Early testing of schoolchildren has absolutely zero effect
Early testing of schoolchildren has no impact on their educational attainment down the line. Nor do the tests have any effect on pupils’ mental health or how happy they are at school.
In brief
- The study includes data from 7500 children aged 7 to 9, born in 2008 and 2009, from 81 Norwegian primary schools. The researchers linked results from difficult maths tests at lower primary level with results from the national numeracy tests taken by the same pupils in Year 5.
- These data were also linked with data from the annual Pupil Survey, which maps factors such as well-being, bullying, school environment and motivation. Screening and tests at lower primary level are controversial in Norway.
- Supporters argue that the tests encourage greater effort and performance, and help even out differences by identifying pupils with additional learning needs at an early stage. Critics claim that testing is detrimental to mental health and creates stress and pressure that negatively affect well-being and the school environment.
- The study shows that early testing has absolutely no impact on educational attainment whatsoever, regardless of sex, background, parental level of education, or whether the children were born early or late in the year.
“We found absolutely no effect,” said NTNU Professor Colin Peter Green, who has investigated whether early testing affects the learning, well-being and school environment of Norwegian children aged 7 to 9.
Screening and tests at lower primary level are controversial in Norway. Supporters argue that they are useful tools for early identification of pupils with additional learning needs. Some people also believe that testing can inspire pupils to focus more on their schoolwork and perform better. Critics argue that testing can be detrimental to children’s mental health and that stress, pressure and a narrow focus on results negatively affect their well-being and the school environment.
A unique chance to look at effect
Hardly any research has been done on the impact testing actually has on young children. The work done by Colin Peter Green, Kari Vea Salvanes and Ole Henning Nyhus shows what effect the introduction of a series of relatively difficult maths tests had on Norwegian children aged 7 to 9. The study involves two cohorts of approximately 7500 pupils born in 2008 and 2009 at 81 primary schools.
From 2016 to 2018, the pupils born in 2008 were tested twice and the 2009 cohort was given three tests. These types of tests are not usually given in Norway. The fact that Norwegian primary schools do not give grades, homework or formal assessments provided a golden opportunity to study the effects of introducing testing in a setting where tests are usually absent.
The PISA test
Apart from the screening tests in Years 1 and 3, Norwegian pupils are not formally tested until they take the first national tests in reading, numeracy and English in Year 5. The system is now changing, including the discontinuation of the Year 1 screening test.
However, according to the education news site Utdanningsnytt(in Norwegian) the number of tests has grown over the past twenty years. This increase has been driven by national authorities, local authorities and the schools themselves. The PISA test for 15-year-olds, developed by the Organisation for Economic Co-operation and Development (OECD), has also had a major impact on Norwegian schools for many years.
Linking maths, tests and well-being
- The study links the results of maths tests taken at lower primary level by cohorts born in 2008 and 2009 with their results from the national tests taken in Year 5.
- The data were also linked with the annual Pupil Survey, which maps factors important for pupils’ learning and well-being at school. The questions focus on well-being, bullying, school environment, motivation and teacher support.
- The researchers have used a statistical method called ‘difference in differences’. It involves identifying an effect by comparing a cohort at a given school that was tested with younger or older cohorts at the same school that were not tested.
- These are then compared with other cohorts at different schools that were not tested. This provides an overview of how the tests, for example in mathematics, affect pupils’ educational attainment when they later take the national tests in the same subject.
- The study was conducted by Professor Colin Peter Green from NTNU’s Department of Economics, along with researchers Ole Henning Nyhus from NTNU Social Research and Kari Vea Salvanes from the Nordic Institute for Studies in Innovation, Research and Education (NIFU)/Department of Economics.
Testing does not make a difference
The study shows that taking two to three fairly challenging maths tests during the first years of primary school had no impact on how the children later performed on the national numeracy tests.
This was the case regardless of sex, background, parental level of education, whether the children were born early or late in the year, or any other factors.
“We also find no evidence that early testing affects well-being at school,” added Green, a professor at NTNU’s Department of Economics.
The big questions school authorities must ask themselves are what they actually intend to use test results for, says Professor Colin Peter Green at NTNU. “Do the tests mean anything? Can we use them to adjust teaching methods? Can they help teachers improve instruction?” Photo: Sølvi W. Normannsen, NTNU
No miracle cure
“The most important finding, at least from this study, is that there is no miracle cure that dramatically improves educational attainment in maths or increases well-being at school,” said Green.
The study nevertheless identified some weak trends: The pupils, especially girls, felt that the tests slightly improved their interaction with the teachers and somewhat increased their engagement. This could mean that concerns about tests reducing pupils’ well-being are unfounded.
“But this is something we haven’t measured. We used the pupils’ own assessments taken from the Pupil Survey that the same cohorts participated in later,” emphasized Green.
Little evidence of what works
When asked if the study shows that Norway differs from other countries, the school researcher said:
“Despite much debate, there is little evidence of what actually happens when you increase the testing of young children. The reason is that the variations we observe are primarily between countries, while we lack data on what happens over time. In East Asia, children are extensively tested from an early age, but you cannot compare the academic performance of 8- and 9-year-olds in Japan with that of their peers in Norway or Finland. We have not found any causal relationships.”
The most important finding, at least from this study, is that there is no miracle cure that dramatically improves educational attainment in maths or increases well-being at school.
So, what is the point?
“Why test young schoolchildren when there is absolutely no evidence that it works?”
“Some people want to discontinue the screening tests in Years 1 and 3, but I believe these tests do have some merit. Their implementation is cost-effective and they don’t hurt anyone. They can also help even out differences by identifying pupils who need additional support. It is easier to help young children who are struggling than 15-year-olds who have been struggling in school for years,” explained Green.
He believes the big question the school authorities need to ask is what they actually use the results for: Are the results important? Can we use them more directly to adjust teaching methods? In what way can they help teachers identify where and how to improve their teaching?
Zero stress
Some of the pupils in the study had already taken two or three tests at lower primary level. At that age, the test probably felt more like a game, and the pupils probably did not think much about why they were doing the tests. When it was time for the national tests in Year 5, some of the pupils had just completed the final test from lower primary school. Some of them had therefore been tested several times, with shorter intervals between tests than others.
According to Green, it can be assumed that some pupils then began to reflect more on the purpose of it all, or feel expectations and increased pressure.
“But it appears to have no effect on them whatsoever. I was expecting to see a little more variation and a greater impact. Perhaps our findings would have looked different if we had used a sample, and the study had been scaled up to 300 schools instead of 81. But to be honest, I do not think the findings would have looked any different then, either.
Source:
Colin P. Green, Ole Henning Nyhus, Kari Vea Salvanes: ‘Does testing young children influence educational attainment and wellbeing?’, Journal of Population Economics. https://doi.org/10.1007/s00148-025-01060-z
