CLASSROOMS THAT MIX AGE GROUPS HAVE MIXED RESULTS: Evidence from Norway of pupil achievement in ”combination classes”

Children can benefit from being in classrooms that mix pupils from different age groups, but it depends crucially on the balance of lower and higher grades. That is the central conclusion of research by Edwin Leuven and Marte Ronning, which is published in the June 2016 Economic Journal.

Their study suggests a cautionary tale for learning outcomes in so-called ''combination classes'': pupils can be worse off if negative effects from the lower grades cannot be countered with positive effects channelled by the presence of higher grades.

The researchers note that many children around the world find themselves in classrooms that group pupils from different ages and/or grades. These combination classes are common in many poor developing countries but they are also often found in industrialised countries.

Although combination classes are sometimes advocated from an educational point of view, they typically arise because of economic constraints. When confronted with an increase or drop in enrolment, schools often group pupils from different grade levels to avoid an extra (costly) classroom.

There are several ways in which combination classes can affect pupil achievement. Classrooms constitute natural peer groups and grouping pupils from different grades in a single classroom changes the peer group relative to a single grade classroom. This may lead to direct negative or positive spillovers due to the presence of more or less able peers since pupils'' grades are positively correlated with their age and length of schooling, and therefore with cognitive development and achievement.

In addition, peers from higher grades can serve as role models in terms of non-academic behaviour, which can feed back to school achievement. Finally, classrooms'' grade composition can also significantly affect teacher inputs and teaching methods.

There is surprisingly little evidence about the impact of combination classes on pupil achievement. Some studies find positive effects, some find negative effects and others find no effects at all. This lack of knowledge is mainly due to the difficulty of finding appropriate comparison classrooms. To isolate the impact of classroom grade composition, these comparison classrooms need to be identical in all other respects.

This study sets out to estimate how classroom grade composition affects pupil achievement using Norwegian data. It uses a novel approach that exploits institutional features in Norway that significantly change the grade composition of classrooms, while also delivering comparison classrooms.

Norwegian junior high schools are bound by national regulation to determine classroom grade composition based on enrolment thresholds at the grade level. This means that some schools need to run combination classes, while comparable schools with slightly higher enrolment levels can run regular classrooms. These rules allow the researchers to estimate the impact of grade composition as well as class size.

Following this approach, they find evidence that a one-year exposure to a classroom that combines two grade levels increases exam performance by about 4% of a standard deviation. Further analysis shows that this effect is driven by pupils benefiting from sharing the classroom with more mature peers from higher grades, but there is also evidence that the presence of a lower grade is detrimental to achievement.

By the time they matriculate from junior high school, most pupils in mixed grade classrooms in Norway have spent time with both higher and lower grades. The average effect is therefore the sum of these positive and negative effects. Since the positive effect of sharing the classroom with a higher grade is somewhat larger in size than the negative effect of sharing the classroom with a lower grade, the average effect is small and positive.

These results illustrate that depending on the type of exposure, average effects of grade mixing can be negative, positive or close to zero. These results go a long way towards explaining the contradictory findings in previous research.

''Classroom Grade Composition and Pupil Achievement'' by Edwin Leuven and Marte Ronning is published in the June 2016 issue of the Economic Journal. Edwin Leuven is at the University of Oslo. Marte Ronning is at Statistics Norway.