International Common Assessment of Numeracy - International Common Assessment of Reading 2025
2025
Get MicrodataIdentification
int-pal-ican-icar-2025-v1
International Common Assessment of Numeracy - International Common Assessment of Reading 2025
ICAN-ICAR 2025
| Type | Identifier |
|---|---|
| DOI | https://doi.org/10.25828/ah42-xw77 |
The PAL Network’s International Common Assessment of Numeracy (ICAN) and Reading (ICAR) is a one-on-one oral assessment for children aged 5–16 that measures foundational numeracy and reading skills. It includes 36 numeracy items and 30 reading items. The numeracy items assess spatial reasoning, patterns, measurement, time, data, number knowledge, and arithmetic problem solving. The reading items assess listening comprehension, letter-sound knowledge, word reading, and reading comprehension across progressively more difficult texts.
From July to October 2025, the PAL Network used ICAN and ICAR in a nationally representative, household-based survey of 89,185 children in 56,913 households across 2,917 enumeration areas. The survey covered both in-school and out-of-school children. It enables low- and middle-income countries to report on Sustainable Development Goal (SDG) indicator 4.1.1(a), which measures foundational learning proficiency. It also provides evidence that countries can use to shape policy and track progress in early foundational learning.
Sample survey data [ssd]
Households and their individuals
Version
Version 1.1: Edited, anonymized data for public distribution
2026-04-01
DataFirst
Scope
Coverage
The study had national coverage of the countries covered. Assessments were conducted in Africa (Kenya, Mali, Mozambique, Senegal, Tanzania and Uganda); Asia (Bangladesh, Nepal, and Pakistan); and Americas (Mexico and Nicaragua).
The lowest level of geographic aggregation in the data is anonymised village
The universe of the study was households with children aged 5 - 16 years old.
Producers and sponsors
| Name |
|---|
| People's Action for Learning Network |
| Name | Abbreviation |
|---|---|
| Foreign, Commonwealth and Development Office | FCDO |
| Echidna Giving | |
| Hempel Foundation | |
| Gates Foundation |
Sampling
Sample frame
For each country, the sampling frame was constructed using the most recent national population and housing census.
Government statistics agencies
56 913 households in 2 917 enumeration areas in total
The survey used a multi-stage, stratified probability sampling design to produce nationally representative estimates for children aged 5-16 years, including both in-school and out-of-school children. In the first stage, Enumeration Areas (EAs) were selected as the primary sampling units (PSUs) from the national sampling frame using probability proportional to size (PPS). The sample was stratified by geographic area and urban/rural residence. In the second stage, 20 households were selected in each sampled EA. Where a household list was available, the list was updated and households were selected systematically. Where no list was available, a spatial selection approach was used, including the 5th household rule; in Mexico, every 3rd household was selected because of higher nonresponse. Only households with at least one child aged 5-16 years were eligible. The planned sample size was 222 EAs per country and approximately 4,440 households per country. More information is provided in the documentation for the dataset.
Data Collection
| Start | End |
|---|---|
| 2025-07-14 | 2025-10-19 |
Cross-section [cross section]
- Face-to-face computer-assisted interviews [CAPI]
Data processing
DataFirst used R software to apply labels to all variables and their associated value categories. The data was reviewed for direct identifiers and other confidential information, including respondents’ names, phone numbers, addresses, and similar sensitive fields. Variable value ranges were checked for consistency and validity. DataFirst also verified the completeness of the data file by cross-checking it against the questionnaire. Variable data types were converted where necessary to ensure that each variable was stored in the correct format. The data file was renamed according to a consistent naming convention to make its contents easier for researchers to identify. The final cleaned data file was saved in Stata format.
quality_standards
Benchmarking:
The technical robustness of the ICAN-ICAR tool has been confirmed by the UNESCO Institute for Statistics (UIS) as meeting the criteria for reporting on SDG 4.1.1(a).
In collaboration with the Australian Council for Educational Research (ACER), PAL Network established global proficiency thresholds using the Pairwise Comparison Method (PCM).
Data on learning conditions and child learning difficulties was collected using the Washington Group Child Functioning Module.
Psychometric Validation:
The survey applied Item Response Theory (IRT) and rigorous checks for Differential Item Functioning (DIF) to ensure measurement fairness across age, gender, and location.
Data Access
| Name | Affiliation | URL | |
|---|---|---|---|
| DataFirst Support | University of Cape Town | http://www.support.data1st.org | support@data1st.org |
| Muhammed Usman | PAL Network (Assessment Unit) | https://www.palnetwork.org/assessment | musman@palnetwork.org |
| Is signing of a confidentiality declaration required? | Confidentiality declaration text |
|---|---|
| yes | The data re-user agrees to make no attempt to re-identify data subjects from the data |
CC-BY 4.0 attribution license
PAL Network. International Common Assessment of Numeracy - International Common Assessment of Reading 2025 [dataset]. Version 1. Nairobi: PAL Network [producer], 2025. Cape Town: DataFirst [distributor], 2026. DOI: https://doi.org/10.25828/ah42-xw77
The data re-user agrees to send DataFirst a link to research publications that are based on the data.
Metadata production
int-pal-ican-icar-2025-v1
| Name | Affiliation | Role |
|---|---|---|
| DataFirst | University of Cape Town | Metadata producer |
2026-04-20
Metadata version
Version 1