'Final test has too many functions'
The final test in primary education selects students, measures the quality of the school and must provide teachers with feedback. Sounds useful, but the question is: for whom exactly?
image: Rosa Snijders
Behind a dam of final test results, three locks open and close rhythmically. The scores roll towards the sorting bins like bingo balls: 50 percent VMBO, 30 percent HAVO, 20 percent VWO. When one container threatens to fill up too quickly, a screw is quickly turned so that the corresponding lock remains closed for a little longer. “We don't have a talent model at all in Dutch education,” notes a testing expert Karen Hey, who expresses the course of events visually. “All students are lumped together and selected in such a way that the distribution across the school types has been virtually stable for many years.”
Research collective Spit tried, among other things, for Education magazine in collaboration with regional media such as the City source en Fresh Concrete to map out which tracking systems and final tests are used. The collaborating journalists soon came to the conclusion that there is more of the same on the market and their inventory ended up against a thick wall of system criticism. No one is against monitoring children's development, yet there are few fervent advocates for the way this is currently done. Or rather: manners.
The final test puts too much pressure on children
The political intention behind the mandatory student monitoring system since 2015 sounds sympathetic: the final test puts too much pressure on children, but also on schools that then have to deal with ambitious parents. By making student monitoring mandatory, schools would get a better picture of development, in addition to the already mandatory final test (since 2013). “Eight years at primary school is a film, and the final test is a photo,” is how SP MP Tjitske Siderius expressed this prevailing idea in the House of Representatives in early 2016.
image: Rosa Snijders
In practice, the student monitoring system appears to be more of the same, only earlier in the school career. This is mainly due to the use of so-called 'placement guides' and agreements between secondary and primary education. To ensure that children receive reliable advice, not only that one final test is taken into account, but also their previous school performance. This shifts the stress of the final test to an earlier moment - when children are younger. Primary school teacher Jeroen Goes spoke prophetic words in a guest column in the magazine didactic just before its introduction in 2015: 'No headmaster will come clean about it, but categorical grammar schools will really look at the results from group 5 of primary school onwards.'
Developmental psychologist and educational psychologist Anke Klein, affiliated with Leiden University as a senior lecturer and co-founder of the Knowledge Center for Anxiety and Stress in Youth, recognizes this from her own experience: “We often hear that teachers tell students that Cito tests from group 5 count towards the final advice, that's quite a bit of pressure you put on a child.” From a development perspective, she wonders whether this pressure on young children is 'wise and desirable'.
Children depend on the performance of other children for their position
Officially, children are selected at the age of 11/12, but due to mutual agreements between schools, they feel this pressure much earlier. “The tests are taken from group 3 onwards, and then the comparison with each other starts,” says Heij. Comparing children with each other as a means of selection has side effects. “Children are organized in relation to each other. To climb to a higher level, another child must come down from the sorting box above you.” In other words: children are dependent on the performance of other children for their position.
Look under the hood
Not only is there criticism of the use of the final test as a selection tool, the way in which the selection is done also raises questions. It is difficult to look under the hood - and probably a well-kept trade secret - but thanks to the brightly colored graphs and claims made in the glossy advertising videos, we can say something about what is measured and what expectations are raised. Take the student monitoring system, for example Student in the picture from Cito. A demonstration video shows how teachers and parents can see at a glance whether children are below, above or at 'average growth' for, for example, arithmetic, reading comprehension and vocabulary. "Not only the teacher, but also the student himself, gets a complete picture of where he stands with Student in Focus," is how primary school director and co-developer of Cito products Willem van Till describes the product in Trots, the professional magazine for primary and special education.
The Netherlands suffers from a collective measurement obsession
“Such a beautiful dashboard, with sleek graphs and figures, people are quickly impressed by it and it is sold as a complete picture. That is exactly the problem,” says Berend van der Kolk, researcher of measurement systems at the Vrije Universiteit. In his book Measuring company he examined the 'collective measurement obsession' of the Netherlands. An obsession that he also sees reflected in the student monitoring systems. While measuring performance - such as vocabulary - can be a useful tool, he worries that the limitations of such a graph are often overlooked.
Puns
Vocabulary is one of the many components on which you can measure language skills. Student monitoring systems and final tests mainly focus on vocabulary and reading comprehension. That is a fairly limited summary of the concept of 'language'. Being able to make puns, write creatively or discuss stories you have just read: aren't these also part of language? “The decision was made to test skills that can be administered with multiple choice and are therefore easy to compare,” Heij explains.
The testable business side of language education has gained the upper hand
The way skills are measured has unintended side effects. For example, children who are raised bilingually may know 75 words in Dutch and 75 words in another language. If you only look at the Dutch vocabulary to determine how good a child is at language, the bilingual child will unfairly lag behind a monolingual child with a vocabulary of, for example, 120 Dutch words. “While the bilingual child is actually ahead, he still falls behind,” says organizational scientist Van der Kolk.
Experts Van der Kolk and Heij are certainly not alone in their criticism. The Education Board already dryly noted in 2022 that 'the testable business side of language education has gained the upper hand'. It immediately raises questions about the second sympathetic promise of the introduction of the student monitoring system: it would also help the teacher to identify delays in a timely manner. But behind on what? Language proficiency or testable elements of language proficiency? “As a teacher you naturally compare the results with your own observations. For example, some tests have linguistic arithmetic sums and the child may not understand the context due to reading comprehension, but can still do the sum," says teacher Masja Straetemans. According to her, it is mainly about expectation management: not only towards the children, but also towards the parents.
Children on sedatives
Media reports about the tests do not always help. Straetemans: “Then as a teacher you try to relieve the stress and get into it Youth news reports about children on sedatives needing places to relax.” Or what about RTL News, which for years gave 'report marks' to schools based on Cito scores. “In the media, test results are often linked to quality,” says director Remco Prast. “The image that tests are an absolute fact of the state of education and the children is prevalent throughout society.”
The Education Inspectorate is not in favor of intensive practice with tests
When asked whether so many roles for the same final test are possible, the Education Inspectorate states that they use different data for every other purpose. After all, quality measurement consists of more than just final test results. Despite the beautifully formulated processes, the Central Planning Bureau found in 2019 that the dual function creates perverse incentives: for example, there are schools that let students practice the final test so that they receive a good assessment. Students may be good test takers, but the final score says (even) less about their skills. Skills such as writing and reasoning with numbers are pushed to the background: it is literally impossible to score with them. Also, not all final tests are completely comparable, which means that schools can choose the final test that gives the highest school recommendations. The Education Inspectorate says it is not in favor of intensive practice with tests, but does not monitor schools for this. Choosing the test for which a higher final recommendation is expected is recognizable to the inspectorate, but according to them, the introduction of standardization in 2023 has put an end to this.
Children are excluded
“Certain children can be exempt from taking the test because they lower the average score,” says director Marten Elkerbout of Public Education Spaarnesant. A score on which the schools are partly assessed by the Education Inspectorate. “That is so contradictory. On the one hand, we are told that we must provide inclusive education, but on the other hand, children are excluded when it is not convenient.” When requested, the Education Inspectorate states that it will include the results of 'all students' in the assessment, except for 'the test results of students who meet the grounds for exemption'.
Standardize, select, qualify. There are a lot of functions that are linked to that one test method. “We have started to confuse indicators with goals. People no longer think about what the nature of education is, but calculate it based on how we score,” Elkerbout notes.
