Teacher-coaching linked to a well-known teaching framework paid dividends for student achievement in the secondary grades, according to a study published today in Science magazine.
In all, the study found a 0.22 standard deviation increase in the scores of students taught by teachers who received a special form of teacher-coaching—roughly the equivalent of an increase from the 50th to the 59th percentile—relative to the students taught by teachers in a control group.
"This study shows dramatically, clearly, when you implement a [teacher] measure rigorously and couple it closely to a PD system, you get dramatic improvements in student achievement," said Robert C. Pianta, a professor at the Curry School of Education at the University of Virginia.
For the study, Pianta and four colleagues, all psychology researchers, randomly assigned 78 middle or high school teachers to participate in the special coaching intervention, or to their regular in-service training. There were no significant differences in the characteristics of the treatment or control group or in the population of students they taught. The study covered some 2,200 students in all.
Professional development was keyed to the Classroom Assessment Scoring System-Secondary, or CLASS-S. It is essentially a modified version of Pianta's well-known CLASS framework.
The key aspect of CLASS is that it focuses heavily on specific observable interactions between teachers and students, such as behavior management, productivity, and conceptual development. Originally designed for P-3 teachers, the researchers modified and tested CLASS for teachers of secondary students. In particular, they tailored it to respond better to research on adolescent learning needs, which include opportunities for them to make decisions about what they're learning and chances to work with their peers, according to Joseph P. Allen, a psychology scholar at UVA and one of the study's authors.
The training was delivered via a Web-based approach called My Teaching Partner, again devised by Pianta and team for use with CLASS. Under the system, each teacher taped his or her instruction and then uploaded it to an online portal. The tapes were then viewed by "coaches" trained on the CLASS-S domains. Then, the coaches would discuss particular interactions with teachers in phone conversations, including how they aligned to the CLASS-S framework and ideas about how to improve those interactions.
The study found that, while there was no effect on student scores in the first year of the intervention, students taught by the teachers receiving the CLASS-S support outperformed those who received regular in-service training in the following year. Further, the study found that some of the improvement could be directly linked to changes in teachers' behavior caused by the extra support.
There are a few reasons to pay attention to this study. For one, effective professional development, in general, remains a very tough education nut to crack. A random-assignment study such as this is important because it demonstrates not only that the PD is linked to student achievement, but also that it caused some degree of that achievement. It is especially noteworthy at the secondary level where research on effective professional development is quite sparse.
As I reported last year, rigorous studies of PD approaches are generally few and far between. Professional development is challenging to study. There are all kinds of potentially confounding factors, like differences in funding and implementation. And professional development is inherently a complex endeavor. Any teacher training affects students only indirectly, after it is filtered through a teacher's own practices.
It's important to note that the My Teaching Partner approach is a very specific way of analyzing and discussing teaching practices. The lesson here is that educators must devise an effective way to help teachers embody new practices and behaviors; merely selecting a set of teaching standards is not enough on its own.
"It is a model for coaching that is different from a lot of other models in that it's very prescribed, very focused on a way of constructing the coaching session, and what the coach does to identify behaviors they work on, and how the coach gives feedback," Pianta said.
Second, the study also tells us a bit more about teaching frameworks, which are being used as the basis of new teacher-evaluation systems. (CLASS is one of several models now being used.)
Scholars have been exploring whether teachers ought to be coached and/or assessed using a general teacher-behavior rubric, like CLASS-S, or one that's customized for each teacher's specific discipline or content area. This study found that the improved teacher-student interactions predicted student achievement regardless of the content area in question.
Pianta told me teachers in the study taught in four different content areas and gains were seen in all of them. Again, this is an interesting finding, especially in light of debates about whether secondary professional development should focus on additional content acquisition or on ways of better teaching content.
A variety of other scholars are also exploring this issue. At the University of Michigan, researchers are looking at math-specific teaching frameworks, while others at Stanford University are reviewing English ones. Much of the study is part of the Bill & Melinda Gates Foundation's Measures of Effective Teaching Study. (The nonprofit that publishes Education Week is a former recipient of Gates Foundation funding.)
Finally, the UVA study also raises some interesting conceptual questions about the very nature of professional development that are worth outlining. The CLASS-S approach here was purely a professional-development tool and not linked to any evaluative purpose.
Some of the most recent research on teaching frameworks, however, have been in districts such as Cincinnati, where the framework doubles as part of a formal teacher-evaluation system.
I bring this up merely to point out that the line between professional development and evaluation is not one that's been well defined or illuminated in current discussions about teaching. But it's poised to emerge as another tension point for the field, especially as more time and energy are spent on teacher evaluation.
Take the case of the District of Columbia, for instance. There, the teachers' union insisted on a formal separation between "master educators" who do some of the conversations in the teacher-evaluation system, and the district's professional-development coaches.
In general, Pianta said he thinks that much more attention needs to be paid to studying teacher-evaluation frameworks and ensuring a good link between what teachers are evaluated on and what supports they receive.
"We should be treating performance assessment with the same rigor that we treat assessment of student achievement," he said. "The risk here is that there's too much looseness on these assessments of teacher performance."