Building Strong Foundations for Life

Reader’s guide

What is IELS?

The OECD International Early Learning and Child Well-being Study (IELS) is comparative survey that assesses how children are faring at age five. IELS provides evidence to help policy makers, education leaders and early childhood practitioners in improving young children’s experiences and outcomes. IELS measures multiple domains of children’s early learning and development and collects a range of contextual information from their parents and teachers.

IELS uses two complementary assessment approaches:

• Direct assessment of children’s skills, delivered through developmentally appropriate, interactive stories and games on tablet devices. This approach provides a common basis for cross-jurisdictional comparisons. Careful design, testing and analysis minimise cultural and other biases, supporting the comparability of results. For IELS 2025, analyses included differential item functioning by gender, country, and language, as well as item-level, latent trait, and validity analysis. Tablet-based assessments also improve reliability by reducing transcription and coding errors.

• Indirect assessment through parent and early childhood education and care (ECEC)/school teacher reports which provide complimentary perspectives on children’s skills. Parents contribute observations across everyday settings, while teachers provide evaluations informed by comparisons with other children of the same age. These perspectives add depth and support the triangulation of results from the direct assessment.

Together, these components allow IELS to provide a robust and multi-dimensional picture of children’s early learning, development and well-being across participating jurisdictions.

What aspects of learning and development are on focus in IELS?

IELS conceptualises early learning and development as holistic, involving cognitive as well as social and emotional skills that are closely interrelated and mutually reinforcing. The study focuses on three dimensions that strongly predict later educational outcomes and well-being: foundational learning, executive function, and social and emotional development. Across these three dimensions, 10 specific domains are assessed. Results for each dimension are distinguished by colour in figures and tables.

 

Dimension	Domains	Dimension	Domains	Dimension	Domains
Foundational learning	Emergent literacy Emergent numeracy	Executive function	Inhibition Working memory Mental flexibility	Social and emotional development	Emotion identification Emotional attribution Trust Pro-social behaviour Non-disruptive behaviour

Who participated in IELS?

Ten national and sub-national entities (“jurisdictions”), participated in IELS 2025, with eight included in this report (Switzerland participated in the field trial only while Hangzhou (China) results were still undergoing analyses and validation). The target population for the study was children aged five to six years enrolled in a registered early childhood education and care (ECEC) centre or school. Samples were designed and weighted to be representative of these populations. For simplicity, this report uses “five-year-olds” to refer to the target population, and “ECEC centre/school” to refer to the setting attended at the time of assessment.

Parents and teachers also participated by providing contextual information about children’s characteristics, home environments, ECEC or school attendance, and skills. “ECEC teacher” refers to the staff member who knew the child best and completed the Teacher questionnaire while “parents” refers broadly to parents, guardians or other individuals who completed the Parent questionnaire.

Participation rates

Participation rates are a key indicator of survey quality and the reliability of results. Under the IELS 2025 Technical Standards, the quality thresholds for centre-level and child participation rates were set at 75%, indicating adequate representativeness of the target population.

Most jurisdictions met this standard before replacements. Three jurisdictions – the Flemish Community (Belgium), Korea and the Netherlands – conducted centre-level non-response bias analysis (NRBA) which indicated a low risk of bias. After replacements, all jurisdictions met the standard for child participation rate (see table below).

Parent and teacher participation rates showed greater variation across jurisdictions and were lower than child participation rates, although still relatively high compared with other international surveys. For jurisdictions with lower parent and teacher participation rates, caution is therefore advised when interpreting results based on questionnaire data with lower participation rates.

Participation rates for IELS 2025

Jurisdiction	Centre-level participation rate (%)	Number of participating children	Children participation rate (%)	Parent participation rate (%)	Teacher participation rate (%)
Baku, Sumgait (Azerbaijan)	98.5	2731	96.6	98.5	99.5
Ceará, Pará, São Paulo (Brazil)*	92.0	2598	83.8	92.0	92.2
England (United Kingdom)	92.1	2523	86.6	47.7	42.2
Flemish Community (Belgium)	94.3	2363	92.2	72.4	81.2
Korea	95.6	2894	85.9	96.4	99.5
Malta	99.1	1644	96.8	68.3	90.3
Netherlands	82.1	2821	77.6	65.9	74.5
United Arab Emirates	97.2	4122	91.9	84.3	90.0
Abu Dhabi (United Arab Emirates)**	98.6	1919	92.0	75.4	79.0

Note: Participation rates are weighted. Centre and children participation rates are after replacements. Parent and teacher data availability are based on participating children. *Estimates for Brazil correspond to averages across the three participating Brazilian states. **IELS Adjudicated Region.

Source: IELS 2025 Technical Report (OECD, 2026[1]).

A guide to interpreting findings in this report

Data underlying the report

IELS results are based on direct and indirect assessments of children’s skills across ten domains of early learning and development. IELS scores are numbers that express a location on a continuum of a latent trait: an unobserved aspect, such as emergent literacy. Because the IELS scales are defined in relation to the observed variation across all participating children, they have no theoretical minimum or maximum.

For reporting purposes, results are scaled to approximately normal distributions, with the international means set at 500 and standard deviations around 100. Therefore, a one-point difference corresponds to a 0.01 of a standard deviation and a 10-point difference of 0.1. Results for child subgroups are reported only when estimates are based on at least 30 children from at least five ECEC centres or schools within a jurisdiction.

Contextual information on children’s lives and skills was collected from parents and teachers, with some information provided by one source only and some by both. Comparisons between parent and teacher reports are based on the subsample of children for whom data from both sources are available.

Participating jurisdictions

Baku and Sumgait (Azerbaijan)

The sample for Azerbaijan was drawn from ECEC centres in the national capital and largest city, Baku, and the nearby city of Sumgait. Together, these two cities represent approximately 25% of Azerbaijan’s total population. The results for Azerbaijan are representative of the population of five-year-olds attending ECEC in these two locations.

Ceará, Pará, São Paulo (Brazil)

Three Brazilian states participated in IELS 2025: Ceará, Pará, and São Paulo. Therefore, the reported estimates pertain only to these three states, rather than the whole country. The results for Ceará, Pará, and São Paulo (Brazil) are representative of the population of five-year-olds attending ECEC in these locations and are presented as an average; weights were standardised to ensure that each of the states contributes to the estimates equally.

England (United Kingdom)

England is one of the countries of the United Kingdom. It has its own education system that is independent of the education systems of the other countries of the United Kingdom. The results for England are representative of the population of five-year-olds in primary schools in England.

Flemish Community (Belgium)

The Flemish Community (Belgium) is one of the three institutional communities of Belgium. The results for the Flemish Community (Belgium) are representative of the population of five-year-olds in ECEC in the Flemish Region as well as in schools under the authority of the Flemish Community in the bilingual Brussels-Capital region.

Hangzhou (China)

Hangzhou (China) was one of two jurisdictions to join the study at a point where the regular field trial was not feasible. Instead, an Operational Field Trial (OFT) was conducted to test administration procedures in preparation to the main study. Further analysis and evaluation of the collected data are required for validation; therefore, results are not included in this report.

Korea

Korea was one of two jurisdictions to join the study at a point where the regular field trial was not feasible. Instead, an Operational Field Trial (OFT) was conducted to test administration procedures in preparation to the main study. The results for Korea are representative of the country’s population of five-year-olds in ECEC.

Malta

Unlike other jurisdictions where a sample of ECEC centres/schools was drawn, all ECEC centres in Malta participated in IELS 2025. Within each centre/school, children were sampled at random in the same way as in other jurisdictions. The results for Malta are representative of the country’s population of five-year-olds attending ECEC or primary schools.

Netherlands

The results for the Netherlands are representative of the country’s population of five-year-olds in ECEC.

Switzerland

Switzerland participated in the IELS 2025 field trial but not in the main study. Therefore, there are no final results to be included in this report.

United Arab Emirates

The results for the United Arab Emirates are representative of the country’s population of five-year-olds attending ECEC.

Abu Dhabi (United Arab Emirates)

ECEC centres and children in the Emirate of Abu Dhabi (Abu Dhabi Region, Al Dhafra Region and Al Ain Region) that are not administered by the Ministry of Education (i.e. private and charter centres) participated as an adjudicated region within the United Arab Emirates. Its results are reported separately and referred to as Abu Dhabi (United Arab Emirates).

Overall IELS averages

Where cross-jurisdiction averages are provided in the OECD’s IELS 2025 International Report, these averages correspond to the arithmetic mean of the estimates for participating jurisdictions.

Composite scores

The ten domains of assessment in IELS 2025 are conceptually organised into three broader dimensions of early learning and development. IELS uses a multi-dimensional model in its analysis, enabling the arithmetic mean of domain estimates within a dimension to represent the overall magnitude in that dimension. For this report, this simple approach was adopted to ensure that the results are more easily replicable. However, there are other ways to construct composite scores, including those in which domains are weighted differently, and researchers are encouraged to explore these methods in secondary analysis.

Socio-economic status measures

In the IELS 2025 Public Use Files, there are two indices of socio-economic status (SES), described below. This report uses the SES index derived from Item Response Theory (IRT) in all of its analyses.

Socio-economic status index derived from Item Response Theory

The first index is derived using an IRT approach that combined an item response model and a latent regression model. To derive the item response model, three variables from the Parent questionnaire were used: household income, parental occupation, and parental education. In addition, the latent regression model used other variables explanatory of the missing data structure: household composition, child outcomes across domains, and jurisdiction. This conditioning aims to ensure that the values of the SES index are appropriate for secondary analysis. The main advantage of the IRT approach is that it uses all available data to construct the index, making it unnecessary to choose the highest level of a variable when data are available for both parents for a child. This is a new approach for this cycle, and therefore, several checks for robustness and consistency across jurisdictions were conducted. For further details, please refer to the IELS 2025 Technical Report (OECD, 2026[1]).

The SES IRT index is standardised to have a mean of 0 in each jurisdiction, while its standard deviation is allowed to vary across jurisdictions. This enables comparisons between jurisdictions in terms of the dispersion of SES in the population. In IELS 2018, no SES IRT index was produced.

Socio-economic status index derived from Principal Component Analysis

The second is an index derived from a Principal Component Analysis (PCA) of three indicators from the parent questionnaire: household income, highest parental occupation, and highest parental education. The term ‘highest’ for parental occupation and parental education indicates that when data are provided for both parents/guardians, the higher of the two levels is selected for the child. For Baku and Sumgait (Azerbaijan), no household income indicator was collected; however, a substituted proxy measure based on household resources (e.g. a separate room for the child, appliances) was used.

The SES index derived from PCA is standardised within each jurisdiction, with a mean of 0 and a standard deviation of 1. Trends analysis for England (United Kingdom) between IELS 2018 and IELS 2025 involving socio-economic status measures should be made using the SES PCA Index.

Statistically significant differences

Unless otherwise stated, a difference reported as statistically significant is significant at the .05 level. This means there is less than a 5% probability that the reported difference occurred by chance; a statistical test has been carried out to establish this. In figures in this report, statistically significant differences in this report are denoted by darker tones.

Interpreting correlations

A correlation coefficient is a measure of the degree to which two variables tend to move together. The coefficient ranges from +1 to -1, indicating the strength and direction of association. If a correlation is positive, it means that as one variable increases, the other increases as well. If a correlation is negative, it means that as one variable increases, the other decreases. In this report, a correlation coefficient with an absolute value between 0 and .19 is interpreted as weak, between .20 and .40 as moderate and already meaningful, and above .40 as substantial.

Standard deviation

The standard deviation is a measure of the dispersion of a set of data around its mean. The more spread out the data, the higher the standard deviation. In a normal distribution, 68% of the scores fall within one standard deviation of the mean, 95% within two standard deviations, and 99% within three. As mentioned above, all IELS scales have an approximate standard deviation of 100.

Standard error

Scores reported in this volume are population estimates based on a sample of selected children. However, even after applying appropriate weights, it is unlikely that the sample mean is exactly equal to the ‘true’ population mean. Some degree of variation – known as sampling error – is to be expected. Thus, each estimate has an associated standard error, which indicates how accurately the sample estimate reflects the ‘true’ population estimate.

Rounding figures

As a result of rounding, some figures in tables may not add up exactly to the stated totals. Totals, averages and differences are calculated using exact (unrounded) numbers and rounded only after calculation. Percentages and mean scores are rounded to whole numbers, and standard errors are rounded to two decimal places.

Additional technical information

Readers interested in additional technical detail regarding IELS 2025 are directed towards the short technical note (Annex B) at the end of this volume and to the IELS 2025 Technical Report (OECD, 2026[1]).

This report uses the OECD StatLinks service, meaning that tables with the data underlying the report are assigned a stable URL leading to an Excel workbook; for a list of tables available online, see Annex C.

References

[1] OECD (2026), IELS 2025 Technical Report, http://oecd.org/en/about/projects/international-early-learning-and-child-well-being-study.