In school reform, the chasm between establishment advice and what the data show keeps on growing. In exchange for a “Performance Promise,” voters approved a $20 million bond issue for Jefferson County (Colorado) Public Schools to be used on projects that, according to the District’s web site, “have been proven to increase student achievement – smaller classes, classroom coaches, staff development, extended learning and individualized attention.”
But contrary to Jeffco’s claims, reducing teacher workloads does not improve student achievement. Between 1950 and 1994, the pupil-teacher ratio in American schools fell by 35%. Student achievement deteriorated. The achievement decline is not explained by changes in family structure, poverty, special education, or increasing numbers of immigrants. Some studies suggest that class size reductions may result in small achievement gains in special situations. In general, however, the more thorough the study, the more likely it is to find that class size reductions produce no gains in student achievement.
Project STAR, which followed Tennessee kindergartners assigned to classes of different sizes through high school, is often cited as proof that small classes raise achievement. A re-analysis of the data by Princeton professor Alan Krueger suggests that any class size effect was limited to kindergarten and first grade. Unfortunately, the quality of the underlying data is suspect. More than 50% of the children in the initial kindergarten classes had dropped out of the experiment by the end of the first 4 years. Project STAR also did not control for variations in teacher quality.
Teacher quality, not class size, is what school districts should improve. Especially teacher quality defined in terms of increases in student performance, rather than by years of teacher education or experience. In one large city school district, good teachers have raised student performance by 1½ grade equivalents in a single academic year. (Bad
