When teachers have used scarce professional learning time to analyze student-assessment data, achievement results have often been mixed. And that’s despite the fact that teacher collaboration itself has a record of improving student learning. As Harvard scholar Heather Hill noted somewhat delicately in 2020, “The research in this area suggests that district and school leaders should rethink their use of state and interim assessments as the focus of teacher collaboration.”
Should school leaders throw out the practice? I can understand why that would be tempting not only from the research findings but from personal experience.
As a data coach for over 30 years, I still too often see teacher teams spend lots of time giving reasons for poor performance that are way outside their ability to have much impact, such as a difficult home life, students’ failure to study, or the unfair demands society puts on schools. There is little attention to students’ assets and their specific misunderstandings or to improving future instruction. No wonder analyzing student-assessment data hasn’t lived up to what some would call its hype but I would call its potential.
Data analyses often fail to have positive impact on student learning because district directors of assessment (I was one) and university faculty members (I am one) have not given educators the words or structure needed to redirect conversations to what the data show that students know, what students do not know, and what the adults on the team are going to do about it. In other words, leaders have too often told teachers to conduct data analyses but very seldom taught them exactly how to do it.
In her seminal 1982 study “Workplace Conditions of School Success,” sociologist Judith Warren Little found that in successful schools, teachers engage in “frequent, continuous, and increasingly concrete and precise talk” about “what teachers do, with what aims, in what situations, with what materials, and with what apparent results.” But “concrete and precise talk” does not occur automatically. Here are a few lessons the teachers in my educational leadership class and I have learned to move teams to a comfort level of being concrete and precise with one another about instruction.
First, a protocol for the conversation must be in place. Using “data dialogue” protocols helps school teams navigate difficult and uncomfortable conversations by providing a predictable series of questions or topics for members to discuss sequentially. Predictability builds teachers’ trust with the school leader who could be part of the conversation and with each other.
Leaders have too often told teachers to conduct data analyses but very seldom taught them exactly how to do it.
There are several good protocols in use. Some of the best are based on the premise that teachers are not responsible for their data but are responsible for their response to their data. My leadership students report that this slight wording change makes a world of difference in teachers’ attitudes and their commitment to their work.
Effective protocols are action-oriented, with predetermined facilitative moves that avoid the “blame game,” triangulate test scores with teachers’ and students’ observations (sometimes called “street data” or “local expertise”), and require teachers to follow up in their classrooms and report results at the next meeting. Classroom follow-up is often the weakest part of the process, my students say.
In the most helpful protocols, teams address classwide patterns of student performance first. The idea is to avoid fixating on the needs of individual (such as “watch list”) students too soon. Once teams head down the path of considering individuals, there is usually no end to the list, collaborative time runs out, and only a few students get attention, while team members whose students aren’t being discussed zone out.
My leadership students report that the best protocols aim for team consensus on the most effective, evidence-based responses—the type, timing, and extent of the interventions. Is the concept that was missed by many students so important that new instruction must stop right away and the concept retaught? Or should teachers integrate additional instruction on the weak content seamlessly over time into the next topic? Do students have a basic understanding of the concept but need more practice? Or is the next step reaching out to students ready for enrichment and to those who need additional time on the weak skills?
To empower dialogue participants to act, leaders must grant teams the autonomy not to be bound by standardized pacing charts that do not provide teachers flexibility for differentiated instruction. If a concept is essential for further learning and teachers have multiple data points indicating a widespread need and have identified a different way to teach it, it is, in my view, educational malpractice not to address it.
Once there is team consensus on the most appropriate follow-up, team members use their combined expertise to help each other plan instructional strategies, in as much detail as possible, so individual teachers don’t have to do this on their own time.
Lack of closure is often the missing ingredient to make meetings impactful, according to management expert Paul Axtell. Therefore, effective protocols include a three- to five-minute closing segment to validate individuals’ contributions, review the takeaways, and agree on what each person will do and by when.
As teams use practices such as these, they move toward concrete and precise dialogue. Members feel empowered, gain instructional insights from each other, and increase their respect for the protocol and for their professional colleagues. And when teachers return to the next dialogue session with evidence of student improvement, their positive feelings about collaboration and the value of data analysis grow even more.
So throw out team analysis of student-assessment data? Not so fast. School leaders have found that the practice works for teachers and students when teachers are provided with the structure and the autonomy to do it well.
