Quality in ECE

Key Findings

• Key Finding 1

  High-quality early education lacks a single definition, but it can be conceptualized in terms of structural inputs, interactions between students and teachers, and instructional quality.

  Structural inputs include factors such as class sizes, child–teacher ratios, teacher education/experience, and teacher compensation. Interactional quality includes emotional support and organization, while instructional quality refers to what is taught and how. Structural quality is typically the easiest of the dimensions to address via policy and to monitor.

  Explore the evidence
• Key Finding 2

  Just as in K-12, the quality of children’s experiences in early education programs varies widely.

  Most licensed programs in the U.S. have solid structural factors such as class sizes, child–teacher ratios, and screenings, but there is variation across states. Compensation, education, and training requirements are typically lower for early education teachers than for K-12 teachers. In the preschool years, average quality is good in terms of classroom organization and emotional climate, but instructional quality tends to be lower on average (i.e., at just above or just below adequate). Programs for younger children (0–3) tend to be of lower quality, on average, than those for preschoolers.

  Explore the evidence
• Key Finding 3

  Structural quality inputs usually improve interactional and instructional quality, but they do not always do so.

  Factors such as smaller class sizes and better-trained teachers make a higher-quality learning experience more likely, but these factors alone are not sufficient to guarantee such a learning experience.

  Explore the evidence
• Key Finding 4

  An evidence-based curriculum implemented with coaching appears to be a “good bet” for improving instruction in preschool classrooms and children’s readiness for kindergarten.

  Curricula that are focused on particular domains of child learning and that have a scope and sequence have repeatedly outperformed curricula that purport to cover all skill areas but that lack a scope and sequence. This is particularly the case when curricula are supported by regular in-classroom mentoring by an expert mentor. However, these more effective curricula are not commonly used.

  Explore the evidence
• Key Finding 5

  More marginalized groups have less access to high-quality early education programs

  Such groups include children from families with low incomes, children of color, children who speak a home language other than English, and children with disabilities.

  Explore the evidence
• Key Finding 6

  Policy-level efforts to regulate and improve early education quality through licensing standards, accountability systems, and accreditation have had some success in improving many dimensions of quality.

  Explore the evidence

Introduction

Early education programs improve children’s readiness for kindergarten, and their effects can—but do not invariably—last into elementary school and beyond.¹ This evidence, alongside the practical realities of family life in the U.S. today, has led to a focus on identifying the active ingredients that best ensure high-quality experiences for young children that best promote their development.²

In this chapter, we explain how quality in early education programs is typically defined, review evidence on both summary measures of quality and specific quality elements/inputs, and describe the track record of policy efforts to promote quality in early education settings. We focus primarily on center-based, publicly funded early education settings for preschoolers (aged 3–5). However, we include evidence, when available, on care for infants and toddlers 0–3 where appropriate and on other kinds of care types.

Just as in K-12, the quality of children’s experiences in early education programs varies widely. Also as in K-12, young children learn more in higher-quality programs. To date, large-scale, policy-level efforts to improve quality through accountability and incentives have increased quality along multiple dimensions, but there is scant evidence that these efforts have also improved child outcomes. The field has multiple evidence-based “good bets” for improving quality in existing programs and for creating/scaling new programs, such as implementing targeted curricula supported by coaching. However, these “good bets” have not yet been widely implemented.

Evidence

Key finding #1: High-quality early education lacks a single definition, but it can be conceptualized in terms of structural inputs, interactions between students and teachers, and instructional quality.

Quality in early childhood education has generally been conceptualized along two dimensions: structural quality and process quality. Structural quality refers to foundational inputs such as class sizes, teacher-child ratios, teacher training and education requirements, teacher compensation, and the length of the instructional day/year.⁴ Process quality refers to interactional quality, meaning the warmth of teacher–child relationships and peer relationships, as well as classroom instructional quality.⁵ Based on conceptual frameworks/theory and available research, researchers have commonly referred to structural quality as setting the stage for high process quality to occur; however, it does not guarantee that it will occur.⁶

Recently, a team of researchers offered a new conceptual model of process quality in the preschool years specifically, with a delineation of interactional quality and instructional quality.⁷ These researchers’ working definitions for these dimensions are as follows:

• “INTERACTIONAL QUALITY refers to the quality of children’s interactions with teachers and other children in the classroom and the ways in which the classroom climate is positive, responsive, and predictable. It includes teachers’ warmth and sensitivity, as well as their overall classroom management and organizational skills.

• INSTRUCTIONAL QUALITY refers to what is being taught in the classroom and how.”

In Table 1, we summarize specific elements of quality often hypothesized as subdomains of these larger dimensions in research⁸ and often emphasized in early education standards and rating systems such as those of the National Association for the Education for Young Children (NAEYC) and the National Institute for Early Education Research (NIEER). We intend this table to be illustrative only and to focus on the preschool years. A fully exhaustive list of all potential quality inputs for all early education ages is beyond the scope of this review. Note, too, that for simplicity of presentation in Table 1, we combine process and instructional quality but differentiate them in our evidence discussion following recent developments in the field.⁹ Process/instructional quality inputs refer to those hypothesized to more directly affect the teaching and learning environment experienced by children. Structural elements have more indirect links.

More recent research has also highlighted cultural responsiveness and inclusion for students with disabilities as understudied but potentially important quality inputs.¹⁰ Cultural responsiveness captures the degree to which children’s home cultures are reflected and focused on in classroom practice.¹¹ Inclusion refers to the degree to which children with disabilities are fully included in learning experiences and the classroom community.¹² In Table 1, we group cultural responsiveness as a process element and inclusion as a structural element, but we acknowledge that inclusion encompasses teaching strategies that are very much process oriented.¹³

Regarding measurement, structural quality elements have proven much easier to assess than other quality dimensions.¹⁴ Many structural quality inputs are commonly available in administrative data or in program documents, or they can be obtained via surveys of teachers or administrators. Other quality elements have typically required observations of classrooms in action by trained, reliable observers who rate the classrooms according to the criterion of the focal measure(s). Researchers sometimes use survey data as proxies for process quality. For example, studies have asked teachers about how often they teach certain content or how much time they spend on particular learning formats.¹⁵

In general, measures of these quality inputs have focused on the whole-class level, averaging across individual children’s experiences. However, research has shown that quality can vary across individual young children in the same preschool classroom and by child demographics characteristics.¹⁶ The inclusion of variation at the individual child and demographic subgroup levels accordingly is a newer direction in the field.¹⁷

• Table 1

  Example preschool quality elements/inputs

  Structural quality elements	Process/instructional quality elements
  Compensation	Curriculum
  School day and year length	Teacher coaching
  Class size	Assessments
  Teacher-child ratios	Alignment with Kindergarten
  Inclusion for Students with disabilities	Cultural Responsiveness
  Universal Screening
  Age mix (i.e., 3's and 4's versus single-age classrooms)
  Suspension/expulsion policies
  Protected time for teacher planning

  Footnotes

  This table is based on Chaudry et al. (2021) and Weiland et al. (2023).¹⁸

Key finding #2: Just as in K-12, the quality of children’s experiences in early education programs varies widely.

• Unsurprisingly given the fragmented approach to early education in the U.S., programs vary widely in quality. To illustrate some of the policy variation that helps drive this variation, in Figure 1, we display quality proxies for 60 state-funded PreK programs in the 2022–2023 school year. As shown,

• The vast majority of programs (>90%) have early learning standards; have some kind of curriculum approval process and require child assessment.

However, there is considerable variation in other inputs and markers of quality. For example, in 2023, just over half of the 60 state-funded Pre-K programs required a BA minimum for lead teachers; just over 30% required pay parity between state Pre-K teachers in public schools and K-12 teachers; 44% provided a list of approved curricula for programs to choose from; and 29% required regular coaching for Pre-K teachers.

As discussed below, these proxies are imperfect. Many state curriculum lists, for example, include curricula that have been regularly outperformed by other options.¹⁹ Nonetheless, these data help illustrate how programs in the same category—state-funded Pre-K —differ greatly across the country in their policies and implementation.

Just like state-funded Pre-K programs, community-based providers, private providers, and family child care homes vary widely in their policies and quality inputs.²⁰ Overall, Head Start is more uniform in its policies than state-funded Pre-K programs are, but there is variation in important factors such as dosage and teacher education in Head Start.²¹ Some of the variation in Head Start is planned, according with its mission of being responsive to local needs.

• Table 2

  Quality metrics of state-funded Pre-K programs

  	No. of
  programs"
  Early learning and development standards	58
  Curriculum approval process and supports	56
  Child assessment	57
  Teacher-child ratio of 1:10 or lower	48
  Teacher has specialized Pre-K training	47
  Class size of 20 or lower	46
  Continuous quality improvement	42
  Screening and referral	41
  Five days per week	37
  BA minimum for teachers	31
  Pay parity: public school Pre-K and K-12	28
  Assistant teacher CDA	19
  Teacher and staff 15 hours training/year; individual PD plan; coaching	19
  Full day	18
  List of approved domain-specific curricula	13
  Pay parity: CBO Pre-K and K-12	6

  Footnotes

  The number of programs meeting standards for the NIEER's 10 benchmarks come from the NIEER 2023 State Preschool Yearbook. Other quality indicators were taken from the 2022 NIEER State Preschool Yearbook appendices (the 2023 NIEER State Preschool Yearbook Appendices were not yet available when this table was compiled).

Unsurprisingly given this policy variation in inputs, there is variation in the quality of young children’s experiences within early learning programs. To illustrate this variation, in Figure 2, we display the quality scores for four large-scale systems from two different observational tools in classrooms serving (4-year-old) children, preschool-age children (3- and 4-year-olds), or both across sectors (see the figure notes for details) On the two measures shown, “3” is considered adequate, and “5” is good. On the largely structural measure of quality shown in the figure (scores from the Early Childhood Environmental Rating Scales (ECERS)),²³ these large-scale programs averaged approximately 4. The process quality scores on the Classroom Assessment Scoring System (CLASS)²⁴ were more variable, but for two dimensions (Emotional Support and Classroom Organization), the systems were close to or over 5. For Instructional Support, the scores were much lower, with only Boston’s Pre-K program averaging over 4.

Overall, the wide variation in policies and in the quality of experiences of young children in early learning settings necessitates nuance in discussions of early care and education (ECE). The models for delivering ECE vary greatly in the U.S. When weighing the evidence on the effects of public preschool, for example, it is important to be clear regarding what kind of preschool program children experienced.

• Figure 1

  Observational quality scores in four large-scale preschool systems

  

  Footnotes

  This figure appeared in our research brief on preschool quality measurement.²⁵ Site data come from Boston Public Schools school-based Pre-K classrooms for 4-year-olds (N=83 classrooms);²⁶ California center-based arrangements including Head Starts, public and private preschool, and private child care centers serving 3- and 4-year-olds (N=248 classrooms);²⁷ New York universal (UPK) providers including school-based Pre-K, independent Pre-K centers, and New York City Early Education Centers for 4-year-olds (N=1,869 providers);²⁸ and a national sample of 370 Head Start classrooms serving 3- and 4-year-olds.²⁹

Key finding #3: Structural quality inputs usually improve interactional and instructional quality, but they do not always do so.

Having explained how quality is defined and how it varies in the U.S., we now turn to the decades of research that has examined the relationship between the quality of ECE settings and the learning and development of young children. This research has included studies of summary measures of quality and individual quality inputs. Overall, this research has mainly been descriptive and correlational, but as we summarize, there have been randomized controlled trials (RCTs) of some elements, particularly preschool curricula.

One challenge in this area of research is that in real-world programs, inputs tend to be bundled together—e.g., teacher pay and teacher education tend to be linked—making it difficult to determine their separate value add. Even when a study can isolate the effects of a particular input, it often cannot answer how the effects of that input may vary by levels of other important inputs. For example, due to policy in a given program, all lead Pre-K teachers may have to have, at minimum, a BA. Accordingly, a study of preschool class size in such a program cannot identify whether the effects of having a smaller class vary based on teacher education level. This real-world constraint matters because in regard to preschool quality elements, synergistic effects are likely—that is, the whole may be more than the sum of the individual parts.

With this important context front and center, we turn first to the evidence on structural quality and then to that on process/instructional quality.

Structural quality evidence

Summary measures of structural quality. For research on summary measures of structural quality, many studies use composite scores, subscale scores, or factor analysis-derived scores³⁰ from the ECERS.³¹ Across these studies, in Head Start programs, community-based preschool programs, and Pre-K programs in public schools, the ECERS-R total score and structural quality scales/composites have shown a mix of small positive and null associations with preschool children’s gains in language, literacy, mathematics, executive function (EF), and social-emotional skills.³² There has been no clear pattern by child skill domain. For example, although an 11-state PreK study found that higher total ECERS-R scores predicted child vocabulary gains (but not literacy, mathematics, or social-emotional skill gains), work in Boston Pre-K³³ and San Francisco Pre-K³⁴ found null associations for child language measures in terms of both the ECERS-R total score and subscale composites of structural measures only. However, this evidence is entirely correlational and may reflect the roles played by omitted factors rather than these structural measures of quality.

Specific structural quality inputs

Preservice teacher education and training. The evidence linking preservice teacher education and training to overall quality and children’s learning is correlational and mixed. As researchers have pointed out, studies of community child care programs in the 1990s generally found consistent links between higher teacher education and both higher classroom quality and higher child school readiness skills.³⁵ However, since the 2000s, as both public Pre-K programs and teacher training programs expanded, the evidence has been more mixed. Some studies have shown links between higher levels of preschool teacher education and higher instructional quality, but the findings of other studies have mostly been null.³⁶ Similarly, regarding links between teacher education and gains in children’s kindergarten readiness skills, some studies suggest benefits, whereas others obtain null findings.³⁷ This recent, more mixed pattern is generally found in state-funded PreK programs (in both public schools and community-based settings) and in Head Start programs, where other inputs are less variable than in child care programs, i.e., other supports may compensate for lower teacher education.³⁸

Notably, multiple leading institutions in the field recommend a BA minimum for lead preschool teachers, including the National Academies of Sciences, Engineering, and Medicine and the NIEER.³⁹ In 2022, 33 out of 62 state Pre-K programs required a BA for lead teachers and 19 required a child development associate (CDA) credential or equivalent for assistant teachers.⁴⁰ Part of the rationale for higher teacher education requirements also concerns workforce retention and compensation as well as professionalizing the field. When teacher education requirements are lower for preschool than for K-12, preschool teachers may be incentivized out of early childhood programs. That is, because education and compensation tend to be highly coupled, a preschool teacher who meets K-12 requirements is likely to move into K-12 teaching if there is a large compensation benefit from doing so.⁴¹

Teacher compensation

As the Workforce chapter highlights, early educators are frequently compensated at a considerably lower level than their K-12 counterparts, fueling turnover. This can be the case even in systems that require a BA minimum for Pre-K and K-12 teachers. For example, K-12 teachers in Michigan average $58K in annual salary. Although Michigan state Pre-K teachers are required to have a BA, until recent reforms, they were paid 31% less than K-12 teachers in public school settings ($40K). In community-based settings, Pre-K teachers were paid 43% less ($33K) than their K-12 counterparts.⁴² These kinds of pay gaps by sector and setting are common across the U.S.⁴³

Some correlational research has linked higher early educator compensation with lower turnover,⁴⁴ higher classroom quality, and higher school readiness skills.⁴⁵ Other correlational research has revealed negative links between higher teacher turnover and preschoolers’ language gains (but not literacy, mathematics, or social skill gains) in Head Start.⁴⁶ However, as others have noted, a challenge in this literature is that early educator compensation is often confounded with other program characteristics. For example, programs that pay higher wages tend to attract and hire more highly trained staff and offer more teacher supports and development opportunities.⁴⁷ The only experimental study of early educator compensation found that a relatively small financial incentive ($1,500 per teacher) led to a large drop in teacher turnover for teachers at child care centers, with no effect for teachers in school-based programs.⁴⁸ Notably, lead teachers in school-based programs earned nearly twice the annual salary of those in child care programs. The effect of the incentive on classroom quality and child gains was outside the scope of the study.

Dosage

Another important structural characteristic is dosage, i.e., whether a program is full day or half day, whether a program is full week vs. part week, and whether wraparound care is offered. Here, the correlational evidence is mixed, with two experimental studies pointing in the direction of larger benefits for children’s learning with increased preschool dosage. For example, one correlational study found that attending full-day versus half-day preschool in centers near or in public schools was associated with improved school readiness across language, mathematics, social-emotional, and physical development indicators.⁴⁹ However, a correlational Head Start study found no difference in academic and social-emotional skills by full-day vs. half-day dosage.⁵⁰ The first of the two RCTs of half-day vs. full-day preschool found larger impacts for children randomized into an 8-hour/day preschool program for 45 weeks than for children randomized into a 2.5-3 hour/day program for 41 weeks.⁵¹ However, the two groups were not equivalent at baseline, which means that the findings are only suggestive. A more recent RCT study in public school-based PreK in a Colorado district without such design flaws found larger impacts on children’s vocabulary and literacy skills for preschoolers randomized to a six-hour school day, five days per week than for those randomized to a three-hour day, four days per week.⁵² Detailed time use data from the study revealed that across the school year, children in full-day classrooms spent an estimated 177 hours more in instructional activities than did their half-day peers.⁵³

Class sizes and teacher-child ratios

For class sizes and teacher-child ratios, the correlational evidence is mixed. Some studies suggest small, positive associations between smaller class sizes and smaller teacher-child ratios and higher classroom quality and improved cognitive, language, and social emotional learning for preschoolers; others have null findings.⁵⁴ There is no clear pattern suggesting that these findings differ by sector (i.e., Head Start, child care, Pre-K programs). A small experimental study of 10 schools and 22 classrooms in Chicago Pre-K found benefits of smaller class sizes (i.e., 15 max vs. 20 max) for children’s literacy skills but no impacts on child vocabulary or mathematics, nor on classroom quality.⁵⁵

Inclusion

Preschool programs vary widely in their policies and supports for the inclusion of students with disabilities. Approximately two-thirds of preschoolers with disabilities attend preschool alongside their typically developing peers.⁵⁶ Overall, as reviewed by the lead author of this chapter, experimental and observational studies of children’s interactions in inclusive versus segregated settings have found that children with disabilities engaged in more positive peer interactions, fewer negative peer interactions, and more advanced levels of play in inclusive versus non-inclusive groups.⁵⁷ Children with disabilities appear to benefit in terms of their language, literacy, mathematics, social-emotional, and EF skills in inclusive settings.⁵⁸ Evidence also suggest that typically developing children appear to make the same gains in inclusive settings as in non-inclusive settings,⁵⁹ with some evidence of improved knowledge about and attitudes toward disabilities in inclusive settings.⁶⁰

Other structural quality elements

There is much less evidence on the other structural quality elements that we highlighted in Table 1. Moreover, the evidence that exists is largely correlational/descriptive and should be viewed with caution.

Screening: Most public preschool programs require children to be screened for disabilities and vision/hearing (i.e., 41/60 state-funded Pre-K programs did so in 2022–2023 in public school or community-based settings; all Head Start programs require it).⁶¹ Screening has long been shown to be beneficial for detecting developmental differences and needs for supports.⁶² However, we are not aware of studies that have examined differences in classroom quality or child learning by whether or not screening is required in preschool.

Mixed-age classrooms: Some programs serve 3- and 4-year-olds in the same classroom, while others serve single-year age groups only. Correlational evidence has shown lower-quality teacher–child interactions in mixed-aged classrooms in Head Start and public PreK programs.⁶³ Research suggests that 4-year-old children also appear to make fewer academic gains in mixed-aged Head Start classrooms of 3- and 4-year-olds than do children in 4-year-old-only Head Start classrooms.⁶⁴

Suspension and expulsion: Children from historically marginalized groups and children with disabilities are more likely than their peers to be suspended and expelled from early learning programs.⁶⁵ Many states have policies aimed at curbing preschool suspension/expulsion, but they have largely not been evaluated.⁶⁶

Protected planning time: Preschool and K-12 teachers in unionized public schools generally have protected planning time, while ECE teachers outside public schools do not. In a recent study of Boston UPK classrooms located in community-based settings, the lack of protected planning time was identified as a barrier to instructional quality.⁶⁷

Process and instructional quality evidence

Process quality summary measures. The most commonly used measure of process quality is CLASS.⁶⁸ Versions are available by child age group (e.g., Pre-K and school aged) and setting (e.g., schools and family child care homes). Trained assessors observe a typical day in the classroom or code from videos of a typical day. CLASS measures the quality of interactions between teachers and children and between children and their classmates in three areas: Emotional Support, Classroom Organization, and Instructional Support. Many U.S. programs use CLASS to gauge quality, in either formative or high-stakes systems. As an example of the latter, CLASS is used to allocate public funding in Head Start and in some state quality rating systems.⁶⁹

As we highlighted in a previous research brief, CLASS has several key strengths.⁷⁰ For example, because it is widely used, it provides a common metric for understanding the strengths and weaknesses of early learning programs around the country. Furthermore, CLASS can and has been used to highlight disparities in access to quality programs by sector, family income, geography (i.e., rural vs. urban), and race/ethnicity. Interventions based on the CLASS theoretical model and classroom quality measure have been shown to improve teacher practice in state-funded Pre-K in public school and community-based settings.⁷¹ However, an RCT of coaching on CLASS in Head Start, public-school Pre-K, and community-based programs found small benefits on children’s inhibitory control but none for literacy or language.⁷² In correlational studies, the associations between the CLASS domains and children’s academic gains in preschool are usually modest or null, with no consistent patterns by program type/setting.⁷³ Some studies have found Instructional Support to be a statistically significant predictor of preschoolers’ gains in language, literacy, mathematics, and inhibitory control skills.⁷⁴ However, meta-analyses across a broad range of preschool program types/settings have found no associations between Instructional Support and children’s gains in academic and developmental outcomes.⁷⁵ Some researchers have also examined whether the association between process quality and children’s gains in preschool is nonlinear or whether there are thresholds of process quality that must be met to see corresponding gains in students’ skills.⁷⁶ This modeling approach has also yielded mixed results, including moderate-sized⁷⁷ and null or small associations between the CLASS domains and gains in student outcomes at higher levels of quality.⁷⁸

Instructional quality summary measures

The field has also approached quality assessment and improvement from a domain-specific approach, collecting summary measures of the degree to which instruction is aligned with the science of learning and development in specific content areas. For example, in language and literacy, the Early Language and Literacy Classroom Observation (ELLCO)⁷⁹ is an observational tool focused on classrooms’ physical literacy environment (i.e., the availability, display, and arrangement of materials such as books, print sources, and writing instruments) and language and literacy instruction (i.e., the depth and length of developmentally appropriate language, reading, and writing activities). This instrument has been shown to effectively identify variation in the type and dosage of language and literacy instruction that children experience in private, Head Start, public PreK, and university lab school settings,⁸⁰ and research has also shown that it is an effective professional development (PD) tool (also across many setting/program types).⁸¹ However, the evidence on the association between ELLCO scores and children’s language and literacy gains is mixed. In particular, the quality of the language and literacy environment has been linked to null to small gains in children’s emergent literacy outcomes (b ranging from 0.00 to 0.07)⁸² and to modest gains in children’s vocabulary.⁸³ A comprehensive meta-analysis found that PD programs that included the ELLCO tool led to moderate to large improvements in classroom literacy (average g = 0.76) and modest effects of on children’s language and literacy outcomes (average g = 0.14).⁸⁴

Another content-oriented observational quality measure is the Classroom Observation of Early Mathematics Environment and Teaching (COEMET) which focuses specifically on the quality of mathematics instruction in preschool classrooms. It captures the specific content of instruction (i.e., numeral recognition, addition and subtraction strategies), the amount of time spent on mathematics instruction, and the quality of mathematics teaching strategies. For this tool, research has reported evidence of predictive validity for mathematics achievement and EF competencies in urban public school districts.⁸⁵ Studies that have found large effects of mathematics-specific curriculum, PD, and coaching on preschoolers’ mathematics skills have also shown large effects on COEMET subscales (g ranging from 0.78 for the quality of mathematics activities to 1.23 for the mathematics culture in the classroom).⁸⁶

Time use summary measures

Classroom time use is another summary measure that signals instructional differences across classrooms and that can predict young children’s learning gains.⁸⁷ Researchers have found systematic differences in the proportion of time that preschool classrooms allocate to learning activities (e.g., language, literacy, and mathematics instruction) and learning formats (i.e., whole class, small groups, and centers). The evidence pertaining to how time is used in early education centers focuses on preschool classrooms, including public and private school, center-based, and Head Start programs. On average, classrooms spend a larger proportion of time in arts, language, and literacy activities than in other content areas.⁸⁸ For example, time spent on literacy and language ranged from 14% to 46% of the observed instructional block across studies, while time spent on mathematics activities ranged from 4% to 21%. For learning formats, most studies in this area suggest that children spend approximately one-third of their day in whole-group or large-group activities, one-third of the school day in free-choice activities or interest centers, and the remaining one-third in routines and meals.⁸⁹ There is variation here as well. For example, across studies, the percentages have ranged from 4% to 28% for small-group activities and from 1% to 16% for individual learning activities.

Regarding links between time use measures and preschoolers’ learning, some studies have found that more time spent on whole-group learning than on free-choice or independent learning predicted child learning gains in language and literacy.⁹⁰ However, findings can differ by child learning domain. For example, other studies have found that more time spent in free-choice learning and less time spent in whole-group learning in preschool was associated with higher gains in social-emotional skills⁹¹ and in mathematics.⁹² For content areas, several studies have found links between more time spent on language and literacy and children’s gains in those skill areas.⁹³ In studies that have created summary measures of time use, being in a classroom with a “light academic profile” (i.e., more free play) appears to be associated with smaller gains in literacy and mathematics skills compared to the gains achieved under other instructional profiles, such as those with more time dedicated to whole-group or individual instruction.⁹⁴ Research in this area is correlational, and accordingly, the findings should be viewed with caution.

Key finding #4: Evidence-based curriculum implemented with coaching appears to be a “good bet” for improving instruction in preschool classrooms and children’s readiness for kindergarten.

Curriculum

Most public preschool programs use some kind of formal curriculum, with over 200 published preschool curricula available in the U.S.⁹⁵ Some preschool curricula, called global or comprehensive curricula, purport to cover all child skill areas. Others target one or two areas (i.e., a literacy-specific curriculum). Most public preschool programs use either Creative Curriculum or HighScope, which are two global curricula.⁹⁶ There have been many RCTs and quasi-experimental trials over the last 20 years that have compared the effects of different curricular approaches in Pre-K in public schools, Head Start, and community-based centers. These studies have found that global curricula sometimes improve classroom process quality more than teacher-created curricula do and approximately equally as well as domain-specific curricula do.⁹⁷ However, domain-specific curricula lead to increased time and quality of instruction in the targeted skill area. Domain-specific curricula also tend to produce more child learning gains in the targeted area⁹⁸ and sometimes in other skills areas as well.⁹⁹ One reason why domain-specific curricula may be more effective is that they tend to include a specified scope and sequence for activities that follows the science of how children learn in a particular skill area. Many comprehensive curricula lack a specified scope and sequence.

Recognizing this evidence, some localities/researchers have had success in combining different domain-specific curricula to cover multiple child learning domains, in Pre-K in public schools, Head Start, and community-based centers.¹⁰⁰ In practice, this has meant providing teachers with guidance on how to coherently integrate the selected curriculum via centrally created sample schedules, training supports, and coaching. In general, more successful curriculum studies have included training and coaching focused on curriculum implementation (and not only on general best teaching practices), i.e., features that have supported fidelity of implementation.¹⁰¹

Regarding whether preschool curriculum effects differ by child demographics, the evidence is mixed. Some studies find larger impacts on children from marginalized groups, while others find that all children benefit equally from a given curricular approach.¹⁰² The National Academies of Sciences, Engineering, and Medicine report on preschool curricula summarized this evidence as suggesting that there is a “potential for curricula, particularly some domain-specific curricula, to have greater effects for marginalized groups”.¹⁰³ According to the same report, the evidence promisingly suggests that with supports, a broad range of early educators, defined in terms of training, education, and sector, implement preschool curricula approximately equally well.¹⁰⁴

In-service professional development

A common quality support for early educators is in-service PD, which often takes the form of trainings and in-classroom coaching by an expert mentor. Just as in the K-12 literature, there is little evidence that one-shot workshops/trainings on their own change early educators’ practices or improve young children's skills, whether in Pre-K in public schools, Head Start, or community-based centers.¹⁰⁵ Multiple large-scale RCTs have shown that coaching on general best practices improves quality in preschool classrooms but does not affect children’s school readiness skills.¹⁰⁶ However, coaching when tied explicitly to implementing domain-specific curricula has been shown in multiple randomized trials across Pre-K in public schools, Head Start, and community-based settings to improve both classroom quality and children’s school readiness.¹⁰⁷

One recent review suggested that the features that make PD for early educators more effective include sufficient intensity, an emphasis on evidence-based teacher practices, and active teacher engagement.¹⁰⁸ More research comparing PD dosages, types, contents, and formats is needed.

Child assessments

Regular assessments of children’s skills and progress have been another common investment meant to improve quality in early education classrooms. Head Start and most state PreK programs require teachers to complete formative assessments on children’s development at multiple points in the year.¹⁰⁹ Many of these assessments require teachers to observe children in the classroom and to rate each child’s skill level across school readiness domains. There are also some direct assessments in which a teacher directly tests a child in a given domain (e.g., letter knowledge and letter sounds), but they are less commonly used. The psychometric validity of some formative assessment systems has been called into question.¹¹⁰ For both types of assessment, to the best of our knowledge, no studies have examined the degree to which assessments improve classroom quality or child learning.

Curriculum alignment with kindergarten. Another element of process and instructional quality is the degree of alignment between children’s preschool and early elementary experiences, particularly their kindergarten experiences. As others have detailed, this includes alignment of curriculum content and learning formats.¹¹¹ Some descriptive research has found that kindergarten instruction can be repetitive with the skills that children already have and that more advanced instruction can promote children’s learning in kindergarten.¹¹² To date, the strongest studies of the alignment question have been mathematics curriculum studies in which there was randomization to a mathematics curriculum enhancement in Pre-K and then re-randomization to an aligned mathematics condition in kindergarten. These studies (in Pre-K in public schools, Head Start, and community-based centers) found lasting benefits only for children who experienced the aligned Pre-K and kindergarten condition.¹¹³

Cultural responsiveness

According to a National Academies of Sciences, Engineering, and Medicine preschool curriculum report, cultural responsiveness refers to an “asset-based approach to classroom cultural diversity”.¹¹⁴ As the report also points out, Head Start and some states emphasize cultural responsiveness in their early learning program standards,¹¹⁵ and interviews with 31 leaders in the field also point to a broad consensus on the important cultural relevance of early childhood curricula and classroom practice.¹¹⁶ However, although there are examples of culturally responsive practices, empirical research linking cultural responsiveness to improved classroom quality or child learning outcomes in the early education years is currently lacking.¹¹⁷

Key finding #5: More marginalized groups have less access to high-quality early education programs.

As discussed in Chapter X (Taryn – ACCESS CHAPTER), families do not have equal access to the types of care for their young children that they would prefer due to factors such as high costs, limited supply, and poor fit with work schedules. As we review below, even when families can access the type of care that they prefer, children of color, children with disabilities, dual language learners, and children from families with low incomes often experience lower-quality programs than do their peers.

Race/ethnicity and home language gaps in quality

Studies in public preschool programs have found that White and monolingual children generally attend classrooms with higher process and structural quality compared to the classrooms attended by their Black and Hispanic peers. There are, however, two studies that found no variation in process and structural quality by school racial/ethnic composition, nor in the quality of the language environment based on children’s dual language learner status.¹¹⁸ Evidence from public Pre-K programs in 11 states showed that, on average, White children attend classrooms with higher CLASS scores compared to the classrooms of their Black (0.65 standard deviation (s.d.)) and Hispanic (0.37 SD) peers, with differences concentrated in classroom Emotional and Instructional Support dimensions of quality. In particular, classrooms with a larger proportion of Black and Hispanic children have shown lower emotional support (-0.66 SD and -0.24 SD, respectively) and lower instructional support (-0.33 SD and -0.31 SD, respectively). This same study showed gaps as large as 0.35 SD between the structural quality of classrooms with a larger proportion of dual language learners and the structural quality of classrooms with a majority of monolingual children.¹¹⁹ A study including private and public UPK providers in Georgia found that classrooms with a high concentration of Black children had lower Emotional (-0.35 SD) and Instructional Support (-0.28 SD) scores, as measured by CLASS, than did classrooms with a higher concentration of White children.¹²⁰ These findings are consistent with evidence from the New York UPK program, where the Black–White gaps in quality were as large as 0.37 s.d. for Emotional Support and 0.42 s.d. for Instructional Support and the Hispanic–White gaps reached 0.16 s.d. and 0.36 s.d., respectively.¹²¹

Income-based quality gaps.

There are also large gaps in quality by family income. Based on a nationally representative sample of ECE center-based providers serving children from birth to 5 years old, a team of researchers concluded that children attending centers located in areas with a high poverty density were less likely to be in classrooms with small teacher–child ratios. ¹²² These classrooms were also more likely to be served by teachers with fewer years of education and to be in centers where fewer benchmarks of workforce support (e.g., PD opportunities and coaching) were met. Similarly, in North Carolina, licensed ECE programs in affluent neighborhoods were more likely than programs in neighborhoods where residents had lower incomes to obtain higher quality ratings.¹²³ A study of Georgia’s UPK program found that programs located in higher-poverty ZIP codes had more experienced teachers but lower levels of emotional support and instructional support than did those located in lower-poverty ZIP codes.¹²⁴ In New York City, lower-quality Pre-K programs were disproportionately found in poorer neighborhoods.¹²⁵

Disability-based quality gaps

Evidence on the quality of center-based settings serving children with disabilities is scarce. However, caregivers of young children ages 0 to 5 with disabilities have reported greater difficulties finding high-quality early education that meets their children’s unique needs, higher costs of care, and an increased administrative burden to coordinate care across service providers.¹²⁶ Specifically, compared with caregivers of children ages 0 to 5 who do not have a disability, a larger proportion of caregivers of children with disabilities experienced difficulty finding care (34% of parents of children with disabilities reported difficulties vs. 25% of parents of children without disabilities) based on findings from the 2016 Early Childhood Program Participation (ECPP) Survey and the 2016–2018 National Survey of Children’s Health (NSCH).¹²⁷ Beyond the availability of slots, interview data from the same study revealed that parents were concerned about child care providers’ experience or knowledge with regard to serving children with disabilities. For example, parents reported a lack of providers where teachers were fluent in American Sign Language; concerns about children’s well-being and physical safety due to large teacher–child ratios; or concerns about institutional policies that did not align with the principle of inclusion, such as barring children from specific activities or inappropriately holding children back with younger peers. A study of 33 Midwestern classrooms with children ages 3–5 found that teachers reported a high level of self-efficacy with respect to including children with disabilities and expressed positive attitudes toward the benefits of inclusion. However, these teachers also expressed the need for training and support on specific inclusion practices.¹²⁸

Sector quality differences

As highlighted in the introductory chapter, early education quality differs by sector. For example, a study with a nationally representative sample found evidence of lower quality in community-based centers on indicators such as book reading frequency, television dosage, and observed structural quality than in public-school PreK and Head Start.¹²⁹ The study also found that family child care homes scored lower than centers on structural quality metrics for 4-year-olds.

Importantly, these sector differences tend to persist even in mixed-delivery Pre-K systems, in which there has been some attempt to equalize policies and investments across care types. (i.e., across Pre-K classrooms in public schools and in community-based programs).¹³⁰ Specifically, research has shown that 1) where there were quality and investment differences in these systems, they tended to favor public schools; and 2) within these systems, families marginalized by race/ethnicity, home language, and/or income were more likely than their peers to select into community-based settings.¹³¹ Interestingly, mixed-delivery Pre-K systems with more similar policies by sector had fewer differences in quality and child gains.

For children younger than 4, the evidence on quality differences by setting type is more scant and more mixed. For example, at age 2, family child care homes scored lower than centers on reading frequency, observed and reported structural quality measures, safety, and teacher education and training, and had more television dosage. However, family child care homes also reported more frequent visits to cultural and scientific sites of interest compared to centers.¹³² Underscoring the quality challenges for younger children, even within the same sector, teachers of young children (i.e., toddlers) show higher turnover rates than do preschool teachers. At least in part, this is likely due to lower compensation.¹³³

Overall, more research is needed to 1) identify the causes of these differential enrollment patterns across demographic groups, 2) determine the best policy levers for ensuring equally high-quality experiences across settings, and 3) examine the causal effects of enrollment in different settings on children’s development and learning.

Key finding #6: Policy-level efforts to regulate and improve early education quality through licensing standards, accountability systems, and accreditation have had some success in improving many dimensions of quality.

Policymakers have implemented a range of policies to improve quality in ECE settings. Below, we summarize three of the most widespread policy initiatives—regulations, accountability via quality rating systems, and accreditation—and review evidence on their effects on child development, system-level intended outcomes such as population-level quality improvement, and unintended outcomes such as increased operating costs.

Regulations

Federal and state policymakers have used regulations, including licensing, as a lever for ECE quality improvement. The primary purpose of licensing standards is to set a floor for programs to be considered safe environments for young children and to enforce the requirements for providers to legally offer child care services based on state, local, or tribal law.¹³⁴ Licensing is a complex process with the potential to offer information about programs along a defined quality continuum, given that all states and U.S. territories have systems in place dedicated to monitoring compliance. These systems are generally funded through a combination of federal and state resources.¹³⁵ To obtain and renew a license, programs generally need to meet a set of standards for their physical environment, administration, operations, personnel, and community engagement. All licensed programs are visited regularly to ensure continued compliance with standards.

Despite the comprehensive data collected by licensing and monitoring systems, the impacts of and challenges in implementation (e.g., potential systematic differences in compliance across communities, responsiveness to rapid demographic changes, or new directions in developmental and education sciences) and the child-level effects of these nationwide policies are understudied. However, researchers have examined the effects of more stringent state regulations on the quality of services. Evidence has shown that the implementation of stringent regulations can increase the structural quality of services in nonprofit and for-profit licensed child care centers serving children aged 0–5, specifically via improvements in the conditions necessary to keep children safe (i.e., physical space and supervision), improved requirements for staff characteristics and ratios, and increases in directors’ education and administrative skills.¹³⁶ However, increased regulations can also decrease teacher wages. For example, in nonprofit and for-profit licensed child care centers serving children aged 0–5, increasing education requirements by one year and raising the hours of training required at the time of hire by 100 hours resulted in wage decreases of 9% and 16%, respectively, due to the increases in costs required for programs to meet these new quality requirements.¹³⁷ To the best of our knowledge, studies have not yet examined how licensing standards relate to centers’ process or instructional quality.

In Head Start, as others have reviewed, there have been multiple regulatory changes aimed at improving program quality and children’s development via the Head Start reauthorization process.¹³⁸ For example, in 1998, the Head Start with the Community Opportunities, Accountability, and Training and Educational Services (COATES) Act raised the minimum teacher education requirement to an AA degree for 50% of Head Start teachers nationally. In addition, an increased proportion of federal Head Start funding was set aside for quality improvements. In the 2007 reauthorization, the teacher education requirement increased to 50% of Head Start teachers with at least a BA. In 2011, Head Start also started requiring observational quality assessments using CLASS, with the lowest-scoring programs losing funding.¹³⁹ Research on these changes overall is descriptive and somewhat limited; hence, it should be interpreted with caution.¹⁴⁰ It is also difficult to isolate the impacts of any particular policy/resource change since there were multiple overlapping changes. On the positive side, Head Start exceeded the education requirement thresholds nationally and well ahead of schedule, with little change in teacher demographics. However, CLASS scores have remained relatively flat, as we showed in prior work and reprint here in Figure 2.¹⁴¹

• Figure 2

  Head Start Grantees’ CLASS scores, 2012–2018

  

  Footnotes

  2019 and 2020 data are not available due to the COVID-19 pandemic. This figure is reprinted from Weiland and Guerrero Rosada (2021).

Quality rating and improvement systems

Quality rating and improvement systems (QRISs) are a widespread policy strategy designed to improve the quality of early childhood education settings through two mechanisms: a) tiered financial incentives supporting program operations and PD; and b) information about program quality that is publicly available to families. Combined, these mechanisms aim to improve quality at the system level, once early education programs are rated and engage in improvement activities. Currently, 44 states, the District of Columbia, and the Commonwealth of the Northern Mariana Islands have fully implemented at least one QRIS. Four states (South Dakota, Kansas, Wyoming, and West Virginia) have a QRIS in development phases, and there is no public information on the Missouri and Mississippi systems.¹⁴² Two states (Maine and Tennessee) have mandatory participation for all licensed programs, and nine additional states (Colorado, Illinois, Louisiana, North Carolina, Oklahoma, Pennsylvania, South Carolina, Vermont, and Virginia) automatically rate licensed programs, with participation rates of 99–100%. Among the remaining states with voluntary participation, participation ranges from 9% (Iowa) to 96% (Utah). These figures might change rapidly as QRISs continuously evolve to increase participation and coverage. Importantly, research with nationally representative data has shown that QRIS participation is lower in predominantly Black communities and higher in low-income communities and across centers that blend funding mechanisms, potentially due to child care subsidy incentives.¹⁴³ Most states (all but 11 with QRISs) use tiered reimbursements, meaning that a higher reimbursement rate is allocated to higher-rated programs as an incentive to improve quality.¹⁴⁴

QRISs rate programs on a variety of structural or process quality indicators that are either conceptually or empirically associated with children’s development. Although QRISs vary widely in terms of the scoring strategy and weights that the systems attribute to quality indicators, most systems measure the extent to which the physical environment and learning materials are developmentally appropriate, staff qualifications, teacher–child ratios, the use of assessments to inform instruction, and opportunities for family involvement, among others. Twenty programs include a standardized observational measure as part of their scoring system (14 require a specific Environmental Rating Scale (ERS) tool and 6 require the CLASS-PreK).

Quasi-experimental research shows that licensed programs serving children aged 0–5 respond to QRISs by improving on the structural and process quality metrics captured by these systems and that lower-rated programs experience enrollment declines,¹⁴⁵ with parents opting for programs with higher ratings.¹⁴⁶ However, the associations between QRIS ratings and children’s development and learning are null and inconsistent; notably, however, the available evidence focuses only on preschool-aged children.¹⁴⁷ Some studies have examined whether the components captured by these systems are more consistently associated with children’s gains than are QRIS overall ratings.¹⁴⁸ However, the existing research on QRISs and children’s development remains correlational, and the findings of this research should be viewed with caution.

Accreditation

Accreditation from the NAEYC is a nationally recognized endorsement of high quality. The accreditation process entails self-assessment by centers, direct observations, and guided improvement across ten areas of quality, including relationships with children, curriculum, teaching approaches, child assessment, nutrition and health, staff qualifications, relationships with children's families, relationships with the community, the physical environment, and program leadership and management. Centers that have decided to pursue accreditation initiate the process and assume the administrative costs of membership and assessment. Although all centers are assessed in reference to a predefined set of quality standards, each program receives support from NAEYC in developing and implementing an improvement plan based on its baseline conditions. Accreditation status is valid for five years. Notably, centers tend to select into NAEYC accreditation. However, it can be a policy lever; Boston, for example, has pursued NAEYC accreditation for all its UPK programs as part of its strategy to increase quality.¹⁴⁹

Descriptive studies have found that local licensing policies and state regulations relate to centers’ likelihood of pursuing and obtaining NAEYC accreditation. In particular, the stringency of state licensing standards is positively associated with the total number of preschool programs involved in the NAEYC accreditation process.¹⁵⁰ NAEYC accreditation can also affect program quality through centers’ organizational climate, work conditions, or staff selection. A study examining differences in the quality of work experiences for staff in accredited vs. nonaccredited early education centers serving children aged 0–5 across 33 states found that staff in accredited centers reported higher scores on domains such as professional growth, innovativeness, goal consensus, and clarity.¹⁵¹ There is correlational evidence that accredited centers have lower staff turnover and pay staff higher salaries than do nonaccredited centers.¹⁵² Moreover, experimental evidence shows that NAEYC-accredited providers are more likely to interview job applicants with specific early education work experience, higher education levels, and other professional credentials, while QRIS participants are not.¹⁵³

NAEYC accreditation is also hypothesized to increase classroom quality and thereby improve child outcomes. Some research shows that programs can obtain NAEYC accreditation but fall short of meeting guidelines for a developmentally appropriate preschool curriculum, as measured by a widely used instrument of structural quality (ECERS-R).¹⁵⁴ However, other work has shown positive associations with improved structural and instructional quality and with gains in the vocabulary scores of 4-year-olds.¹⁵⁵ Overall, most of the evidence on NAEYC accreditation is correlational; hence, it should be interpreted with caution.

Future directions

Given the broad consensus and evidence base showing that high-quality early education can benefit children in both the short run and the long run, there is increasing attention to identifying and measuring the active ingredients of effective, scalable early learning programs. Although most research in this area has focused on the preschool years, there is evidence of socioeconomic disparities in the quality of services for all ages. Some key areas for future research include improvements in defining and measuring quality in early learning settings; rigorous comparisons of the effects of different kinds of quality inputs on children and teachers; comprehensive studies of the early education experiences of younger children (aged 0–3); and evidence on new, potentially promising policy reforms in this area, including efforts to increase teacher pay, licensing and QRIS reforms, and new scholarship programs for early education teachers to increase their education and training.

Endnotes and references