Longer-Term Effects of Low-Dosage Tutoring in Mathematics

Tutoring is considered one of the most effective educational interventions, but its longer-term effects are less understood. We leverage data from a large-scale field experiment to examine the effects of small-group instruction in mathematics for students aged 7–9. The intervention shares many features with other high-impact tutoring programs but had a lower dosage, targeted the full student population and used a pull-out strategy keeping the instruction time fixed. We find no significant effects of the intervention 3.5 to 4.5 years after it concluded, although the point estimates suggest fadeout of approximately 47% to 58%, depending on the specification.