Mathematics Learned by Young Children in an Intervention Based on Learning Trajectories: A Large-Scale Cluster Randomized Trial

42 schools / 1,375 preschoolers

US preschoolers in low-resource settings (TRIAD scale-up of the Building Blocks curriculum).

Cluster randomized trial of a guided-constructivist, learning-trajectories preschool mathematics curriculum with substantial teacher professional development. The outcome is REMA, a measure built by the curriculum's own developers and closely aligned to the taught trajectories — which is why the magnitude should not be compared directly with standardized-measure effects.

A well-run cluster RCT producing the single largest curriculum effect in the early-math literature (g = 0.72) — on a developer-built, trajectory-aligned measure, at the end of treatment. Read strictly as an end-of-treatment, aligned-measure result; the follow-ups are where the practical meaning is decided.

Cluster-randomized, so the immediate effect is causal. What the design cannot tell you is whether the gain reflects durable skill or aligned-test familiarity.

Followed up by Bailey et al. 2016 and Watts et al. 2026, both of which document large fadeout. The effect itself replicated across TRIAD sites.