Skip to main content
Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Child-Staff Ratios in Early Childhood Education and Care Settings and Child Outcomes: A Systematic Review and Meta-Analysis

  • Michal Perlman ,

    Affiliation Applied Psychology and Human Development, University of Toronto/OISE, Toronto, Ontario, Canada

  • Brooke Fletcher,

    Affiliation Department of Gastroenterology, Alberta Children's Hospital, Calgary, Alberta, Canada

  • Olesya Falenchuk,

    Affiliation Applied Psychology and Human Development, University of Toronto/OISE, Toronto, Ontario, Canada

  • Ashley Brunsek,

    Affiliation Applied Psychology and Human Development, University of Toronto/OISE, Toronto, Ontario, Canada

  • Evelyn McMullen,

    Affiliation Applied Psychology and Human Development, University of Toronto/OISE, Toronto, Ontario, Canada

  • Prakesh S. Shah

    Affiliations Department of Pediatrics, Mount Sinai Hospital, Toronto, Ontario, Canada, Department of Pediatrics, University of Toronto, Toronto, Ontario, Canada, Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, Ontario, Canada


Child-staff ratios are a key quality indicator in early childhood education and care (ECEC) programs. Better ratios are believed to improve child outcomes by increasing opportunities for individual interactions and educational instruction from staff. The purpose of this systematic review, and where possible, meta-analysis, was to evaluate the association between child-staff ratios in preschool ECEC programs and children’s outcomes. Searches of Medline, PsycINFO, ERIC, websites of large datasets and reference sections of all retrieved articles were conducted up to July 3, 2015. Cross-sectional or longitudinal studies that evaluated the relationship between child-staff ratios in ECEC classrooms serving preschool aged children and child outcomes were independently identified by two reviewers. Data were independently extracted from included studies by two raters and differences between raters were resolved by consensus. Searches revealed 29 eligible studies (31 samples). Child-staff ratios ranged from 5 to 14.5 preschool-aged children per adult with a mean of 8.65. All 29 studies were included in the systematic review. However, the only meta-analysis that could be conducted was based on three studies that explored associations between ratios and children’s receptive language. Results of this meta-analysis were not significant. Results of the qualitative systematic review revealed few significant relationships between child-staff ratios and child outcomes construed broadly. Thus, the available literature reveal few, if any, relationships between child-staff ratios in preschool ECEC programs and children’s developmental outcomes. Substantial heterogeneity in the assessment of ratios, outcomes measured, and statistics used to capture associations limited quantitative synthesis. Other methodological limitations of the research integrated in this synthesis are discussed.


Early childhood is a critical period in shaping children’s developmental trajectories [1]. Research has shown that high quality early childhood education and care (ECEC) programs can enhance child development [2,3]. Given that enrollment in ECEC settings has become the norm for preschool aged children [4,5], it is important to empirically identify aspects of the ECEC environment that are key in supporting children’s development. Organizations such as the American Academy of Pediatrics (AAP), the American Public Health Association (APHA), the National Institute of Early Education Research (NIEER), and the National Association for the Education of Young Children (NAEYC) have identified indicators of high quality ECEC programs. These guidelines are intended to assist parents, policy makers, funders and other key stakeholders in making informed decisions about ECEC arrangements. The ratio of children to staff is one relatively quantifiable aspect of structural quality identified as a key quality indicator in the AAP’s Policy Statement on Quality Early Education and Child Care [610].

The perceived importance of low child-staff ratios in ECEC programs is demonstrated by the fact that all Quality Rating and Improvement Systems (QRIS) in North America include child-staff ratios in their ratings of quality [11]. QRIS are accountability systems that revolve around multicomponent assessments designed to improve the quality of ECEC programs. Regulations regarding child-staff ratios in licensed ECEC settings are implemented at various levels of government in the U.S (Table 1 for a summary of ratio recommendations); however, there is wide variability between jurisdictions with few jurisdictions in the US and Canada meeting these recommended standards [12,13].

Table 1. Recommended Standard for Child-Staff Ratios in Center-Based ECEC Settings for Preschool-Aged Classrooms.

Research findings on associations between child-staff ratios and outcomes have been contradictory. Some studies show that better ratios are associated with improved child outcomes [1416] while other studies have not found such linkages [1720] or report mixed results [21]. Many practitioners working with children and families are in a position to provide parents with information about ECEC for their children. By way of just one example of this, in its policy statement on quality early education and child care, the AAP notes that pediatricians have a significant role to play in helping parents identify high quality ECEC environments for their young children [10]. However, to do this effectively, practitioners must have access to evidence-based recommendations about which aspects of the ECEC environment are important. To our knowledge, a synthesis of available literature to identify appropriate or ideal ratios in relation to children’s outcomes has not been conducted. To provide such guidance, our objective in this study was to systematically review and, if possible, meta-analyze associations between child-staff ratios in ECEC classrooms that serve preschool-aged children and child outcomes.


Types of Participants and Setting

This study focused on classrooms that serve preschool-aged children as these serve the largest number of children in ECEC settings [22]. Preschool-age was defined as ranging from 30 to 72 months. ECEC settings included child care centers, preschool programs, nursery schools, pre-kindergarten programs, and Head Start programs. Studies that only examined home-based child care, or those in which results for home-based and center-based care could not be separated were excluded.

Types of Studies

Studies reporting associations in cohort, cross-sectional or longitudinal analyses were included in this review. Studies reporting a statistical link between an aggregate ECEC quality variable that consisted of several measures of quality and child outcomes were not included if the specific effect of ratios could not be separated. Case-series, reviews, editorials and letters to editors were read to identify articles but were not included in the review.


Outcomes were operationalized broadly and included measures of children’s cognitive, pre-academic, social, emotional, behavioral, and motor functioning. Outcome measures included data that were gathered from direct testing of children as well as teacher and parent reports. Measures that focused on dyads (e.g., staff-child attachment) were excluded because it is difficult to separate “caregiver/program” effects from child characteristics using such measures.

Search Strategy

An extensive search of the electronic databases PsycINFO, Medline, and ERIC was conducted for English language studies published until July 3, 2015. Two separate searches were performed within each of the three databases. One search combined search terms specific to child-staff ratios AND child outcomes. The other combined search terms related to more global ECEC quality indicators AND child outcomes. This increased the likelihood that we would capture studies in which ratios were used as a control variable. Specific keywords used in electronic searches are provided in Tables A-D in S1 File.

Second, websites of the Cost, Quality, and Outcomes Study (CQO) [23], Early Childhood Longitudinal Study (ECLS) [24]; Effective Provision of Pre-School Education (EPPE) Project [25]; Head Start Impact Study (HS) [26]; National Center for Early Development and Learning’s (NCEDL) Multi-State Study of Pre-Kindergarten and State-Wide Early Education Program Study (SWEEP) [27]; Head Start’s Family and Child Experiences Survey (FACES) [28] and the National Institute of Child Health and Human Development’s (NICHD) [29] Study of Early Child Care and Youth Development were reviewed to retrieve relevant studies. Lastly, reference lists of those studies that met inclusion criteria were manually searched to identify additional relevant studies (see Table 2 for a complete list of all inclusion criteria).

Table 2. Inclusion Criteria for Systematic Review and Rationale.

Selection Strategy

The title and abstract of each paper located through the literature searches was reviewed for relevance. Abstracts identified as potentially relevant to the current study underwent full-text review to determine if inclusion criteria were met (Table 2). Each step was conducted by a pair of independent raters.

Data Extraction

Information was extracted from each study by two independent raters and then compared for accuracy. Discrepancies were resolved through discussion. If consensus could not be reached, the third author was consulted to make the final decision. Since raters extracted objective information (e.g., correlation statistics, descriptive statistics) disagreements were very rare.

Meta Analyses

Studies that could be meta-analyzed were selected from the pool of studies selected for the systematic review. To be meta-analyzed studies had to use identical child outcome measures and had to use an identical operationalization of ratios. We adopted a minimum requirement of three studies to conduct a meta-analysis on a particular child outcome. To increase homogeneity among studies that were meta-analyzed, only studies that ensured children’s exposure to the program were included. Thus, we only meta-analyzed studies that used child pre-scores as a covariate, or used gain scores in analyses, or in which the authors stated explicitly that only children who had been in the program for a period of time prior to their assessment were included. In addition, only statistics that accounted for covariates (e.g., child and family characteristics) were combined within a single meta-analysis. Finally, to avoid overlap in samples where subsamples were drawn from the same dataset, only the study with the largest sample size was selected for inclusion in meta-analyses. Thus, within a given meta-analysis, only one coefficient from each sample was included in any one analysis.

Statistical models with quadratic terms assume non-linear associations between the variables. Given the statistics extracted for most studies only test for linear relationships (correlation coefficients and linear regression coefficients), associations in models using quadratic terms were excluded and only results examining linear relationships were used in the meta-analyses. We used random-effects models for meta-analyses. All meta-analyses were conducted using the Comprehensive Meta-Analysis Version 3 statistical package (see S2 File for formulas for converting statistics) [30]. Finally, for each meta-analysis I2 was used to test statistical heterogeneity.


The details of the search results and reasons for exclusion of articles are presented in Fig 1. In total, 29 studies met our inclusion criteria; 23 of the studies were peer reviewed journal articles, five were reports and one study reported their results in a book chapter. Descriptive information for the 29 studies is presented in Table 3. Seven studies contained samples that were drawn from both the NCEDL’s Multi-State Study and SWEEP study [20,3136], three studies were from the CQO project [21,37,38], two studies utilized the Head Start FACES 2000 Cohort sample [39,40], and two studies included samples drawn from the NICHD Study of Early Child Care [15,41]. Furthermore, two studies [42,43] were drawn from the same dataset as were two additional studies using a separate sample of preschool-aged children from Bermuda [44,45]. Thus, many studies had overlapping samples.

Fig 1. Flow diagram for study selection.

Adapted from Moher et al.[46].

Table 3. Description of Studies Meeting Inclusion Criteriaa.

Across studies the average child-staff ratios ranged from 5 [6] to 14.5 [41] preschool-aged children per adult with a mean of 8.65. The average standard deviations across the studies ranged from 1.43 to 4.1 but were only reported in 20 of the 31 samples (see Table 3 for more information). To assess the variability of ratios within studies, we computed a difference score, M = 8.35, and median = 13 (calculated using the range for that study as the maximum score minus the minimum score). Range scores for ratios was reported in only 13 of the 31 samples. The lowest variance (difference score of 2.2) was for ratios ranging from 4.2 to 6.4 preschool-aged children per adult [52], and the highest variance (difference score of 33) was for ratios ranging from 1 to 34 preschool-aged children per adult [41]. These finding suggest substantial variance within and across studies, with some classrooms having ratios well above the standard guidelines (see Table 1).

Four studies [41,44,45,47] included ECEC programs located outside of the United States. Five studies [21,44,45,49,51] used a cross-sectional design. Twenty-four studies used a longitudinal design [14,15,1720,3143,47,48,50,52,53]. Of the independent samples, (i.e., excluding overlapping datasets by retaining only the largest sample size), the total sample size was 15,191 preschool-aged children (range 51 to 3584, median = 532).

Measurement of Child-Staff Ratios

The way child-staff ratios were assessed varied between studies. In 15 of the studies [14,15,19,21,31,36,3943,49,50,52] child-staff ratios were collected by research assistants who counted (often multiple times, at set intervals) the number of children and adults present in each classroom during classroom observations. Other studies relied on either staff [17,20,32,34,35,38,44,45,47,48] or maternal [51] reports. In one study, information on ratios was collected through both observations and staff reports and the highest number obtained for each classroom was utilized in the analyses [18]. In another study, ratios were initially collected via staff reports; however, after concerns surfaced about reliance of self-reported ratio counts, a sign-in/sign-out procedure for staff in the remaining classrooms was implemented [53]. One study did not present information as to how child-staff ratios were obtained [33]. Some studies included all adults present in a given classroom in their calculations of child-staff ratios regardless of whether the adult was an aide, volunteer, or parent [15,39,40]. Others used ambiguous terms such as counting “caregivers” [52] or “staff” [42,43], while others only included staff who were “actively interacting with children” [39,40] and one study [15] limited calculations to “all adults who regularly worked more than 10 hours per week”. However, most studies did not provide details about which adults were included in the ratio counts.

Seven studies [14,15,1921,36,37] used a dichotomous child-staff ratio variable in their analyses to indicate whether or not a specific standard for ratios had been met. Point-biserial correlation coefficients were utilized for analyzing dichotomous variables. This type of correlation is mathematically equivalent to the Pearson (product moment) correlation that is used for continuous and ordinal variables. Thus, studies using either absolute values or binary variables were included within a given meta-analysis. Finally, for all samples used in this analysis, we ensured that lower child-staff ratios indicate less children per staff member (i.e., more favorable conditions). For those studies in which the original data reflected the opposite (i.e., a higher ratio represented less children per staff) [15,18,19,21,31,32,3638,45,5153], the direction of the statistical associations were adjusted following data extraction to ensure consistency across the reporting of all results.

Systematic Review

All 29 eligible papers were systematically reviewed. The associations between child-staff ratios and child outcomes are reported using various statistical approaches and are reported in full in Tables A-D in S3 File. A snapshot of the results is also provided below in Fig 2, which displays the results obtained for only those outcomes that were used in three or more samples. In the 29 included studies, 209 distinct statistical analyses quantifying the association between child-staff ratios and measures of child outcomes were reported. These statistical analyses included 83 unique child outcome measures (S4 File). These were associated to ratios [17,3335,3944,4750,53] operationalized in a variety of ways including number of staff divided by the number of children [18,31,32,38,45,52] the number of children divided by the number of staff or met/did not meet a specific standard (6:1, 7:1, 8:1 or 10:1) [14,15,20,21,36,37]. In addition, one study compared groups of classrooms with the highest (1:9.5) to lowest (1:7.2) ratio scores [19], and another compared groups of classrooms for ratios at different levels (1–6, 7–9, and 10–16) [51].

Fig 2. Systematic review results for the associations between child-staff ratios and child outcomes in 3 or more samples.

a Abbreviations: Symbols bolded are significant and positive, symbols bolded and italicized are significant and negative, and symbols in grey are non-significant. Star = Zero Order Pearson’s Correlation, Unfilled circle = Beta, Filled square = Unstandardized Coefficient, Black diamond minus white X = T-Test, Key clover = Partial Correlation, Downward arrow = Effect Size, Filled circle = F-Ratio. aTo improve the readability of this complex table, 9 papers [15,17,19,37,4143,48] that had an outcome that appeared in only that one paper were omitted from this table. Several analyses from other papers that had idiosyncratic outcomes are also excluded. For a comprehensive display of all of the data for all of the child outcomes see Tables A-D in S3 File. bRatio scores have been adjusted to be consistent across all data. In keeping with how ratios were operationalized in most of the papers we reviewed, we reverse scored when necessary so that lower ratio scores indicated fewer children per adult across all studies. Thus, negative relationships reflect an association between better ratios and better outcomes. In the case of problem behaviors, we expected a positive association as this reflects a correlation between better ratios and lower rates of problem behaviors [18,21,31,32,3638,45,51,52].cThis paper is one of a series of Meta-Analyses and Systematic Reviews assessing the relationship between child care quality and children’s outcomes; therefore, superscript letters below are in reference to various large databases that samples in these papers were drawn from. These letters have been kept consistent across the series for our readers. dSamples within papers are described in more detail in Table 2. eAcronyms for child outcomes are listed in S4 File. fIdentifying Letters (also refers to Identifying Letters, Naming Letters, and Letter-Naming Test). ANational Center for Early Development and Learning Dataset (NCEDL, 2002, 2004); CBermuda Preschool Study (1980); DCost, Quality and Outcomes Study (CQO, 1993–1994); KHead Start Family and children Experiences Survey (FACES, 2000) Cohort; S8-County Region of North-Central Indiana (Year NR); ZColorado QRIS.

Approach to learning outcomes

Two studies [45,53] reported associations between child-staff ratios and outcomes that look at children’s approach to learning (see Table A in S3 File). Zellman et al. [53] used five subscales of the Child Behavior Inventory measure and Phillips et al. [45] only used the Dependence and Task Orientation subscales from this measure. Both studies showed nonsignificant associations between ratios and this type of child outcome.

Cognitive outcomes.

Ten studies reported associations between child-staff ratios and cognitive outcomes measured using 7 different instruments (see Table A in S3 File). Most of these measures were only used in a single study with the exception of the Intelligence subscale of the Child Behavior Inventory (n = 3) and Preschool Inventory (n = 3). The majority of these studies found no association between child-staff ratios and this type of child outcome. Only 2 studies reported significant relationships, with better child-staff ratios related to better cognitive child outcomes [14,50].

Physical outcomes.

Zill et al. [40] reported on the relationship between child-staff ratios and a child physical development outcome. The outcome in this study was measured using the Design Copying instrument. The association between the variables of interest was not significant.

Combination outcomes.

One study reported the relationship between child-staff ratios and the Child Behavior Rating Scale [50]. The results of this study were nonsignificant.

Math outcomes.

Associations between child-staff ratios and 5 different math outcomes were reported in 12 studies (see Table A in S3 File). The WJ AP measure of child competency in mathematics is the only outcome that was reported in a large number of studies (n = 10), see Fig 2. The rest of the math outcomes were reported in only one study. Most of the results reported across these studies were nonsignificant suggesting a lack of relationship between child-staff ratios and math outcomes. Two studies reported significant results. However, while Colwell et al. [48] reported that higher child-staff ratios were related to better math outcomes, Anders et al. [47] reported that lower child-staff ratios were related to better math outcomes.

Language outcomes.

The associations between child-staff ratios and child language outcomes were reported for 21 studies included in this systematic review (see Table B in S3 File). The results for 19 different measures of language development were reported. Most of these measures were used in a single study. Only 7 measures were used in 2 or more studies (see Fig 2). The PPVT was reported in the largest number of studies (n = 16).

The vast majority of the results in these studies showed nonsignificant associations between child-staff ratios and child language outcomes. Among the small number of significant results that were reported, some [14,21,31,32,44] showed that lower child-staff ratios were related to better child language outcomes, while others [33,39] showed that higher child-staff ratios related to better child language outcomes.

Social-emotional: positive behavior outcomes.

Studies looking at the relationship between child-staff ratios and child positive behavior reported findings for 22 different outcomes in 18 studies (see Table C in S3 File). Most of these outcomes were only used in a single study. Most studies involved multiple indicators of positive behavior and child-staff ratios. Four outcomes (CBI–Considerateness and Sociability, SPSSP—Prosocial Responses, and Teacher Child Rating Scale–Social Competence) were reported in 2 and 4 studies respectively.

For most positive behavior outcomes reported in a single study the associations were nonsignificant. Of the significant results some [15,41,45,48] indicated that the classrooms with the smaller number of children per staff had children with more positive behavior outcomes (see Table C in S3 File).

Social-emotional: problem behavior outcomes.

Problem behavior was measured in 16 studies with 23 different outcome variables (see Table D in S3 File). With the exception of the Behavior Problems subscale from the TCRS that was used in 4 studies, the TRF CBCL—Externalizing and Internalizing subscales that was found in 2 studies and the PBQ–Hyperactivity scale that was reported in 2 studies, all other outcomes were used in a single study.

The majority of the reported results showed no significant association between child-staff ratios and problem behavior outcomes. A few significant results revealed fewer child behavior problems (or less children with problem behavior) in classrooms with lower ratios [15,18,32,41]. In contrast, one study reported more children with behavior problems in classrooms with lower ratios [45].

Overall, the vast majority of the results reported in the 29 studies reviewed as part of this systematic review suggest small or no associations between child-staff ratios and children’s cognitive, language, and social-emotional outcomes. Our examination of study characteristics such as publication year, study design and operationalization of child-staff ratios showed no patterns of association across the studies. Significant associations were reported for statistics that do (e.g., beta coefficients from regression) and do not (e.g., Pearson correlations and F-ratios) control for covariates.


Eighty-three different child outcomes were reported in the included studies (see S4 File for a complete list of outcome measures across all 29 studies). Outcome measures varied substantially in terms of the skill/ability being assessed (e.g., inattention; receptive language; counting task), informant (e.g., child assessment, teacher report, parent report), and psychometric properties (e.g., standardized norm-referenced measures vs. tasks developed by authors). Of the 83 child outcome measures reported in our sample of papers only the PPVT was utilized in three or more studies with statistics that enabled us to conduct meta-analyses (see Fig 3 and S4 File).

Fig 3. Meta-analysis results for the associates between child-staff ratios and child language outcome.

The association between child-staff ratios and the PPVT was not significant (pooled correlation coefficient 0.03 95% CI: 0.00 to .05). While this meta-analysis is based on only three studies, together they represent 5780 participants. The lack of association was consistent across the included studies with a very low value for the index of heterogeneity (I2 = 0).


Researchers have been examining early childhood education and care settings for decades. Yet, we still lack an acceptable empirically-based directive for stakeholders regarding the effects of quality of care on children’s development [54]. Conclusions concerning which aspects of quality matter most for children are often deduced from a few select intensive early intervention projects conducted in the 1960’s (i.e., The High Scope Perry Preschool Project) and 1970’s (i.e., The Carolina Abecedarian Project). The populations served and scope of services provided in these studies call into question their generalizability and applicability to today’s ECEC programs. In the several decades that have passed since these seminal longitudinal studies were undertaken, extensive research has documented mixed results. We conducted very comprehensive searches that lead to a review of 29 studies. This systematic analysis revealed that within the range of ratios found in the literature, variations in child-staff ratios for pre-school aged children classrooms have small, if any, associations with concurrent or subsequent child outcomes. Ratios showed substantial variability both within the studies that we included and across them. Therefore, the lack of associations between ratios and child outcomes found in our review are not explained by limited ranges of the ratios observed in the samples. Based on the small number of studies available for meta-analysis, no association was found between child-staff ratios and children’s receptive language as measured by the PPVT. However, it is important to note that the studies included in this review only included ratios that met local regulations. Thus, findings from this analysis cannot be used to argue for relaxing existing child-staff ratio regulations.

To our knowledge this is the first attempt to systematically review and meta-analyze this highly complex and heterogeneous literature. Strengths of our review include a very comprehensive search strategy, strict inclusion and exclusion criteria to screen the appropriate literature with a primary aim to systematically review and only meta-analyze homogeneous studies. However, this review also suffers from a number of limitations that mainly stem from the methodological issues of the primary research covered in this review.

A number of conceptual explanations that may contribute to the lack of associations between child-staff ratios and children’s outcomes are described below. The association in question may be curvilinear [55] or “J” shaped whereby thresholds must be met for quality to impact children’s outcomes. None of the studies we reviewed empirically explored this question. However, a systematic comparison of the results from those studies whose mean child-staff ratios fell within the lowest, highest, and mid-range of values in our systematic review sample did not produce evidence to support a curvilinear trend between ratios and child outcomes (results not shown).

It is possible that some children are more affected than others by environmental circumstances, a concept referred to as differential susceptibility [5658]. Thus, an interaction may exist between child characteristics (e.g., temperaments, family demographics) and the impact of ratios on child outcomes. Unfortunately, data were not available to enable us to assess this hypothesis. However, our qualitative synthesis did not suggest that findings systematically varied by the composition of the sample (i.e., children from at-risk backgrounds), publication year, or study design (i.e., longitudinal vs. cross-sectional). Nonetheless, researchers are encouraged to utilize study designs and statistical analyses that allow for the investigation of moderation, mediation, and ecological transactional models [59] in order for the ECEC literature to go beyond investigating the effect of ratios on the ‘average’ child.

Future research is needed to determine whether teacher characteristics (e.g., level of education; years of experience) and classroom/program variables moderate the effect of ratios on child development.

Finally, there is a mismatch in the unit of measurement in this literature as predictors (i.e., quality indicators) are almost always measured at the classroom level while child outcome variables are examined at the individual level [60]]. This may introduce imprecision in the linkages between quality and child outcomes. Advanced statistical techniques can account for different levels of measurement. However, the conceptual issue as to whether a classroom aggregate measure that estimates the average child’s experience of quality would be expected to impact the outcomes of an individual child remains.

As noted above, this area of research suffers from a number of methodological limitations. These methodological limitations within the ECEC literature may explain the lack of statistical associations between child-staff ratios and child outcome measures. In fact, due to these methodological issues and to limited variability in methodological quality between studies, a risk of bias assessment yielded non-informative results. As described below, methodological issues included: i) how child-staff ratios were operationalized; ii) the domains of child outcomes measured; iii) the psychometric properties of the outcome measures; and iv) study design.

There was enormous variability in the method of ratio data collection (e.g., via observations or staff/parent reports), the adults included in ratio counts, and how ratios were calculated. Previous research suggests that the validity of child-staff ratio measurements can be improved by lengthening the observation period [61] as ratios fluctuate over the course of the day [61,62]. Until consistent, empirically-based methods to measure ratios are adopted by ECEC researchers (and fully described in their papers), differences in ratios across studies may simply be an artifact of discrepancy in measurement approaches rather than actual variations in the numbers of children and adults present in each room.

Eighty-three different child outcome measures were identified in the studies we reviewed. Despite the large number of outcomes found in this literature, many key domains of child development are under-represented. This is consistent with a review of sixty-five studies of ECEC quality published between 1979 and 2005 [60], which like the current study, found an emphasis on socioemotional, language and cognitive development with limited consideration of motivational aspects of learning (“approaches to learning” such as task persistence or enthusiasm), physical development, or health outcomes. We argue that future research on the impact of ECEC quality on outcomes should focus on established measures with good psychometric properties to assess specific aspects of child development that are conceptually linked to the specific aspect of ECEC quality in question. Such an approach will go a long way towards helping researchers synthesize information across research in this area in the future.

The studies included in this review exhibited many measurement problems. Simply reporting information related to the reliability and validity of the child outcome measure(s) would be an important first step in addressing the methodological problems present in this area of research [60].

Finally, all but one of these studies used an observational design and therefore causal conclusions regarding the effect of ratios on child outcomes cannot be made. The one exception, the California Staff/Child Ratio Study [19] involved random assignment of children into groups with different ratio levels. Staff reports of antisocial, depressed, attention-deficit, and immature/dependent behaviors for each child in the classroom were not linked to better/worse ratios after controlling for baseline behavior scores. However, the period in which classrooms operated at the different ratios levels was relatively short in this study. It is imperative that researchers design studies that are more methodologically rigorous in the future. Conducting research on this highly regulated quality indicator is not easy and may require partnerships with policy makers and regulatory agencies involved in the oversight of ECEC programs.


Despite the substantial limitations of research in this area, the current study suggests that, within the range of permissible child-staff ratios, variations in ratios have small, if any, associations with concurrent and subsequent child outcomes. The small number of significant associations between child-staff ratios in preschool-aged classrooms and children’s developmental outcomes that were reported may reflect selection biases. Specifically, they may reflect family-level factors that play a role in child care selection, such as maternal education or family income, rather than true child care effects. However, as noted above, the research available for this systematic review included studies with significant methodological limitations and only studies with child-staff ratios that fell within current regulations. Thus, our findings should not be interpreted as indicating that regulation of ratios can be relaxed in any way. Rather, we emphasize that within the range of what is currently permissible by licensing regulations, better ratios in preschool-aged ECEC classrooms are not associated with better outcomes for children. This is consistent with findings from large-scale meta-analyses examining teacher-student ratios and student achievement in formal education systems [63]. While we stress that these findings should not be overstated, they do suggest that other areas of investment in quality improvement in ECEC programs, such as staff professional development, may yield better payoffs for the many stakeholders impacted by ECEC quality. In addition, our results indicate a strong need for comparative effectiveness type of research designs on this issue in multiple settings. These include prospective cohorts or cluster randomized studies, with different ratios and prospective criterion-specific evaluations of consumer (parents)-centric outcomes to guide future practice in this area, which affects wide-scale populations.


Special thanks to Joseph Beyene, Charles Pascal, Elaine Baxter-Trahair, Debra Mathews, Christa Japel, Lynn Karoly, Deborah Phillips and Elizabeth Uleryk. We thank Jenaya Webb, OISE/UT’s Public Services Librarian for her invaluable assistance throughout this project. Thank you also to Stephanie Aglipay, Teresa Pietropaolo, Anastasia Bukowski and Nadina Mahadeo for all of their help with the huge organizational task associated with this project.

Author Contributions

  1. Conceptualization: MP BF OF AB EM PSS.
  2. Data curation: BF OF AB EM.
  3. Formal analysis: MP BF OF AB EM.
  4. Funding acquisition: MP.
  5. Investigation: BF AB EM.
  6. Methodology: MP BF OF AB EM PSS.
  7. Project administration: MP.
  8. Resources: MP.
  9. Supervision: MP PSS.
  10. Validation: MP BF OF AB EM.
  11. Visualization: AB EM BF OF.
  12. Writing – original draft: MP BP.
  13. Writing – review & editing: MP BF OF AB EM PSS.


  1. 1. National Research Council and Institute of Medicine. From neurons to neighborhoods: the science of early childhood development. Shonkoff JP, Phillips DA, editors. Washington, DC: National Academies Press; 2000.
  2. 2. Ladd G, Herald S, Andrews R. Young children’s peer relations and social competence. In: Spodek B, Saracho O, editors. Handbook of research on the education of young children [Internet]. 2nd ed. Mahwah, NJ: Lawrence Erlbaum Associates Publishers; 2006. p. 23–54. Available from:
  3. 3. Seifert K. Cognitive development and the education of young children. In: Spodek B, Saracho O, editors. Handbook of Research on The Education of Young Children [Internet]. Mahwah, NY, US: Lawrence Erlbaum Associates; 2006. p. 9–22. Available from:
  4. 4. Laughlin L. Who’s minding the kids? Child care arrangements: spring 2011. Household economic studies. [Internet]. United States: United States Census Bureau; 2013 p. 1–23. Available from:
  5. 5. Statistics Canada. Child care: an eight-year profile. The Daily [Internet]. 2006 Apr 5 [cited 2015 Jul 1]; Available from:
  6. 6. Clarke-Stewart K, Allhusen V, Clements D. Nonparental caregiving. In: Bornstein N, ed Handbook of Parenting. Hillsdale, NJ, England: Lawrence Erlbaum Associates; 1995. p. 215–52. (Status and Social Conditions of Parenting; vol. 3).
  7. 7. Lamb M. Nonparental child care: context, quality, correlates, and consequences. In: Damon W, Sigel IE, Renninger KA, editors. Handbook of Child Psychology Vol 4: Child Psychology in Practice. 5th ed. Hoboken, NJ: John Wiley & Sons; 1998. p. 73–133.
  8. 8. Scarr S. Why child care has little impact on most children’s development. Curr Dir Psychol Sci. 1997;6(5):143–8.
  9. 9. Vandell D. Early child care: the known and the unknown. Merrill-Palmer Q. 2004;50(3):387–414.
  10. 10. American Academy of Pediatrics, Committee on Early Childhood, Adoption, and Dependent Care. Quality early education and child care from birth to kindergarten. Pediatrics. 2005;115(1):187–91. pmid:15630001
  11. 11. Zellman G, Perlman M. Child-care quality rating and improvement systems in five pioneer states: implementation issues and lessons learned [Internet]. Santa Monica, CA: RAND Corporation; 2008 [cited 2015 Jul 13] p. 1–67. Available from:
  12. 12. American Federation of State, County & Municipal Employees. We can do better: Child care aware of America’s ranking of state child care center regulations and oversight 2013 update [Internet]. 2013. Available from:
  13. 13. Friendly M, Ferns C, Prabhu N. Ratios for four and five year olds: what does the research say? what else is important? Univ Tor Childcare Resour Res Unit [Internet]. 2009; Available from:
  14. 14. Burchinal MR, Roberts JE, Riggins R, Zeisel SA, Neebe E, Bryant D. Relating quality of center-based child care to early cognitive and language development longitudinally. Child Dev. 2000 Apr;71(2):338–57.
  15. 15. NICHD Early Child Care Research Network. Child outcomes when child care center classes meet recommended standards for quality. Am J Public Health. 1999;89(7):1072–7. pmid:10394318
  16. 16. NICHD Early Child Care Research Network. Child care structure—> process—> outcome: direct and indirect effects of child care quality on young children’s development. Psychol Sci. 2002;13(3):199–206. pmid:12009038
  17. 17. Clarke-Stewart K, Gruber C, Fitzgerald L. Children at home and in day care. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc; 1994.
  18. 18. Dunn L. Proximal and distal features of day care quality and children’s development. Early Child Res Q. 1993;8(2):167–192.
  19. 19. Love J, Ryer P, Faddis B. Caring Environments: Program Quality in California’s Publicly Funded Child Development Programs: Report on the Legislatively Mandated 1990–91 Staff/Child Ratio Study. Portsmouth, NH: RMC Research Corporation; 1992.
  20. 20. Mashburn A, Pianta R, Hambre B. Measures of classroom quality in prekindergarten and children’s development of academic, language, and social skills. Child Dev. 2008;79:732–49. pmid:18489424
  21. 21. Howes C. Children’s experiences in center-based child care as a function of teacher background and adult: child ratio. Merrill-Palmer Q. 1997;43(3):404–25.
  22. 22. U.S. Department of Education, National Center for Education Statistics. The Condition of Education 2007 (NCES 2007–064) [Internet]. Washingon, DC: National Center for Education Statistics; 2007 [cited 2015 Jul 3]. Available from:
  23. 23. Cost, Quality and Outcomes Study (CQO). Web site. [Internet]. Frank Porter Graham Child Development Institute. [cited 2015 Jul 8]. Available from:
  24. 24. Early Childhood Longitudinal Program (ECLS). Web site. [Internet]. National Center for Education Statistics. [cited 2015 Jul 1]. Available from:
  25. 25. Effective Pre-School, Primary and Secondary Education (EPPSE). Web site. [Internet]. Institute of Education, University College London. [cited 2015 Jul 1]. Available from:
  26. 26. Head Start Impact Study (HS). Web site. [Internet]. Administration for Children and Families. [cited 2015 Jul 1]. Available from:
  27. 27. National Center for Early Development & Learning (NCEDL) multi-state study of pre-kindergarten. Web site. [Internet]. Frank Porter Graham Child Development Institute. [cited 2015 Jul 8]. Available from:
  28. 28. Head Start Family And Child Experiences Survey (FACES) series. Web site. [Internet]. Inter-University Consortium for Political and Social Research. [cited 2015 Jul 1]. Available from:
  29. 29. Study of Early Child Care and Youth Development (NICHD). Web site. [Internet]. U.S. Department of Health and Human Services, National Institutes of Health. 2015 [cited 2015 Jul 8]. Available:
  30. 30. Comprehensive Meta-Analysis. Engelwood, NJ: Biostat; 2005.
  31. 31. Dotterer A, Burchinal M, Bryant D, Early D, Pianta R. Universal and targeted pre-kindergarten programmes: a comparison of classroom characteristics and child outcomes. Early Child Dev Care. 2012;183(7):931–50.
  32. 32. Downer JT, López ML, Grimm KJ, Hamagami A, Pianta RC, Howes C. Observations of teacher–child interactions in classrooms serving Latinos and dual language learners: Applicability of the Classroom Assessment Scoring System in diverse settings. Early Child Res Q. 2012 Jan;27(1):21–32.
  33. 33. Howes C, Burchinal M, Pianta R. Ready to learn? children’s pre-academic achievement in pre-kindergarten programs. Early Child Res Q. 2008;23:27–50.
  34. 34. Mashburn A, Justice L, Downer J, Pianta R. Peer effects on children’s language achievement during pre-kindergarten. Child Dev. 2009;80(3):686–702. pmid:19489897
  35. 35. Sabol T, Soliday Hong S, Pianta R, Burchinal M. Can rating pre-k programs predict children’s learning? Science. 2013;341(6148):845–6. pmid:23970684
  36. 36. Reid J, Ready D. High-quality preschool: the socioeconomic composition of preschool classrooms and children’s learning. Early Educ Dev. 2013;24(8):1082–111.
  37. 37. Burchinal M, Cryer D, Clifford R, Howes C. Caregiver training and classroom quality in child care centers. Appl Dev Sci. 2002;6(1):2–11.
  38. 38. Burchinal M, Nelson L. Family selection and child care experiences: Implications for studies of child outcomes. Early Child Res Q. 2000;15(3):385–411.
  39. 39. Zill N, Resnick G, Kim K, O’Donnell K, Sorongon A, McKey R, et al. Head start FACES 2000: a whole-child perspective on program performance [Internet]. Washington DC: Administration for Children and Families, U.S. Department of Health and Human Services; 2003. Available:
  40. 40. Zill N, Resnick G, Kim K, O’Donnell K, Sorongon A, Ziv Y, et al. Head start performance measures enter family and child experiences survey (FACES 2000): Technical report [Internet]. Washington, DC: U.S. Department of Health and Human Services, Administration for Children and Families, Office of Planning, Research and Evaluation; 2006. Available from:
  41. 41. Clarke-Stewart K, Lee A, Allhusent V, Kim M, McDowell D. Observed differences between early childhood programs in the U.S. and Korea: reflections of “developmentally appropriate practices” in two cultural contexts. J Appl Dev Psychol. 2006;27(5):427–43.
  42. 42. Holloway S, Reichhart-Erickson M. The relationship of day care quality to children’s free-play behavior and social problem-solving skills. Early Child Res Q. 1988;3(1):39–53.
  43. 43. Holloway S, Reichart-Erickson M. Child-care quality, family structure, and maternal expectations: relationship to preschool children’s peer relations. J Appl Dev Psychol. 1989;10(3):281–98.
  44. 44. McCartney K. Effect of quality of day care environment on children’s language development. Dev Psychol. 1984;20(2):244–60.
  45. 45. Phillips D, McCartney K, Scarr S. Child-care quality and children’s social development. Dev Psychol. 1987;23(4):537–43.
  46. 46. Moher D, Liberati A, Tetzlaff J, Altman D, The PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6(7).
  47. 47. Anders Y, Rossbach H, Weinert S, Ebert S, Kuger S, Lehrl S, et al. Home and preschool learning environments and their relations to the development of early numeracy skills. Early Child Res Q. 2012;27(2):231–44.
  48. 48. Colwell N, Gordon R, Fujimoto K, Kaestner R, Korenman S. New evidence on the validity of the Arnett Caregiver Interaction Scale: results from the Early Childhood Longitudinal Study-Birth Cohort. Early Child Res Q. 2013;28(2):218–33. pmid:24058264
  49. 49. Owen M, Klausli J, Mata-Otero A, Caughy M. Relationship-focused child care practices: quality of care and child outcomes for children in poverty. Early Educ Dev. 2008;19(2):302–29.
  50. 50. Seppanen P, Godin K, Metzger J, Bronson M, Cichon D. Observation Study of Early Childhood Programs. Final Report Volume II: Chapter 1-Funded Early Childhood Programs [Internet]. 1993 [cited 2015 Jul 3]. Available from:
  51. 51. Studer M. Quality of center care and preschool cognitive outcomes: differences by family income. Sociol Stud Child Dev. 1992;5:49–72.
  52. 52. Travers J, Goodson BD, Singer JD, Connell DB. Research results of the national day care study. Final report of the national day care study. Volume II. [Internet]. Cambridge, MA: Abt Associates, Inc; 1980. Available from:
  53. 53. Zellman G. Assessing the validity of the Qualistar early learning quality rating and improvement system as a tool for improving child-care quality [Internet]. Santa Monica, CA: RAND Education; 2008 [cited 2015 Nov 24]. Available from:
  54. 54. Shpancer N. The effects of daycare: persistent questions, elusive answers. Early Child Res Q. 2006;21(2):277–237.
  55. 55. Le V, Schaack D, Setodji C. Identifying baseline and ceiling thresholds within the qualistar early learning quality rating and improvement system. Early Child Res Q. 2015;30(Part B):215–26.
  56. 56. Belsky J, Pluess M. Beyond diathesis stress: differential susceptibility to environmental influences. Psychol Bull. 2009;135(6):885–908. pmid:19883141
  57. 57. Berry D, McCartney K, Petrill S, Deater-Deckard K, Blair C. Gene-environment interaction between DRD4 7-repeat VNTR and early child care experiences predicts self regulation abilities in prekindergarten. Dev Psychobiol. 2014;56:373–91. pmid:23460366
  58. 58. Berry D, Blaire C, Ursache A, Willoughby M, Garett-Peters P, Veron-Feagans L, et al. Child care and cortisol across early childhood: context matters. Dev Pscyhology. 2014;50(2):514–25.
  59. 59. Lynch M, Cocchetti D. An ecological-transactional analysis of children and contexts: the longitudinal interplay among child maltreatment, community violence, and children’s symptomatology. Dev Psychopathol. 1998;10(2):235–57. pmid:9635223
  60. 60. Zaslow M, Halle T, Martin L, Cabrera N, Calkins J, Pitzer L, et al. Child outcome measures in the study of child care quality. Eval Rev. 2006;30(5):577–610. pmid:16966677
  61. 61. Le V, Perlman M, Zellman G, Hamilton L. Measuring child-staff ratios in child care centers: balancing effort and representativeness. Early Child Res Q. 2006;21(3):267–79.
  62. 62. Setodji C, Le V, Schaack D. Accounting for movement between child care classrooms: does it change teacher effects interpretations? J Appl Dev Psychol. 2012;33(1):1–12. pmid:22389546
  63. 63. Hattie J. Visible learning: a synthesis of over 800 meta-analyses relating to achievement. New York, NY: Routlege; 2009.