To improve teachers’ mathematics instruction, enlist a team of coaches. But to find out what actually makes for effective coaching, ask researchers.
That’s the premise of an instructional coaching program in Tennessee. Over the last four years, select coaches have been working with math teachers in grades 3 through 8 to promote high-quality teaching that’s rooted in complex thinking and aligned with state standards. Researchers with the University of Pittsburgh have been evaluating the coaches to better understand how to help them do their jobs better.
“As our state was transitioning to more rigorous standards, we needed a way to support our teachers,” said Laura Booker, the executive director of research for the Tennessee education department. At the same time, “we realized there was a real lack of training on how to be an instructional coach and work with adult learners.”
The coaching program tackles both of those challenges. Funded by a federal Institute of Education Sciences grant, the program uses a continuous improvement process, in which researchers analyze the coaching data periodically and then refine the guidance they provide to coaches.
The ongoing feedback loop helps the researchers determine the best coaching practices and create a model for guiding math coaches, said Jennifer Lin Russell, an associate professor of learning sciences and policy at the University of Pittsburgh, and the lead researcher on the coaching program.
So far, more than 70 coaches across 31 districts in Tennessee have been trained in the model, which emphasizes higher-level thinking in mathematics. Researchers are still analyzing the data collected. Results on student achievement are expected later this summer.
But so far, researchers have found that when coaches and teachers had deep and specific conversations while planning lessons, the teachers were better able to orchestrate high-level and open-ended mathematics tasks in the classroom. They became more skilled at helping students better understand math concepts.
Teachers also became more comfortable allowing students to do most of the thinking about math problems, rather than jumping in and providing assistance as soon as students started to struggle, Russell said.
Booker said she remembers getting “cold chills” when she realized the process was working. "[One teacher said], ‘I feel like my teaching has so dramatically improved,’ ” she said. “It was like, this is why we’re doing this.”
‘Grappling in the Learning’
The coaching model asks coaches to provide evidence-based feedback to teachers; establish mathematics and pedagogical goals before lessons; and engage teachers in deep and specific discussions of the “instructional triangle,” which is made up of content, pedagogy, and student thinking. Coaches work with their partner teachers before, during, and after a lesson a minimum of three times per school year.
This level of intensive coaching requires time and resources. Still, high-level coaching can transform the instruction of a teacher who, for example, is great at language arts but struggles with teaching math, said Jim Knight, a senior partner at the Instructional Coaching Group, a consulting firm that partners with states and districts to train coaches.
“If you keep that great teacher in the system, it’s worth that kind of support,” said Knight, who is not involved in Tennessee’s coaching work.
![BRIC ARCHIVE](https://epe.brightspotcdn.com/dims4/default/9783fdd/2147483647/strip/true/crop/600x920+0+0/resize/218x335!/quality/90/?url=https%3A%2F%2Fepe-brightspot.s3.us-east-1.amazonaws.com%2F8b%2Fb6%2F97c731f6f80253feeb9981f8ac7b%2F35-mathcoaching-chart-600px.jpg)
To evaluate the coaching, the Pittsburgh researchers looked at planning documents and transcripts of videos of the coaching cycle: the pre-lesson conference, the lesson observation, and the post-lesson reflection.
“We would go out into the field, work with the coaches, and then we’d listen to what the coaches are doing and study their videos. That would call into question things that needed to be modified in the coaching project,” said Victoria Bill, a senior fellow at the Institute for Learning at the University of Pittsburgh, which provides the professional development for the coaches.
For example, coaches were asked to set learning goals with the teachers before lessons. But when Bill’s team watched coaches’ videos, the researchers realized that the coaches had a different understanding of math learning goals than they’d intended.
“We showed them some examples, but it wasn’t enough for them to learn it,” Bill said. “They needed to grapple in the learning with us.”
Coaches were trained to push the teachers to think and reason deeply about the mathematical concepts, as well as consider how to make sure students understood the concepts.
“There was all this educating of pedagogy, student learning,” Bill said. “And these are the people who are ... already the expert teachers. But I think they learned what they didn’t know, and we learned about what they need to learn more deeply.”
Jamelie Johns was part of the first cohort of math coaches in the program. She said the researchers taught her how to have conversations with teachers that press for both depth and specificity.
“I was able to get feedback [that said], ‘When you did this, we saw the teacher change her practice in this way.’ That’s a move I want to continue to make,” Johns said. “If we saw something that wasn’t as impactful, [I would] revise.”
The continuous improvement model was a “big departure” from previous research projects, said Pittsburgh’s Russell, who is also a research scientist in the Learning Research and Development Center at the university. Mainly, it was faster.
“We’re used to these long cycles of data collection and analysis, and here we were, ... engaged in much more rapid analysis,” she said. “The way we were engaging with [the coaches] in a very regular way helped us understand what was happening in the data.”
Researchers could see what the challenges of the implementation were, in close to real time, she said.
For example, the researchers wanted coaches to be having deep and specific conversations with teachers about mathematics, pedagogy, and student thinking. As the school year went on, the researchers were able to better refine the training and guidance they gave to coaches.
By the end of the year, the researchers could see noticeable improvements in the quality of coaches’ discussions, Russell said.
Learning to Be a Coach
For Johns, who is now the director of elementary math and science for Hamilton County schools, the biggest challenge in becoming a coach was learning how to work with adult learners, rather than students.
“As adults, we’re more fixed in our ways,” she said. “I think the perception is that teachers are expected to know everything because we’re the ones teaching, but it’s not that at all: Teachers are learners.”
That learning curve is why coaches need coaching, too, said Knight, who is also a research associate at the University of Kansas Center for Teaching and Learning.
Coaches need professional development to learn how to set goals with teachers, gather data, and pass their knowledge on to other adults, he said, adding that it’s tough to create lasting change in teachers’ instruction.
“To dramatically change the way you do your work ... it’s like changing your personality,” Knight said. “It’s going to take some time, it’s going to take some support.”
Ultimately, the three teachers Johns coached through the model had gains in both student achievement and student growth.
Now, there are 12 elementary math coaches in the Hamilton County district, which includes Chattanooga. A few have been trained in the state model, and the rest have learned the process from their peers.
“The selling point we often make [to teachers] is even professional athletes have coaches,” Johns said. “Coaching isn’t for weak people. It’s for strong people who want to get better.”
The $2.5-million federal grant that has funded the research in Tennessee runs out at the end of December, and the state education department is strategizing on how to keep the program going past the life of the grant.
Booker, of the Tennessee education department, said the plan is to divide the work into three regions of the state, with math consultants from the education department convening networks of coaches in each region. This could eventually scale up the program, she said, as the consultants extend the training to new coaches.