Includes updates and/or revisions.
Reducing class sizes—a popular policy among parents, teachers, and lawmakers—has long been viewed as a way to increase student achievement.
But while shrinking the number of students in a class can lead to higher test scores overall, it might not necessarily reduce the achievement gaps that exist between students in a given classroom, a new study suggests.
Reviewing data from Project STAR—a longitudinal research study on class-size reduction in Tennessee and the most famous experiment on the topic—Spyros Konstantopoulos, an assistant professor of education and social policy at Northwestern University, in Evanston, Ill., said that it’s “tempting” to think that having fewer students assigned to a teacher will reduce the achievement gaps between students.
Instead, he found, “manipulating class size” doesn’t appear to narrow those gaps. In fact, the range from the lowest achievers to the highest achievers—what he calls “variability”—was greater in the smaller classes of 13 to 17 children than it was in larger classes of 22 to 26 students. He came to that conclusion after looking at the performance of all students in the STAR study on the Stanford Achievement Test.
“In the present study, I found that high achievers benefit more from being in small classes than low achievers,” Mr. Konstantopoulos said in an e-mail. “This indicates that the achievement gap is larger in small classes than in regular-size classes.”
He suggests in the article, which is being published in the March issue of Elementary School Journal, that the higher achievers, perhaps, are better at taking “advantage of the opportunities or teacher practices that take place in small classes.”
Small classes in Tennessee’s Project STAR study produced greater differences—“variability”—between lower- and higher-achieving students than regular-size classes, an analysis suggests.
SOURCE: Elementary School Journal
Mr. Konstantopoulos’ paper is the latest analysis of the STAR study data, and likely to generate discussion among supporters of class-size-reduction policies.
Other than Project STAR, which was paid for by the Tennessee legislature and involved roughly 7,000 students from 79 schools, from 1985 to 1989, research on the effects of class-size reduction has been limited. While the children in the small and regular-size Project STAR classrooms have been the subjects of follow-up research, a similar large-scale experiment has not been undertaken.
“I suspect the current [U.S.] Education Department has other priorities,” Alan B. Krueger, a Princeton University economist who has also conducted research using Project STAR data, said about the Bush administration.
During the Clinton administration, a federal class-size-reduction initiative provided up to $1.6 billion annually for states to pay for smaller classes. The funding, however, was rolled into other efforts to improve teacher quality when Congress passed the No Child Left Behind Act in 2001.
Eric A. Hanushek, a senior fellow at the Hoover Institution, a think tank at Stanford University—and one of the scholars who have maintained that the benefits of reducing class size are minimal—suggested that supporters of class-size reduction are satisfied with the STAR results.
“People got the answer they wanted, and didn’t want to jeopardize it,” he said. He called it “outrageous” that California, for example, spends close to $2 billion a year on a voluntary class-size-reduction program without a rigorous study of its effects.
As with many school improvement methods, the results—and interpretation of results—from class-size-reduction efforts are mixed. While some researchers say that smaller produce increases in student achievement, others argue that once class sizes drop, teachers must also alter their practices for students to benefit.
Regardless of the debate, many policymakers and the public favor—and expect results from—class-size reductions.
“Once it’s in, it’s hard to take away,” Mr. Hanushek said. “It’s an idea that has such popular appeal.”
In Utah, a bill that would continue an elementary school class-size-reduction program—but add new accountability requirements for districts—is moving through the legislature. A recent legislative audit showed that $460 million over the past seven years was used to maintain class sizes, but not lower them.
Some states are even extending such efforts into the upper grades, though the bulk of the research, and practice, focuses on the elementary school years. In his article, Mr. Konstantopoulos notes that the academic benefits of small classes even seem to be the greatest during kindergarten and 1st grade, and fall off in 2nd and 3rd grades.
In North Carolina, a pilot program was approved last year to cut class sizes in middle schools to 17 students as part of a program to strengthen those grades, reduce high school dropout rates, and increase college attendance. The competitive-grant program, which is estimated to cost close to $1 million per middle school, is expected to start with three schools in the 2009-10 school year.
In addition to California’s incentive program for reducing K-3 class sizes, a $2.9 billion grant program—financed through the settlement of a lawsuit brought against Gov. Arnold Schwarzenegger by the state teachers’ union—is being used in part to bring the average class size down to 25 students in grades 4-12.
And in Florida, where voters passed a class-size-reduction amendment in 2002, schools are seeking flexibility to deal with class-size changes when students transfer in during the school year. (“Leaner Class Sizes Add Fiscal Stress to Florida Districts,” Feb. 20, 2008.)
Calls for More Research
Tennessee’s Project STAR, short for Student/Teacher Achievement Ratio, began in 1985, with children randomly assigned to small classes, regular-size classes, or regular classes with an aide for each teacher.
After four years, researchers Jeremy Finn, a counseling and school psychology professor at the State University of New York at Buffalo, and Charles M. Achilles, an education professor at Seton Hall University, in South Orange, N.J., found a significant relationship between smaller classes and higher academic achievement.
Follow-up studies have shown that children who were in the smaller classes continued to outperform those from larger classes.
In 2005, Mr. Finn, who is still conducting research on the subject, published a paper showing that students assigned to the smaller classes for the full four-year period—especially those from low-income homes—were more likely to graduate from high school than were those in larger classes.
Princeton’s Mr. Krueger found in a 2001 study that African-American students and students from poor families had larger gains in scores than other students when they were assigned to Tennessee’s small classes.
“Since these groups of students consistently have lower achievement scores than white and [non-low-income] students, I conclude from the experiment that smaller classes benefit those in lower-achieving groups the most,” Mr. Krueger wrote in an e-mail.
Even if gaps persist among the students within a class, as Mr. Konstantopoulos suggests, others argue that reducing class sizes has an overall positive effect on groups of students that usually trail behind in academic achievement.
“We’re still behind the class-size-reduction program,” said Shannon Farrell, an education fiscal services assistant with the California Department of Education. “It would be a mistake to let [K-3 classes] go back to 30 kids.” Mr. Konstantopoulos notes that previous studies he has worked on produced “inconclusive evidence” that reducing class size can significantly improve performance among disadvantaged and minority students.
Even those who differ over the implications of the research agree that more scientific research on class size is necessary.
In his article, Mr. Konstantopoulos calls for a “new randomized experiment” that would include observations of teaching practices and interactions between students and teachers in small classes.
And Mr. Krueger said, “I think it is very unfortunate that the federal government has not sponsored a large-scale experiment like [Project STAR.]” He added that “the nation should not have to depend” only on one study from Tennessee to determine whether class-size reduction is an effective strategy for improving student achievement.
Coverage of education research is supported by a grant from the Spencer Foundation.
A version of this article appeared in the February 27, 2008 edition of Education Week