New exam format reduces social disparities in grades in Danish schools

In the wake of the major public school reform of the 2010s, the grading of written final exams in Danish and mathematics was changed. Previously, the written test was examined by both the student’s teacher and an external examiner.

Today, written exams in subjects such as Danish and mathematics are graded solely by an external examiner who does not know the student.

New research from the University of Copenhagen shows that this has had a marked effect on grade differences between students from different socio-economic backgrounds.

“We find that inequality in grades between students from more and less advantaged socio-economic origins drops quite significantly after the reform,” says Assistant Professor Jesper Fels Birkelund from the Department of Sociology.

Teachers influence grading

It is well documented that students from resourceful families tend to perform better in school than classmates from less advantaged backgrounds. But now that teachers are no longer involved in exam grading, these differences have decreased significantly.

“Removing the teacher from the equation reduces inequality in grading between students from more and less resourceful families by 15 to 20 percent in the written Danish exam. That is quite remarkable. We rarely find effects this large and evident in our work,” says Jesper Fels Birkelund.

To understand why this development has occurred following the reform, one might assume that teachers generally display socio-economic bias and have given higher grades to students from “better-off” families.

“One could imagine that teachers give higher marks to children of highly educated parents because these students are better at addressing the teacher in ways that make them appear more academically bright than they actually are,” Jesper Fels Birkelund says.

However, the results of the study show that the explanation is not that simple. It is not about direct favoritism toward certain groups of students, but rather a more indirect mechanism.

“What explains most of the reform effect is that teachers tended to take students’ previous academic performance into account when grading exam assignments. In the exam situation, earlier performance ends up influencing the teacher’s evaluation, even though only the actual exam paper is supposed to be assessed,” he explains.

In other words, teachers cannot entirely set aside their knowledge of how students have performed in the past.

“If a student underperforms on the exam day, the teacher may have a tendency to compensate by awarding a higher grade than the actual performance merits,” says Jesper Fels Birkelund.

Students from more advantaged backgrounds generally have stronger academic records throughout their schooling and therefore a higher chance of benefitting from this tendency. Now that teachers are no longer part of the grading process, this effect disappears, and the external examiner can only assess the performance shown in the exam itself.

What is the purpose of an exam?

Some may argue that it was actually preferable when teachers were involved in grading and could take the students’ overall performance into account. In that way, capable students - regardless of socio-economic background - could avoid being penalized for a single off day.

“This raises a fundamental question about what an exam is supposed to measure. Some may believe the old model was beneficial because it provided a more complete picture of a student’s ability, allowing the teacher to include more data points than just the exam paper,” says Jesper Fels Birkelund.

“But exam regulations clearly state that it is the exam itself that must be assessed, not other factors. And even if teachers did not have a direct socio-economic bias, there appears to have been a reputational bias, where students were rewarded based on their academic reputation - something that particularly benefited students from certain social origins,” he adds.