Inside the Black Box of Schools
Policy briefing dissects research on classrooms, teachers, and school leaders

Fall 2005, Volume 27, Number 1

	James Spillane discusses distributed leadership in schools with an attendee.

Observers seem to have few problems in identifying lackluster student achievement in American schools. Yet pinpointing the school-level factors that contribute to students’ achievement—and avoiding those that do not—is difficult. At a February 4 IPR policy briefing, supported by a Joyce Foundation grant, a trio of experts delved into the “black box” of schools, exploring the impact of classroom size, teacher quality, and school leadership on school and student success.

Evidence about school leaders

Does school leadership have a direct effect on student achievement?

Past studies of school leadership have found only small, mostly indirect, effects on student learning, noted IPR Faculty Fellow James Spillane, professor of human development, social policy, and learning sciences. But these effects can account for up to 25 percent of all variation explained by school-level variables, and tend to be stronger for the nation’s most troubled schools. The available evidence has its share of problems, especially when it comes to making causal claims. According to Spillane, these problems are a function of both the research design and the definition of school leadership.

Having led the nation’s largest study on distributed leadership, which examines how school leadership is distributed among formally designated and informal leaders, Spillane pointed out that the principal’s job is too big for one person. School principals are expected to work on three sets of organizational functions: setting direction for their school, managing teacher and staff development, and supporting the school’s organizational structure and culture. Yet it “makes no sense” to expect just the principal, and perhaps the assistant principal, to accomplish all of these tasks effectively, he said.

“It’s not just that there’s too much to do. It’s that in order to do this job effectively, there’s too much to know for one person,” Spillane said. In fact, leadership involves many people in schools, including teachers. One recent study of 120 schools found between three and seven formally designated leaders per elementary school.

The challenge is learning more about—and developing—leadership as a practice through evidence-based research. “We know a lot about processes, structure, and roles, but the how of leadership is not understood,” he emphasized.

Evidence from the classroom

IPR Faculty Associate Spyros Konstantopoulos, assistant professor of human development, social policy, and learning sciences, discussed class size and its effect on class achievement.

From his study of the state of Tennessee’s Project STAR* data, Konstantopoulos finds strong evidence to support the idea that students in small classes do better than those in larger ones: Each year, they tested higher in reading and math than their peers in larger classes.

But the more pressing question, he points out, is whether the benefits of smaller class sizes persist over time and help the students beyond elementary school.

It would seem so. From the results of a follow-up study, Konstantopoulos demonstrated that the students, including minority students, were more prepared to take standardized college tests and scored higher. This, in turn, gives them a higher probability of being admitted to—and attending—college.

While encouraging, more work remains to be done on issues such as whether small classes are cost effective and how much they can help to close the racial achievement gap, he said.

*Project STAR (Student Teacher Achievement Ratio) was a $12 million large-scale, four-year randomized experiment, covering 11,000 students in 79 Tennessee schools.
Web site: www.heros-inc.org/star.htm

Evidence about teachers

Economist Kim Rueben considered whether teachers matter and whether their performance can affect student achievement. Rueben is a senior research associate at the Urban Institute in Washington, D.C.

“This seems like it should almost be a no-brainer,” she said. “Of course, teachers matter. If you are a parent or a principal, you know there are good teachers and bad teachers in your school. You know you want your kids in certain classrooms and not others.”

But it is extremely difficult to identify the indicators that will reveal which teachers are going to be more effective, Rueben said.

Rueben cited that across the board, experience seems to matter. The most effective teachers have passed the initial three-to-five-year learning curve.

Other possible indicators of teacher quality include the teacher’s alma mater, verbal test performance, and state certification—though it is difficult to untangle whether the teachers are effective because of certification or their experience, Rueben said.

As to whether the best teachers are distributed equitably, Rueben said they are not—mainly because teacher transfers are based on seniority. Thus more experienced teachers tend to be in—and can transfer to—schools in more affluent areas, leaving less experienced teachers in high-poverty, high-minority areas where skilled instructors are needed the most.

To view the presentations, please go to
www.northwestern.edu/ipr/events/briefingFeb05.htm