pa., March. Prescott, E., and David, T. G. (1976) Concept Paper on the Effects of the Physical Environment on Day Care. Unpublished paper prepared for the Assistant Secretary for Planning and Evaluation, U.S. Department of Health, Education, and Welfare.
159 Prescott, E., Jones, E., and Kritchevsky, S. (1967) Group Day Care as a Child-Rearing Environment. Report prepared for the Children's Bureau, U.S. Department of Health, Education, and Welfare. Pacific Oaks College, Pasadena, Calif. Prescott, E., Jones, E., Kritchevsky, S., Milich, C., and Haselhoef, E. (1975) Assessment of child-rearing environments. Part II. An Environmental Inventory. Pacific Oaks College, Pasadena, Calif. Prescott, E., Kritchevsky, S., and Jones, E. (1972) The Day Care Environmental Inventory. Pacific Oaks College, Pasadena, Calif. Prescott, E., and Milich, C. (1974) School's Out: Group Day Care for the School- Age Child. Pacific Oaks College, Pasadena, Calif. Provence, S., and Lipton, R. (1962) Infants in Institutions. New York: International Universities Press. Ramey, C., Dorval, B., and Baker-Ward, L. (1981) Group day care and socially disadvantaged families: effects on the child and the family. In S. Kilmer, ea., Advances in Early Education and Day Care. Greenwich, Conn.: JAI Press, Inc. Ramey, C., Farran, D., and Campbell, F. (1979) Predicting IQ from mother-infant interactions. Child Development 50:804-814. Ramey, C., and Mills, J. (1975) Mother-Infant Interaction Patterns as a Function of Rearing Conditions. Paper presented at the biennial meeting of the Society for Research in Child Development, Denver, Colo., April. Ricciuti, H. N. (1974) Fear and development of social attachments in the first year of life. In M. Lewis and L. A. Rosenblum, eds., The Origins of Human Behavior: Fear. New York: John Wiley & Sons, Inc. Riccioti, H. N. (1976) Effects of Infant Day Care Experience on Behavior and Development: Implications for Social Policy. Unpublished paper prepared for the Assistant Secretary for Planning and
160 Evaluation, U.S. Department of Health, Education, and Welfare. Rivlin, A. (1973) Child care. In C. L. Schultz et al., eds., Setting National Priorities: The 1973 Budget. Washington, D.C.: The Brookings Institution. Rowe, R., et al. (1972) Child Care in Massachusetts: The Public Responsibility. Report prepared for the Massachusetts Advisory Council on Education. Cambridge, Mass.: Harvard University Press. Rubenstein, J. L., and Howes, C. (1979) Caregiving and infant behavior in day care and in homes. Developmental Psychology 15:1-24. Rubenstein, J. L., Pedersen, F. A., and Yarrow, L. J. (1977) What happens when mother is away: a comparison of mothers and substitute care givers. Developmental Psychology 13:143-154. Ruopp, R., Travers, J., Coelen, C., and Glantz, F. (1979) Children at the Center. Volume I of the final report of the National Day Care Study. Cambridge, Mass.: Abt Books. Schaefer, E. S. (1961) Converging conceptual models for maternal behavior and for child behavior. In J .C. Glidewell, ea., Parental Attitudes and Child Behavior. Springfield, Ill.: Charles C Thomas. Schwarz, J. C., Strickland, R. G., and Krolick, G. (1974) Infant Day Care: Behavioral Effects at Preschool Age. Developmental Psychology 10:502-506. Shapiro, S. (1975) Preschool ecology: mental variables. 12:236-241. a study of three environ Reading Improvement Shaw, L. B. (1974) The Utilization of Subsidized Child Care in the Gary Income Maintenance Experiment: A Preliminary Report. Washington, D.C.: Office of Economic Opportunity. Spitz, R. A. tl945) Hospitalism: an inquiry into the genesis of psychiatric conditions in early childhood. Psychoanalytic Study of the Child 1:53-74.
161 Sroufe, L. A. (1979) The coherence of individual development: early care, attachment and subsequent developmental issues. American Psychologist 34:834-841. Stallings, J., and Porter, A. (1980) National Day Care Home Study Observation Component. Final report to the Day Care Division, Administration for Children, Youth, and Families, U.S. Department of Health, Education, and Welfare. Menlo Park, Calif.: SRI International. Le ~ and Green, C. Three types of day care: causes, concerns and consequences. Unpublished paper, Program in Social Ecology, University of California, Irvine. Steinberg, (1978) Steinberg, L., and Green, C. (1979) How Parents May Mediate the Effects of Day Care. Paper presented at the biennial meeting of the Society for Research in Child Development, San Francisco, Calif., March. Thomas, A., Chess, S., Birch, H. G., Hertzog, M. E., and Korn, S. (1969) Behavioral Individuality in Early Childhood. New York: New York University Press. Travers, J., Coelen, C., and Ruopp, R. (1977) National Day Care Study Second Annual Report 1975-1976. Cambridge, Mass.: Abt Associates, Inc.. UNCO, Inc. (1975) National Child Care Consumer Study: 1975. Report prepared for the Office of Child Development, U.S. Department of Health, Education, and Welfare. U.S. Department of Health, Education, and Welfare (1978) Social Services U.S.A.: July-September 1976. Administration for Public Services. Washington, D.C.: U.S. Government Printing Office. Vaughn, B. E., Gove, F. L., and Egeland, B. (1980) The relationship between out-of-home care and the quality of infant-mother attachment in an economically disadvantaged population. Child Development 51:1203-1214.
162 Walker, D. K. (1973) Socioemotional Measures for Preschool and - Kindergarten Children San Francisco, Calif.: Jossey-Bass, Inc., Publishers. Westinghouse Learning Corporation and Ohio University (1969) The Im act of Head Start An Evaluation of P the Effects of Head Start Experience on - Childrenls Cognitive and Affective Development. Westinghouse Learning Corporation and Ohio University. Westinghouse Learning Corporation and Westat Research, Inc. (1971) Day Care Survey--1970; Summary Report and Basic Analysis. Report prepared for the Office of Economic Opportunity, Washington, D.C. White, B. L. (1969) Child development reseach: an edifice without a foundation. Merrill-Palmer Quarterly 1:47-78. Woolsey, S. (1977) Pied piper politics and the child-care debate. Daedalus (Spring):127-145. Zigler, E., and Trickett, P. K. (1978) IQ, social competence and evaluation of early childhood intervention programs. American Psychologist 33:789-798.