When it comes to high school size, smaller might not be better, concludes a national study by researchers from Michigan State University.
The study, which was presented May 22 at a conference sponsored by the Washington-based Brookings Institution, raises questions about high-profile efforts taking root across the country to reshape the nation鈥檚 high schools.
Spurred by major donations from the Bill & Melinda Gates Foundation and other private and public funding sources, school districts in New York City, Chicago, Houston, and other cities are replacing large high schools with smaller, more personal learning environments for students.
But Barbara Schneider, the study鈥檚 lead researcher, said her data suggest those efforts are misguided.
鈥淚n an effort like this, you are dismantling large high schools and putting money into creating small high schools,鈥� Ms. Schneider, an education professor at Michigan State University in East Lansing, said in an interview. 鈥淲e can鈥檛 afford to continue down this path without serious and rigorous assessment of this thing.鈥�
Ms. Schneider and her co- authors, Adam E. Wysse and Vanessa Keesler, based their conclusions on data from the Education Longitudinal Study of 2002, a federal survey that tracks students beginning in 10th grade. The 1,000-plus students in the researchers鈥� study were surveyed twice, once in 10th grade and again as seniors.
Of the 660 schools in the study sample, the smallest tended to be in rural and suburban areas and to enroll mostly white students. The largest schools, most often located in urban areas and in some suburbs, enrolled higher-than-average numbers of poor and minority students.
Methodology Debated
Using a technique pioneered in the 1970s by Harvard University statistician Donald Rubin, the researchers tried to put all of the schools on a more equal footing by matching students on 98 different characteristics. Those characteristics included the kinds of courses the students had taken and the extracurricular activities in which they participated, as well as traditional socioeconomic traits such as race and family-income levels.
To measure students鈥� academic progress, the researchers looked at test scores in mathematics, whether students planned to attend college or had applied, and whether they chose a two- or four-year college or some other postsecondary institution.
The only students who performed better in small schools, the researchers found, were those who were most likely to attend them鈥攊n other words, mostly white rural and suburban students. For the urban and minority students in large schools, the smaller settings would have offered no significant advantages for the outcomes the researchers tracked, according to the authors.
鈥淢y thought really is that size doesn鈥檛 matter,鈥� Ms. Schneider said. 鈥淚t鈥檚 also about what goes on in schools.鈥�
Marie Groark, a spokeswoman for the Gates Foundation, which is underwriting improvement efforts in more than 1,600 high schools, did not disagree.
Papers and additional materials from the conference will be available from the in the coming weeks.
鈥淲e see really great results with our new small schools, but our existing schools have been slower to improve,鈥� she said, referring to large existing high schools that have been restructured into small schools or learning communities.
鈥淲e do know size is not a panacea for improving high school outcomes,鈥� Ms. Groark said. 鈥淚t requires a focus on personalization, on ensuring kids won鈥檛 fall through the cracks, increasing expectations for all, and on improving instruction and curriculum so that the classroom is a lively, engaging, thought-provoking experience for kids.鈥�
But other experts attending the May 22-23 conference questioned the study methods. They noted, for instance, that the survey began in 10th grade, a year after students had entered high school and after many students had already dropped out.
鈥淚t鈥檚 very hard to talk about school effects when kids have already experienced half of high school at the beginning of the study,鈥� said Valerie E. Lee, an education professor at the University of Michigan in Ann Arbor.
It would have been better, Ms. Schneider conceded, to follow students beginning in 8th grade. She got the same results, however, from analyzing the data in different ways as well.
Long a proponent of smaller high schools, Ms. Schneider said she reconsidered her position after meeting with Chicago-area high school sports coaches who complained that the small-schools movement threatened their programs. The coaches told her that sports were key to keeping marginal students in school and in generating college scholarships for students who might not otherwise see college as an option.
鈥淚鈥檓 afraid we have done a terrible disservice to kids,鈥� Ms. Schneider added.
Other Studies Presented
Her study was among 11 discussed at the meeting, which focused on examining the effects of school and class size. In a study conducted with Douglas D. Ready, an assistant professor of education at the University of Oregon, Ms. Lee also found that school size did not seem to directly affect learning for kindergartners and 1st graders. That was not the case, though, for changes in class size, Ms. Lee and Mr. Ready said. Using federal data from a nationally representative sample of 7,740 children, the pair found that kindergartners and 1st graders learned at the same pace in both medium-size classes鈥攖hose with 17 to 25 pupils鈥攁s they did in classes with fewer than 17 students. But being in a class of 25 or more students appeared to negatively affect children鈥檚 mathematics and literacy learning, the study found.
Because most children already experience medium-size classes in the early grades, the researchers said, their results raise questions for efforts going on around the country to shrink primary classes to fewer than 17 pupils.
Another study, meanwhile, highlighted some noneducational benefits linked with small schools. Analyzing federal survey data on 15,362 high school sophomores and their parents, Thomas S. Dee, an associate professor of economics at Swarthmore College in Swarthmore, Pa., found that parents whose children attended high schools with fewer than 800 students were more likely than parents from larger schools to know the parents of their children鈥檚 friends and to take part in PTA activities.
On the other hand, the study also found, the small-school parents were less likely than those from schools of 1,600 or more to belong to the PTA鈥攁 result that Mr. Dee attributed to more formalized recruiting efforts in the large schools.
