Gates Teacher Study Opens Video Library to Researchers
The Bill and Melinda Gates Foundation's Measures of Effective Teaching study is complete, but researchers will have a chance to comb through thousands of hours of classroom videos for additional research.
More detailed information about what makes particular teachers effective in specific subjects, grades, or lessons may be unearthed in the MET longitudinal database and video library, maintained at the Institute for Social Research at the University of Michigan. Beginning in January, researchers can request access to quantitative, qualitative and video data from the MET extension study, which recorded more than 340 math classes and more than 200 language arts by 350 teachers in grades 4-9, in six districts during the 2011-12 and 2012-13 school years. Some of these lessons covered the Common Core State Standards, and all of the more than 15,000 video segments have been tagged to identify various teaching practices.
"We have an incredibly rich set of videos and observation protocols," as well as additional materials such as lesson plans and class observations, said Brian Rowan, a research professor at the Institute.
The $45 million original study involved video recordings, student surveys, value-added test scores, and other data collection on 3,000 teachers and their students in in six large urban districts: Charlotte-Mecklenberg, N.C.; Dallas; Denver; Hillsborough County, Fla.; Memphis, Tenn.; and New York City. It produced a series of reports on new ways to evaluate teacher effectiveness. (Full disclosure: the Gates Foundation supports some of Education Week's business and innovation coverage.)
However, some scholars have argued the original study's suggested teacher-effectiveness measures did not adequately predict how students would perform on more cognitively complex tasks such as those required under new common math and reading standards.
More extensive analysis of the video data could provide better context to those measures, said Daniel McCaffrey during a discussion of the initiative at the National Academy of Education's annual meeting in October.
"All the measures we use are error prone, to both individual and systematic errors," he said. For example, he noted that principals observing teachers were better at judging how well instructors handled their class's behavior and attention than how well they prepared them academically. "Maybe what people know about instruction is less than what they know about management."
While data from the extension study is still in its pilot phase of secondary research, the institute and the National Academy of Education have announced 10 early-career researchers who will each receive $25,000 from the the William T. Grant and the Spencer foundations to use the MET longitudinal database for the next year. Among the projects:
- Tanner LeBaron Wallace, education psychologist at the University of Pittsburgh, will try to use differences in teaching practices and student response to tease out differences in teacher proficiency among instructors working with high-performing students, who might show little growth because of a "ceiling effect" in the tests used to measure them.
- Leslie C. Dietiker, a mathematics assistant professor at Boston University, will use the videos to identify characteristics of math lessons interesting to students.
- New York University researchers Peter Halpin and Michael J. Kieffer will study English differentiation in middle school classes.
Want more research news? Follow @SarahDSparks on Twitter for the latest studies, and join the conversation.