Sorting and Supporting: Why Double-Dose Algebra Led to Better Test Scores but More Course Failures
Nomi T, Allensworth EM · 2013
grade Cquasi-experimentindependentunreplicated
Sample
Chicago Public Schools 9th-grade cohorts under the 2003 double-dose policy
Population
CPS 9th graders in double-dose and regular algebra classrooms.
Design
Decomposes the double-dose policy into its two simultaneous effects: extra instructional SUPPORT, and the de facto SORTING it created because assigning below-median students to a second period regrouped classrooms by skill.
Key findings
Same policy, two opposite mechanisms. Extra instructional time raised test scores; the sorting it silently created lowered the peer environment and raised course failures. Any targeted-support policy that groups by need is running both at once.
Genetic confound
Peer-composition effects here are induced by an administrative rule rather than by family choice, which is what makes them interpretable.
Replication notes
The mechanism paper for the double-dose result; complements the short-run outcome paper and the long-run regression-discontinuity follow-ups.
Effects
| Outcome | Metric | Value | Measure | Timing | Vs | Horizon | Class |
|---|---|---|---|---|---|---|---|
| De facto tracking induced by the policy | SD | average classroom peer skill fell ~0.16 SD below the cutoff | standardized | 9th grade | none | end-of-treatment | domain-skill |
| Joint outcome of support plus sorting | pattern | better test scores but MORE course failures — the support raised learning while the peer regrouping raised failure rates | standardized | 9th grade | unclear | end-of-treatment | domain-skill |
Cited by
- Algebra timing — acceleration mandates, delay mandates, and readiness-based placementmixedconf: highgc: low