Ask the Scholar
Document scope · 1 page
Scholar
Ask about this object, its catalog metadata, its source description, or the page inventory.
For page-specific OCR and visual context, open one of the page chats.
Scholar Source Context
Document identity
localId
26415381
label
Child Care Quality [1]
core
doc
dtoType
document
citationUrl
pageCount
1
Source metadata
id
26415381
sourceUrl
contentType
document
title
Child Care Quality [1]
citationUrl
collections
Records of the Domestic Policy Council (Clinton Administration)
Jennifer Klein's Files
imageCount
1
hasImages
yes
source
import
hasTranscription
no
Source extras
naId
26415381
levelOfDescription
fileUnit
otherTitles
42-t-7367456-20140209S-005-002-2015
recordType
description
ocrSource
nara-archive
Single page context
seq
1
pageIndex
0
type
document
mediaId
5f0f394b5286aab5
ocrText
CC
QUALITY
I
11/25/97
17:11
202 690 6562
DHHS/ASPE/HSP
001
DEPARTMENT OF HEALTH AND HUMAN SERVICES
ASSISTANT SECRETARY FOR PLANNING AND EVALUATION
HEALTH
&
OF
HUMAN
DEPARTMENT
SEPTIMES
USA
FAX: (202) 690-6562
Date 11/25/97
From:
Ann Segal
To: Jen Klein
Division:
Agency:
Phone Number: 690-8410
Phone Number: 456-2599
Fax Number: 690-6562
Fax Number: 456-2878
Number of Pages + Cover 21
REMARKS Here is one quality/research paper Martha
Moorehouse is bringing a longer, perhaps better
one tomorrow for the 8:30 meeting. Please let
me know if you need more or different.
Que
11/25/97
17:11
202 690 6562
DHHS/ASPE/HSP
002
RESEARCH MONOGRAPHS OF THE NATIONAL ASSOCIATION
FOR THE EDUCATION OF YOUNG CHILDREN, VOLUME 1
Quality in Child
Care: What
Does Research
Tell Us?
Deborah A. Phillips, Editor
National Association for the Education of Young Children
Washington, D.C.
11/25/97
17:12
202 690 6562
DHHS/ASPE/HSP
003
Chapter 1
Indicators of Quality in Child Care:
Review of Research
Deborah A. Phillips and Carollee Howes
A
S OF 1986, the majority of children younger than age 6, including more
than half of those younger than 1 year of age, were in need of child
care while their mothers worked. This demographic fact had a profound
effect on the major issues addressed In child care research. The question
of whether or not children should be in child care has become obsolete. We
have also been able to move beyond this question because 20 years of
research on child care allayed our worst fears that nonmaternal care was
inevitably harmful to children. To the contrary, the overwhelming mes-
sage was that children in good quality child care show no signs of harm,
and children from low-income families may actually show improved
cognitive development (Clarke-Stewart & Fein, 1983; Rutter, 1981; Zigler &
Gordon, 1982).
The key to this basic conclusion lies in the term good quality. Most of
the supporting research involved high quality often university-based child
care centers, which are not representative of the child care options
available to most parents. Just as homes vary in the experiences they
afford children, so do child care arrangements. Accordingly researchers
moved on to capture this diversity.
The National Day Care Study
The National Day Care Study launched this next phase of empirical
research (Ruopp, Travers, Glantz, & Coelen, 1979). The federal government
initiated the study to guide the construction of national child care
standards. The task was to identify key provisions that best predict good
outcomes for children and to develop cost estimates for offering these
provisions.
1
11/25/97
17:12
202 690 6562
DHHS/ASPE/HSP
004
The major results both contradicted and confirmed the intuitive wis-
dom of the field. Staff-child ratios, long heralded as a significant quality
indicator, contributed only minimally to developmental effects for pre-
schoolers in center-based care. Group size and specialized caregiver
training emerged as the most potent predictors of positive classroom
dynamics and child outcomes. In classrooms with smaller groups, lead
teachers engaged in more social interaction with children; children were
more cooperative, innovative and involved in tasks, and talkative; and
children made greater gains on cognitive tests. Lead teachers with
child-related education spent more time in social interaction with the
children; the children in their classes showed more cooperation and
greater task persistence; and, in centers with higher proportions of
trained caregivers, preschoolers made greater cognitive test score gains.
Additionally preschoolers made more rapid gains on cognitive tests in
centers whose staff and directors voiced concern about cognitive develop-
ment and emphasized individual development rather than group experi-
ences. The investigators interpreted this finding as buttressing the hy-
pothesis that cognitive outcomes in child care are particularly responsive
to children's interactions with their caregivers, rather than with materials
and other children.
The results of the National Day Care Study were a little different for
center-based infant and toddler care. Both staff-child ratios and group size
emerged as significant influences on caregiver behavior and child devel-
opment. Infants in programs with fewer caregivers per child and larger
groups showed more overt distress and apathy than infants in programs
with high ratios and small groups. Low ratios were also associated, as were
large groups, with increased staff time spent in management or control
situations and less social interaction and cognitive/language stimulation
with children.
So few of the infant caregivers had received specialized training that the
effects of this factor could not be determined. However, overall years of
education were positively related to the amount of social interaction and
cognitive/language stimulation in toddler groups and to lower ratings of
child apathy and potential danger in infant groups.
In sum, the National Day Care Study identified group size and special-
ized caregiver training as significant elements of child care quality in
center-based programs for preschoolers and added staff-child ratios to
these elements for infant and toddler care. This study also proposed that
the association between these regulatable variables and children's devel-
opment in child care is largely a function of their facilitating effect on
2
INDICATORS OF QUALITY IN CHILD CARE
11/25/97
17:13
202 690 6562
DHHS/ASPE/HSP
005
caregivers' efforts to interact in positive, stimulating ways with the chil-
dren in their care.
Contemporary research on child care quality
Has subsequent research confirmed these broad conclusions of the
National Day Care Study called them into question, or extended them to
other settings, other elements of quality, or other outcomes? The answer to
each of these questions is "yes." Since 1979, research that has attempted
to reproduce the results of the National Day Care Study has confirmed
some of the results, contradicted others, and forged new areas of Inquiry.
During the last 10 years, researchers have increasingly acknowledged
the complexity of defining quality in child care. In response, they have
adopted multiple methods to assess quality, examined a wider range of
child care arrangements, and placed these questions about quality in a
broader context that considers the interdependence of child care and
family environments.
Quality, by its nature, is a fuzzy concept. Nowhere is this felt more
acutely than by parents who ask, "How can I know what's right for my
child?" In research, quality has been viewed in several ways. First. global
assessments of quality have been used to capture the overall climate of a
program. Second, efforts to extract the specific dimensions of child care
quality have emphasized (a) structural aspects of child care, such as group
composition and staff qualifications, (b) dynamic aspects of child care
that capture children's daily experiences, and (c) contextual aspects of
child care, such as type of setting and staff stability. A third, and relatively
new, perspective encompasses the joint effects of child care quality and
children's family environments.
Global assessments of child care quality
Most observers of child care will readily acknowledge that good chings
go together. Vandell and Powers (1983), for example, found that high
quality university-run centers had high levels of teacher training, large
amounts of space per child, and good staff-child ratios. This inevitable
confounding of individual quality measures led some researchers to treat
quality as a global construct.
Three approaches have been used. The first combined discrete indica-
tors of quality into a composite measure by which programs are evaluated
as either high or low in quality. Howes and Olenick (1986), for example.
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
3
11/25/97
17:14
202 690 6562
DHHS/ASPE/HSP
006
divided child care centers into high and low quality groups using as
criteria for high quality (a) adult-child ratios no lower than 1:4 for children
2 years old or younger and 1:7 for children aged 30 months or older, (b) the
presence of caregivers with formal training in child development, and (c)
staff turnover that did not exceed two teachers per year. They found that
toddlers in high quality centers were significantly more compliant in child
care and were better able to resist the temptation to play with forbidden
toys and to eat forbidden food in a laboratory session.
Using a similar method that relied on ratios, staff training, and space as
quality criteria, Vandell and Powers (1983) found that children in high
quality centers were more likely to engage in positive social interactions
and behaviors than children enrolled in moderate and low quality centers,
who displayed more solitary and unoccupied behavior.
A second approach to obtaining a global assessment of quality relies on
a rating scale that taps multiple areas of program quality. Using observa-
tions of centers, the Early Childhood Environment Rating Scale, com-
monly called ECERS, (Harms & Clifford, 1980) leads to scores on seven
dimensions of quality: (1) personal care, (2) creative activities, (3) lan-
guage/reasoning activities, (4) furnishings/display (5) fine/gross motor
activities, (6) social development, and (7) adult facilities/opportunities.
Summing scores across these seven dimensions generates an overall
quality assessment.
This is the approach taken by the Bermuda Study (see Chapter 3). After
the researchers eliminated the adult facilities subscale, the other six
dimensions were found to be highly interrelated and were thus used to
create a summary score. This score predicted children's intellectual,
language, and social development, such that children in higher quality
centers showed more advanced communication skills and verbal intelli-
gence (McCartney, 1984) and more positive social behavior and task
orientations (Phillips, McCartney & Scarr, 1987).
Finally important policy questions have been raised about whether good
quality child care can achieve the effects that have been demonstrated for
early childhood intervention programs. Efforts to answer these questions
also require that programs be evaluated using global assessments of
quality. The Bermuda data, for example, were reanalyzed to address the
question of whether child care can serve as an effective intervention for
low-income children (McCartney Scarr, Phillips, & Grajek, 1985). One of
the child care centers examined in Bermuda was substantially higher in
quality than the other centers. It was run by the government and served
primarily low-income families. In comparisons with both a subgroup of
4
INDICATORS OF QUALITY IN CHILD CARE
11/25/97
17:14
202 690 6562
DHHS/ASPE/HSP
007
Good quality child care can serve as an effective intervention
for children from low-income families.
low-income children and the entire sample of children attending the
nongovernment, lower quality centers, the low-income children attending
the government-run center were found to have significantly better lan-
guage skills and to be more considerate and sociable. Apparently good
quality child care can serve as an effective intervention for children from
low-income families (see also Ramey & Haskins, 1981).
In sum, global assessments of quality have confirmed common sense
knowledge that better child care is better for children. This conclusion is
not Insignificant, however, in light of the telling qualification it places on
questions of whether child care is detrimental, neutral, or beneficial for
children's development. Without attention to the quality of the child care
in which children spend their days, answers to the either/or question of
sheer enrollment in child care are not only obsolete, but also uninforma-
tive.
Structural dimensions of quality
While it is true that good things co-occur in child care, global assess-
ments of quality are of little use to practitioners and policymakers who
seek to influence specific program features that predict positive outcomes
for children. Moreover, some of the good things may have a more powerful
impact on children's development than others.
These issues were addressed in research that examined specific struc-
tural dimensions of child care quality. The greatest attention has been paid
to the dimensions identified by the National Day Care Study - adult-child
ratio, group size, and caregiver training and experience. This review thus
focuses on these dimensions.
Adult-child ratio. Ratio is considered an important quality indicator
on the basis of assumptions that adult caregivers mediate children's
contact with the social and physical world. Through social games, verbal
interaction, and physical contact, caregivers offer children opportunities
to practice and enjoy social exchanges, learn about the properties of
objects, and acquire a sense of security and self-worth.
The number of children with whom each caregiver can engage in a
stimulating and sensitive fashion is obviously limited. With too many
children to care for, the caregiver's interactions with each child are likely
to become brief and cursory. What does research tell us?
The majority of studies have found that the ratio has a significant effect
on adult and child behavior in child care. Among the outcomes affected
are the amount of adult-child imitation (Francis & Self, 1982), children's
verbal interaction (Field, 1980; Howes & Rubenstein, 1985; Smith &
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
5
11/25/97
17:15
202 690 6562
DHHS/ASPE/HSP
008
Staff-child ratios have a significant effect on adult and child
behavior in child care.
Connolly, 1981), children's engagement in play (Bruner, 1980; Field, 1980:
Howes & Rubenstein, 1985), and nurturant, nonrestrictive caregiver be-
havior (Howes, 1983; Howes & Rubenstein, 1985; Smith & Connolly 1981).
These results are from research studying infant and toddler as well as
preschool-age child care incenters and family day care homes.
Group size. As with ratio, interest in group size derives from both
developmental considerations about the critical socializing function of
child care providers and practical considerations about the demands on
caregivers' time. The results of the National Day Care Study also spurred
interest in group size as a critical structural feature of child care.
The research evidence is quite clear. Smaller groups appear to facilitate
constructive caregiver behavior and positive developmental outcomes for
children. Howes (1983) found that larger groups were associated with less
social stimulation and responsiveness in both center and family day care
settings and more negative affect and restriction on the part of family day
care providers. Howes and Rubenstein (1985) further found that children
in small groups were more talkative. Stith and Davis (1984) studied family
day care homes and also found that larger groups were associated with less
positive affect and less responsiveness to infant distress on the part of
caregivers.
In studies of center-based care, similar results emerge. Bruner (1980)
viewed more pretend play and more elaborate play by children in smaller
centers (fewer than 26 children). Similarly Cummings and Beagles-Ross
(1983) found that children in small centers (8 to 12 children) showed more
positive affect and less avoidance upon entering their child care centers
than children in larger centers (20 to 25 children).
Clarke-Stewart and Gruber (1984) present a more complex picture of the
effects of group size. Consistent with the negative associations observed
in other studies, children in family day care homes, centers, and classes
with large enrollments were less sociable and cooperative with strangers,
especially unfamiliar peers, than children in child care settings with small
enrollments. But children in large classes were also more knowledgeable
about the stranger's social perspective and less likely to behave negatively
with the unfamiliar peer. Large enrollments may have positive as well as
negative consequences.
Caregiver training, education, and experience. The skills and
experience that child care providers bring to their jobs, as in any
profession, are presumed to affect the quality of their performance.
Accordingly experience, education, and training are often used as indica-
tors of caregiver competence. A central controversy in this area is whether
6
INDICATORS OF QUALITY IN CHILD CARE
11/25/97
17:16
202 690 6562
DHHS/ASPE/HSP
009
Smaller groups appear to facilitate constructive caregiver
behavior and positive developmental outcomes for children.
the sheer amount of education or the substance of the education is the
more potent predictor of good quality care. The value of experience, as
opposed to education and training, has also been a topic of debate.
With respect to the question of the amount versus the content of
education and training, the picture is mixed. Unlike the conclusions of the
National Day Care Study that clearly implicated caregivers' child-related
education, and not total years of education, as a determinant of pre-
schoolers' social and cognitive development in child care, subsequent
research points to both dimensions as contributing to quality child care.
There is ample evidence that specialized training is associated with
good quality care. Howes (1983) found that caregivers in centers and family
day care homes with more child-related training engaged in more social
stimulation and responsiveness than other caregivers. In centers, trained
caregivers also showed less negative affect. A national study of family day
care homes (Stallings & Porter, 1980) reported similar results for caregiver
training. Training was associated with more teaching, helping, dramatic
play, and activity that involved interaction with children. Trained family
day care providers also showed more comforting behavior and spent less
time away from the children than untrained providers. In this study total
years of education showed few relationships with caregiver behavior.
Arnett (1987) found associations between specialized caregiver training
and more positive Interactions with children, lower levels of detachment,
and less punitiveness.
Other evidence (Berk, 1985; Clarke-Stewart & Gruber, 1984) suggests
that more education is better than less and that the amount and nature of a
caregiver's preparation may augment each other such that more highly
educated adults who have also received specialized training may be
among the most proficient caregivers.
Berk (1985). for example, found that caregivers with at least 2 years of
college were more likely than less educated caregivers to display encour-
agement, teacher direction, and promotion of verbal skills. They were also
lower in restrictive behavior. She also found, however, that college-
educated caregivers with a child-related major showed more indirect
guidance, less restriction, and more encouragement of children's self-
initiations and verbal expression.
Clarke-Stewart and Gruber (1984) similarly report that the caregiver's
formal education and knowledge of child development are associated with
higher social and cognitive competence in children attending family day
care homes. No significant effects were found for specialized training in
child development. Moreover, children in centers with more highly trained
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
7
11/25/97
17:16
202 690 6562
DHHS/ASPE/HSP
010
There is ample evidence that specialized training is associated
with good quality care.
staff were found to be less independent and socially competent than
children in centers with less highly trained staff.
The evidence on the contribution of experience is also mixed. Care-
givers with more years phexperience have been found to engage in less
social interaction and cognitive stimulation with infants and toddlers
(Ruopp et al., 1979). On the other hand, Howes (1983) found that experi-
enced caregivers were more responsive to children's bids for attention.
Stallings and Porter (1980) found no effects for caregiver experience.
Experience is a multifaceted construct. More sensitive measures that
are capable of deciphering beneficial features of experience and exploring
their relation to competent caregiving are needed, as is substantial
refinement of measures of education and training. For example, whereas
most studies find a relation between training of caregivers and child
outcomes, the content and extent of the training that produces these
outcomes are virtually unexplored. This research has also not distin-
guished the value of education, training, and experience for different
levels of child care staff, such as the director of a center, the classroom
teachers, and the teacher assistants.
Dynamic measures of classroom quality
While evidence about structural indicators of quality that can be
addressed in child care regulations is directly pertinent to licensing
authorities and program directors who establish child care policies, these
indicators offer few insights into children's actual experiences in child
care. Why do more-staff-per-child ratios and small groups promote posi-
tive social and cognitive development? What beneficial processes in child
care are set in motion in well-structured programs?
The results of the National Day Care Study suggest that structural
predictors of quality serve to facilitate constructive interactions between
caregivers and children. Several of the studies summarized above imply
this as well. What other evidence exists on this issue?
Rubenstein, Howes, and Boyle (1979) followed a sample of 10 children
who attended infant care centers. At age 3½, those who had attended
centers characterized by high frequencies of social play with caregivers
responded more favorably to their mothers following a brief separation
than those who were in centers with highly directive caregivers.
Similarly Carew (1980) followed 23 children who attended child care
centers that varied in quality and found that language mastery experi-
ences provided by their caregivers predicted children's performance on IQ
and receptive language tests. Golden and his colleagues (1978) also found
8
INDICATORS OF QUALITY IN CHILD CARE
11/25/97
17:17
202 690 6562
DHHS/ASPE/HSP
011
that 2-year-olds who experienced high levels of cognitive and social
stimulation from their child care providers scored higher on measures of
social competence and language comprehension when they were 3.
These results are corroborated by those reported by McCartney (1984)
in her study of center-based child care in Bermuda. The degree of verbal
stimulation provided to the children by their caregivers predicted chil-
dren's test performance on three measures of language development. In
contrast, conversations initiated with peers had a negative influence on
language development, leading McCartney to hypothesize that peer talk
replaces the more important caregiver talk when fewer adults are on the
staff.
In sum, given associations between structural features of child care and
caregiver behaviors, the results that link caregivers' social, cognitive, and
language interactions with children to child outcomes suggest that the
influence of regulatable variables such as ratios and group size is mediated
by their effects on caregivers. Structural features of child care appear to
affect the dynamic environment that captures children's actual experi-
ences in child care, which in turn predicts children's development in child
care.
Contextual features otchild care
A relatively recent emphasis in the research on child care quality has
expanded the empirical lens to include a variety of child care settings and
aspects of quality such as staff stability that are not reflected when
observations are restricted to single points in time.
The child care setting. Whereas center-based care was studied almost
exclusively in the early research on child care, family day care homes are
now beginning to be studied, as are in-home care arrangements. This
expansion of the child care settings selected as sites for research is highly
important. As of 1982, center-based child care constituted 15% of all
arrangements used by employed mothers. Family day care, in contrast,
constituted 40% of child care arrangements (split about evenly between
that provided by a relative and by a nonrelative) and in-home care provided
by a nonrelative added another 5.5% (U.S. Bureau of the Census, 1983). It is
important to note, however, that center-based care is the most rapidly
growing form of care for children of all ages (Hofferth & Phillips, 1987).
Comparing the results of research conducted in different types of care
is treacherous given that the measures, the ages and characteristics of the
children, and the goals of the research may differ along with the child care
setting. Only a few studies have integrated different types of care into a
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
9
11/25/97
17:18
202 690 6562
DHHS/ASPE/HSP
012
The loss of an attachment figure can be very painful to a
young child.
single research effort (e.g., Benn, 1986; Clarke-Stewart & Gruber, 1984:
Howes, 1983; Howes & Rubenstein, 1985). These studies offer the most
valid sources of comparison across types of care.
Benn (1986) compared the quality of mother-son attachment for chil-
dren in family day care homes and in-home arrangements. No differences
were found for type of child care setting. Howes (1983) and Howes and
Rubenstein (1985) compared children in center and family day care. Both
similarities and differences were found. In both types of care smaller
groups, higher staff-child ratios, and trained caregivers were associated
with better caregiving and child development. Clarke-Stewart and her
colleagues (Clarke-Stewart & Gruber, 1984) examined four types of care in
Chicago: centers, nursery schools, family day care homes, and in-home
care. The results from this study are described fully in Chapter 2, but
briefly Clarke-Stewart concludes that the various types of care present
children with qualitatively distinct environments, ranging from home-like
to institutional settings with varying degrees of exposure to other children
and to educational programs. Very few results were uniform across the
four different types of care examined in the Chicago Study.
Staff stability. Developing secure attachment relationships is among
the most important developmental tasks for young children. Evidence is
clear that children in child/care do not replace their attachments to their
parents with attachments to their child care providers (Ainslie & Ander-
son, 1984; Farran, Burchinal, Hutaff, & Ramey 1984; Kagan, Kearsley, &
Zelazo, 1978). At the same time, however, children do get attached to their
caregivers (Ainslie & Anderson, 1984; Cummings, 1980; Ricciuti, 1974) and
use them as a secure base during the day.
Attachment formation is based in part on the availability and predict-
ability of the caregiver. The loss of an attachment figure can be very
painful to a young child. When these observations are juxtaposed with the
40% annual turnover among center-based child care providers and 60%
turnover among home-based providers (NAEYC, 1985), there is tremen-
dous cause for concern.
Research on infant and toddler care suggests that very young children
differentiate between stable and nonstable caregivers. Rubenstein and
Howes (1979) found that twice as much interaction took place in center
care between infants and/head teachers as between infants and less stable
volunteers. Cummings (1980) observed infants during their morning entry
into center-based child care. Infants were less resistant to transference
from the mother to a stable caregiver and exhibited more positive affect
when the mother left, as compared to infants who were transferred to
10
INDICATORS OF QUALITY IN CHILD CARE
11/25/97
17:19
202 690 6562
DHHS/ASPE/HSP
013
nonstable caregivers.
Howes and Stewart (1987) found that infants and toddlers (age range 11
to 30 months) who had experienced more changes (number of changes
ranged from none to five) in child care arrangements were less likely to
engage in competent playiwith peers and objects when observed in their
current family day care homes. Moreover, in a study of first-grade chil-
dren's school adjustment, the stability of prior child care arrangements
predicted academic progress (Howes, 1988).
Two studies, however, failed to find effects for caregiver stability. Benn
(1986) examined caregiver stability in family day care homes and in-home
arrangements. No association was found between the number of caregiver
changes (ranging from one to eight for boys aged 17 to 21 months) and the
quality of the mother-son attachment relationship. Everson, Sarnat, and
Ambron (1984) also examined stability in center and family day care home
arrangements and found no effects on a broad range of child competence
measures.
In sum, when stability is examined within center-based care, there
appears to be an association between the consistent presence of an adult
caregiver and infants' development in child care. In two studies in which
stability was defined as the total number of changes in child care, no
association was found Changes in arrangements and changes in care-
givers are quite distinct measures of stability the first being far more
extensive in the degree of change involved. Clearly this is a very new area
of research with much need of further study and clarification.
Joint effects of child care and family environments
Home-rearing (no regular use of other supplemental child care arrange-
ments) has often been used as an implicit standard against which the use
of child care has been compared (McCartney & Phillips, 1988). Alterna-
tively, child care has frequently been studied as a separate socialization
environment apart from children's homes. In reality, childrearing has
become a collaborative endeavor with children moving back and forth
many on a daily basis between their homes and child care. The effects
of these two environments may be additive; they may compensate for each
other, or some aspects of one may override aspects of the other in positive
or negative ways. A full, understanding of child development thus requires
that both environments be examined.
In addition, there is an important methodological reason to assess the
joint effects of child care and family environments. Parents select their
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
11
11/25/97
17:19
202 690 6562
DHHS/ASPE/HSP
014
It is likely that parents with different values, finances, and
family structures choose child care that varies in form and
quality.
children's child care arrangements. It is likely that parents with different
values, finances, and family structures choose child care that varies in
form and quality.
Howes and Olenick (1986), for example, found that families enrolling
their children in low quality child care had more complex and presumably
more stressful lives than the families using higher quality care. Moreover,
both parents and caregivers of the children in low quality centers were less
involved and invested in assuring that their children complied with their
requests. This evidence demonstrates that family and child care environ-
ments are not independent, making it difficult to attribute child outcomes
exclusively to child care or exclusively to family factors. Efforts to tease
apart these two realms of influence require, of course, that measures of
each be included in research designs. In the absence of this approach the
effects of child care on child development may be overestimated (Howes &
Olenick, 1986).
Clarke-Stewart and Gruber (1984), for example, report that associations
found between children's competence and features of their child care
settings, such as group composition and caregiver characteristics, were
substantially weakened when variance due to family socioeconomic status
(SES) was removed. Kontos's work relating regulatable characteristics of
child care centers to quality and children's development similarly revealed
that family background variables (SES and family values) significantly
predicted developmental outcomes, while structural characteristics of the
centers made virtually no contribution to development outcomes (Kontos,
1987; Kontos & Fiene, this volume, pp. 57-,80). Goelman and Pence
(1987) have also reported that family variables superseded center
quality variables in predicting child language outcomes in a large study
of child care in Canada, whereas quality variation in family day care
homes was a significant predictor of children's language development.
Alternatively, substantial effects of the quality of children's child care cen-
ters remained in the Bermuda Study (McCartney, 1984; Phillips,
McCartney, & Scarr, 1987) after the influence of the parents' childrear-
ing values was statistically removed.
One possible explanation for this disparate pattern of results concerns
the relative range of variation in the family variables versus program
quality variables. When the range of families included in the research is
more extreme than the range of quality represented by the child care
programs, family factors emerge as the more salient influence, whereas
the opposite pattern of results appears to emerge when an ample range of
child care quality arrangements is included (see subsequent chapters and
12
INDICATORS OF QUALITY IN CHILD CARE
11/25/97
17:20
202 690 6562
DHHS/ASPE/HSP
015
especially Chapter 7 by Clarke-Stewart for further discussion). This
finding has led several investigators to recognize the importance of
examining the interrelations between family and quality measures rather
than treating them as independent influences on child development.
The Chicago Study provides an example of examining interactive
effects. Clarke-Stewart (1984) presents evidence that a combined measure
of toys in the home and in child care was more predictive of child
development than measures that reflected only the home or child care
environments. Howes and Olenick (1986) also found that analyses that
incorporated both child care and family influences were more predictive
of several child outcomes (e.g., compliance in girls, task resistance in boys)
than analyses that took into account only one set of factors.
In a longitudinal study of children in home care, family day care, and
center care in Sweden, Cochran (1977), Gunnarsson (1978), and Cochran
and Robinson (1983) examined the interaction of structure and process
variables in both in-home and out-of-home child care settings. Children's
scores on the Griffiths Scale of Mental Development were strongly
influenced by the interaction of child care structure variables (type of
care), child care process and family process variables (social interactions
with caregivers and peers) family structure variables (maternal marital
status), and child sex. A factor that somewhat restricts the generalizability
of these findings is that the subject pool was characterized by unequal
proportions of children from one-parent families in the family day care
(8%) and center care groups (33%). Nonetheless, this study represents one
of the few attempts to examine systematically the interaction of child care
and family variables. The investigators concluded that while previous
studies
have tended to view day care as an independent, causal agent operating on the
lives of young children the day care experience is better conceptualized as an
intervening variable [their emphasis) which mediates certain family types on the
one hand (two working parents, single parent) and long-term developmental
outcomes on the other. (Cochran & Robinson, 1983, p. 61)
Recently assessments of family influences that relied on broad socio-
economic classifications, childrearing values, and measures of the home
environment have been supplemented by more subtle, but perhaps more
directly pertinent measures of maternal attitudes toward the use of child
care (Everson, Sarnat, & Ambron, 1984; Hock, 1984; Hock, DeMeis, &
McBride, 1987). Hock has presented convincing evidence that mothers'
attitudes about separation from their children are associated with differ-
ent patterns of child care use. Employed mothers who have children
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
13
11/25/97
17:21
202 690 6562
DHHS/ASPE/HSP
016
It is important to understand parental feelings and attitudes
when assessing the effects of different types of care on children.
enrolled in child care centers are significantly less concerned about the
consequences of maternal separation than are employed mothers who use
other forms of care. Hock concludes that it is important to understand
parental feelings and attitudes when assessing the effects of different
types of care on children (Hock, DeMeis, & McBride, 1988).
This approach was taken by Everson and his colleagues (Everson. 1981;
Everson, Sarnat, & Ambron, 1984). They examined the mediating intlu-
ence of mothers' positive or negative disposition to use child care on
children's adjustment to child care in both center and family day care
arrangements. After the children— all toddlers had been enrolled in
child care for 5 months, the results suggested that the congruence
between maternal attitudes and use of child care was a highly significant
predictor of children's adjustment. Mothers who relied on child care but
were uncomfortable with it and mothers who felt comfortable with the use
of child care but were not using it (called inconsistent mothers) had
children who were more easily upset by a frustrating task, showed greater
distress at maternal separation, and were less compliant with their
mothers' requests while playing. The inconsistent mothers were also
quicker to become angry and impatient with their children.
After 10 months in child c are, a different picture emerged. Attitude-
behavior consistency was no longer the issue. Attitudes alone predicted
child outcomes. Specifically mothers who were positively disposed toward
the use of child care, compared to those who were negatively disposed,
had children who were less cooperative with adults, were less compliant
with their mothers, and displayed inferior approaches to a problem-solving
task, regardless of whether they were in child care or not. Everson
concludes, "The specific effects of day care may depend in large measure
on maternal attitudes toward day care and other family characteristics"
(Everson, Sarnat, & Ambron, 1984, pp. 90-91).
In sum, the combined effects of child care quality and type. the
children's child care experience, and their family context need to be
considered in future studies of child care. It is entirely possible that family
factors (such as parental attitudes about the use of child care) mediate
child care choices that, in turn, have differing effects on children. At the
very least, inclusion of family-related measures in the study of child care
drives home the complexity of identifying where, when, and how quality of
care makes a difference in the lives of children.
14
INDICATORS OF QUALITY IN CHILD CARE
11/25/97
17:21
202 690 6562
DHHS/ASPE/HSP
017
Conclusions
Research on child care quality has accumulated a vast collection of
results during the last 10 years. More than any other aspect of child
development research, this literature has driven home the true complexity
of child care and the realchallenges faced by those who seek to assess its
effects on children.
The first challenge for researchers involves selecting a measure of
quality global or discrete, regulatable or more dynamic, a static snap-
shot measure or one that captures children's and caregivers' movement in
and out of child care. The second challenge involves measuring other
factors, particularly aspects of the family environment, that affect child
development and may interact with, compensate for, or operate com-
pletely independently of the influence of child care quality.
In the chapters that follow, five groups of researchers present their
efforts to confront these challenges and the results their work has yielded.
They have examined different types of child care, in different locations in
and out of the United States, and with different populations of children
and families. The measures of quality used in these collective studies
encompass the full rangejof options described in this review. Each study
placed the developmental consequences of variation in the quality of child
care environments in the context of children's home environments. Alison
Clarke-Stewart then addresses the central question of how the results of
these recent studies confirm, contradiet, and extend those of the research
reviewed here.
References
Ainslie, R.C., & Anderson, C.W. (1984). Day care children's relationships to their
mothers and caregivers: An inquiry into the conditions for the development of
attachment. In R.C. Ainslie (Ed.), Quality variations in daycare (pp. 98-132).
New York: Praeger.
Arnett, J. (1987, April). Training for caregivers in day care centers. Paper
presented at the biennial meeting of the Society for Research in Child
Development, Baltimore, MD.
Benn. R. (1986). Factors promoting secure attachment relationships between
employed mothers and their sons. Child Development, 57, 1224-1231.
Berk, L. (1985). Relationships of educational attainment, child-oriented attitudes.
job satisfaction, and career commitment to caregiver behavior toward chil-
dren. Child Care Quarterly, 14, 103-129.
Bruner, J. (1980). Under five in Britain. Ypsilanti. ML: High/Scope.
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
15
11/25/97
17:22
202 690 6562
DHHS/ASPE/HSP
018
Childrearing has become a collaborative endeavor with children
moving back and forth between their homes and child care.
16
INDICATORS OF QUALITY IN CHILD CARE
17:23
202 690 6562
DHHS/ASPE/HSP
019
Carew, J. (1980). Experience and the development of intelligence in young
children. Monographs of the Society for Research in Child Development, 45
(6-7, Serial No. 187).
Clarke-Stewart, A. (1984). Day care: A new context for research and development.
In M. Perlmutter (Ed.), The Minnesota Symposium on Child Psychology, Vol.
17 (pp. 61-100). Hillsdale, NJ: Erlbaum.
Clarke-Stewart, A., & Fein. G.G. (1983). Early childhood programs. In M. Haith &J.
Campos (Eds.). Handbook of child psychology: Vol. 2. Infancy and develop-
mental psychobiology (pp: 917-1000). New York: Wiley.
Clarke-Stewart, K.A., & Gruber, C. (1984). Daycare forms and features. In R.C.
Ainslie (Ed.), Quality variations in daycare (pp. 35-62). New York: Praeger.
Cochran, M.M. (1977). A comparison of group day and family childrearing pat-
terns. Child Development. 48, 702-707.
Cochran, M.M., & Robinson, (1983). Day care, family circumstances and sex
differences in children. In S. Kilmer (Ed.), Advances in early education and
day care (Vol. 3, pp. 46-67). Greenwich, CT: JAI Press.
Cummings, E.H. (1980). Caregiver stability and day care. Developmental Psy-
chology, 16, 31-37.
Cummings, M., & Beagles-Ross, J. (1983). Towards a model of infant daycare:
Studies of factors influencing responding to separation in daycare. In R.C.
Ainslie (Ed.), Quality variations in daycare (pp. 159-182). New York: Praeger.
Everson, M.D. (1981). The impact of day care on the attachment behavior of 12
to 24 month olds. Unpublished doctoral dissertation, Stanford University Palo
Alto. California.
Everson. M.D., Sarnat. L., & Ambron, S.R. (1984). Day care and early socialization:
The role of maternal attitude. In R.C. Ainslie (Ed.), Quality variations in
daycare (pp. 63-97). New York: Praeger.
Farran, D.C., Burchinal, M., Hutaff, S.E., & Ramey C.T. (1984). Allegiances or
attachments: Relationships among infants and their day care teachers. In R.C.
Ainslie (Ed.), Quality variations in daycare (pp. 133-158). New York: Praeger.
Field, T. (1980). Preschool play: Effects of teacher child ratio and organization of
classroom space. Child Study Journal, 10, 191-205.
Francis, P., & Self, P. (1982). Imitative responsiveness of young children in day care
and home settings: The importance of the child to caregiver ratio. Child Study
Journal, 12, 119-126.
Goelman, H., & Pence. A. (1987). Some aspects of the relationships between
family structure and child language development in three types of day care. In
D. Peters & S. Kontos (Eds.), Annual advances in applied developmental
psychology, Vol 2: Continuity and discontinuity of experience in child care
(pp. 129-146). Norwood, NJ: Ablex.
Golden, M., Rosenbluth, L., Grossi, M., Policare, H., Freeman, H., & Brownlee, M.
(1978). The New York City Infant Day Care Study. Medical and Health
Research Association of New York City 40 Worth St., New York, NY 10013.
Gunnarsson, L. (1978). Children, day care and family care in Sweden: A
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
17
202 690 6562
DHHS/ASPE/HSP
020
follow-up. Gothenburg, Sweden: University of Gothenburg.
Harms, T., & Clifford, R.M. (1980). Early Childhood Environment Rating Scale.
New York: Teachers College Press, Columbia University.
Hock. E. (1984). The transition to day care: Effects of maternal separation anxiety
on infant adjustment. In R.C. Ainslie (Ed.). The child and the day care setting
(pp. 183-206). New York: Praeger.
Hock, E., DeMeis, D., & McBride, S. (1988). Maternal separation anxiety: Its
role in the balance of employment and motherhood in mothers of infants. In
A. Gottfried & A. Gottfried (Eds.), Maternal employment and children's
development: Longitudinal research. New York: Plenum.
Hofferth, S.L., & Phillips, D.A. (1987). Child care in the United States, 1970 to 1995.
Journal of Marriage and the Family, 49(3), 559-571.
Howes, C. (1983). Caregiver behavior in center and family day care. Journal of
Applied Developmental Psychology, 4, 99-107.
Howes, C. (1988). Relations between early child care and schooling. Developmental
Psychology, 24, 53-57.
Howes, C., & Olenick, M. (1986). Family and child care influences on toddlers'
compliance. Child Development, 57, 202-216.
Howes, C., & Rubenstein, J. (1985). Determinants of toddlers' experience in
daycare: Age of entry and quality of setting. Child Care Quarterly, 14,
140-151
Howes, C., & Stewart, P. (1987). Child's play with adults, toys and peers: An
examination of family and child care influences. Developmental Psychology,
23, 423-430.
Kagan, J., Kearsley R.B., & Zelazo, P.R. (1978). Infancy: Its place in human
development. Cambridge, MA: Harvard University Press.
Kontos, S. (1987, April). Day care quality, family background and children's
development. Paper presented at the biennial meeting of the Society for
Research in Child Development, Baltimore, MD.
McCartney K. (1984). The effect of quality of day care environment upon
children's language development. Developmental Psychology, 20, 244-260.
McCartney, K., & Phillips, D. (1988). Motherhood and child care. In B. Birns &
D. Hay (Eds.), Different faces of motherhood. New York: Plenum.
McCartney K., Scarr, S., Phillips, D., & Grajek, S. (1985). Day care as intervention:
Comparisons of varying quality programs. Journal of Applied Developmental
Psychology, 6, 247-260.
National Association for the Education of Young Children. (1985). In whose
hands? A demographic factsheet on child care providers. Washington. DC:
NAEYC.
Phillips, D., Scarr, S., & McCartney K. (1987). Child care quality and children's
social development. Developmental Psychology, 23, 537-543.
Ricciuti, H. (1974). Fear and the development of social attachments in the first
year of life. In M. Lewis & L. Rosenblum (Eds.), The origins of human behavior:
Fear (pp. 73-106). New York: Wiley
18
INDICATORS OF QUALITY IN CHILD CARE
11/25/97
17:24
202 690 6562
DHHS/ASPE/HSP
021
Rubenstein, J.L., & Howes, C. (1979). Caregiving and infant behavior in day care
and in homes. Developmental Psychology. 15, 1-24.
Rubenstein, J., Howes, C., & Boyle, P. (1979). A two year follow-up of infants In
community based day care. Journal of Child Psychology and Psychiatry, 22,
209-218.
Ruopp, R, Travers, J., Glantz, F., & Coelen, C. (1979). Children at the center: Final
results of the National Day Care Study. Cambridge, MA: Abt Associates.
Rutter, M. (1981). Social-emotional consequences of day care for preschool
children. American Journal of Orthopsychiatry, 51, 4-28.
Smith, P., & Connolly K. (1981). The behavioral ecology of the preschool.
Cambridge, England: Cambridge University Press.
Stallings, J., & Porter, A. (1980). National Daycare Home Study. Palo Alto, CA:
Stanford Research Institute.
Stich, S.M., & Davis, A.J. (1984). Employed mothers and family day care: A
comparative analysis of infant care. Child Development, 55, 1340-1348.
U.S. Bureau of the Census. (1983, November). Child care arrangements of
working mothers: June 1982. Special Studies, Series P-23, No. 129. Washing.
ton, DC: U.S. Department of Commerce.
Vandell, D.L., & Powers, C.P. (1983). Day care quality and children's free play
activities. American Journal of Orthopsychiatry, 53, 493-500.
Zigler, E., & Gordon, E. (1982). Day care: Scientific and social policy issues.
Boston: Auburn House.
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
19
Children whose development was advanced not only had the advan-
tage of being in high quality child care programs but also came from
families who gave them support, stimulation, and education.
RESEARCH MONOGRAPHS OF THE NATIONAL ASSOCIATION
FOR THE EDUCATION OF YOUNG CHILDREN, VOLUME 1
Quality in Child
Care: What
Does Research
Tell Us?
Deborah A. Phillips, Editor
National Association for the Education of Young Children
Washington, D.C.
Chapter 1
Indicators of Quality in Child Care:
Review of Research
Deborah A. Phillips and Carollee Howes
A
S OF 1986, the majority of children younger than age 6, including more
than half of those younger than 1 year of age, were in need of child
care while their mothers worked. This demographic fact had a profound
effect on the major issues addressed in child care research. The question
of whether or not children should be in child care has become obsolete. We
have also been able to move beyond this question because 20 years of
research on child care allayed our worst fears that nonmaternal care was
inevitably harmful to children. To the contrary the overwhelming mes-
sage was that children in good quality child care show no signs of harm,
and children from low-income families may actually show improved
cognitive development (Clarke-Stewart & Fein, 1983; Rutter, 1981; Zigler &
Gordon, 1982).
The key to this basic conclusion lies in the term good quality. Most of
the supporting research involved high quality often university-based child
care centers, which are not representative of the child care options
available to most parents. Just as homes vary in the experiences they
afford children, so do child care arrangements. Accordingly researchers
moved on to capture this diversity.
The National Day Care Study
The National Day Care Study launched this next phase of empirical
research (Ruopp, Travers, Glantz, & Coelen, 1979). The federal government
initiated the study to guide the construction of national child care
standards. The task was to identify key provisions that best predict good
outcomes for children and to develop cost estimates for offering these
provisions.
1
The major results both contradicted and confirmed the intuitive wis-
dom of the field. Staff-child ratios, long heralded as a significant quality
indicator, contributed only minimally to developmental effects for pre-
schoolers in center-based care. Group size and specialized caregiver
training emerged as the most potent predictors of positive classroom
dynamics and child outcomes. In classrooms with smaller groups, lead
teachers engaged in more social interaction with children; children were
more cooperative, innovative and involved in tasks, and talkative; and
children made greater gains on cognitive tests. Lead teachers with
child-related education spent more time in social interaction with the
children; the children in their classes showed more cooperation and
greater task persistence; and, in centers with higher proportions of
trained caregivers, preschoolers made greater cognitive test score gains.
Additionally preschoolers made more rapid gains on cognitive tests in
centers whose staff and directors voiced concern about cognitive develop-
ment and emphasized individual development rather than group experi-
ences. The investigators interpreted this finding as buttressing the hy-
pothesis that cognitive outcomes in child care are particularly responsive
to children's interactions with their caregivers, rather than with materials
and other children.
The results of the National Day Care Study were a little different for
center-based infant and toddler care. Both staff-child ratios and group size
emerged as significant influences on caregiver behavior and child devel-
opment. Infants in programs with fewer caregivers per child and larger
groups showed more overt distress and apathy than infants in programs
with high ratios and small groups. Low ratios were also associated, as were
large groups, with increased staff time spent in management or control
situations and less social interaction and cognitive/language stimulation
with children.
So few of the infant caregivers had received specialized training that the
effects of this factor could not be determined. However, overall years of
education were positively related to the amount of social interaction and
cognitive/language stimulation in toddler groups and to lower ratings of
child apathy and potential danger in infant groups.
In sum, the National Day Care Study identified group size and special-
ized caregiver training as significant elements of child care quality in
center-based programs for preschoolers and added staff-child ratios to
these elements for infant and toddler care. This study also proposed that
the association between these regulatable variables and children's devel-
opment in child care is largely a function of their facilitating effect on
2
INDICATORS OF QUALITY IN CHILD CARE
caregivers' efforts to interact in positive, stimulating ways with the chil-
dren in their care.
Contemporary research on child care quality
Has subsequent research confirmed these broad conclusions of the
National Day Care Study, called them into question, or extended them to
other settings, other elements of quality, or other outcomes? The answer to
each of these questions is "yes." Since 1979, research that has attempted
to reproduce the results of the National Day Care Study has confirmed
some of the results, contradicted others, and forged new areas of inquiry.
During the last 10 years, researchers have increasingly acknowledged
the complexity of defining quality in child care. In response, they have
adopted multiple methods to assess quality, examined a wider range of
child care arrangements, and placed these questions about quality in a
broader context that considers the interdependence of child care and
family environments.
Quality, by its nature, is a fuzzy concept. Nowhere is this felt more
acutely than by parents who ask, "How can I know what's right for my
child?" In research, quality has been viewed in several ways. First, global
assessments of quality have been used to capture the overall climate of a
program. Second, efforts to extract the specific dimensions of child care
quality have emphasized (a) structural aspects of child care, such as group
composition and staff qualifications, (b) dynamic aspects of child care
that capture children's daily experiences, and (c) contextual aspects of
child care, such as type of setting and staff stability. A third, and relatively
new, perspective encompasses the joint effects of child care quality and
children's family environments.
Global assessments of child care quality
Most observers of child care will readily acknowledge that good things
go together. Vandell and Powers (1983), for example, found that high
quality university-run centers had high levels of teacher training, large
amounts of space per child, and good staff-child ratios. This inevitable
confounding of individual quality measures led some researchers to treat
quality as a global construct.
Three approaches have been used. The first combined discrete indica-
tors of quality into a composite measure by which programs are evaluated
as either high or low in quality. Howes and Olenick (1986), for example.
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
3
divided child care centers into high and low quality groups using as
criteria for high quality (a) adult-child ratios no lower than 1:4 for children
2 years old or younger and 1:7 for children aged 30 months or older, (b) the
presence of caregivers with formal training in child development, and (c)
staff turnover that did not exceed two teachers per year. They found that
toddlers in high quality centers were significantly more compliant in child
care and were better able to resist the temptation to play with forbidden
toys and to eat forbidden food in a laboratory session.
Using a similar method that relied on ratios, staff training, and space as
quality criteria, Vandell and Powers (1983) found that children in high
quality centers were more likely to engage in positive social interactions
and behaviors than children enrolled in moderate and low quality centers,
who displayed more solitary and unoccupied behavior.
A second approach to obtaining a global assessment of quality relies on
a rating scale that taps multiple areas of program quality. Using observa-
tions of centers, the Early Childhood Environment Rating Scale, com-
monly called ECERS, (Harms & Clifford, 1980) leads to scores on seven
dimensions of quality: (1) personal care, (2) creative activities, (3) lan-
guage/reasoning activities, (4) furnishings/display (5) fine/gross motor
activities, (6) social development, and (7) adult facilities/opportunities.
Summing scores across these seven dimensions generates an overall
quality assessment.
This is the approach taken by the Bermuda Study (see Chapter 3). After
the researchers eliminated the adult facilities subscale, the other six
dimensions were found to be highly interrelated and were thus used to
create a summary score. This score predicted children's intellectual,
language, and social development, such that children in higher quality
centers showed more advanced communication skills and verbal intelli-
gence (McCartney 1984) and more positive social behavior and task
orientations (Phillips, McCartney & Scarr, 1987).
Finally important policy questions have been raised about whether good
quality child care can achieve the effects that have been demonstrated for
early childhood intervention programs. Efforts to answer these questions
also require that programs be evaluated using global assessments of
quality. The Bermuda data, for example, were reanalyzed to address the
question of whether child care can serve as an effective intervention for
low-income children (McCartney Scarr, Phillips, & Grajek, 1985). One of
the child care centers examined in Bermuda was substantially higher in
quality than the other centers. It was run by the government and served
primarily low-income families. In comparisons with both a subgroup of
4
INDICATORS OF QUALITY IN CHILD CARE
Good quality child care can serve as an effective intervention
for children from low-income families.
low-income children and the entire sample of children attending the
nongovernment, lower quality centers, the low-income children attending
the government-run center were found to have significantly better lan-
guage skills and to be more considerate and sociable. Apparently good
quality child care can serve as an effective intervention for children from
low-income families (see also Ramey & Haskins, 1981).
In sum, global assessments of quality have confirmed common sense
knowledge that better child care is better for children. This conclusion is
not insignificant, however, in light of the telling qualification it places on
questions of whether child care is detrimental, neutral, or beneficial for
children's development. Without attention to the quality of the child care
in which children spend their days, answers to the either/or question of
sheer enrollment in child care are not only obsolete, but also uninforma-
tive.
Structural dimensions of quality
While it is true that good things co-occur in child care, global assess-
ments of quality are of little use to practitioners and policymakers who
seek to influence specific program features that predict positive outcomes
for children. Moreover, some of the good things may have a more powerful
impact on children's development than others.
These issues were addressed in research that examined specific struc-
tural dimensions of child care quality. The greatest attention has been paid
to the dimensions identified by the National Day Care Study - adult-child
ratio, group size, and caregiver training and experience. This review thus
focuses on these dimensions.
Adult-child ratio. Ratio is considered an important quality indicator
on the basis of assumptions that adult caregivers mediate children's
contact with the social and physical world. Through social games, verbal
interaction, and physical contact, caregivers offer children opportunities
to practice and enjoy social exchanges, learn about the properties of
objects, and acquire a sense of security and self-worth.
The number of children with whom each caregiver can engage in a
stimulating and sensitive fashion is obviously limited. With too many
children to care for, the caregiver's interactions with each child are likely
to become brief and cursory. What does research tell us?
The majority of studies have found that the ratio has a significant effect
on adult and child behavior in child care. Among the outcomes affected
are the amount of adult-child imitation (Francis & Self, 1982), children's
verbal interaction (Field, 1980; Howes & Rubenstein, 1985; Smith &
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
5
Staff-child ratios have a significant effect on adult and child
behavior in child care.
Connolly, 1981), children's engagement in play (Bruner, 1980; Field, 1980;
Howes & Rubenstein, 1985), and nurturant, nonrestrictive caregiver be-
havior (Howes, 1983; Howes & Rubenstein, 1985; Smith & Connolly, 1981).
These results are from research studying infant and toddler as well as
preschool-age child care in centers and family day care homes.
Group size. As with ratio, interest in group size derives from both
developmental considerations about the critical socializing function of
child care providers and practical considerations about the demands on
caregivers' time. The results of the National Day Care Study also spurred
interest in group size as a critical structural feature of child care.
The research evidence is quite clear. Smaller groups appear to facilitate
constructive caregiver behavior and positive developmental outcomes for
children. Howes (1983) found that larger groups were associated with less
social stimulation and responsiveness in both center and family day care
settings and more negative affect and restriction on the part of family day
care providers. Howes and Rubenstein (1985) further found that children
in small groups were more talkative. Stith and Davis (1984) studied family
day care homes and also found that larger groups were associated with less
positive affect and less responsiveness to infant distress on the part of
caregivers.
In studies of center-based care, similar results emerge. Bruner (1980)
viewed more pretend play and more elaborate play by children in smaller
centers (fewer than 26 children). Similarly Cummings and Beagles-Ross
(1983) found that children in small centers (8 to 12 children) showed more
positive affect and less avoidance upon entering their child care centers
than children in larger centers (20 to 25 children).
Clarke-Stewart and Gruber (1984) present a more complex picture of the
effects of group size. Consistent with the negative associations observed
in other studies, children in family day care homes, centers, and classes
with large enrollments were less sociable and cooperative with strangers,
especially unfamiliar peers, than children in child care settings with small
enrollments. But children in large classes were also more knowledgeable
about the stranger's social perspective and less likely to behave negatively
with the unfamiliar peer. Large enrollments may have positive as well as
negative consequences.
Caregiver training, education, and experience. The skills and
experience that child care providers bring to their jobs, as in any
profession, are presumed to affect the quality of their performance.
Accordingly experience, education, and training are often used as indica-
tors of caregiver competence. A central controversy in this area is whether
INDICATORS OF QUALITY IN CHILD CARE
Smaller groups appear to facilitate constructive caregiver
behavior and positive developmental outcomes for children.
the sheer amount of education or the substance of the education is the
more potent predictor of good quality care. The value of experience, as
opposed to education and training, has also been a topic of debate.
With respect to the question of the amount versus the content of
education and training, the picture is mixed. Unlike the conclusions of the
National Day Care Study that clearly implicated caregivers' child-related
education, and not total years of education, as a determinant of pre-
schoolers' social and cognitive development in child care, subsequent
research points to both dimensions as contributing to quality child care.
There is ample evidence that specialized training is associated with
good quality care. Howes (1983) found that caregivers in centers and family
day care homes with more child-related training engaged in more social
stimulation and responsiveness than other caregivers. In centers, trained
caregivers also showed less negative affect. A national study of family day
care homes (Stallings & Porter, 1980) reported similar results for caregiver
training. Training was associated with more teaching, helping, dramatic
play, and activity that involved interaction with children. Trained family
day care providers also showed more comforting behavior and spent less
time away from the children than untrained providers. In this study total
years of education showed few relationships with caregiver behavior.
Arnett (1987) found associations between specialized caregiver training
and more positive interactions with children, lower levels of detachment,
and less punitiveness.
Other evidence (Berk, 1985; Clarke-Stewart & Gruber, 1984) suggests
that more education is better than less and that the amount and nature of a
caregiver's preparation may augment each other such that more highly
educated adults who have also received specialized training may be
among the most proficient caregivers.
Berk (1985), for example, found that caregivers with at least 2 years of
college were more likely than less educated caregivers to display encour-
agement, teacher direction, and promotion of verbal skills. They were also
lower in restrictive behavior. She also found, however, that college-
educated caregivers with a child-related major showed more indirect
guidance, less restriction, and more encouragement of children's self-
initiations and verbal expression.
Clarke-Stewart and Gruber (1984) similarly report that the caregiver's
formal education and knowledge of child development are associated with
higher social and cognitive competence in children attending family day
care homes. No significant effects were found for specialized training in
child development. Moreover, children in centers with more highly trained
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
7
There is ample evidence that specialized training is associated
with good quality care.
staff were found to be less independent and socially competent than
children in centers with less highly trained staff.
The evidence on the contribution of experience is also mixed. Care-
givers with more years of experience have been found to engage in less
social interaction and cognitive stimulation with infants and toddlers
(Ruopp et al., 1979). On the other hand, Howes (1983) found that experi-
enced caregivers were more responsive to children's bids for attention.
Stallings and Porter (1980) found no effects for caregiver experience.
Experience is a multifaceted construct. More sensitive measures that
are capable of deciphering beneficial features of experience and exploring
their relation to competent caregiving are needed, as is substantial
refinement of measures of education and training. For example, whereas
most studies find a relation between training of caregivers and child
outcomes, the content and extent of the training that produces these
outcomes are virtually unexplored. This research has also not distin-
guished the value of education, training, and experience for different
levels of child care staff, such as the director of a center, the classroom
teachers, and the teacher assistants.
Dynamic measures of classroom quality
While evidence about structural indicators of quality that can be
addressed in child care regulations is directly pertinent to licensing
authorities and program directors who establish child care policies, these
indicators offer few insights into children's actual experiences in child
care. Why do more-staff-per-child ratios and small groups promote posi-
tive social and cognitive development? What beneficial processes in child
care are set in motion in well-structured programs?
The results of the National Day Care Study suggest that structural
predictors of quality serve to facilitate constructive interactions between
caregivers and children. Several of the studies summarized above imply
this as well. What other evidence exists on this issue?
Rubenstein, Howes, and Boyle (1979) followed a sample of 10 children
who attended infant care centers. At age 3½, those who had attended
centers characterized by high frequencies of social play with caregivers
responded more favorably to their mothers following a brief separation
than those who were in centers with highly directive caregivers.
Similarly Carew (1980) followed 23 children who attended child care
centers that varied in quality and found that language mastery experi-
ences provided by their caregivers predicted children's performance on IQ
and receptive language tests. Golden and his colleagues (1978) also found
8
INDICATORS OF QUALITY IN CHILD CARE
that 2-year-olds who experienced high levels of cognitive and social
stimulation from their child care providers scored higher on measures of
social competence and language comprehension when they were 3.
These results are corroborated by those reported by McCartney (1984)
in her study of center-based child care in Bermuda. The degree of verbal
stimulation provided to the children by their caregivers predicted chil-
dren's test performance on three measures of language development. In
contrast, conversations initiated with peers had a negative influence on
language development, leading McCartney to hypothesize that peer talk
replaces the more important caregiver talk when fewer adults are on the
staff.
In sum, given associations between structural features of child care and
caregiver behaviors, the results that link caregivers' social, cognitive, and
language interactions with children to child outcomes suggest that the
influence of regulatable variables such as ratios and group size is mediated
by their effects on caregivers. Structural features of child care appear to
affect the dynamic environment that captures children's actual experi-
ences in child care, which in turn predicts children's development in child
care.
Contextual features of child care
A relatively recent emphasis in the research on child care quality has
expanded the empirical lens to include a variety of child care settings and
aspects of quality such as staff stability that are not reflected when
observations are restricted to single points in time.
The child care setting. Whereas center-based care was studied almost
exclusively in the early research on child care, family day care homes are
now beginning to be studied, as are in-home care arrangements. This
expansion of the child care settings selected as sites for research is highly
important. As of 1982, center-based child care constituted 15% of all
arrangements used by employed mothers. Family day care, in contrast,
constituted 40% of child care arrangements (split about evenly between
that provided by a relative and by a nonrelative) and in-home care provided
by a nonrelative added another 5.5% (U.S. Bureau of the Census, 1983). It is
important to note, however, that center-based care is the most rapidly
growing form of care for children of all ages (Hofferth & Phillips, 1987).
Comparing the results of research conducted in different types of care
is treacherous given that the measures, the ages and characteristics of the
children, and the goals of the research may differ along with the child care
setting. Only a few studies have integrated different types of care into a
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
9
The loss of an attachment figure can be very painful to a
young child.
single research effort (e.g., Benn, 1986; Clarke-Stewart & Gruber, 1984:
Howes, 1983; Howes & Rubenstein, 1985). These studies offer the most
valid sources of comparison across types of care.
Benn (1986) compared the quality of mother-son attachment for chil-
dren in family day care homes and in-home arrangements. No differences
were found for type of child care setting. Howes (1983) and Howes and
Rubenstein (1985) compared children in center and family day care. Both
similarities and differences were found. In both types of care smaller
groups, higher staff-child ratios, and trained caregivers were associated
with better caregiving and child development. Clarke-Stewart and her
colleagues (Clarke-Stewart & Gruber, 1984) examined four types of care in
Chicago: centers, nursery schools, family day care homes, and in-home
care. The results from this study are described fully in Chapter 2, but
briefly Clarke-Stewart concludes that the various types of care present
children with qualitatively distinct environments, ranging from home-like
to institutional settings with varying degrees of exposure to other children
and to educational programs. Very few results were uniform across the
four different types of care examined in the Chicago Study.
Staff stability. Developing secure attachment relationships is among
the most important developmental tasks for young children. Evidence is
clear that children in child care do not replace their attachments to their
parents with attachments to their child care providers (Ainslie & Ander-
son, 1984; Farran, Burchinal, Hutaff, & Ramey 1984; Kagan, Kearsley &
Zelazo, 1978). At the same time, however, children do get attached to their
caregivers (Ainslie & Anderson, 1984; Cummings, 1980; Ricciuti, 1974) and
use them as a secure base during the day.
Attachment formation is based in part on the availability and predict-
ability of the caregiver. The loss of an attachment figure can be very
painful to a young child. When these observations are juxtaposed with the
40% annual turnover among center-based child care providers and 60%
turnover among home-based providers (NAEYC, 1985), there is tremen-
dous cause for concern.
Research on infant and toddler care suggests that very young children
differentiate between stable and nonstable caregivers. Rubenstein and
Howes (1979) found that twice as much interaction took place in center
care between infants and head teachers as between infants and less stable
volunteers. Cummings (1980) observed infants during their morning entry
into center-based child care. Infants were less resistant to transference
from the mother to a stable caregiver and exhibited more positive affect
when the mother left, as compared to infants who were transferred to
10
INDICATORS OF QUALITY IN CHILD CARE
nonstable caregivers.
Howes and Stewart (1987) found that infants and toddlers (age range 11
to 30 months) who had experienced more changes (number of changes
ranged from none to five) in child care arrangements were less likely to
engage in competent play with peers and objects when observed in their
current family day care homes. Moreover, in a study of first-grade chil-
dren's school adjustment, the stability of prior child care arrangements
predicted academic progress (Howes, 1988).
Two studies, however, failed to find effects for caregiver stability. Benn
(1986) examined caregiver stability in family day care homes and in-home
arrangements. No association was found between the number of caregiver
changes (ranging from one to eight for boys aged 17 to 21 months) and the
quality of the mother-son attachment relationship. Everson, Sarnat, and
Ambron (1984) also examined stability in center and family day care home
arrangements and found no effects on a broad range of child competence
measures.
In sum, when stability is examined within center-based care, there
appears to be an association between the consistent presence of an adult
caregiver and infants' development in child care. In two studies in which
stability was defined as the total number of changes in child care, no
association was found. Changes in arrangements and changes in care-
givers are quite distinct measures of stability the first being far more
extensive in the degree of change involved. Clearly this is a very new area
of research with much need of further study and clarification.
Joint effects of child care and family environments
Home-rearing (no regular use of other supplemental child care arrange-
ments) has often been used as an implicit standard against which the use
of child care has been compared (McCartney & Phillips, 1988). Alterna-
tively, child care has frequently been studied as a separate socialization
environment apart from children's homes. In reality, childrearing has
become a collaborative endeavor with children moving back and forth-
many on a daily basis between their homes and child care. The effects
of these two environments may be additive; they may compensate for each
other; or some aspects of one may override aspects of the other in positive
or negative ways. A full understanding of child development thus requires
that both environments be examined.
In addition, there is an important methodological reason to assess the
joint effects of child care and family environments. Parents select their
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
11
It is likely that parents with different values, finances, and
family structures choose child care that varies in form and
quality.
children's child care arrangements. It is likely that parents with different
values, finances, and family structures choose child care that varies in
form and quality.
Howes and Olenick (1986), for example, found that families enrolling
their children in low quality child care had more complex and presumably
more stressful lives than the families using higher quality care. Moreover,
both parents and caregivers of the children in low quality centers were less
involved and invested in assuring that their children complied with their
requests. This evidence demonstrates that family and child care environ-
ments are not independent, making it difficult to attribute child outcomes
exclusively to child care or exclusively to family factors. Efforts to tease
apart these two realms of influence require, of course, that measures of
each be included in research designs. In the absence of this approach the
effects of child care on child development may be overestimated (Howes &
Olenick, 1986).
Clarke-Stewart and Gruber (1984), for example, report that associations
found between children's competence and features of their child care
settings, such as group composition and caregiver characteristics, were
substantially weakened when variance due to family socioeconomic status
(SES) was removed. Kontos's work relating regulatable characteristics of
child care centers to quality and children's development similarly revealed
that family background variables (SES and family values) significantly
predicted developmental outcomes, while structural characteristics of the
centers made virtually no contribution to development outcomes (Kontos,
1987; Kontos & Fiene, this volume, pp. 57-80). Goelman and Pence
(1987) have also reported that family variables superseded center
quality variables in predicting child language outcomes in a large study
of child care in Canada, whereas quality variation in family day care
homes was a significant predictor of children's language development.
Alternatively, substantial effects of the quality of children's child care cen-
ters remained in the Bermuda Study (McCartney, 1984; Phillips,
McCartney, & Scarr, 1987) after the influence of the parents' childrear-
ing values was statistically removed.
One possible explanation for this disparate pattern of results concerns
the relative range of variation in the family variables versus program
quality variables. When the range of families included in the research is
more extreme than the range of quality represented by the child care
programs, family factors emerge as the more salient influence, whereas
the opposite pattern of results appears to emerge when an ample range of
child care quality arrangements is included (see subsequent chapters and
12
INDICATORS OF QUALITY IN CHILD CARE
especially Chapter 7 by Clarke-Stewart for further discussion). This
finding has led several investigators to recognize the importance of
examining the interrelations between family and quality measures rather
than treating them as independent influences on child development.
The Chicago Study provides an example of examining interactive
effects. Clarke-Stewart (1984) presents evidence that a combined measure
of toys in the home and in child care was more predictive of child
development than measures that reflected only the home or child care
environments. Howes and Olenick (1986) also found that analyses that
incorporated both child care and family influences were more predictive
of several child outcomes (e.g., compliance in girls, task resistance in boys)
than analyses that took into account only one set of factors.
In a longitudinal study of children in home care, family day care, and
center care in Sweden, Cochran (1977), Gunnarsson (1978), and Cochran
and Robinson (1983) examined the interaction of structure and process
variables in both in-home and out-of-home child care settings. Children's
scores on the Griffiths Scale of Mental Development were strongly
influenced by the interaction of child care structure variables (type of
care), child care process and family process variables (social interactions
with caregivers and peers), family structure variables (maternal marital
status), and child sex. A factor that somewhat restricts the generalizability
of these findings is that the subject pool was characterized by unequal
proportions of children from one-parent families in the family day care
(8%) and center care groups (33%). Nonetheless, this study represents one
of the few attempts to examine systematically the interaction of child care
and family variables. The investigators concluded that while previous
studies
have tended to view day care as an independent, causal agent operating on the
lives of young children the day care experience is better conceptualized as an
intervening variable [their emphasis] which mediates certain family types on the
one hand (two working parents, single parent) and long-term developmental
outcomes on the other. (Cochran & Robinson, 1983, p. 61)
Recently assessments of family influences that relied on broad socio-
economic classifications, childrearing values, and measures of the home
environment have been supplemented by more subtle, but perhaps more
directly pertinent measures of maternal attitudes toward the use of child
care (Everson, Sarnat, & Ambron, 1984; Hock, 1984; Hock, DeMeis, &
McBride, 1987). Hock has presented convincing evidence that mothers'
attitudes about separation from their children are associated with differ-
ent patterns of child care use. Employed mothers who have children
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
13
It is important to understand parental feelings and attitudes
when assessing the effects of different types of care on children.
enrolled in child care centers are significantly less concerned about the
consequences of maternal separation than are employed mothers who use
other forms of care. Hock concludes that it is important to understand
parental feelings and attitudes when assessing the effects of different
types of care on children (Hock, DeMeis, & McBride, 1988).
This approach was taken by Everson and his colleagues (Everson, 1981:
Everson, Sarnat, & Ambron, 1984). They examined the mediating intlu-
ence of mothers' positive or negative disposition to use child care on
children's adjustment to child care in both center and family day care
arrangements. After the children - all toddlers - had been enrolled in
child care for 5 months, the results suggested that the congruence
between maternal attitudes and use of child care was a highly significant
predictor of children's adjustment. Mothers who relied on child care but
were uncomfortable with it and mothers who felt comfortable with the use
of child care but were not using it (called inconsistent mothers) had
children who were more easily upset by a frustrating task, showed greater
distress at maternal separation, and were less compliant with their
mothers' requests while playing. The inconsistent mothers were also
quicker to become angry and impatient with their children.
After 10 months in child e are, a different picture emerged. Attitude-
behavior consistency was no longer the issue. Attitudes alone predicted
child outcomes. Specifically mothers who were positively disposed toward
the use of child care, compared to those who were negatively disposed,
had children who were less cooperative with adults, were less compliant
with their mothers, and displayed inferior approaches to a problem-solving
task, regardless of whether they were in child care or not. Everson
concludes, "The specific effects of day care may depend in large measure
on maternal attitudes toward day care and other family characteristics"
(Everson, Sarnat, & Ambron, 1984, pp. 90-91).
In sum, the combined effects of child care quality and type. the
children's child care experience, and their family context need to be
considered in future studies of child care. It is entirely possible that family
factors (such as parental attitudes about the use of child care) mediate
child care choices that, in turn, have differing effects on children. At the
very least, inclusion of family-related measures in the study of child care
drives home the complexity of identifying where, when, and how quality of
care makes a difference in the lives of children.
14
INDICATORS OF QUALITY IN CHILD CARE
Conclusions
Research on child care quality has accumulated a vast collection of
results during the last 10 years. More than any other aspect of child
development research, this literature has driven home the true complexity
of child care and the real challenges faced by those who seek to assess its
effects on children.
The first challenge for researchers involves selecting a measure of
quality - global or discrete, regulatable or more dynamic, a static snap-
shot measure or one that captures children's and caregivers' movement in
and out of child care. The second challenge involves measuring other
factors, particularly aspects of the family environment, that affect child
development and may interact with, compensate for, or operate com-
pletely independently of the influence of child care quality.
In the chapters that follow, five groups of researchers present their
efforts to confront these challenges and the results their work has yielded.
They have examined different types of child care, in different locations in
and out of the United States, and with different populations of children
and families. The measures of quality used in these collective studies
encompass the full range of options described in this review. Each study
placed the developmental consequences of variation in the quality of child
care environments in the context of children's home environments. Alison
Clarke-Stewart then addresses the central question of how the results of
these recent studies confirm, contradict, and extend those of the research
reviewed here.
References
Ainslie, R.C., & Anderson, C.W. (1984). Day care children's relationships to their
mothers and caregivers: An inquiry into the conditions for the development of
attachment. In R.C. Ainslie (Ed.), Quality variations in daycare (pp. 98-132).
New York: Praeger.
Arnett, J. (1987, April). Training for caregivers in day care centers. Paper
presented at the biennial meeting of the Society for Research in Child
Development, Baltimore, MD.
Benn, R. (1986). Factors promoting secure attachment relationships between
employed mothers and their sons. Child Development, 57, 1224-1231.
Berk, L. (1985). Relationships of educational attainment, child-oriented attitudes,
job satisfaction, and career commitment to caregiver behavior toward chil-
dren. Child Care Quarterly, 14, 103-129.
Bruner, J. (1980). Under five in Britain. Ypsilanti, MI: High/Scope.
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
15
Childrearing has become a collaborative endeavor with children
moving back and forth between their homes and child care.
16
INDICATORS OF QUALITY IN CHILD CARE
Carew, J. (1980). Experience and the development of intelligence in young
children. Monographs of the Society for Research in Child Development, 45
(6-7. Serial No. 187).
Clarke-Stewart, A. (1984). Day care: A new context for research and development.
In M. Perlmutter (Ed.), The Minnesota Symposium on Child Psychology, Vol.
17 (pp. 61-100). Hillsdale, NJ: Erlbaum.
Clarke-Stewart, A., & Fein, G.G. (1983). Early childhood programs. In M. Haith &J.
Campos (Eds.). Handbook of child psychology: Vol. 2. Infancy and develop-
mental psychobiology (pp. 917-1000). New York: Wiley.
Clarke-Stewart, K.A., & Gruber, C. (1984). Daycare forms and features. In R.C.
Ainslie (Ed.), Quality variations in daycare (pp. 35-62). New York: Praeger.
Cochran. M.M. (1977). A comparison of group day and family childrearing pat-
terns. Child Development, 48, 702-707.
Cochran. M.M., & Robinson, J. (1983). Day care, family circumstances and sex
differences in children. In S. Kilmer (Ed.), Advances in early education and
day care (Vol. 3, pp. 46-67). Greenwich, CT: JAI Press.
Cummings, E.H. (1980). Garegiver stability and day care. Developmental Psy-
chology, 16, 31-37.
Cummings, M., & Beagles-Ross, J. (1983). Towards a model of infant daycare:
Studies of factors influencing responding to separation in daycare. In R.C.
Ainslie (Ed.), Quality variations in daycare (pp. 159-182). New York: Praeger.
Everson, M.D. (1981). The impact of day care on the attachment behavior of 12
to 24 month olds. Unpublished doctoral dissertation, Stanford University Palo
Alto, California.
Everson. M.D., Sarnat, L., & Ambron, S.R. (1984). Day care and early socialization:
The role of maternal attitude. In R.C. Ainslie (Ed.), Quality variations in
daycare (pp. 63-97). New York: Praeger.
Farran, D.C., Burchinal, M., Hutaff, S.E., & Ramey C.T. (1984). Allegiances or
attachments: Relationships among infants and their day care teachers. In R.C.
Ainslie (Ed.), Quality variations in daycare (pp. 133-158). New York: Praeger.
Field, T. (1980). Preschool play: Effects of teacher child ratio and organization of
classroom space. Child Study Journal, 10, 191-205.
Francis, P., & Self, P. (1982). Imitative responsiveness of young children in day care
and home settings: The importance of the child to caregiver ratio. Child Study
Journal, 12, 119-126.
Goelman, H., & Pence, A. (1987). Some aspects of the relationships between
family structure and child language development in three types of day care. In
D. Peters & S. Kontos (Eds.), Annual advances in applied developmental
psychology, Vol 2: Continuity and discontinuity of experience in child care
(pp. 129-146). Norwood, NJ: Ablex.
Golden, M., Rosenbluth, L., Grossi, M., Policare, H., Freeman, H., & Brownlee. M.
(1978). The New York City Infant Day Care Study. Medical and Health
Research Association of New York City 40 Worth St., New York, NY 10013.
Gunnarsson, L. (1978). Children, day care and family care in Sweden: A
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
17
follow-up. Gothenburg, Sweden: University of Gothenburg.
Harms, T., & Clifford, R.M. (1980). Early Childhood Environment Rating Scale.
New York: Teachers College Press, Columbia University.
Hock, E. (1984). The transition to day care: Effects of maternal separation anxiety
on infant adjustment. In R.C. Ainslie (Ed.), The child and the day care setting
(pp. 183-206). New York: Praeger.
Hock, E., DeMeis, D., & McBride, S. (1988). Maternal separation anxiety: Its
role in the balance of employment and motherhood in mothers of infants. In
A. Gottfried & A. Gottfried (Eds.), Maternal employment and children's
development: Longitudinal research. New York: Plenum.
Hofferth, S.L., & Phillips, D.A. (1987). Child care in the United States, 1970 to 1995.
Journal of Marriage and the Family, 49(3), 559-571.
Howes, C. (1983). Caregiver behavior in center and family day care. Journal of
Applied Developmental Psychology, 4, 99-107.
Howes, C. (1988). Relations between early child care and schooling. Developmental
Psychology, 24, 53-57.
Howes, C., & Olenick, M. (1986). Family and child care influences on toddlers'
compliance. Child Development, 57, 202-216.
Howes, C., & Rubenstein, J. (1985). Determinants of toddlers' experience in
daycare: Age of entry and quality of setting. Child Care Quarterly, 14,
140-151.
Howes, C., & Stewart, P. (1987). Child's play with adults. toys and peers: An
examination of family and child care influences. Developmental Psychology,
23, 423-430.
Kagan, J., Kearsley R.B., & Zelazo, P.R. (1978). Infancy: Its place in human
development. Cambridge, MA: Harvard University Press.
Kontos, S. (1987, April). Day care quality, family background and children's
development. Paper presented at the biennial meeting of the Society for
Research in Child Development, Baltimore, MD.
McCartney K. (1984). The effect of quality of day care environment upon
children's language development. Developmental Psychology, 20, 244-260.
McCartney, K., & Phillips, D. (1988). Motherhood and child care. In B. Birns &
D. Hay (Eds.), Different faces of motherhood. New York: Plenum.
McCartney, K., Scarr, S., Phillips, D., & Grajek, S. (1985). Day care as intervention:
Comparisons of varying quality programs. Journal of Applied Developmental
Psychology, 6, 247-260.
National Association for the Education of Young Children. (1985). In whose
hands? A demographic factsheet on child care providers. Washington. DC:
NAEYC.
Phillips, D., Scarr, S., & McCartney K. (1987). Child care quality and children's
social development. Developmental Psychology, 23, 537-543.
Ricciuti, H. (1974). Fear and the development of social attachments in the first
year of life. In M. Lewis &L. Rosenblum (Eds.), The origins of human behavior.
Fear (pp. 73-106). New York: Wiley
18
INDICATORS OF QUALITY IN CHILD CARE
Rubenstein, J.L., & Howes, C. (1979). Caregiving and infant behavior in day care
and in homes. Developmental Psychology. 15, 1--24.
Rubenstein, J., Howes, C., & Boyle, P. (1979). A two year follow-up of infants in
community based day care. Journal of Child Psychology and Psychiatry, 22,
209-218.
Ruopp, R., Travers, J., Glantz, F., & Coelen, C. (1979). Children at the center: Final
results of the National Day Care Study. Cambridge, MA: Abt Associates.
Rutter, M. (1981). Social-emotional consequences of day care for preschool
children. American Journal of Orthopsychiatry, 51, 4-28.
Smith, P., & Connolly K. (1981). The behavioral ecology of the preschool.
Cambridge, England: Cambridge University Press.
Stallings, J., & Porter, A. (1980). National Daycare Home Study. Palo Alto, CA:
Stanford Research Institute.
Stith, S.M., & Davis, A.J. (1984). Employed mothers and family day care: A
comparative analysis of infant care. Child Development, 55, 1340-1348.
U.S. Bureau of the Census. (1983, November). Child care arrangements of
working mothers: June 1982. Special Studies, Series P-23, No. 129. Washing-
ton, DC: U.S. Department of Commerce.
Vandell. D.L., & Powers, C.P. (1983). Day care quality and children's free play
activities. American Journal of Orthopsychiatry, 53, 493-500.
Zigler, E., & Gordon, E. (1982). Day care: Scientific and social policy issues.
Boston: Auburn House.
DEBORAH A. PHILLIPS AND CAROLLEE HOWES
19
Children whose development was advanced not only had the advan-
tage of being in high quality child care programs but also came from
families who gave them support, stimulation, and education.
As a follow-up to the discussion at the meeting on child care quality on November 25, here is
information on the research on child care quality and children's development:
The review by Deborah Phillips and Carollee Howes covers the literature through the
1980's.
John Love's 1996 review includes more recent studies. It is a very comprehensive
review of the developmental literature. However, despite the title, "Are They in Any
Real Danger", the implications of basic risks to children's health and safety are not
focused on in these developmental studies. Developmental studies have tended to look
beyond these basics to focus on how quality can promote children's learning and
development.
The omission of health and safety issues also came up in the development of the CEA
report on the economics of child care. There is some research on this topic, but it is not
as accessible. ASPE provided the CEA with some information, and a copy of these
comments also is enclosed. The comments are best read in conjunction with the revised
report which has added the health and safety topic.
ASPE Comments on "The Economics of Child Care: A Draft Report"
We appreciate the opportunity to review this draft report. In a very short time, the CEA has
worked hard to synthesize a sizeable and complex literature from multiple disciplines. Though
much has been accomplished, we see a number of serious gaps and issues that are important to
address before the paper is finalized. Our most serious concern with this report is that it does
not address the cost to society of not ensuring the health, safety, and development of children
in child care. Interestingly, Ron Haskins addressed this issue directly in a 1989 article which
is included in the enclosures.
We also recommend that the CEA obtain additional outside review by leading child care
researchers. From our efforts to help with the development of the paper, we know that the
analysis has relied heavily on reference materials that could be gathered very quickly. We are
concerned that important studies and domains of research (health and safety issues for children)
may be omitted or not covered adequately. Also, the representation of forth coming work be
very selective. We also think that vetting the paper is an important part of the process of
developing a report of this kind.
We believe that there is a stronger case to be made for the importance of investing in
our nation's children (see R. Haveman and B. Wolf, 1994, Succeeding Generations: On
the Effects of Investment in Children) and for the value of providing quality child care
that protects children's health and safety and promotes children's learning and
development (see below).
The issue of harm to children from child care that fails to protect children's health and
safety must be given more serious consideration. What is known about the costs of
failing to protect children's health and safety is not represented here. The recent news
article on D.C. child care programs that expose children to rats, roaches, filth,
overcrowding, open windows, and poisons is a vivid illustration of these problems. We
think this is a critical omission and have marked places throughout the text where it
needs to be added or expanded on. We also are providing reference materials, but
believe there is more on this than in the health literature that needs to be considered.
Important findings include the following:
A 1986 study (Bell, Gleiber, Mercer, Phifer, Gunter, Cohen, Epstein, and Narayanan)
[AJPH April 1989, Vol 79, No. 4], which is attached, identified characteristics of day
care which might be risk factors for infection and described the resulting costs. The
following conclusions were derived from this study:
Children in day care centers were 4.5 times more likely to be
hospitalized than those in other settings.
Data indicated that there are certain potentially modifiable characteristics
of day care which are risk factors for illness and that excess illnesses
result in excess financial costs.
1
The mean monthly cost of medical care was $32.94 for children in the
highest risk settings, compared with $19.78 for those in other settings.
Illness of a child accounted for 40% of parental absenteeism from work.
Ron Haskins, in "Acute Illness in Day Care: How Much Does it Cost?" (1989),
which is attached, reported that preschool children attending day care have more
acute illnesses than children reared at home. This excess illness imposes costs
on both families and society.
Costs were divided into three parts: medical costs of treating excessive
acute illness, cost of missed work, and cost of long-term effects that
remain after the acute phase of the illness has passed.
Excess day care illness imposes at least $1.8 billion per year in costs on
American families and society. Over two-thirds, or $1.3 billion, is
attributed to missed work by parents.
Only about 70% of the employees of medium and large firms have paid
sick leave; small firms, where a disproportionate number of women
work, probably have even lower rates of coverage of sick leave. Many
mothers, then, lose money when they are forced to care for ill children
or when they themselves are ill. We would note that this may also put
their jobs at risk.
Note that the review of the child development literature on quality tends to miss these
issues because this research has focused primarily on the question of whether higher
quality care enhances or optimizes children's development and learning. It is also
important to note the methodological issues here. Although there is some
representation of the range of quality of care settings, especially in the large-scale
studies, there also are indications that the poorest quality settings may be under
represented in research studies (from providers refusals and other difficulties in
sampling providers-- see the NRC volume, Child Care for Low-Income Families,
edited by D. Phillips, which covers this).
A related issue is the question of whether some children are differentially affected by
the quality of care. There is some evidence to suggest children from disadvantaged
backgrounds benefit the most from high quality programs. However, children who are
most in need of highly supportive care also may be least likely to get it (see the Helbrun
and Howes article which is already referenced and also the NRC volume, Child Care
for Low-Income Families, edited by D. Phillips which we are enclosing).
Continuity of care is one of the attributes of care that we know is important for
children's development and for mothers' employment. It has a number of implications
for costs and benefits. For example, training, wages, and benefits for providers are
related to job turnover and disruptions in care relationships for children. Informal care
arrangements are also associated with more changes in care arrangements. These
2
changes in care arrangements are associated with disruption in mothers' work force
participation as well as with disruptions in children's relationships with care givers.
We noted that the continuity of care issue is omitted in the review of the literature and
throughout the policy sections.
The report does not make the connection to the White House Early Learning and
Development Conference and the implications of the brain research. The major
message of this conference is that society cannot afford to waste the opportunity of
early childhood and that child care is one of the critical settings for acting on this
opportunity. The CEA paper developed for that conference does not seem to be
referenced.
We suggest indicating that there are special issues for infants and toddlers. They are
especially vulnerable because of their dependence on care givers and susceptibility to
harm, and they also are more likely to be receiving substandard care. This also is an
area where the child care rates have risen rapidly, and under welfare reform, many
states are for the first time implementing policies requiring mothers with very young
infants to meet participation requirements.
We did not see the benefits of enabling mothers to work included in the discussion of
external benefits. The NRC volume edited by D. Phillips, Child Care for Low-Income
Families, covers this.
We suggest using more caveats. We've noted that some broad statements are based on
a single study. Also, strong causal language is frequently used in summarizing the
literature. We are not certain that the referenced studies are able to demonstrate causal
effects. One example is page 23, first paragraph in the section "Effects of subsidies
where results of research by Blau and Hagy are described. The one study we know of
which has examined actual changes in regulations -- C. Howes, E. Smith, & E.
Galinsky, (1995), The Florida Child Care Quality Improvement Study, New York:
Families and Work Institute -- does not appear to be considered here. We are enclosing
a copy of this report.
More specific concerns:
On page 10, the relationship between child care and early education programs is more
complex than indicated in the text. Many times they are one in the same; some early
childhood intervention programs have included a full-day high quality child care
component, and some child care programs deliver high quality child development
services. A number of the reference sources you are consulting lay out the close
relationship between these kinds of programs and the linkages in the research literature.
Note also that economic studies of "child care" subsidies frequently include Head Start.
3
There are a number of blanket statements about the absence of a literature showing
relationships between specific attributes (including the last bullet in the Exec.
Summary). In other places, there is a more nuanced discussion, and we encourage this
treatment. We think it is important that the treatment be consistent and not overstate
the lack of information. Beyond the studies reference here, there is additional research
on specific attributes of care, including structural attributes, that show relationships to
children's learning and development. Most reviewers of this literature have concluded
that structural attributes provide the basic conditions which enable good care giver
interactions to take place. However, as noted above, most of the literature focuses on
what we know about the attributes that enhance children's learning and development
rather than on those that protect basic health and safety. The basic protections for
health and safety need to receive specific consideration. Finally, it is somewhat
misleading to focus on isolating effects of single attributes of care; characteristics of
care are closely interrelated. This is acknowledge in some places, but needs to done
more systematically.
We hope these comments are helpful to you. See also the enclosed copy of the report with
marginal notes and the additional reference materials.
enclosures
4