Principals Aren't Tapping Teacher-Effectiveness Data, Says Study
Despite a trove of data on teacher effectiveness that has accumulated from the rollout of teacher-evaluation systems in recent years, many principals are not using that information to guide decisions about hiring, assignments, and professional development, according to the findings in a report scheduled for release this week by Vanderbilt University researchers.
When principals do avail themselves of that information, they are more likely to rely on classroom-observation data, rather than on value-added measures of students' test scores or parent, student, and teacher surveys. They viewed the surveys, particularly those of parents, as less "valid, specific, and transparent" when compared with other measures, the researchers found. For example, only 14 percent of principals saw parent surveys as valid to a large degree, while 84 percent viewed teacher-observation data to be valid to that extent, and 56 percent said the same about student achievement or growth.
A host of barriers — including access to the data, the availability of value-added measures when decisions are being made, a lack of understanding of the statistical models used in the evaluation systems, and the absence of training in using the data — help explain why teacher-effectiveness data are not more widely used in human-resources decisions, according to the researchers.
"Much of the emphasis has been on student-assessment data helping to drive instructional decisions," said Ellen B. Goldring, a professor of education policy and leadership at Vanderbilt's Peabody College of Education and Human Development, in Nashville.
"There has been much less attention [paid] to teacher effectiveness data for talent-management decisions. It's really important to widen the discussion of how school leaders can use data for school improvement, and we found a number of consistent barriers to why principals are not using the data or are not able to use the data," she said.
Teacher evaluations have been fraught with controversy — over whether and to what extent they should include value-added measures, and whether those evaluations should form the basis for decisions about firing workers.
The study, "Principal Use of Teacher Effectiveness Measures for Talent Management Decisions," by Ms. Goldring, Christine M. Neumerski, Mollie Rubin, Marisa Cannata, Timothy Drake, Jason A. Grissom and Patrick Schuermann, was funded by the Seattle-based Bill & Melinda Gates Foundation. (The foundation also supports some news coverage in Education Week.)
It was conducted during the 2012-13 school year, and focused on six urban districts and two charter-management organizations that were implementing — or had already implemented — teacher-evaluation systems that included multiple measures. The participating school systems were in Shelby County and Metropolitan Nashville, Tennessee; Baltimore city; Hillsborough County, Fla.; Houston; Denver; and the Green Dot and Alliance College-Ready charter school networks in Los Angeles.
In many cases, particularly in districts with newer teacher-evaluation systems, principals faced challenges in accessing and using the data, which the researchers distilled to five major categories: time, timing, technology, training, and a lack of trust. Eighty-three percent of principals, for example, said timing was a minor to strong barrier in using teacher-effectiveness data. Student-achievement data, teacher value-added scores, and survey results, for example, arrived after decisions about renewals and placements had been made. And 75 percent of the principals listed a lack of time as a barrier to using the data.
Researchers also found that principals, in some cases, were already using teacher-effectiveness data to identify their teachers' strengths and weaknesses and to target support and professional development. In those districts, the central office had acknowledged the importance of data use, communicated to staff the kind of data that should be used for specific decisions, and made it easier for staff members to access the data. Some provided professional support to help principals, Ms. Goldring said.
Douglas N. Harris, a professor of economics at Tulane University and the director of the Education Research Alliance for New Orleans, said the results confirmed feedback he had received from other educators about the challenges in using teacher-evaluation systems.
"The data that they are producing and these analyses are useful for understanding how the principals perceive the data — what they understand, what they don't understand, what they trust, and what they don't trust," said Mr. Harris, who has studied teacher-evaluation systems. "And that's really important if we are going to understand how [the evaluation systems] are being implemented and how we might [make] these systems better."
Randi Weingarten, the president of the American Federation of Teachers, said both principals and teachers want "the tools, time, and trust to do their jobs well and to use data properly."
"Great teaching involves rich engagement between teachers and students, which principals can evaluate only if given the time for observations," Ms. Weingarten said in a statement.
Michael J. Petrilli, the president of the Thomas B. Fordham Institute, a Washington-based research and advocacy group, said while earlier reports noted that principals were not widely using valued-added measures in human-resources decisions, the Vanderbilt study provides insight into the reasons why.
"We have these systems that just are not yet working in terms of getting information into the hands of principals," Mr. Petrilli said. "And until that happens, principals are doing the most logical thing, which is using the information they do have to make these decisions."
With the roadblocks identified, efforts should now focus on making it easier for principals to get the information they need to make the best decisions, Mr. Petrilli said.
Overall, the researchers recommend that districts clarify their expectations for how principals should use data and what data sources should be used for specific human-resources decisions. They recommend training for principals on using value-added estimates, openly encouraging discussions about data use, and clarifying the roles of value-added estimates and observation scores.
Vol. 34, Issue 03, Pages 1,12-13