When less is more: Effects of grade skipping on adult STEM productivity among mathematically precocious adolescents
Park G, Lubinski D, Benbow CP · 2013
grade Clongitudinalindependentunreplicatednumbers spot-checked
Sample
363 grade skippers vs 657 matched controls
Population
SMPY participants in the top 1% of quantitative ability, followed ~35 years.
Design
Matched-observational comparison within an already highly selected sample. The paper's own sensitivity analysis is the key number: hidden bias of only OR 1.11-1.34 would erase significance — this is a fragile causal claim by the authors' own accounting.
Key findings
The best long-run data on academic acceleration: skippers hit every career milestone earlier and produced more. The magnitude is selection-fragile by the authors' own sensitivity analysis, so treat the direction as informative and the 2x as an upper bound.
Genetic confound
Skippers were abler at baseline even after matching, and the decision to skip is itself correlated with motivation and family push. Genetic and motivational selection cannot be excluded — read as 'a saved year with no harm and probable compounding', not as a causal doubling of STEM output.
Replication notes
Never independently replicated in a comparable cohort; the sensitivity analysis makes the upside estimate selection-fragile.
DOI / URL
Effects
| Outcome | Metric | Value | Measure | Timing | Vs | Horizon | Class |
|---|---|---|---|---|---|---|---|
| STEM PhD attainment, skippers vs matched controls | % | 16.3 vs 7.9 (~2.1x); more publications, patents and citations by age 50; career milestones reached 0.8-1.6 years earlier | standardized | age 50 | business-as-usual | adulthood | attainment |
| Sex heterogeneity | pattern | pattern strong for males; female skippers at or below controls on STEM PhDs, shifted toward MD/JD paths | standardized | age 50 | business-as-usual | adulthood | attainment |
Cited by
- Acceleration and gifted programs — the label vs the contentmoderate supportconf: highgc: medium
- Placement and mastery diagnosis — deciding what to teach next from evidence of current skillmoderate supportconf: mediumgc: medium