Current information about this program can be found under the Education Research program.

Empirical Education Inc.
Principal Investigator: Denis Newman
Development Project: Low Cost Experiments to Support Local School District Decisions

The movement toward evidence-based practice in education has increased the pressure on educators to enact reforms on the basis of sound research findings. There is little research, however, that measures the effectiveness of many instructional practices, and what evidence there is often does not take into account the specific circumstances faced by education decision-makers in their schools or districts. The purpose of this project is to evaluate the feasibility of helping school districts design and carry out their own experimental studies to guide their decision-making.

The researchers are carrying out this project in collaboration with a set of school districts. The researchers are developing and refining a model to help school districts carry out randomized experiments of education interventions as a guide for decision-making. Three experiments are being conducted by the research team and their district partners in the first year, followed by five in the second year and ten in the third year. The research team is also studying whether the costs of these experiments are reasonable in relation to the cost implications of the school improvement decisions they are supposed to inform. A group of external evaluators are providing feedback about the process, to help refine the model based on these experiences. The goal is to determine whether helping school districts develop and incorporate low-cost, scientifically sound randomized experiments into their program planning and decision-making is a workable, economical model for providing technical support for school district improvement efforts.

MDRC
Principal Investigators: Janet Quint and Jason Snipes
Development Project: Learning from Efforts to Strengthen Educational Leadership in Urban School Districts

The accountability context in which education reform efforts now occur places increasing pressure on education leaders to produce results, especially increases in test scores and reductions in achievement disparities by income and race. This pressure to improve student achievement places a premium on defining effective leadership at various levels in the education system, thinking clearly about the connections and interactions between the actions of leaders at different levels, and focusing on how these actions relate to desired instructional practices and student outcomes. At present, however, there is little evidence regarding what good instructional leaders actually do that has an impact on students' academic achievement. The purpose of this project is to investigate the possible relationships among leadership actions, teacher practices, and student achievement, in order to build research knowledge about how education leaders can become more effective.

The project is being carried out at 80 primary schools in three urban school districts serving a high proportion of students from low-income families and diverse racial and ethnic backgrounds. The researchers are focusing on the actions of school principals and the school districts' intermediary supervisors who oversee the work of those principals. These school leaders have participated in the professional development activities of the Institute for Leadership (IFL), which supports a particular theory of effective instructional leadership. The researchers are using interviews, surveys, observation, and education tests to collect data to address three questions. First, do schools with leaders who demonstrate "effective" leadership actions (that is, actions consistent with the IFL's theory of effective instructional leadership) experience greater improvements in student outcomes over time? Second, do schools with "more effective" leaders experience greater improvements in instructional practice over time? And third, do these changes in teaching practice appear to explain the connection between leadership actions and trends in student outcomes? The researchers are developing reports about the measurement of leadership behaviors and about the relationships between these behaviors, teacher practices, and student outcomes. The overall goal of this study is to develop preliminary information on the empirical connections between specific leadership actions, changes in teaching practices, and improvements in student achievement. By doing so, the study may provide useful information for district and school leaders about how they can act effectively to improve teaching and, ultimately, student learning.

New York University
Principal Investigator: Leanna Stiefel
Identification Project: How Should We Organize Primary Schooling? Grade Span, School Size and Student Academic Achievement

Even though there is significant variation in the organization of schools in terms of overall school size and grade span, little is known about how different ways of organizing schools impact students' academic achievement. Obviously practical considerations at the local level often influence decisions about what grade levels schools serve, but it is also true that future policy and school construction decisions could benefit from knowledge about the effect of school organization on student academic achievement in general, or on particular groups of students. The purpose of this project is to analyze existing data to investigate what relationships might exist between the organization of primary schools-including school size, grade span (the number of grades and particular grades included), and classroom configurations (the number of classrooms per grade and school, size of classes, and composition of classes)-and the academic achievement of the students who attend them. In this study, primary schools are defined as those serving some combination of grades between first and eighth.

This research involves analyses of a database of a large urban school district. The database contains information on more than a million students in first through eighth grades per year over a seven-year period (1995-96 through 2001-02), including extensive data at the individual, classroom, and school levels. The primary focus of this research will be on analyses involving cohorts of third, fourth, and fifth graders followed longitudinally. Statistical modeling techniques will be used to assess whether school organization characteristics, including school size, grade span, or the number of classrooms, and class size at particular grades are related to changes in students' math and reading achievement, as assessed through standardized tests administered annually in third through eighth grades. The large sample size will also allow these analyses to include assessments of whether relationships between school organization characteristics and academic achievement vary depending upon selected student characteristics, including gender, special education status, eligibility for free or reduced price lunch, race/ethnicity, home language, English proficiency, and country of birth. Because the database includes longitudinal information on students as long as they remain within the school system, student progress over time can be tracked and associated with school organization characteristics. The overall goal of this study is to provide new information on school organization characteristics, particularly grade span-a factor that has been largely overlooked in the research literature-and the associations between these characteristics and student academic outcomes.

University of Pennsylvania
Principal Investigator: Jonathan Supovitz
Efficacy Project: Assessing the Impact of Principals' Professional Development: An Evaluation of the National Institute for School Leadership

Principals can make important contributions to the quality of the opportunities to learn provided by schools. But while many programs offer training opportunities to strengthen the capacities of principals and other school leaders, strong empirical evidence about whether and how these programs contribute to improved school leadership practice is lacking. In addition, there is scant research regarding how principals' leadership practices are connected to teachers' efforts to improve their teaching, and to student achievement. The current study is designed to address these gaps in the empirical literature by evaluating the National Institute for School Leadership (NISL) a district-level strategy for improving student achievement by developing principals' knowledge and skills. NISL is designed to develop principals' capacity to lead intensive instructional improvement efforts in their schools. The study will address the following research questions. First, how does participation in the NISL leadership development program influence the knowledge and practice of school principals who participate in the program? Second, how do the beliefs, practices, and opportunities for instructional improvement of teachers in schools led by NISL participants differ from those of teachers in schools not led by NISL participants, and how do these continue to develop over time? Third, in what ways does student achievement change in schools led by NISL participants?

Participants in this study are drawn from a large urban school district serving substantial numbers of students from low income families and diverse racial/ethnic backgrounds. A delayed-treatment experimental design is being used in which 30 primary school principals will be randomly assigned to begin participating in the NISL program while 30 others will be assigned to delay participation for a year. Principals will be followed for three years. The study will also include surveys and qualitative data collection that are intended to complement the randomized trial by supporting the identification of factors that may be associated with the effectiveness or ineffectiveness of NISL. Observations, logs, interviews, and surveys will be used to examine how participation in the NISL program affects the knowledge and practice of participating school principals. Teacher surveys will be used to examine how teachers' beliefs, practices, and opportunities for instructional improvement in schools with NISL participants differ from those of other teachers, using teacher surveys. Finally, effects of principals' participation in NISL on changes in student achievement will be examined using third and fourth grade students' performance on reading and mathematics. The results of the evaluation will provide information on the direct impact of the NISL program on school principals, and how that impact in turn leads to changes in teacher practices and student learning. The overall goal of this study is to increase understanding of how, and under what conditions, principal leadership development programs can work to create organizational conditions that ultimately support improvements in student achievement.

Vanderbilt University
Principal Investigator: Ellen Goldring
Identification Project: Public School Choice: Magnet Schools, Peer Effects, and Student Achievement

The use of magnet schools to accomplish public school improvement has been a controversial policy. While some people argue that magnet schools create competition that encourages schools to improve their quality, others claim that magnet schools attract the best students and teachers in the area, leaving the conventional public schools with poorer teachers and students. The research evidence to date has been inconclusive, in part because of the difficulties created by the fact that magnet schools have been chosen by the parents of the students who attend them, while other public schools acquire their students as a result of family residential location. The purpose of this project is to carry out an unbiased evaluation of the effects of attending magnet schools on student academic achievement, and to explore what factors might produce such an effect, if magnet schools are indeed better.

This project is being carried out in a school district that provides a naturally randomized population of students in magnet and conventional public schools, because more parents request magnet school assignments for their children than the magnet schools can accept, and the school district uses a lottery system to make school assignments. Ten of the school district's thirty-five middle schools are magnet schools, and the school district is located in a racially diverse urban area where half the students are from low income families. The researchers have access to school district data about students before and during their enrollment in middle school. These data are being analyzing to determine whether students attending magnet middle schools make greater academic gains than comparable students enrolled in conventional public schools. They are also examining the data to try to discover what factors contribute to magnet school success, focusing particularly on the possible effects of peers on student achievement, including the possible influence of student attendance and disciplinary histories. The overall goal of this project is to improve instruction and student achievement by providing a rigorous evaluation of the effectiveness of magnet schools and what factors contribute to whatever differences in student performance are found between magnet and conventional public schools.