The Center for Evaluation of the Academy's Initiatives for Children program was founded by Frederick Mosteller (Harvard University) to apply meta-analysis to social policy research. The Center's initial focus on educational research grew out of an examination of the literature, demonstrating that evaluations of many educational interventions do not provide reliable information about whether the interventions are effective. In an influential paper, "Sustained inquiry in Education: Lessons from Skill Grouping and Class Size" (Harvard Educational Review, Winter 1996), Mosteller and his colleagues compared the inconclusive results of poor research on skill grouping with the more precise answers derived from a rigorous, well-designed evaluation of class size: Tennessee's Project Star. The following editorial, reprinted courtesy of The Boston Globe (October 22, 1999), describes the findings of Project Star.
The Benefits of Smaller Classes
Part of the public's frustration with public education stems from the lack of good, hard data on what works best in the classroom. Too much of what passes for research in education is nothing more than "hunches, anecdotes, and impressions," according to the American Academy of Arts and Sciences.
A significant exception is Tennessee's Project Star, a controlled experiment that began in the mid-1980s and involved almost 12,000 students in kindergarten through Grade 3 who were placed either in small classrooms (13 to 17 students) or regular ones (22 to 25 students). Researchers tracked student achievement for the study groups through high school and beyond. They found that the students educated in small classrooms scored significantly better on both standardized and classroom tests in all subjects.
The value of that early experience could be seen years after students rejoined larger classrooms. By eighth grade, the students from the smaller classes exceeded the reading levels of other students by over a year. By graduation, 45 percent of the students from the smaller classes received honors diplomas, as opposed to 29 percent of students from bigger classes.
It is heartening to see the Massachusetts Teachers Association promoting the Tennessee study. The union is using good science to urge good policy. Wisely, union leaders are asking state legislators to consider the need for smaller classes in kindergarten through Grade 3 when funding school building construction and renovation.
Teachers unions should prepare their members to accept that the effective first-, second-, and third-grade teachers of tomorrow may be teaching just 13 students while colleagues in higher grades may need to teach twice that number. The Tennessee study shows that students up to third grade need at least three years in small classrooms in order to benefit academically in later years. Many educators believe small class size is not so crucial in higher grades, but the data are in dispute. A study comparable in quality to Project Star would be desirable.
In fact, the NITA is urging state officials to collect better data on class size, which should also be a national priority. Too often the public is lulled by hearing that the local student-teacher ratio is 15 to 1 or some such agreeable number. That number can be misleading when guidance counselors, assistant principals, and vocational education teachers are counted in the equation. When it comes to class-rooms up to Grade 3, the only statistic worth citing is the one that describes the precise number of students for whom the permanent teacher is responsible.
The most optimistic finding in the report is that reductions in class size can boost the scores of black and Hispanic students. The College Board, which oversees the Scholastic Aptitude Test, offers the latest research on the achievement gap between white and minority students. The gap usually appears in first grade and widens considerably by third grade. After that, high achievement is unlikely, as evidenced by reading and math tests in which only 10 percent of those scoring at the highest level are non-Asian minorities even though they comprise 30 percent of the test takers.
The Tennessee study found that black and Hispanic students from small classrooms up to Grade 3 closed the black-white testing gap by 54 percent. Teachers attribute that advance in part to more opportunities to work individually with students and less time spent on discipline.
So much research in education clouds the issues. The Tennessee class size study begins to illumine this overcast field.