Small Classes: Popular, But Still Unproven
As President Clinton pushes his new national initiative to shrink the size of classes for 1st, 2nd, and 3rd graders, he sounds very sure of the idea.
"We know that students learn more when they are in smaller, more orderly learning environments where teachers and students alike are not distracted from the job at hand," Mr. Clinton said in his recent State of the Union Address.
The president may be certain that reducing class sizes will enhance learning, but he doesn't have much company among education researchers. Despite decades of studies on the subject, scholars remain deeply divided over whether simply reducing class sizes can bring about lasting improvements in achievement.
"Clearly, reducing class size will not hurt educational attainment," said Douglas E. Mitchell, a University of California, Riverside, researcher who is tracking his state's year-old initiative to trim classes in the early grades. "On the other hand, it's one of the most expensive education reforms that you can undertake, and there is some question whether the benefits are worth the cost."
Mr. Clinton, for example, wants to spend $12 billion over seven years to train and hire 100,000 new teachers to staff all the added classrooms his plan requires.
Despite the high price tag, reducing class sizes is now among the most politically popular of school reform ideas. California, Florida, Indiana, Nevada, Tennessee, and Wisconsin are among 20 or more states that have either tried a plan to reduce pupil-teacher ratios or are considering one.
Parents like the idea, and teachers love it. The National Education Association has called for it since the 1960s. And state and local policymakers are embracing it.
The popularity of shrinking class sizes is based in part on its intuitive appeal. Fewer students means more attention for each one, the thinking goes, with less time spent on discipline and administration and more time on teaching.
"I think a child just gets sort of lost in the shuffle in a traditional-sized class,'' said Amy Helm, who teaches a small 1st grade class in Drexel, N.C.
The Clinton administration says it is pegging its proposal on more than opinion polls and conventional wisdom, however. Administration officials point to a landmark study involving 7,000 students in 79 Tennessee schools. Beginning in 1985, those students were randomly assigned to either normal-sized classes or to small classes of 15 to 17 students. The children remained in small classes from kindergarten through 3rd grade, and the researchers have tracked their progress ever since.
Not only did the students learn more while they were in smaller classes, the researchers found, but they also continued to have an edge--albeit a much smaller one--years after returning to normal-sized classrooms.
Members of minority groups and students in urban schools, in particular, seemed to benefit from the smaller classroom settings.
"This is one of the great experiments in education in United States history," said Frederick Mosteller, a Harvard University statistics professor who reviewed the Tennessee study 10 years later. "It definitively answers the question of whether reduction from this size to that size does make a difference, and it clearly does."
No one disputes that the Tennessee study, which was conducted by a consortium of four universities in the state, is the largest, longest-lasting, and most controlled study to date on the class-size question. But other researchers say the study raises as many questions as it answers for an initiative such as Mr. Clinton's.
For one, they ask, why didn't the learning gains build up from year to year? At the end of the 1st and 2nd grades, students in smaller classes were just as far ahead of their peers as they were in kindergarten.
"The assumption is that it's going to change the rate of learning. If so, then every year students who get ahead should get farther ahead," said Mr. Mitchell, an education professor at UC-Riverside and the director of a research consortium of 28 California districts.
Does that mean, other researchers ask, that one year of small classes would be just as good as three?
"There's a difference between doing a fine study and then using results for specific policy design," said Michael W. Kirst, a Stanford University education professor.
In the Tennessee experiment, for example, classes were trimmed from an average of 23 students to 15-17, Mr. Kirst noted. "How do those findings relate to California, where classes were 30 and above and went to 20?" he said. "Is there a magic number?"
In the mid-1980s, 15 schools in Austin, Texas, were given $300,000 apiece that they used to reduce class sizes in varying degrees. Only two reported dramatic gains in student achievement as a result, said Joan McRobbie, the California liaison for WestEd, a federally funded research laboratory in San Francisco. The schools with big improvements, it turned out, were undertaking a host of other reforms at the same time.
"That raises the question," Ms. McRobbie said. "Is class-size reduction all by itself going to make a difference?"
Looking for Alternatives
"There have to be better ways to improve education than reducing pupil-teacher ratios," said Eric A. Hanushek, an economist at the University of Rochester in Rochester, N.Y., who has studied the issue.
Mr. Hanushek examined more than 300 studies, many of which looked at the ratio of students to staff members. Contrary to the popular notion, however, he found that most of those studies suggested that either fewer educators per student is better, or that the positive difference that smaller classes make is too small to be noteworthy.
But other researchers note that Mr. Hanushek's calculations in many instances included, for example, librarians and special education teachers--neither of which figures much in reducing the actual class size in a given school.
A 1989 review by researchers at Johns Hopkins University of 14 studies came to a similar conclusion as Mr. Hanushek. It found positive--but minimal--effects for small classes.
But the Tennessee investigators, for their part, pointed out that proven "better ways" of upgrading schools are hard to find. And, of the small number of education programs with successful track records, such as reform models like Success for All and Reading Recovery, many already employ small classes, small-group work, or individual tutoring for all or part of the day.
Though not nearly as extensive as Tennessee's, a few recent studies bolster the case for small classes. In a study published last summer, for example, the Educational Testing Service's Policy Information Center in Princeton, N.J., found that 4th graders in smaller-than-average classes were about four months ahead of 4th graders in larger-than-average classes. ("ETS Study Supports Value of Smaller Classes," Aug. 6, 1997.)
And a Wisconsin program aimed at smaller classes for young students at high-poverty schools also produced test-score gains in its first year of operation. Students in that program improved more over the course of the year than did students in larger classes who weren't in the program.
But researchers aren't yet sure whether the gains were due entirely to the smaller classes or to some other aspect of the program. ("Research Notes: Studying Smaller Classes," Jan. 14, 1998.)
Lessons From California
California's massive experiment, the largest single effort to reduce class sizes to date, would seem to offer the best chance to answer the question once and for all. Beginning in the 1996-97 school year, the state launched a $1.5 billion effort to trim classes in grades 1 through 3. This year, the program was expanded to 4th grade as well.
But already, complications have arisen that cloud the picture. A preliminary survey of 89 districts suggests that the effort, which districts implemented in just six weeks, is spawning widespread shortages of certified teachers and classrooms.
Many districts are closing child-care centers and libraries to make room for classrooms. Hardest hit, the study found, are urban districts.
"Is it the right thing to do to put a teacher with no qualifications in a classroom with 20 students rather than to put a trained teacher with skills for teaching limited-English-proficient students in a bilingual classroom of 30?" Stanford's Mr. Kirst said.
His organization, Policy Analysis for California Education, is heading up a consortium of six research groups planning to track California's efforts.
But Mr. Kirst also observed that President Clinton seems to have drawn some lessons from California's experience. The administration's plan has a much longer time frame and would require states to, at a minimum, give new teachers competency and basic-skills tests. It would also allow states to use 10 percent of the money for teacher training.
The president also is asking Congress to allow states and localities to issue a total of $22 billion in no-interest bonds for school construction and renovation during the 1999-2000 school year.
Despite California's troubles, teachers there report high levels of satisfaction with the initiative, as have teachers in other states that have experimented with smaller classes. One of the biggest advantages, they say, is the opportunity to better gauge their students' strengths and weaknesses and to get to know them as individuals.
"Now, when I go to fill out report cards I feel I don't even need to refer to my grade book or any of the other documentation we keep," Ms. Helm said of her 1st grade class. "I do, of course, but I know my students better now."