Small‐ and large‐group instruction and a didactic method in mathematics for low‐performing adolescents: results from a randomized field experiment

Abstract

There is an increasing focus on helping low‐performing adolescents in school. We have randomly exposed low‐performing adolescents in mathematics in grade 8 to variations in key school inputs: teacher training, teaching material, and group size. Test scores of students in small groups taught by newly trained teachers provided with teaching material increased by 0.06 standard deviation. We find no impact on students in large groups with trained teachers and no new teaching material. Classroom observations indicate higher implementation fidelity of the didactic method among teachers of small groups, suggesting that targeted instruction with suitable teaching material in small and homogeneous groups can improve results for low‐performing adolescents and reduce achievement gaps in lower secondary education.