The Evidence on Teaching

Experimental Evidence on the Effect of Childhood Investments on Postsecondary Attainment and Degree Completion

Dynarski S, Hyman J, Schanzenbach DW · 2013

grade Arctindependentunreplicatednumbers spot-checked
Sample
Project STAR participants matched to National Student Clearinghouse records
Population
Tennessee STAR participants followed into their late twenties and thirties.
Design
Long-run follow-up of the STAR experiment using National Student Clearinghouse enrolment and degree records, which capture completion rather than attendance alone.
Key findings
Small kindergarten classes raised not only college enrolment but degree COMPLETION, with effects more than twice as large for Black students and largest in the poorest schools. This is the attainment half of the fadeout-with-persistence story.
Genetic confound
Randomized within school; the heterogeneity is by pre-existing disadvantage rather than by selection into treatment.
Replication notes
Added to this archive after adversarial review, because attainment COMPLETION (not just enrolment) is the outcome a founder should care about.

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
College enrolmentpp+2.7standardizedinto the thirtiesbusiness-as-usualadulthoodattainment
Degree completionpp+1.6standardizedinto the thirtiesbusiness-as-usualadulthoodattainment
Heterogeneitypp>2x the average effect for Black students; +7.3pp in the poorest third of schoolsstandardizedinto the thirtiesbusiness-as-usualadulthoodattainment

Cited by