BERKELEY -- The Los Angeles Times has brought into the public eye a topic once debated mostly by policy analysts and bureaucrats: the evaluation of classroom teachers. And it did so with maximum splash, by publishing the names and effectiveness ratings of 6,000 elementary school teachers based on its analysis of students' test score data.
In the weeks since, the newspaper has been vilified and praised for its decision to publish those names. And on Monday, before a lively audience gathered at UC Berkeley, L.A. Times reporter Jason Felch shared the newspaper's motivations for doing so -- as well as plans for further analysis of the raw data obtained from the Los Angeles school district through public records requests.
"There is a culture, not just in Los Angeles schools but across the country, where differences in instruction are ignored," Felch said.
As a result, Felch said, not only do bad apples remain in the classroom, but few teachers receive the help or feedback they need to improve.
"They're essentially teaching in isolation in their classrooms, hoping they're doing a good job," he said.
The Times used a complex model known as "value-added" to rate teachers based on their students' progress on reading and math test scores during the course of a school year. Because such ratings tend to vary widely with a sample as small as 20 to 30 students, the Times used up to seven years' worth of data. Felch said the average teacher included in the study was rated based on the results of 110 students. Teachers didn't receive a rating unless the Times had scores for at least 60 students who had been assigned to them.
The Los Angeles school district has ignored this valuable data, Felch said, but the L.A. Times will continue to mine it for stories. Among the possibilities? A similar analysis to gauge middle and high school teacher effectiveness and another to figure out which children are on track to fail eighth-grade algebra, based on their elementary school math scores.
"School districts can use this data to help them before they fail," he said.
Some of the other panelists were critical of the series. Sophia Rabe-Hesketh, a UC Berkeley statistician, drew applause when she concluded that such data sets shouldn't be used for high-stakes decisions. She said the current "value-added" models don't take into account other factors that affect student test score gains, such as the school's leadership, environment, materials and curricula.
"My solution to all these problems is do not use teacher value-added for high stakes decisions, and do not use them for naming and shaming," Rabe-Hesketh said.
Anthony Cody, a former Oakland middle schoolteacher who now serves as a mentor in the district, said that while evaluations must improve, the series was an example of a growing hostility to teachers.
On Sunday, police found the body of a fifth-grade Los Angeles Unified teacher, 39-year-old Rigoberto Ruelas, under a bridge. While it's not clear what caused Ruelas to take his life, his teachers union -- which had previously organized a boycott of the L.A. Times -- demanded the database come down. Ruelas had received a slightly below-average overall rating. The L.A. Times published a statement on Sunday evening, extending its condolences to the family.
"He may be the first casualty in America's war on teachers," Cody said.
Susan Rasky, a senior lecturer at UC Berkeley's journalism school (which Felch attended), said that had she been an editor in the L.A. Times newsroom, she would have argued to publish only the names of the top teachers. Still, she urged the audience not to "shoot the messenger."
"I want to do nothing to squelch the investigative ardor of the L.A. Times," Rasky said. The newspaper, she added, has "done a very important thing."