The results of the latest PISA survey were published a few days ago. At times like this, everyone and their dog has something to say about why the results are so dire (and here I could quote from Hungarian, Dutch, German and English press reports from the past few days). The PISA figures always provide a good opportunity to start a discussion about schools – and we certainly must talk about them – but they certainly do not allow us to draw any conclusions about the state of the education system. The fact is, we have no idea what they are measuring; the tests are never made public.
It’s as if we were given a ‘fail’ for our language or maths A-levels, but weren’t allowed to see where we’d gone wrong. Anyone who, like me, works regularly with children can say for sure that students haven’t become any less intelligent. This article, originally published in Hungarian in the ÖT magazine, explores the situation from the perspective of Hungary.
First of all, let’s look at what the figures tell us: surveys conducted in 2022 and 2025 show that 15-year-olds’ reading comprehension, maths and science skills have plummeted worldwide – or at least that is what the results of the undisclosed tests suggest. Compared with 2018, the decline by 2022 was massive, and the situation is similar between 2022 and 2025. This is not just the case for Hungarian children, but in almost every country.
Within this international field, Hungary remained in the middle of the pack, close to the OECD average: in 2022, we were 1 point below the OECD average in maths, 3 points below in reading comprehension, and also 1 point below in science, with the OECD average standing at around 480 points in each area. Compared with Hungary’s 2018 results, there was a decline in mathematics, whilst there was no real change in reading comprehension and science; in other words, Hungary’s relative international position remained essentially unchanged. In 2025, again in line with the international trend, results in maths and reading comprehension deteriorated further, whilst performance in science remained essentially stable.
So one thing is certain: these are not the results we really need to be concerned about. Especially because, as I mentioned, it is impossible to know what the problem actually is.
Two years ago, at the congress of the International Confederation of Principals, I had the opportunity to give a presentation to an audience of over 1,000 headteachers, immediately after Andreas Schleicher, the OECD’s Director for Education and the head of the PISA surveys. On such occasions, we usually send each other the outline of our presentations in advance so that we can respond to one another’s work.
I’ve now taken a look at Andreas’s slides because I remembered that he certainly didn’t emphasise these figures at all. In fact, when we were having dinner together, he confided that they suspect the tests weren’t properly calibrated in 2022. In his presentation, he merely mentioned that while the students were able to answer roughly 60 per cent of the questions correctly, ChatGPT was already achieving a score of around 90 per cent in reading comprehension by the end of 2022. In maths, compared to the children’s 50 per cent score, the AI achieved 60 per cent. And just think how much it has improved since then.
What he did talk about at length, however, were the outcomes beyond the knowledge assessment. He also explained that these are far more important, because the greatest challenge is to ensure that school leavers become lifelong learners who are creative, curious, able to adapt to change and eager to learn. Subject-specific knowledge can be acquired later on. Far more important than the three key indicators are the students’ mental and physical wellbeing, the extent to which they are allowed to take control of their own learning, how resilient they are in times of major change or crisis, and the amount of homework they have versus how much free time they are left with.
He highlighted the fact that 15-year-olds’ creative thinking lags far behind that of 10-year-olds, and that schools do a great deal to stifle this skill, which is now indispensable. Interestingly, whilst girls have long since caught up in terms of subject knowledge, they are far better creative thinkers than boys. Perhaps this is also the reason why their prospects for further education are now much better throughout Europe; even in doctoral programmes, male students are slowly becoming a rarity.
Hungary (together with many other countries) performs poorly in all these areas, so these are the issues we should really be discussing. Hungarian 15-year-olds fall well below average in terms of curiosity, helpfulness and perseverance. At the same time, the proportion of those who believe there is only one correct viewpoint in a debate stands out as particularly high. Far fewer than average are prepared to stand up for something that matters to them, and students find it very difficult to perform under pressure.
Hungary also lags significantly behind more developed countries in the conscious use of digital technology in schools, even though every child already has a smart device by the age of 15. According to OECD data, the use of digital technology for learning purposes has a clearly positive impact on pupils’ marks and other outcomes. However, this requires teachers with advanced digital skills and knowledge, which are certainly in short supply.
A few weeks ago, Francesca Rivafinoli reflected on why it is a problem that Hungarian pupils are overburdened, when pupils in other countries are also struggling. She is right: there are countries with a high number of teaching hours and a lot of homework where PISA scores are better, but we do not know whether this is because of the workload or in spite of it. The usual example of this is Singapore, where children work roughly as much or even more than Hungarian students, whose average school-related working time when they are 15 is around 60 hours.
In Hungary, however, at least the suicide rate is not as high as in that Far Eastern country. Another regular example is Finland, which has performed well for many years in every area – not just in the figures – where the number of teaching hours is very low, homework is banned in primary school and there isn’t much of it in secondary school either.
There are certain regularly mentioned PISA “facts” that have already been disproved. For a long time, for example, it was stubbornly maintained – and education policy documents are still based on this misconception today – that the results of migrant children indicate a serious lack of knowledge. Yes, the hundreds of thousands of Hungarian children attending school in other countries do have results that are below the local average.
However, researchers were given the opportunity to have children attending school in a non-native-language environment sit the test in a language they are comfortable using. And lo and behold, the difference disappeared – it made absolutely no difference whether the children were Polish, Hungarian or Syrian. Only if someone can do the maths test in a language they are proficient in, they can demonstrate what they know.
It is therefore a good thing to discuss schools when the PISA data is published. But until we know why the whole world has received a lower mark than three years ago, it would be worth continuing the discussion on the basis of the other data, rather than the scores in the three knowledge areas. In 2014, more than 80 international education experts and academics wrote an open letter asking the OECD to postpone the next PISA testing cycle and, instead, to launch a comprehensive, democratic consultation, whilst making the tests publicly available and ensuring independent professional scrutiny.
According to the signatories of the petition, before continuing with the assessments, a global consensus should first be reached – involving teachers, parents, labour market stakeholders and researchers – on what the true aims of education are, and indeed what and how it is worth measuring in the interests of students’ development. They also noted that the focus on these three figures has a negative impact on learning. This has never actually happened, but it would be a good thing. It would help us understand what we are reacting to when we analyse the figures.
In the meantime, let’s be glad that whenever the latest PISA data is released, schools always look more important and take centre stage in the media, but let’s always start by acknowledging that the figures don’t really mean anything. I wrote an article on the same subject nearly 10 years ago for the Modern Iskola website (a Hungarian education site). It’s truly sad that nothing has changed since then https://moderniskola.hu/2017/01/pisa-fogyaszthatoan-ketchup-es-sajt-nelkul/
Eszter Salamon
