CoChat

What Have Researchers Learned from Project STAR?

Schanzenbach, Diane Whitmore · 2006 · Brookings Papers on Education Policy
Cited by 121

The DOI matches its Crossref record; 5 of the 5 references it cites were found in a registry or an index.

Abstract

What Have Researchers Learned from Project STAR? Diane Whitmore Schanzenbach Project STAR (Student/Teacher Achievement Ratio) was a large-scale randomized trial of reduced class sizes in kindergarten through the third grade. Because of the scope of the experiment, it has been used in many policy discussions. For example, the California statewide class-size-reduction policy was justified, in part, by the successes of Project STAR. Recent (failed) proposals in the Senate that sought federal assistance for class-size reductions were motivated by Project STAR research. Even the recent discussion of small schools often conflates the notion of small schools and smaller classrooms. Because of the importance of Project STAR, it has been studied by many scholars looking at a wide variety of outcomes and even exploiting the randomization to understand variations in inputs and other aspects of the education production function that do not directly relate to class size. This paper provides an overview of the academic literature using the Project STAR experiment. What Was Project STAR? Project STAR was a randomized experiment that assigned students to a small-size class (target of thirteen to seventeen students), a regular-size class (target of twenty-two to twenty-five students), or a regular-size class with a full-time teacher's aide. Teachers were also randomly assigned to class types. Randomization was done within school, so all analysis presented here looks at within-school differences by class size. The experiment took place in seventy-nine [End Page 205] Tennessee public schools for a single cohort of students in kindergarten through third grade in 1985–89. Eventually, 11,600 students and 1,330 teachers took part in the experiment. The experiment was funded by the Tennessee legislature under Governor Lamar Alexander (later the secretary of education under President George H. W. Bush and currently a U.S. senator), at a total cost of approximately $12 million.1 In the ideal implementation of this experiment, students were to remain with the same randomly assigned class type from kindergarten through the end of the third grade. In practice, there were several major sources of deviation from this model. Students who entered a participating school while the cohort was in first, second, or third grades were added to the experiment and randomly assigned to a class type. There was a substantial number of new entrants: 45 percent of eventual participants entered after kindergarten. An especially large group of students entered in first grade—fully one-third of first-grade participants were new in first grade—in part because, at the time, kindergarten was not required in Tennessee. A relatively large fraction of students exited Project STAR schools (45 percent of overall participants), due to school moves, grade retention, or grade skipping, which also caused deviations from the original plan.2 Students who were male, African American, or on free or reduced-price lunch were more likely both to exit and to enter Project STAR. In addition, in response to parental concerns about fairness to students, all students in regular and regular-aide classes were randomized again in the first grade. Finally, a smaller number of students (about 10 percent of participants) were moved from one type of class to another in a nonrandom manner. Most of these moves reportedly were due to student misbehavior and not typically the result of parental requests to move their child to a small class.3 This weakness of the experiment can be addressed through use of an "intent-to-treat" setup—that is, use of the variation caused by initial randomly assigned class instead of the actual (possibly nonrandom) class attended. The experiment only manipulated class size and did not provide additional teacher training, new curriculum, or any other intervention. One exception is that teachers in fifteen schools were offered a three-day training seminar between years two and three of the experiment (that is, as the students were entering second grade) and again the following year. The training was given to all...

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1353/pep.2007.0007

Cites

5 of 5 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 5 references

Cited by

See all 121

Metadata

Authors
Schanzenbach, D. W.
Venue
Brookings Papers on Education Policy
Publisher
Project MUSE
Published
2006
Volume
2006
Issue
1
Pages
205-228
ISSN
  • 1096-2719 (ISSN-L)
  • 1533-4457
OpenAlex
W2040291538
Cited by
121
References
5
FWCI
16.01
Type
journal article
Language
en
Open access
closed

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: What Have Researchers Learned from Project STAR? https://app.cochat.ai/papers/10.1353/pep.2007.0007
The paper
Schanzenbach, D. W. (2006). What Have Researchers Learned from Project STAR? Brookings Papers on Education Policy. https://doi.org/10.1353/pep.2007.0007

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.