What if school districts could speed up and slow down research at the same time?
The 86,000-student Austin, Texas, district is part of an unusual research partnership intended to solve school-level problems quickly, while also helping school leaders and policymakers step back and take the time to study how day-to-day issues fit in a larger system.
In the first major test of “improvement science” for K-12 education,in their first, most challenging years of instruction. They have used a cycle of deep data collection; tailored intervention development; short, narrowly focused testing; and adjustment. The cycle, developed by the Stanford, Calif.-based Carnegie Foundation for the Advancement of Teaching, is garnering interest as a way for districts and researchers to partner on work with an immediate impact in the classroom.
The Austin, Texas, public schools used a cyclical research process developed by the Carnegie Foundation for the Advancement of Teaching to test new ways to give novice teachers feedback.
SOURCE: Carnegie Foundation for the Advancement of Teaching
“A lot of us hold assumptions about what works, but articulating them so they can be tested is critical,” said Sandra Park, a senior associate for improvement science and the director of the Building Teacher Effectiveness Network initiative at the Carnegie Foundation. She introduced researchers toat popular professional-development sessions at the Society for Research on Educational Effectiveness conference here earlier this month.
“If we had been able to test them first, we would have abandoned some of these [school improvement strategies] a lot earlier, before we tried to implement them across the district or across the country,” Ms. Park said.
To respond to a problem in the field, such as new-teacher burnout, Austin officials used a 90-day research cycle, in which school leaders and researchers planned and implemented a single change, based on available research, then studied the effects and tweaked the change. One intervention may go through several of those cycles in a single school year, working from at first only a single teacher or classroom and then moving to districtwide deployment.
“It took me a while to feel that you could ever learn anything from a single case with two people, … but this [research process] gives us an opportunity to look at systems and processes,” Karen M. Cornetto, a senior research associate for the Austin schools, said in a presentation as part of a Carnegie Foundation forum on the initiative here.
“We have a safe space to innovate,” she said. “We would never have three years to [implement a change at] 19 schools—usually it’s more like, let’s start with all the elementary schools.”
David Kauffman, the principal of Lance Corporal Nicholas S. Perez Elementary School in Austin, said he was skeptical when his school was asked to test a new protocol for providing feedback to new teachers. Mr. Kauffman, previously a founding member of Harvard University’s Project on the Next Generation of Teachers, a longitudinal-research project, said he had already “checked all the boxes” for new teachers: induction, observation and feedback, and peer support.
Yet he and Daniel Girard, the principal of neighboring Akins High School, said they still saw too many exhausted new teachers leave their school or the profession.
Both leaders said plans for more frequent classroom walkthroughs or more helpful teacher-feedback sessions often were akin to New Year’s resolutions—"I will get into classrooms more,” Mr. Kauffman said he pledged again and again, before the day-to-day distractions of the principalship got in the way.
“The work really forced us to think about the flow of communication that we have with new teachers,” Mr. Girard said. “What we used to do was basically, ‘Hey, new teacher—good job,’ and we did provide feedback but it wasn’t targeted; we tried to do everything all at once.
“By focusing on one little thing at a time,” Mr. Girard said, “we were able to get a whole lot of bang for the buck,” he said.
By coupling development of the new observation and feedback protocol with brief surveys of new teachers every six weeks, the school leaders were able to identify specific challenges for their teachers and help the newcomers improve their skills.
The research cycle’s broader data analysis also made Mr. Kauffman take a more objective look at his work in the school context and learn how to delegate activities that had distracted him from working with new staff members.
“What in the system of the school is preventing us from having these important conversations with new teachers? " he said. “When we looked at that, it was much better at making things happen than just self-flagellation.”
While the Carnegie Foundation is winding down the Building Teaching Effectiveness Network, school officials in Austin are trying to keep the process going, and other districts and states have expressed interest in the model.
However, the process may be most helpful for school districts if it never gains a national profile, argues Brad Jupp, a senior program adviser to U.S. Secretary of Education Arne Duncan.
Improvement science seems to Mr. Jupp “incredibly relevant at the schoolhouse level,” but he cautioned against “the horrible thought of trying to turn it into a requirement at the federal level or turn it into a law.”
Michael Cohen, the president of the college-readiness group Achieve, based in Washington, agrees that not every district has the capacity to build in research on a regular basis, but he still approves of the federal Institute of Education Sciences’ efforts to encourage research-practitioner partnerships to explore similar research models.
“I think the time is right for a conversation about improvement science in the policy arena,” Mr. Cohen said. “It is all about data, but it is not accountability on steroids—it’s very local data, collected more often to advance local priorities. That’s a very different approach.”
A version of this article appeared in the September 17, 2014 edition of Education Week as Austin School District Tests New R&D Model With Novice Teachers