The Evidence on Teaching

The Effects of Universal Pre-K on Cognitive Development

Gormley, W. T., Gayer, T., Phillips, D., & Dawson, B. · 2005

grade Cquasi-experimentindependentmixed
Sample
1,567 pre-K children + 1,461 kindergarteners
Population
Tulsa, Oklahoma universal pre-K.
Design
Birth-date RD, end-of-treatment only. Same design limitation as Weiland: the comparison includes a year of maturation.
Key findings
Large end-of-treatment effects widely cited as evidence for universal pre-K. Provides no durability evidence, and its design inflates estimates by comparing children a year apart in age.
Counterfactual
Children about to enter the SAME program; alternative care for the untreated group not measured.
Genetic confound
RD design; but see the maturation caveat.
Replication notes
The frequently cited Tulsa middle-school follow-up uses propensity weighting (grade D) and finds math but not reading effects.
DOI / URL
10.1037/0012-1649.41.6.872

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
WJ Letter-Word IdentificationSD+0.79standardizedend-of-treatmentunclearend-of-treatmentdomain-skill
WJ SpellingSD+0.64standardizedend-of-treatmentunclearend-of-treatmentdomain-skill
WJ Applied ProblemsSD+0.38standardizedend-of-treatmentunclearend-of-treatmentdomain-skill

Cited by