Data-driven Duncan, the presiding spirit of public education, has promised us larger classes for the near future. He has  justified increasing class size with research, based on test scores. In a speech to the American Enterprise Institute (November 17) he declared,

“Up through third grade, research shows a small class size of 13 to 17 students can boost achievement. Parents, like myself, understandably like smaller classes. We would like to have small classes for everyone–and it is good news that the size of classes in the U.S. has steadily shrunk for decades. But in secondary schools, districts may be able to save money without hurting students, while allowing modest but smartly targeted increases in class size.”

Class size may or may not alter performance on standardized tests, but it inevitably affects the quality of classroom interaction as it ascends from a class of twenty. I do not have hard data to justify this claim, only twenty years of high school classroom experience with class sizes varying from twelve to thirty-eight. No one but an experienced teacher can verify what happens when five students are added to a class of 25 or 30 or 35, but the climate is significantly changed. The number of students who do not participate in discussion increases, the number of students who can receive individual support decreases, and the community feeling that promotes collaboration in learning dwindles.  These factors affect learning profoundly, especially for the students in the expanding undifferentiated middle of the class.

Many of the undifferentiated middle are subject to absenteeism and dropping out. Urban schools often count on these students dropping out when they assign more students to a classroom than there are seats. They anticipate attrition like the meteorologist anticipates the hurricane season. Many of these students will never take the tests that measure whether class size affects performance, because they won’t be in school that day.

Many of the undifferentiated middle will be chronically weak performers on standardized tests, so their lack of achievement won’t jar the mean test performance in a school, a city or a state. Some of them will actually be superb test-takers, so the size of the class will not affect their performance, but neither will they have learned anything that year.  Some students will transfer from another school system, where the classes were small. They will bring acquired critical thinking habits to the test and keep the scores buoyant.

Teachers know these stories, the lives behind the numbers  that dance in the dreams of state school administrators. Teachers know that mean test scores reflect very little of the learning climate in their classes, but individual students affect them profoundly. The motivated student whose parents decided to move away to a district where peer pressure boosted learning.  The ESL student who worked harder when the class size expanded, but learned less, because the teacher could not attend to individuals as often.  The bright, but disillusioned student who had not been challenged for years in school, because no one noticed he had skills far beyond his classmates.  These are a few of the casualties of class size.

Secretary Duncan clearly does not remember, if he ever knew, how numbers affect classroom interaction, so he resorts to the last refuge of scoundrels: gross data.  Administrators have cited these studies forever to make a virtue of necessity, but they are obviously administrators who have forgotten the ecology of real classrooms. Whatever the data of test performance reveals, it does not reveal the quality or depth of learning in secondary classrooms. It does not reveal who has disappeared from the radar, because of the unwieldy class size.

Perhaps we must all face the hardships of a depressed economy, and schools may have to shoulder their share.  If so, we should not expect school reform to advance under an increasing burden of enrollment and smaller faculty. Our alleged leader, the Secretary of Education, should deplore the conditions brought on by declining resources, not justify them with superficial data.

If the devil can quote scripture to his purpose, then so can a bureaucrat cite data. It doesn’t mean he should stoop that low.