Plato Knew
- Saturday, April 1, 2000
Executive Summary
Although school reformers have not gone so far as attempting to reach the Platonic ideal student-teacher ratio of one to one, tutor to pupil, they have been enthusiastic advocates of reducing class size. It is a proposal Which has enormous popular and political appeal, one of those innovations that just "feels" like it should work. Some two decades of research studies have indicated a relationship between classes of fewer than 20 students in the early elementary grades and achievement gains for the children fortunate to be in them.
Dr. Novello's paper focuses on two guiding questions: first, has all this research really proved anything, and second, what has happened in Nevada?
In response to the first question, the research results have, for the most part, been mixed. Some have suggested dramatic and lasting achievement gains from reduced class size, some have produced effects which had no staying power, and some have shown no relationship whatsoever between class size and achievement. Quite a bit of the research has indicated that teaching technique was an essential ingredient for success.
Most of the large state-funded studies, such as those in Wisconsin, Indiana, and North Carolina have suffered from obvious research flaws, but Tennessee's Project STAR (Student-Teacher Achievement Ratio) has been touted in all the literature for its accuracy. Kindergarten, first-, second-, and third-grade pupils and teachers in 79 schools were randomly assigned to three different kinds of classes (13-l7 students, 22-26 students without or with an instructional aide). The children in the small classes were reported to have dramatically outperformed their counterparts in the larger classes. The presence of a second teacher, or instructional aide, had no effect. Studies tracking these same students all the way through high school showed that the ones who had been in the small classes continued to achieve at a higher level than the others.
While these results were very impressive to school administrators, politicians, and lay people, a trained researcher could immediately spot problems with the studies themselves. For example, failure to control for certain variables such as differences or similarities in the classes or teachers, misrepresentation of correlation research as causal, and considering a result to be significant when it was not of statistical significance might have seriously corrupted the data.
Responding to the second question posed earlier, Nevada has spent around $254 million to achieve nothing. We still rank 44th out of 51 (the states and the District of Columbia) in teacher-student ratio. The main difficulty has been the lack of sufficient rooms to accommodate smaller classes, since funding for capital expenditures is a district issue. Team teaching has been as unsuccessful as Project STAR predicted it would be.
Dr. Novello's recommendation is to abandon the class-size reduction plan in Nevada and institute strict academic standards with high-stakes testing to see if the standards are being met. The money that we have been spending on trying to achieve small classes could be used to reward teachers whose students meet the standards and succeed at the tests.
