Small-group instruction to improve student performance in mathematics in early grades: Results from a randomized field experiment

•RCT using additional teachers to provide small-group instruction in mathematics.•Pull-out strategy with all students treated and instruction time held fixed.•Treatment: grades 2–4, 4–6 students per group, 4–6 weeks, 2 periods per school year.•Led to increased performance on national test by0.06of a standard deviation.•No differential effect by pre-ability level, suggesting that all students benefit. We investigate whether small-group instruction improves student performance in mathematics in the early grades using a large-scale RCT covering 159 Norwegian schools over four years. The students − 7–9 years old - are pulled out from their regular mathematics classes into small, homogenous groups of 4–6 students for mathematics instruction for 3 to 4 h per week, for two periods of 4–6 weeks per school year. Unlike many other recent tutoring experiments, all students are pulled out, not only struggling students. In our intention-to-treat analysis, we find that students in treatment schoolsincreased their performance by0.06of a standard deviation in national tests, with no differential effect by baseline test score level, parental education, or gender. Our study is particularly relevant for policy-makers seeking to use additional teaching resources to target a heterogeneous student population efficiently.