The Effect of Class Size on Achievement

Glass and Smith (1979)

Glass and Smith examined 77 containing 900,000 students over 7 decades and ended up with this graph Meta-analysis of research on class size and achievement:

The methodology behind this analysis was criticized as being "sensible" but 'suboptimal". Re-analysis using newer techniques found much the same Hedges:

The analysis was criticized on an number of other points Class size research: A critique of recent meta-analyses, but one of the authors stood by their results (even in court) On criticism of our class size/student achievement research: No points conceded, and I find their defense generally solid. The one exception is that they only classified 14 trials as "well-controlled" and many of them were not in normal public school environments.

That being said, I do think this meta-analysis is a good examination of the literature in the late 1970s - it's just the literature wasn't amazing.

A meta-analysis of 10 studies of Indiana's project PRIME TIME found an effect size of d = 0.34 ± 0.25 on test scores when class sizes shrank from 26.4 to 19.1 McGiverin.

Project STAR

Project STAR is probably the largest and most famous RCT examining the effect of class size - to the extent that it is literally the only study the handbook of economic field experiments references on the topic.

Roughly 12,000 students and teachers were randomly assigned to one of three groups: (a) a class with 13-17 students, (b) a class with 22-25 students, or (c) a class with 22-25 students and a teacher aide. The study proceeded from kindergarten to 3rd grade.

A long-term followup Krueger found significant test score gains among students in the smaller classes, but those gains largely faded out once class sizes returned to normal in 4th grade. This is generally consistent with the ~75% fade-out in test scores usually seen after an educational intervention ends Bailey.

The authors also found that 43.7% of students in the small classes took the ACT or SAT - compared to 40.0% in the regular classes (the aid had no effect). The effect on black students was larger. The effect on ACT/SAT score was small (d~0.02) and insignificant (SE~0.03).

Goldstein et al

A meta-analysis in 2000 found that average test score decreases by d~0.02 per student, generally agreeing withh the STAR study's pre-fadeout results Goldstein.

All things considered, the best evidence is probably the STAR project and this most recent meta-analysis: shrinking classes from ~23.5 to ~15 students improves test scores by d~0.15 in the short-run and d~0.04 in the long-run.

The effects are probably stronger as class sizes shrink - a result generally confirmed by the known power of high-dosage tutoring.

Chetty et al

Chetty et al's treatise on income mobility also touches on schooling in section "VI.D School Quality" Chetty. They find a strong correlation between class size and mobility (r~-0.3), but not in urban areas.


Malcolm Gladwell suggests there is a U-shaped curve where the optimal class size has ~20 students. Unfortunately, he doesn't provide any trials arguing this - relying instead on teacher anecdotes and one correlational study

Glass, G. V., & Smith, M. L. (1979). Meta-analysis of research on class size and achievement. Educational evaluation and policy analysis, 1(1), 2-16. Hedges, L. V., & Stock, W. (1983). The effects of class size: An examination of rival hypotheses. American Educational Research Journal, 20(1), 63-85. Educational Research Service. (1980). Class size research: A critique of recent meta-analyses. Phi Delta Kappan, 239-241. Glass, G. V. (1980). On criticism of our class size/student achievement research: No points conceded. The Phi Delta Kappan, 62(4), 242-244. Krueger, A. B., & Whitmore, D. M. (2001). The effect of attending a small class in the early grades on college‐test taking and middle school test results: Evidence from Project STAR. The Economic Journal, 111(468), 1-28. Bailey, D., Duncan, G. J., Odgers, C. L., & Yu, W. (2017). Persistence and fadeout in the impacts of child and adolescent interventions. Journal of research on educational effectiveness, 10(1), 7-39. McGiverin, J., Gilman, D., & Tillitski, C. (1989). A meta-analysis of the relation between class size and achievement. The Elementary School Journal, 90(1), 47-56. Goldstein, H., Yang, M., Omar, R., Turner, R., & Thompson, S. (2000). Meta‐analysis using multilevel models with an application to the study of class size effects. Journal of the Royal Statistical Society: Series C (Applied Statistics), 49(3), 399-412. Gladwell, M. David and Goliath: Underdogs, Misfits & the Art of Battling Giants. Gladwell, M. (2013). David and Goliath: Underdogs, misfits, and the art of battling giants. Little, Brown. Chetty, R., Hendren, N., Kline, P., & Saez, E. (2014). Where is the land of opportunity? The geography of intergenerational mobility in the United States. The Quarterly Journal of Economics, 129(4), 1553-1623.