DN Debate Replicas. "The shortcomings of today's national tests do not give reason to reject them completely"

In a reply to our debate article, Minister of Secondary Education Lotta Edholm (L) believes that today's national exam has serious flaws. We share that criticism, but the shortcomings are largely due to previous government decisions. Therefore, the tests at the primary level primarily measure the knowledge of low-achieving students.

In middle school, the tests are better adapted to the students' knowledge levels, but the Swedish National Agency for Education collects only the test grades, not the actual score results, which greatly limits the informational value of the tests. The fact that the Swedish National Agency for Education has not yet digitized the tests has also prevented the introduction of external corrections that would counteract joy corrections.

Research has developed increasingly better methods for establishing causal relationships between interventions and school results. Such evaluations require data on students' knowledge, but rarely comparable test results over time. Despite the shortcomings, the national tests have therefore been decisive for evaluations of several major school reforms, which has increased our understanding of what works in schools.

In order for future reforms to also be evaluated, access to broader measures of knowledge must be secured in the long term. We therefore propose that the national tests at primary and middle school be retained until a workable replacement is found. Without measures of all students' knowledge, future reforms cannot be evaluated, regardless of whether the Swedish National Agency for Education receives the assignment or not.

We fear that the deficiencies in today's national tests that Edholm points to risk being repeated in the design of the standardized tests. The government has decided that the purpose of the tests should be to identify students with support needs. The decision controls the design so that the tests do not capture all students' knowledge. According to our information, the Swedish National Agency for Education also plans to only collect data on which students achieve a certain level of requirements. It would significantly impair the possibilities of evaluating the school's development. Edholm writes that "the government's ambition is that the tests should be able to be used to follow the knowledge development of all students, regardless of whether they are high or low achievers". It is an ambition we share, but it is unfortunately incompatible with the decisions made by the government and the National Board of Education regarding the design of the tests.

It is possible to design standardized tests that both measure knowledge in the entire student group and identify students who need support. However, this is not how the government has chosen to design the new exam system. We therefore call on the government to broaden the purpose of the tests so that they measure the knowledge of all students. The Norwegian School Board must also collect the students' actual score results and not just information about who passed the tests.

And it's in a hurry. Results that are not measured and collected now cannot be recreated later. The Government and the Swedish National Agency for Education should also consult with experts on impact evaluation when the tests and the collection of results are designed. This reduces the risk that today's shortcomings will be passed on to the new system. In this way, knowledge of the school's development can be strengthened in the long term and contribute to an education that rests on a scientific basis.