Reform of our schools founders on a lack of reliable knowledge. To introduce new programs in schools and to get them funded, programs are typically oversold. This overselling appears in terms of what outcomes can be expected and what evidence supports the program. As a result, school policymakers and the public are frequently disappointed by the results. Although not politically sexy, correcting the lack of information about program effectiveness should be a national priority.
The Elementary and Secondary Education Act, the blueprint for the federal policy, is up for reauthorization by Congress. It lacks a coherent strategy for improving our knowledge base. Nothing currently in the bill is likely to do as much for the education of our nation’s children as would an extensive program of educational experimentation.
In education, it is common to argue that urgency precludes taking the time to develop an extensive evidentiary base. Moreover, isn’t experimentation on kids undesirable, perhaps immoral? Thus, new programs are introduced more on the plausibility of the underlying idea than on evidence about effectiveness in improving student achievement. As a result, the typical high school senior today achieves at essentially the same level as the high school senior of 1970 even though real spending per student is up more than 80 percent.
Education research grounded in basic scientific principles is rare. At a basic level, most research fails to look at any objective measures of student outcomes or what causes what. For example, if we were interested in knowing the effect of a reading program designed for disadvantaged learners and we simply compared achievement of those in the program with those not in the program, we would have difficulty inferring the effectiveness of the program itself. The students in the program are selected for it because they need extra help. So are the differences in student performance a reflection of the program or the underlying selection mechanism?
An appealing way to circumvent such problems is to employ a random-assignment experiment, where students are placed into a program by chance. The extensive and informative social experimentation in the 1970s—including welfare, housing, and health insurance—has strongly influenced current policy. Medicine and agriculture have made dramatic strides through random-assignment experimentation. In education, only one experiment can be located (class-size reduction in the mid-1980s under Tennessee’s Project STAR), with continuing controversy about its implications. Instead of designing experiments to resolve the policy dilemma, states across the country have simply followed California’s example of rushing in with full-blown and expensive programs—never facing up to the possibility that nothing might happen to student achievement.
The fact is that schools are continually experimenting with students, although it is not called that and it lacks an evaluation component. We should experiment more systematically across a range of reform areas so that we can improve our knowledge base. A strategy for systematic innovations in the reauthorized Elementary and Secondary Education Act could be revolutionary.