To help districts and individual schools make those crucial decisions about whether to adopt a comprehensive school reform model, the American Institutes for Research has published a study rating 22 widely used elementary school reform programs. Produced by the institutes’ Comprehensive School Reform Quality Center, the report pulls together a large body of data and then rates each reform model—from zero to very strong—on how well it demonstrated improved student achievement.

“Our purpose in providing the ratings is not to pick winners and losers but rather to clarify options for decisions-makers,” says Steve Fleishman, who oversaw the study. (Fleishman is a former AFT educational issues staffer.) “This report is being issued in the hopes that the information and analysis it provides contributes to making research relevant in improving education.”

The reform models reviewed include the most prominent national programs, some serving thousands of students, mostly in high-poverty, low-performing schools. Even the smallest models are in at least 20 schools in three states.

The ratings indicate that there are significant differences among the models when it comes to raising achievement. Only two—Direct Instruction and Success for All—received “moderately strong” ratings in the category titled “evidence of positive effects on student achievement.” (None of the models received “very strong” ratings.)

At the other end of the spectrum, seven reform models received a “zero” rating; the report indicates that zero means evidence was found to provide a rating but none of it was of sufficient quality to be counted as reliable. See the table above for ratings of all 22 models.

While student achievement is the driving force under NCLB and school reform in general, the report also rates the reform models in other categories, including nonacademic outcomes such as student discipline, school climate and teacher satisfaction; positive effects on parents, family and community involvement; links between research and the model’s design; and services and supports to schools to help them implement the model successfully.

Despite the large amount of data that was reviewed in compiling the report, the authors caution about its limitations. For example, implementation is vital, which means that even good models can show no results when done badly, and models with lower ratings might be a good option if implemented well. In addition, some of the information used to rate the programs was supplied by the service providers themselves, which can make it harder to verify their claims.

Ultimately, the authors say their goal is to provide “consumer-friendly reports based on the best available evidence and scientific thinking.” Educators, policy-makers and the public, according to the authors, “cannot be expected to do ‘what works’ until they actually know what works.”

The full report, which includes extensive information on each model, is available online at www.csrq.org/reports.asp.