The Bill & Melinda Gates Foundation will embark this fall on an ambitious research effort to analyze—and provide some initial answers to—a perennially vexing question in education: What are the best indicators of excellent teaching?
The foundation’s research partners intend to videotape and examine the teaching practices of 4,000 teachers in New York City, Charlotte-Mecklenburg, N.C., and several other yet-to-be named districts, to arrive at an understanding of the correlation between those practices and student learning.
In addition, the foundation will look at the relationship between student achievement and pupils’ perceptions of their instructors’ effectiveness; teachers’ content knowledge and ability to find the right pedagogical tools to teach that content; and “value added” estimates of teacher effectiveness based on test scores.
The research agenda makes up a component of the foundation’s five-year, $500 million push to define and promote effective teaching practices, and will help shape its approach to the teacher-effectiveness plans it intends to fund this fall. Foundation officials hope the research will help settle widely divergent opinions about what constitutes effective teaching—a lack of clarity that policy groups, administrators, and teachers’-union leaders alike say has contributed to poorly calibrated evaluation methods and perfunctory judgments that have left administrators frustrated and teachers bristling.
“One of the critiques teachers have—which we think is totally valid and fair—is that their performance shouldn’t be judged on one single measure of anything, whether test scores or observations,” said Vicki L. Phillips, the director of the foundation’s education division. “We want to take a very independent look at what indicators, when taken together, give you a fair, reliable, clear view of teacher effectiveness that both teachers and researchers can support and embrace.”
Among its other education philanthropy, the Seattle-based foundation provides grant support for Education Week.
‘Providing the Headlights’
The research project will occur alongside Gates’ “intensive partnerships for teacher effectiveness”—plans to retool recruitment, professional development, pay, and evaluation structures in selected districts. Four districts, and one consortium of charter schools, are finalists for that funding. The foundation will unveil its decisions in November. (“Finalists to Vie For Grants On Teaching,” Aug. 26, 2009.)
The foundation will invite teachers in some of those districts, in addition to New York and Charlotte, to participate in the research. The work in the selected sites will be done with the blessing of local leaders, educators, and teachers’ unions.
Gates officials could not provide an estimate of the overall amount they will spend on the research, but say the figure will be in the tens of millions of dollars. New York City will receive a $2.6 million grant to participate, while Charlotte is poised to receive $1.4 million.
The partnership districts have pledged to use the results of the research to adjust their own plans, Gates officials said.
“Almost all foundations think of research as evaluation that helps them look in the rear-view mirror; very few think of research as providing the headlights to help identify where are the things we could be doing that can have transformational impacts,” said Thomas Kane, the deputy director of research and data for the foundation’s education program.
The research will have several key parts, but its major thrust will be to study a variety of measures to identify effective teaching, and then to determine which combinations of those measures produce the most reliable judgments of individual teachers.
Evaluating Videos
Digital videos of teachers’ practices will be a central feature of the research. The taped observations will be scored by trained observers using a variety of performance rubrics that describe escalating levels of teacher competency. They will include the Framework for Teaching, developed by teacher-evaluation consultant Charlotte Danielson, and the Classroom Assessment Scoring System, developed by Robert C. Pianta, the dean of the education school at the University of Virginia, in Charlottesville.
Where appropriate, the videos will also be scored against several content-specific frameworks in math, English language arts, and biology.
Each teacher will be taped four times. Two of the observations will occur during lessons on common topics, and the other two will take place during randomly selected lessons. The point of that system, Mr. Kane explained, is so that researchers can determine whether similarity in lesson design helps observers’ ability to agree on a teacher’s effectiveness at presenting that lesson, or whether observers are capable of identifying high-quality teaching regardless of the context in which it occurs.
The observational data will be supplemented with student-provided information about teacher attributes that the students credit with accelerating their learning.
Overcoming Fears
Much of the foundation’s research initiative addresses teachers’ concerns about the use of standardized student test scores as a measure of teacher effectiveness.
That use of student scores has risen in prominence as an issue in part because the proposed guidelines for the $4.35 billion federal Race to the Top grant program ask states to incorporate such testing data into teacher-evaluation systems. (“Rich Prize, Restrictive Guidelines,” Aug. 12, 2009.)
Although the Gates Foundation will look at results from state reading and math tests in grades 4-8, it will supplement that information with data generated from assessments that the researchers will give to students in the various study sites and that contain open-ended, performance-based questions.
“We want to be able to take seriously the concern that people have raised that teachers that are successful at promoting achievement on the state tests aren’t necessarily the teachers that are good at promoting higher-level conceptual understanding,” Mr. Kane said.
Scholars have also raised questions about the ability of value-added assessment systems, which purport to isolate teachers’ contributions to student growth, to screen for factors that might skew estimates of teacher performance. The foundation will tackle that problem in a research component that will begin in the 2010-11 school year.
After gathering one year of effectiveness data from the various sources, researchers will randomly assign those teachers who appear to be the most effective, and those who appear to be the least effective, to specific classrooms with similar students. (The composition of the classrooms will continue to be set by school policies.)
Then, student-achievement results will be compared. The randomized design should be able to screen out statistical “noise,” organizers of the initiative say, and to verify whether the measures used accurately determine which teachers are most effective in the classroom.
Participation in the Gates research by principals and teachers will be voluntary. The teachers who participate will receive a stipend.
Douglas N. Harris, a professor of education-policy studies at the University of Wisconsin-Madison, praised the foundation’s research plans. But, he added, an equally important consideration for researchers will be to determine the appropriate use of the measures—whether for teacher improvement or for accountability purposes.
“Even if the measures are accurate, and as a result of the Gates Foundation’s work we have a better idea of who the best teachers are, it doesn’t mean that merit pay based on those measures will be effective,” he said.
