Rwanda’s first Programme for International Student Assessment (PISA) results show mean scores of 317 in science, 312 in reading and 301 in mathematics. Compared to the OECD/ PISA 2025 average scores of 482 for science, 463 for mathematics and 461 for reading.

But these figures are not examination marks or percentages. PISA uses a scoring scale that takes into account the difficulty of the questions students answer, rather than simply counting how many questions they get right.

ALSO READ: Experts weigh in on Rwanda’s first PISA results

Organised by the Organisation for Economic Co-operation and Development (OECD), PISA assesses how well 15-year-olds can apply what they have learnt to unfamiliar and challenging situations.

How the scores are calculated

PISA reports results using mean scores and proficiency levels. The mean score shows the average position of students on the PISA scale, while proficiency levels describe what students at different points on the scale are able to do.

The difficulty of questions is an important part of the scoring process. Students do not all answer exactly the same questions. Different versions of the assessment contain some common questions, allowing responses to be placed on a common scale.

The questions are analysed according to their difficulty, and students are placed on the scale based on the level of questions they are able to answer. As a result, the number of correct answers alone does not determine a student’s PISA score.

At the launch of Rwanda’s 2025 results in Kigali on September 8, officials used two hypothetical students, Aline and Bosco, to illustrate how the system works.

Aline could answer eight questions correctly out of 10, compared with six for Bosco. In a conventional examination, Aline would score higher.

However, if Aline answered mostly easier questions while Bosco answered more difficult ones, Bosco could demonstrate a higher level of proficiency despite answering fewer questions correctly. PISA is designed to capture this difference.

What Level 2 means

PISA results are divided into proficiency levels, with Level 2 used as the baseline. Level 2 is not a pass mark. Rather, it indicates that students have demonstrated basic knowledge and skills in a subject.

In science, a student at Level 2 can recognise a correct explanation for a familiar scientific phenomenon and, in simple cases, use information from data to judge whether a conclusion is valid.

In mathematics, students at this level can recognise how a simple real-life situation can be represented mathematically, including situations involving distances or currency.

In reading, they can identify the main idea in a moderately long text, find information based on explicit criteria and reflect on the purpose or form of a text when guided to do so.

Thirteen percent of Rwandan students reached Level 2 or above in science. The proportion was eight percent in mathematics and four percent in reading.

By comparison, the OECD averages were 74 percent in science, 65 percent in mathematics and 69 percent in reading.

This means about 87 percent of Rwandan students were below Level 2 in science, 92 percent in mathematics and 96 percent in reading.

Rwanda had almost no students at Levels 5 or 6 in science and mathematics, and almost no students at Level 5 or above in reading.

Students at these higher levels are expected to handle more complex and unfamiliar situations, apply knowledge independently and, depending on the subject, evaluate information or develop and assess different approaches to a problem.

What the three percent means

Across the three subjects, three percent of Rwandan students reached at least Level 2 in science, reading and mathematics.

This figure differs from the individual subject results because a student had to reach the baseline level in all three subjects to be included.

For example, a student could reach Level 2 in science but remain below that level in reading or mathematics and therefore not be counted among the three percent.

The three percent is therefore not Rwanda’s overall PISA score, nor does it mean students answered three percent of the questions correctly.

Who took the assessment?

A total of 6,622 students from 213 schools took part in Rwanda’s assessment. The sample was designed to represent the country’s population of 15-year-olds, estimated at about 99,000.

A total of 7,455 students had been sampled, with the assessment conducted between April and June 2025.

Of those who took the assessment, 93 percent were in lower secondary and seven percent in upper secondary. Public-school students accounted for 91 percent and private-school students nine percent. Eighty-two percent were from rural areas and 18 percent from urban areas.

The OECD said Rwanda’s data met its quality standards and were fit for reporting.

Rwanda used the paper-based version of PISA in 2025, making it one of six participants to use the format.

The assessment format is relevant when interpreting some of the results. The OECD says paper- and computer-based performance is broadly comparable, although some comparisons require caution.

In mathematics, proficiency for paper-based participants is described using the PISA 2012 proficiency levels rather than the PISA 2025 levels.

Rwanda did not participate in the separate Learning in the Digital World assessment, which examined computational problem-solving and self-directed learning.

PISA is not a school examination

PISA does not give individual students, teachers or schools a grade. Instead, a representative sample of 15-year-olds is assessed to provide an indication of how an education system is performing.

The assessment also collects information about students’ backgrounds and school experiences, allowing performance to be examined alongside factors linked to learning.

In Rwanda, students in the highest socioeconomic quarter scored 30 points more in science than those in the lowest quarter. The OECD average gap was 85 points.

Socioeconomic status accounted for three percent of the variation in science performance among Rwandan students, compared with 12 percent across the OECD.

Some disadvantaged students also performed strongly. Twenty-two percent of disadvantaged Rwandan students were among the top quarter of science performers in the country.

Students were also asked about their experience at school. More than 70 percent reported a sense of belonging.

The assessment gathered information on bullying, cyberbullying and family support for learning. More than one in five students reported experiencing cyberbullying at least monthly.

Digital-device use was another area covered by the assessment. Rwandan students reported high levels of use at school for learning and leisure.

The OECD found that moderate use for learning was associated with better science performance, while excessive leisure use was associated with weaker performance.

A first PISA baseline

Rwanda’s results come against a wider decline in reading performance in many education systems. The OECD reported that the decline in several systems began before the Covid-19 pandemic.

Some countries have recorded improvements. Cambodia, for example, increased its science score by 35 points between 2022 and 2025.

The government plans to use the results to identify areas for improvement and prepare for the next PISA assessment in 2029.