A new Think Twice review released today finds that a recent report on the effect of Florida's class-size reduction reform on student achievement does not actually study the impact of class-size reduction.
The Impact of a Universal Class-Size Reduction Policy: Evidence from Florida's Statewide Mandate, written by Matthew M. Chingos for the Program on Education Policy and Governance at Harvard University's Kennedy School, was reviewed for the Think Twice think tank review project by Professor Jeremy Finn of the University at Buffalo-SUNY. Finn, a statistics expert, was a lead researcher of Tennessee's Project STAR, a large, randomized experiment in class-size reduction (CSR). In 2002, Florida voters passed a constitutional amendment mandating CSR throughout that state's schools.
The Chingos study compares student test scores in districts that already had average class sizes smaller than required by the amendment with student scores in districts with average class sizes larger than required by the amendment. Districts that already had smaller class sizes received the same additional funding but could use the money as they saw fit, while those with larger class sizes were required to use the state CSR funds to reduce class sizes. Chingos concludes that "mandated CSR in Florida had little, if any, effect on cognitive and non-cognitive outcomes" in the students examined. Finn, however, points out that the study doesn't actually address the effect of CSR on student achievement. Instead the study compares the results of schools that reduced class size with a group of schools that received monies to use as they wished. Both sets of districts in the study had small class sizes. According to Finn, the study's finding would more accurately be stated as "administrative discretion in spending state class-size reduction funds did not affect students' academic performance."
Finn's review also points out that there are other flaws in the Chingos study: It uses the broad brush of school and district averages rather than student-level information about class sizes and test scores. Also, the actual class-size differences between the two groups were too small to make an educational difference; both of the groups had small average class sizes.
Finn concludes, "Despite its title, this report does not address the issue of class-size reduction. By being presented as an evaluation of Florida's mandated class size limits, it may lead parents, educators, or policy makers to draw faulty conclusions about the impact of the program."