How PISA works
PISA is an international study by the Organisation for Economic Co-operation and Development (OECD). Since 2000, every three years, it has assessed whether 15-year-olds can apply what they know in real life.
Who takes part
Students aged roughly between 15 years 3 months and 16 years 2 months, enrolled in at least grade 7. In many countries this is close to the end of compulsory education.
Sampling
Not every school takes part; a scientifically drawn sample does. First schools are selected, then students within them. The sample is enough to describe the whole system. In 2025, 22,680 students from 643 schools and colleges in all regions of Kazakhstan took part. Nazarbayev Intellectual Schools take part with a separate sample and are not included in the national figure.
How the test runs
- The test is computer-based and takes about two hours.
- Items are grouped into units: one stimulus (text, table, simulation) and several questions.
- Response formats: multiple choice, short answer and open response. Open responses are scored by trained coders using a guide.
- Reading and mathematics use multistage adaptive testing: the next block depends on how the student did on the previous one.
- After the test, students complete a questionnaire about family, school, attitudes to learning and technology. School principals also complete a questionnaire.
Why there are no individual scores
Each student answers only part of the items, so one student's responses are not enough for an individual score. To calculate group results correctly, PISA uses plausible values. Results are anonymous and analysed only at country and group level: no student, parent or teacher ever sees an individual result.
Scale and levels
Each literacy has its own scale, set so that in the year it was introduced the OECD average was 500 and the standard deviation 100. The scale is divided into levels, each with a description of what students can do. Level 2 is the baseline.
Comparing results
When comparing countries or cycles, the standard error and link error are taken into account. Two numbers may differ while the difference is not statistically significant. OECD reports show this for every comparison.
Why preparing for PISA makes no sense
PISA has no consequences for students or schools. Items are new each time and the sample is random. Targeted preparation distorts the results, so they give a less accurate picture of the system. The best “preparation” is teaching students to understand texts, reason and work with data in every lesson.
Kazakhstan and PISA
Kazakhstan has taken part since 2009, in six cycles: 2009, 2012, 2015, 2018, 2022 and 2025. The OECD judged the 2015 results not comparable. Results: Data.