This chapter compares average scores and variation in children’s early learning and development outcomes across the jurisdictions participating in IELS 2025. It then looks at associations between average scores and jurisdictions’ contextual factors. The chapter concludes by examining correlations between children’s scores and parents’ and teacher’s perceptions of their skills, and between children’s outcomes across domains of assessment.
Building Strong Foundations for Life
Results from the 2025 Early Learning and Child Well-being Study
Abstract
IELS 2025 finds significant differences in early learning and development outcomes across participating jurisdictions. In foundational learning and executive function domains, children in Korea consistently achieve high scores. Children in England (United Kingdom), in both emergent literacy and emergent numeracy, and in Baku and Sumgait (Azerbaijan), in emergent numeracy, also have comparatively high scores. In executive function domains, jurisdictions cluster more often around the international mean. In social and emotional development, several jurisdictions combine scores above and below the international mean in different domains.
In foundational learning domains, gaps between high- and low-performing children are relatively large in the United Arab Emirates and Baku and Sumgait (Azerbaijan), and relatively small in Korea, the Flemish Community (Belgium) and the Netherlands. These gaps are more homogeneous between jurisdictions in executive function domains. In social and emotional development domains, differences between children with the highest and lowest scores are smallest in the Netherlands, and largest in the United Arab Emirates and Malta.
Age differences in children’s scores as they grow from age 5 to age 6 are more pronounced in the domains of emergent numeracy and inhibition, followed by emergent literacy, working memory and mental flexibility, and smaller in social and emotional development domains.
On average, 36% of the variation in foundational learning scores, and 22% of the variation in both executive function and social and emotional development scores occur between early childhood education and care (ECEC) centres or schools. This suggests that, within jurisdictions, the characteristics of ECEC centres/schools play a substantive role in explaining differences in children’s early learning and development outcomes.
ECEC teachers’ perceptions about children’s skills show stronger correlations with children’s assessed early learning and development outcomes than parents’ perceptions.
Correlations between children’s outcomes in different domains of early learning and development are consistently positive but vary in strength. A strong correlation exists between scores in the foundational learning domains of emergent literacy and emergent numeracy. Moderate correlations exist between executive function domains, particularly with inhibition. In social and emotional development, directly and indirectly assessed domains correlate weakly. Correlations are sometimes higher between domains in different dimensions than between domains in the same dimension. Scores in emergent literacy and emergent numeracy correlate strongly with scores in working memory and mental flexibility, and emotion identification correlate more with those than with most other social and emotional development domains.
There is strong similarity in the relative position of children in the performance distributions across dimensions of early learning and development. Large shares of children with high or low scores in foundational learning have also high and low scores in executive function, respectively. The overlap between the distributions of children’s scores is less pronounced between foundational learning and social and emotional development. Overall, a minority of five-year-olds combine relatively strong skills in one dimension with relatively weak skills in others, but cross-over is higher in intermediate layers of performance.
IELS assesses the learning and development outcomes of 5-year-olds as a window into their acquisition of basic skills that matter for their future educational success and well-being. By focusing on an age at which children are close to starting primary school, IELS provides participating jurisdictions with reliable, valid and comparable data about the levels and drivers of these outcomes and about gaps between different groups of children. This information is essential to evaluate the performance of jurisdictions’ early years policies in providing a strong start in life for all children.
This chapter describes jurisdictions’ average performance in IELS 2025, setting the stage for the rest of the report and for further analyses of IELS 2025 data. This system-level perspective serves to identify jurisdictions’ main strengths in promoting early learning and development outcomes as well as areas where additional investments may be needed, both for future cohorts of young children and, in the near term, for primary school policies.
The first part of the chapter presents average scores for foundational learning, executive function and social and emotional development domains and compares performance across jurisdictions. Wide differences in social and economic conditions between the jurisdictions in IELS 2025 are the background for these comparisons. The second part of the chapter describes within-jurisdiction variation in early learning and development outcomes, preparing the ground for more detailed analyses of the factors driving differences among various groups of children.
The chapter concludes by looking at correlations between children’s scores in the 10 domains of IELS’ multidimensional assessment model and parents’ and teachers’ perceptions of children’s early skills, combining complementary perspectives about how five-year-olds are learning and developing. Lastly, the chapter presents correlations between children’s scores across those domains of assessment as well as their positions in joint performance distributions in different dimensions of early learning and development, providing insights into how different types of early skills relate to each other.
IELS assesses children’s outcomes in 10 early learning and development domains using the same metric, with international means standardised to 500 score points. Table 3.1 presents the mean scores of each of the jurisdictions participating in IELS 2025, and Table 3.2 shows how jurisdictions cluster when their means are not statistically significantly different. Mean scores are presented for all assessed outcomes, while performance clusters are based on the composite scores for the three broad dimensions of early learning and development (see the Reader’s guide). It is important to emphasise that no single ranking can do justice to the richness of information that IELS provides and, more importantly, to the variety of goals that jurisdictions pursue in their early years policies. This section also highlights the statistical uncertainty in IELS results when comparing jurisdictions.
IELS 2025 finds significant differences in early learning and development outcomes across participating jurisdictions. In the foundational learning and executive function dimensions, children in Korea stand out for consistently achieving higher scores than those in other jurisdictions. Children in England (United Kingdom), in both emergent literacy and emergent numeracy, and children in Baku and Sumgait (Azerbaijan), in emergent numeracy, also have comparatively high scores in the foundational learning dimension. There is generally less differentiation across executive function domains, with jurisdictions clustering around the international mean.
Note: OECD members are listed in black font. Partner countries/economies are listed in blue font. Sub-national jurisdictions are listed in italics. *Estimates for Brazil correspond to averages across the 3 participating Brazilian states. **IELS Adjudicated Region. Blue (grey) shading indicates scores are statistically significantly above (below) the IELS 2025 international averages, which are standardised to 500 score points in each domain. For more information, see the IELS 2025 Technical Report (OECD, 2026[1]). Jurisdictions are ranked in descending order of average composite scores for each dimension.
Source: OECD, IELS 2025 Database, Tables B.3.1, B.3.2 and B.3.3.
Note: OECD members are listed in black font. Partner countries/economies are listed in blue font. Sub-national jurisdictions are listed in italics. *Estimates for Brazil correspond to averages across the 3 participating Brazilian states. Regions are excluded from comparisons. Jurisdictions listed together horizontally have average composite scores not statistically significantly different from each other. Blue (grey) shading indicates scores are statistically significantly above (below) the IELS 2025 international averages, which are standardised to 500 score points in each domain. For more information, see the IELS 2025 Technical Report (OECD, 2026[1]). Jurisdictions are ranked in descending order of average composite scores for each dimension.
Source: OECD, IELS 2025 Database, TablesB.3.1, B.3.2 and B.3.3.
In the social and emotional development dimension, children in England (United Kingdom) achieve high average scores in three out of the five domains of assessment, which is also the case in two domains for children in Korea, Malta and the United Arab Emirates. Overall, results appear less consistent in this dimension, with children in four jurisdictions – the Flemish Community (Belgium), Malta, the Netherlands and Korea – combining average scores above and below the international mean in different domains. When comparing mean scores between jurisdictions (Table 3.2), it is important to note that only statistically significant differences should be considered (see the Reader’s guide).
How skills are distributed across the child population complements information provided by jurisdiction averages. Two similar average performances can result from different distributions within jurisdictions, for instance, when the majority of children have scores near the average and when equal shares of children have scores distant from the average. In the first case, a compressed distribution means less differentiation in the skills that children attain by age 5, whereas in the second case, a wide distribution implies different groups of children attain rather unequal levels of skills. These two scenarios can have different policy implications.
Overall variation in outcomes within jurisdictions
In the foundational learning domains of emergent literacy and emergent numeracy, children in the Flemish Community (Belgium), Korea and the Netherlands have the smallest variability in scores, while children in the United Arab Emirates and Baku and Sumgait (Azerbaijan) have the largest variability, as measured by standard deviations1 (Table B.3.1). Scores in the executive function domains display less variation overall, with no clear pattern of results across jurisdictions (Table B.3.2). Variation is again comparatively low among children in the Netherlands across the five social and emotional development domains, as it is among children in Ceará, Pará and São Paulo (Brazil) in three domains and children in the Flemish Community (Belgium) in two domains. By contrast, variation in this dimension tends to be large among children in the United Arab Emirates and Baku and Sumgait (Azerbaijan), albeit not in all domains. In jurisdictions such as England (United Kingdom) and Korea, children’s scores vary less than international averages in the directly assessed domains of emotion identification and emotional attribution, but more in some of the indirectly assessed domains (Table B.3.3). Overall, across all early learning and development outcomes, there is a weak and inconsistent association between jurisdictions’ mean performance and the degree of variability in children’s scores.
Another measure of variation in outcomes is the difference that separates the children with the highest and lowest scores within a jurisdiction (i.e. 90th-10th inter-decile range)2 (Figure 3.1). For the 10 domains of early learning and development assessed in IELS 2025, the international averages for these differences range from 220 to 250 score points. This means that the scores of children in the upper and lower portions of the performance distribution tend to differ by approximately two to two and a half standard deviations. This level of variation is broadly consistent with the inter-decile ranges observed in PISA 2022 among 15‑year-olds (OECD, 2023[2]).
In foundational learning domains, the widest gaps between high- and low-performing children are found in the United Arab Emirates and Baku and Sumgait (Azerbaijan), where inter-decile ranges are above 280 score points in both emergent literacy and emergent numeracy. By contrast, these differences remain below 200 score points in Korea (in emergent literacy) and in the Flemish Community (Belgium) and the Netherlands (in emergent numeracy). Inter-decile ranges tend to be smaller in jurisdictions with higher average scores in emergent literacy, but no such association exists in emergent numeracy.
Distances between children in the upper and lower parts of the performance distribution are more homogeneous across jurisdictions in executive function domains, reflecting the lesser overall variation in this dimension of early development. In inhibition and mental flexibility, differences between children tend to be larger in jurisdictions with higher average scores on these domains, while the opposite holds in working memory.
Regarding social and emotional development domains, differences between children with the highest and lowest scores are smallest in the Netherlands, being below 200 score points in four of these domains, and largest in the United Arab Emirates and Malta, and generally mirror the overall level of variation in each jurisdiction. In trust, pro-social behaviour and non-disruptive behaviour, inter-decile ranges are larger in jurisdictions with higher average scores in these domains.
Differences between the 90th and 10th percentiles of the performance distribution in foundational learning, executive function, and social and emotional development domains
Note: *Estimates for Brazil correspond to averages across the 3 participating Brazilian states. For more information, see the IELS 2025 Technical Report (OECD, 2026[1]). Jurisdictions are ranked in ascending order of the difference in scores between the 90th and 10th percentiles of their performance distribution.
Source: OECD, IELS 2025 Database, Tables B.3.1, B.3.2 and B.3.3.
Variation by age
Children learn and develop rapidly during their early years. Age differences in early learning and development outcomes primarily reflect naturally occurring increases in children’s abilities, driven by cognitive and physical maturation processes that start from conception and continue into the pre-school years (Brown and Jernigan, 2012[3]) (Figueroa and An, 2016[4]). Differences in scores by month of age between ages 5 and 6 provide an estimate of the rate of expected learning and developmental progressions and help examine whether gaps between younger and older children vary across the three dimensions of early learning and development assessed by IELS.
Results in Figure 3.2 show that children’s scores in early learning and development outcomes increase as children grow from age 5 to age 6. However, the gains associated with age are not consistent across outcomes. The increase is most pronounced in the domains of emergent numeracy and inhibition, with scores increasing by approximately 70 score points over this age range on average across jurisdictions. In emergent literacy, working memory and mental flexibility, the average increase ranges from 43 to 58 score points. By contrast, scores raise more moderately in the social and emotional development dimension, particularly in the domains of trust, pro-social behaviour, and non-disruptive behaviour, where the average increase is between 15 and 20 score points and where scores level off or decrease slightly between 5 years and 10 months and 6 years of age. Gains in emotion identification and emotional attribution are more similar to those observed in the foundational learning and executive function domains, with average increases of 42 and 32 score points, respectively.
Variation in children’s outcomes by month of age is not consistent across jurisdictions. In England (United Kingdom), score differences between younger and older children are larger than international averages across all domains of assessment. In the foundational learning and executive function dimensions, these differences are also large among children in the Netherlands while small among children in Baku and Sumgait (Azerbaijan), Ceará, Pará and São Paulo (Brazil) and Malta, relative to international averages. Across other jurisdictions, the size of age differences varies by domain, without displaying a clear pattern (Tables B.3.5 and B.3.6). In the social and emotional development domains of emotion identification and emotional attribution, these age differences are also large among children in the United Arab Emirates and the Flemish Community (Belgium), whereas they are small among children in Baku and Sumgait (Azerbaijan), Ceará, Pará and São Paulo (Brazil) and Korea. However, in indirectly assessed domains, differences between jurisdictions tend to be narrow (Table B.3.7).
Mean scores in foundational learning, executive function, and social and emotional development domains by month of age of five-year-olds
Note: Estimates exclude children younger than 5 years and older than 6 years, who represent less than 4% of children in IELS 2025 samples, on average across jurisdictions. Red lines correspond to the IELS average, and grey lines to jurisdictions. *Estimates for Brazil correspond to averages across the 3 participating Brazilian states. For more information, see the IELS 2025 Technical Report (OECD, 2026[1]).
Source: OECD, IELS 2025 Database, Tables B.3.5, B.3.6 and B.3.7.
Variation between and within ECEC centres and schools
The early learning and development outcomes of five-year-olds in IELS 2025 vary widely, and that variation can be broken down into differences at the child, ECEC centre/school and jurisdiction levels. This decomposition analysis builds on the design of IELS, which samples participating children through ECEC centres/centres (see the Reader’s guide)3. Identifying the relative share of different components of the overall variation in children’s outcomes can provide valuable policy insights. If a large share of the variation is attributable to differences in performance between jurisdictions, this suggests that their system-level characteristics (e.g. economic and social conditions, early years policies) strongly influence children’s outcomes. Likewise, if differences between ECEC centres/schools account for a large share of the overall variation in outcomes within a jurisdiction, then differences in their characteristics are important for policy to consider.
Ensuring consistently high quality through ECEC centres/schools is a major challenge for any education system. Within a jurisdiction, performance differences between early education settings may be attributable to how general features of ECEC systems (e.g. governance and funding models) or specific quality standards (e.g. curriculum frameworks, staff-to-child ratios, qualification requirements for ECEC teachers) apply to different segments of the system (OECD, 2025[5]). Performance differences can also result from the socio-economic composition of settings, which in turn may reflect factors such as residential segregation, and thus from disparities in the resources that ECEC centres/schools and the families that they serve can mobilise. In this vein, the share of the variation in outcomes occurring between ECEC centres/schools, rather than within them, can be seen as a proxy measure for the variation in the quality of provision between ECEC centres/schools as well as for the concentration of socio-economic disadvantage across settings. Recent data from TALIS Starting Strong 2024 show that pre-primary ECEC centres with very small and large shares of children from socio-economically disadvantaged homes coexist in many countries (OECD, 2025[6]).
In IELS 2025, about 6% of the variation in foundational learning composite scores, 13% of the variation in executive function composite scores, and 2% of the variation in social and emotional development composite scores is linked to average differences in children’s scores between participating jurisdictions. These results indicate that system-level contextual characteristics influence children’s outcomes, while also implying that most of the variation in children’s outcomes occurs within jurisdictions, rather than between them.
Figure 3.3 shows the share of this within-jurisdiction variation in outcomes that, in turn, occurs between ECEC centres/schools in IELS 2025. On average across jurisdictions, 36% of the variation in foundational learning composite scores, and 22% of the variation in both executive function composite scores and social and emotional development composite scores is attributable to performance differences between ECEC centres/schools. This suggests that the characteristics of ECEC centres/schools play an important role in explaining children’s early learning and development outcomes. This includes system or institutional factors relating to the quality of ECEC and early schooling (see Chapter 6), as well as the extent to which children are concentrated in different centres/schools according to their socio-economic background. Across the three dimensions of early learning and development, the between-centres/schools percentage of variance in outcomes is strongly correlated (with coefficients in the range of .50 to .67) with the share of variation in socio-economic status at the centre/school level (Table B.3.8).
Importantly, the average variance split in Figure 3.3 masks substantial differences between jurisdictions. Across the three dimensions of early learning and development, the share of between-ECEC centres/schools variation in children’s scores is systematically larger than the IELS average in Baku and Sumgait (Azerbaijan), Ceará, Pará and São Paulo (Brazil) and the United Arab Emirates, accounting for instance for between 45% and 65% of the variation in foundational learning composite scores. These are the three jurisdictions with more variation in the socio-economic composition of centres/schools in IELS 2025. By contrast, the share of variance in outcomes between ECEC centres/schools is smallest, across the three dimensions, in England (United Kingdom) and the Netherlands, two jurisdictions where five-year-olds attend the final years of ECEC in school-based settings, and with comparatively low variation in the socio-economic makeup of these settings. Differences between jurisdictions in Figure 3.3 are therefore likely to reflect the heterogeneity that exists within them in both the quality of ECEC/schools and the socio-economic composition of ECEC centres/schools. Meanwhile, most of the within-jurisdiction variation in children’s outcomes remains attributable to differences in children’s characteristics, including their individual backgrounds and home environments. Subsequent chapters in the report explore those factors.
Percentage of within-jurisdiction variation in composite scores in foundational learning, executive function and social and emotional development attributable to differences in outcomes between ECEC centres/schools
Note: The variance decomposition analysis between vs. within ECEC centres/schools is based on a two-level regression model (i.e. child and ECEC centre/school levels), with normally distributed residuals, and estimated through maximum likelihood. *Estimates for Brazil correspond to averages across the 3 participating Brazilian states. For more information, see the IELS 2025 Technical Report (OECD, 2026[1]). Jurisdictions are ranked in descending order of the between-ECEC centre/school variation in composite scores in each area of early learning and development as a percentage of the total variation in these outcomes across jurisdictions.
Source: OECD, IELS 2025 Database, Table B.3.8.
Analyses of children’s average outcomes across diverse countries and education systems poses numerous challenges because the social, economic and cultural context of jurisdictions being compared are often very different. Each jurisdiction’s contextual factors provide the background for its specific findings, as these affect the family and the wider social environments in which the five-year-olds included in this study were growing up. For instance, the relative prosperity of some jurisdictions allows them to allocate more funding to early years policies and enrol more children in high-quality ECEC services at an earlier age, whereas others are constrained by a lower national income. For children, opportunities for participating in ECEC and experiencing supportive home environments are limited in high poverty contexts. However, it is not only economic conditions that matter for promoting early childhood outcomes at the system-level; for instance, the level of education of previous generations also influences the societal context in which young learn and develop. The participating jurisdictions in IELS 2025 differ in scale, economic development, demographic composition and in the features of the early care and education systems and other policies aimed at supporting young children and their families. Annex A presents a selection of statistical indicators and descriptors of basic features of these jurisdictions’ ECEC systems. This information provides context for interpreting the results that are presented in the report.
As an illustration, Figure 3.4 shows the system-level correlation between jurisdictions’ average performance in the foundational learning dimension and the percentage of 25-34-year-olds who had attained tertiary education in 2020 (see Annex A). This population group corresponds loosely to the cohort of parents of the 5-year-olds assessed in IELS 2025. According to this elementary analysis, variation in the share of tertiary-educated young adults would account for 43% of the variation in the foundational learning composite score between jurisdictions. A stronger correlation exists between this contextual indicator and jurisdictions’ average composite scores in executive function (53% of variation), whereas the association with the composite score in the social and emotional development dimension is slightly weaker (30% of variation). The individual-level association between a child’s scores in early learning development outcomes and their parents’ level of education is explored later in the report (see Chapter 5).
Percentage of 25–34-year-olds with tertiary education in 2020 and average foundational learning composite score of five-year-olds in 2025
Note: *Estimates for Brazil correspond to averages across the 3 participating Brazilian states.
Source: Annex A and OECD, IELS 2025 Database, Table B.3.1.
IELS 2025 not only provides children’s scores in 10 different domains of early learning and development but also information on the perceptions that their parents and ECEC/school teachers have about children’s skills (see Chapter 2). This enables IELS to compare information from direct and indirect assessments and also gauge children’s skills across a broader scope of areas, based on the competences and behaviours that parents and teachers observed at home or at the ECEC centres/schools that children attend. Research documents moderate correlations between assessments of children’s early learning and development outcomes and adult reports of children’s developmental competencies, with the concordance generally being higher regarding early literacy and early numeracy skills than executive function and self-regulation skills (Li, Rao and Gong, 2026[7]).
In IELS 2025, parents and ECEC/school teachers reported on how children were developing regarding cognitive and motor skills and social and emotional skills, while teachers reported also about children’s global capabilities. This information was used to construct three indices summarising the perceptions from familiar adults, using an IRT estimation method consistent across contextual and outcomes measures (see the Chapter 2 and the IELS 2025 Technical Report (OECD, 2026[1])).
Correlations between assessments and parent and ECEC/school teacher perceptions about children’s skills
Table 3.3 presents the international averages for correlations between assessed early learning and development outcomes and parents’ and ECEC/school teachers’ perceptions of children’s skills. Correlations of children’s scores in foundational learning and executive function domains and in the two directly assessed domains of social and emotional development are stronger with the index of global capabilities than with the rest of measures based on adult reports. This index synthetises information from teachers about children’s ability to complete a series of tasks involving emergent literacy or emergent numeracy commonly encountered in early education settings (e.g. sort a group of objects by shape, size or colours; count in multiples; speak using multiple sentences to explain something) – as well as tasks related to empathy (e.g. draw inferences about how a character felt after listening to a story). Substantial correlations also exist between the teacher-based indices of cognitive and motor skills and of socio-emotional skills, on the one hand, and children’s scores in the indirectly assessed domains of trust, pro-social behaviour and non-disruptive behaviour, on the other, a result that is likely explained by both sets of measures being based on reports from ECEC/school teachers.
Overall, across the three dimensions, parents’ perceptions about children’s skills appear less consistent with children’s scores than those of teachers, as indicated by lower correlations.
Average correlation coefficients across jurisdictions
Note: All coefficients are statistically significant at the 5% level (p<.05). The term ECEC/school teacher is shorthand for the person completing the Staff questionnaire (see the Reader’s guide). For more information on indices based on parent and ECEC teacher reports, see Chapter 2 and the IELS 2025 Technical Report (OECD, 2026[1]).
Source: OECD, IELS 2025 Database.
For young children, the different dimensions of their early learning and development are inter-related and mutually reinforcing. Notwithstanding the positive correlations extensively documented between these dimensions, research continues to explore potential trade-offs between different types of skills – based, for instance, on concerns about the relative balance between direct instruction and play in early education settings (Le et al., 2019[8]) – as well as the dynamics of early skills transfer and co-development (Thijssen et al., 2024[9]) (Hart et al., 2026[10]) (see also Chapter 1). Because IELS data are cross-sectional rather than longitudinal, they cannot illustrate potential skills complementarities or trade-offs over time. However, they provide a unique opportunity to examine how multiple dimensions of early learning and development, and specific domains within these, relate to each other in large samples of five-year-olds internationally.
Across the jurisdictions participating in IELS 2025, the average correlations between children’s composite scores in different dimensions of early learning and development are consistently positive but vary significantly in strength (Figure 3.5). These scores are more strongly correlated between the dimensions of foundational learning and executive function (r=.64), but the associations are weaker with those for social and emotional development (r=.46 and r=.40, respectively).
Within dimensions, a very strong correlation exists between children’s scores in the foundational learning domains of emergent literacy and emergent numeracy (r=.72). In executive function domains, correlations are still substantial, particularly between working memory and mental flexibility (r=.52), but slightly lower with inhibition. In social and emotional development, scores in trust and pro-social behaviour are also very strongly correlated (r=.79), and a more modest association exists between the two empathy-related domains (emotion identification and emotional attribution) and the two behavioural domains (pro-social and non-disruptive behaviours). Notably, though, within this dimension, correlations between children’s scores in directly and indirectly assessed domains are weak.
An important insight from IELS 2025 is that associations between different domains or early learning development are sometimes stronger across rather than within some of its dimensions. For instance, correlations between children’s scores in emergent literacy and emergent numeracy are higher with their scores in working memory and mental flexibility than with scores in inhibition. Meanwhile, scores in emotion identification correlate more closely with scores in most foundational learning and executive function domains than with scores in most other social and emotional development domains.
In some jurisdictions, relationships between performance in different domains deviate significantly from these averages. For instance, the correlation between scores in the foundational learning domains of emergent literacy and emergent numeracy is stronger than the IELS average among children in England (United Kingdom) and the Netherlands, and weaker among children in Ceará, Pará and São Paulo (Brazil). In turn, correlations between scores in executive function domains are stronger than average among children in England (United Kingdom), and weaker among children in Baku and Sumgait (Azerbaijan). In turn, children’s scores in social and emotional development domains are generally more strongly correlated with each other in the Flemish Community (Belgium).
Average correlation coefficients across jurisdictions between children’s scores in foundational learning, executive function and social and emotional development domains
Note: All coefficients are statistically significant at the 5% level. For more information, see the IELS 2025 Technical Report (OECD, 2026[1]).
Source: OECD, IELS 2025 Database.
IELS 2025 provides further insights into the early skill sets of young children by enabling an exploration of where five-year-olds stand regarding particular combinations of early learning and development outcomes (Figure 3.6). This analysis suggests a strong similarity in the relative positions of children when the distributions of composite scores of two dimensions of early learning and development are considered together. This means that the majority of children with high or low scores in one area tend to have similarly high or low scores in other areas, and that only a minority of five-year-olds combine relatively strong skills in one dimension with relatively weak skills in others.
On average across jurisdictions, 56% of the five-year-olds in the bottom quarter of the distribution of foundational learning composite scores are also in the bottom quarter of the same distribution in executive function (right-hand side of Figure 3.6). At the other end of the skills range, 57% of the children in the top quarter of the distribution of foundational learning scores are also top scorers in executive function. More generally, more than eight out of ten of these children with either the highest or lowest composite scores in foundational learning are in the upper or lower halves of the performance distribution in executive function, respectively. This implies that the cross-over between relatively high performance in one dimension and relatively low performance in the other is very limited. Nonetheless, almost 40% of children in any of the two intermediate quarters of the distribution of foundational learning scores are in the opposite half of the performance distribution in executive function, which suggests more diverse skill profiles for children with intermediate levels of skills in these dimensions (Table B.3.9).
Children’s positions in the distribution of foundational learning skills are less strongly correlated with their relative level of social and emotional development skills (left-hand side of Figure 3.6). In the bottom quarter of the performance distribution in foundational learning, 49% of children have scores within the same range in social and emotional development, whereas among top performers, this proportion is 44%. Inversely, up to 25% of five-year-olds with the lowest foundational learning composite scores are in the upper half of the performance distribution in social and emotional development, and 28% of the children with the highest scores in the first dimension are in the bottom half of the distribution in the second dimension (Table B.3.10).
While these results depict only aggregate patterns, as they rely on composite scores and international averages, IELS 2025 data represent an opportunity to extend this analysis to jurisdiction-specific patterns and to particular combinations of specific early learning and development domains, which may yield important insights for ECEC and primary education policies, as well as for early years policies in other sectors. For instance, knowing whether child populations have relatively low levels of skills in a single dimension or domain of early learning and development or in several of those simultaneously can inform the design of both universal policies and of more targeted interventions when the characteristics of children with specific skill sets are identified. Similarly, jurisdictions gaining a better understanding of how the early learning and development outcomes of five-year-olds combine in different areas can use this information to design curricular programmes and pedagogical practices in the early years of primary school.
Distribution of five-year-olds across quarters of the performance distributions of composite scores in foundational learning, executive function and social and emotional development, on average across jurisdictions
Note: The graph shows the joint distribution of children across quarters of performance distributions. Flows between columns represent the percentage of children in a specific quarter of the distribution of foundational learning composite scores (central column) who are also in different quarters of the distributions of social and emotional development composite scores (flows to the left column) and of executive function composite scores (flows to the right column). For more information, see the Reader’s guide and the IELS 2025 Technical Report (OECD, 2026[1]).
Source: OECD, IELS 2025 Database, Table B.3.9 and B.3.10.
References
[3] Brown, T. and T. Jernigan (2012), “Brain Development During the Preschool Years”, Neuropsychology Review, Vol. 22/4, pp. 313-333, https://doi.org/10.1007/s11065-012-9214-1.
[4] Figueroa, R. and R. An (2016), “Motor Skill Competence and Physical Activity in Preschoolers: A Review”, Maternal and Child Health Journal, Vol. 21/1, pp. 136-146, https://doi.org/10.1007/s10995-016-2102-1.
[10] Hart, E. et al. (2026), Using experimental variation to examine the (co-)development of cognitive and social-emotional skills in early childhood, Annenberg Institute at Brown University:, https://doi.org/10.26300/sb70-rd25 (accessed on 25 January 2026).
[8] Le, V. et al. (2019), “Advanced Content Coverage at Kindergarten: Are There Trade-Offs Between Academic Achievement and Social-Emotional Skills?”, American Educational Research Journal, Vol. 56/4, pp. 1254-1280, https://doi.org/10.3102/0002831218813913.
[7] Li, Z., N. Rao and J. Gong (2026), “Concordance between direct assessment and adult report in assessing learning and developmental competencies of young children: A systematic review and meta-analysis”, Early Childhood Research Quarterly, Vol. 74, pp. 253-265, https://doi.org/10.1016/j.ecresq.2025.10.006.
[1] OECD (2026), IELS 2025 Technical Report, http://oecd.org/en/about/projects/international-early-learning-and-child-well-being-study.
[5] OECD (2025), Reducing Inequalities by Investing in Early Childhood Education and Care, Starting Strong, OECD Publishing, Paris, https://doi.org/10.1787/b78f8b25-en.
[6] OECD (2025), Results from TALIS Starting Strong 2024: Strengthening Early Childhood Education and Care, TALIS, OECD Publishing, Paris, https://doi.org/10.1787/20af08c0-en.
[2] OECD (2023), PISA 2022 Results (Volume I): The State of Learning and Equity in Education, PISA, OECD Publishing, Paris, https://doi.org/10.1787/53f23881-en.
[9] Thijssen, M. et al. (2024), “Cross Productivities of Executive Functions: Evidence from a Field Experiment”, Journal of Human Capital, Vol. 18/4, pp. 563-589, https://doi.org/10.1086/732483.
← 1. The standard deviations in early learning and development scores summarise variation in outcomes among 5-year-olds within each jurisdiction. These standard deviations were standardised to have an international average of about 100 score points. If a standard deviation is larger than the international average, it indicates that children’s scores vary more from a particular jurisdictions’ mean performance than they vary internationally. A smaller standard deviation means that children’s scores vary less in a jurisdiction than they vary internationally.
← 2. For each outcome, this is the difference between the 90th percentile of performance (the score above which only 10% of children scored) and the 10th percentile of performance (the score below which only 10% of children scored).
← 3. This analysis was carried out in two steps. The first step estimated the share of the variation in children’s composite scores between jurisdictions. Of the remaining variation, the second step identified the between-centre and within-centre variances. Within-centre variation are differences in scores between children attending the same ECEC centres. The sum of the between- and within centre shares represents the total variation in that jurisdiction as a proportion of the IELS average level of variation in performance; that is why the sum of the two bars may exceed or fall short of 100%.