The monitoring of education quality for 2023–2025 revealed a picture that cannot be reduced to a single average mark. The State Service for Education Quality of Ukraine has published the results of a three-year cycle of research into the academic achievements of pupils in Years 6 and 8 in Ukrainian and mathematics. The data was presented by Ivan Yuriychuk at a panel discussion on 3 June 2026. The system is maintaining the middle of the distribution, and in some areas is raising it, but the top segment of results remains narrow, whilst Year 8 pupils’ performance in mathematics shows a decline across all three indicators — the average mark, the median and the mode.
What the monitoring measured and why
The choice of year groups follows the logic of the educational pathway. Year 6 assesses pupils’ progress following the transition from primary to lower secondary school. Year 8 serves as a diagnostic assessment ahead of the final Year 9, when any gaps can still be addressed. The Ukrainian language is responsible for basic communicative competence, whilst mathematics is responsible for logical-mathematical competence.
The study was initiated by the Service and supported by the Ministry of Education and Science, the SURGe programme and UNICEF, with implementation carried out by the Service’s regional offices, the Institute of Educational Analytics and the Ukrainian Institute for Educational Development. The methodology included test items, pupil questionnaires to identify influencing factors, and data from school heads on operating arrangements, inclusion and the status of internally displaced persons. The test sets were updated annually, whilst maintaining the structure of content areas and cognitive levels — it is precisely this that enables results to be compared across years.
An important methodological detail concerns the standards. Year 6 pupils were taught using curricula based on the 2020 State Standard for Basic Secondary Education, i.e. within the framework of the New Ukrainian School (NUS) reform. Year 8 pupils were taught using curricula based on the 2011 standard. These represent two different generations of educational requirements, so the comparative results should be interpreted correctly within their respective contexts.
Sampling and data reliability
The study covered 226 general secondary schools across five macro-regions and the city of Kyiv. Urban areas were represented by 166 schools, whilst rural areas by 60. A reserve sample of a further 140 schools was established in case of security threats.
The procedure involved four roles in each school: a coordinator, an instructor who does not teach the subject in question, a technical administrator and an independent observer. Pupil attendance exceeded 90 per cent, resulting in over 4,300 participants for each subject in every class. Observers recorded 1,131 observations, of which 54.5 per cent were conducted in person and 45.5 per cent remotely. Eighteen breaches of academic integrity were identified, representing 1.59 per cent. In 90 per cent of cases, observers awarded the highest levels of confidence in the procedure—9 and 10 marks.
The authors outlined the limitations of their interpretation. The sample included only state and local authority schools where the language of instruction is Ukrainian, with class sizes of at least 10 pupils and an internet speed of at least 30 Mbit/s. Participation was voluntary; therefore, pupils living abroad or who were unwell were under-represented. All identified correlations were classified as statistical dependencies rather than proven cause-and-effect relationships.
Read also:
Demining in Ukraine has stopped after an attack on sappers
Four outcome profiles in mathematics and the Ukrainian language
Rather than a single trend, the monitoring revealed four distinct patterns.
Mathematics in Year 6 shows a stable core. The average mark rose from 10.77 in 2023 to 11.52 in 2025, but the median remained unchanged — 10, 10 and 10 again. There is some movement, but the typical result for a Year 6 pupil has remained the same. The ‘Data and Statistics’ topic was completed by 57.4 per cent of pupils, whilst ‘Equations’ accounted for 37.1 per cent and ‘Geometry’ for 37.3 per cent. Only 20.6 per cent of pupils completed the task of calculating the area of a square in a real-life context. In terms of cognitive levels, the picture is structured in stages: knowledge – 58.8 per cent, application – 43.4 per cent, reasoning – 39.8 per cent.
Mathematics in Year 8 provides the most striking indication of the entire study. Here, the entire central portion of the distribution has shifted: the mean score fell from 22.8 to 20.47, the median from 22 to 19, and the mode from 14 to 11. The simultaneous decline in these three indicators points to a shift across the entire set of results. Knowledge scores stood at 60 per cent, reasoning at 45 per cent, and application at 41 per cent. 84 per cent of Year 8 pupils were able to identify a geometric shape, whilst 19 per cent could calculate the area of a trapezium.
Ukrainian language results in Year 6 show a clustering around the centre. The mean score rose from 32.25 to 34.86, whilst the median rose from 30 to 33. At the same time, 94 per cent of participants are concentrated within a narrow range — 5, 6 and 7 marks on a 12-mark scale. The mode and maximum score are falling, meaning the top segment is not expanding. Working with text comes more easily — 53 per cent of tasks were completed, whilst language knowledge accounts for 43–45 per cent. Phonetics was completed by only 33 per cent of pupils.
Ukrainian language in Year 8 conceals two different realities beneath a stable average score. The average mark remains almost unchanged — 50.09, then 49.31, then 50.19. The median is constant: 48, 48, 48. The distribution has two peaks, and in some subtests, two distinct groups of results coexist — lower and higher. Syntax showed an increase from 27.1% to 40.2%. Practical morphology fell from 80.9% to 72.8%, whilst part-of-speech identification was achieved at only 37%.
The authors formulate their overall conclusion unequivocally: the average score is insufficient for management decisions.
Monitoring 2023–2025
Four different profiles of one system
Pick a subject and a grade to see how the distribution of results behaves. Curve shapes are schematic; the figures are actual monitoring data.
Source: State Service of Education Quality of Ukraine, education quality monitoring 2023–2025. Distribution curves are shown schematically.
Factors affecting learning outcomes
The pupils’ questionnaires made it possible to divide the influencing factors into three groups based on the stability of their correlation with the results.
The most significant and consistent factor proved to be the pedagogical quality of the lesson. Clear explanations, understandable requirements, activities encouraging critical thinking, and demonstrations of the practical usefulness of the material were consistently linked to better results across all sections. Next comes the pupil’s internal state: children who feel more confident and have lower levels of anxiety achieve higher marks. The cultural environment is reflected in the presence of books at home — an indicator of a setting where reading is part of everyday life. Pupils who seek help from their teacher or parents achieve better results. The type of settlement acts as an indicator of access to resources and staff stability, and the Year 8 data for the Ukrainian language clearly point to segregation along this line.
Volatile factors do not provide universal solutions. Digital confidence is more important than simply having a device, but IT skills proved effective in Year 6 but not in Year 8 for mathematics. Having one’s own space at home is significant mainly in Year 6. The status of an internally displaced person is not in itself associated with poorer results — in some groups, scores are even slightly higher, so social vulnerability does not equate to academic weakness. The worst results are shown by those who do not complete their homework, whilst a moderate amount is optimal — around 30 minutes for Year 6 and an hour for Year 8. A private tutor is merely an indicator of access to additional resources.
A separate group consists of factors whose expected impact is consistently not confirmed by the data. Gender shows virtually no correlation in mathematics; in Ukrainian language, girls have slightly higher marks, but there is no evidence of a gender gap. A teacher’s general encouragement to study, without a clear pedagogical explanation, does not work. The length of homework assignments is irrelevant in itself — what matters is quality, content and feasibility. The mere availability of learning materials also has no effect unless they are integrated into the child’s real life.
The researchers identified two contexts. 25 per cent of Year 8 pupils speak predominantly a language other than Ukrainian at home, and the authors interpret this as a methodological challenge for schools. Taking tests remotely results in higher marks, but this is classified as a methodological artefact of the procedure.
Read also:
The invisible mental health crisis among Ukrainians and therapies that work
The State of Institutional Audit and the Gap Between Existence and Effectiveness
The second part of the analysis is based on the results of institutional audits. In 2025, 233 unscheduled audits were carried out, whilst the total number of audits for the period 2019–2025 stands at 783. The audit does not directly explain the monitoring results, but it reveals a similar pattern of issues at the institutional level.
Pupils’ difficulties in applying knowledge to new situations are compounded by the predominance of teacher-centred teaching, which is present in 93 per cent of lessons. Applied learning practices remain marginal: project-based work is recorded in 14 per cent of lessons, STEM in 4 per cent, and differentiation in 21 per cent.
The requirement for pupils to reflect on and assess their own progress is hampered by the fact that 53.2 per cent of schools have problems with their assessment system. Self-assessment is present in 47 per cent of lessons, peer assessment in 30 per cent, and the option to choose a level of difficulty in 22 per cent. Assessment criteria are made public in only 40 per cent of cases, whilst assessment based on these criteria takes place in 47 per cent of lessons. Pupils are left without answers to questions about expectations, their own place in the learning process and the next step.
The failure rate in mathematics in Year 8 is reflected in the audit: mathematical competence is developed in lessons in only 46 per cent of cases, and scientific and technical competence in 45 per cent. Data-driven quality management also remains incomplete: 94 per cent of schools plan to carry out monitoring, but only 42 per cent actually track progress.
The pattern observed by the audit at school level mirrors that recorded by monitoring at pupil level. The existence of these measures does not always translate into action. There are documents, plans, criteria and monitoring mechanisms that do not become part of everyday classroom practice.
Block 2 · Institutional audit
The audit mirror: the gap between having and doing
Monitoring tests the student, the audit assesses the school. These are different levels of analysis, yet the structure of the problems repeats itself. Expand any row to see both sides.
At the school level the audit mirrors the same structure found at the student level: what exists does not always turn into action. There are documents, plans, criteria and monitoring exercises, and they do not always become the daily practice of a lesson.
Source: State Service of Education Quality of Ukraine, education quality monitoring 2023–2025 and institutional audit data.
Seven management questions for the education system
The findings of the three-year cycle formed the basis for seven management hypotheses presented for discussion at the panel session. These relate to the transition from stability to development; the early identification of gaps before they accumulate at the start of a specialist school; the alignment of the assessment system with learning objectives; the impact of teachers’ professional development on actual classroom practice; the development of digital literacy alongside device management, the system’s ability to embed effective practices in a typical classroom, and the view of a child’s individual outcome as part of a broader architecture of opportunities.
The authors’ reasoning boils down to several assertions. If the aim of learning is the ability to reason, then measuring merely the reproduction of material is insufficient. Teacher qualifications do not guarantee an improvement in lesson quality, and the real advantage lies in the ability to transform traditional classroom instruction into an environment that fosters thinking. The presence of a gadget is not an advantage without attention management, and the discussion about the digital environment must go beyond mere permissions and restrictions.
Read also:
Ukrainian teenagers have carried out a study — only 30% of their peers seek help
