Email Status

Volume , Number

Improving Teaching and Learning through Instructional Rounds


Principal Chris Saheed and deans at the Cambridge (MA) Rindge and Latin High School review school data as part of the instructional rounds process.

Principal Randall Lewis stood at the front of the school library, where members of his district’s instructional rounds network had gathered for coffee, muffins, and conversation before the official start of the day’s visit. “Welcome to Jefferson Middle School,” he said. “We’re excited to have you here today to help us with our problem of practice. We’re also a little nervous, but that’s okay. I’ve told the teachers that this is about my learning and the network’s learning, and that we’re going to get lots of good information from having so many eyes and ears in our classrooms.”

Randall described the “problem of practice” on which he and the teachers had asked the visitors to focus: “Last spring, we rolled out a new literacy initiative that required a radical shift in teaching strategies for many of our teachers. A year later, we’re trying to understand what we’ve learned and what we haven’t, and whether it’s translating into different kinds of learning for students.” As participants greeted the other members of their observation team and gathered maps and papers for notes, there was a buzz of anticipation, much like a group of scientists about to embark on fieldwork for data collection.

Randall Lewis and his colleagues are about to spend the day doing something that most educators have never done: look at classroom instruction in a focused, systematic, purposeful, and collective way. Along with other principals, teachers, union leaders, and central office personnel, Randall is learning about improving instructional practice by participating in instructional rounds, an idea adapted from the medical rounds model that doctors use (see "Rx for a Profession"). A small but growing number of educators are using instructional rounds to look closely at what is happening in their schools’ classrooms and to work together systematically to try to provide high-quality teaching and learning for all their children.

These teachers and school, district, and union leaders work in networks with one another and in consultation with our team of faculty and students at the Harvard Graduate School of Education. They represent all types of educators: networks of superintendents in Connecticut and Iowa, principals in Massachusetts, and mixed teams (superintendents, chief academic officers, union leaders, teachers, and principals) in Ohio. They spend much of their time in classrooms, looking at instruction in fine detail. They learn to talk in new ways with each other about what they see, replacing vague or judgmental generalizations (“She did a great job of transitioning from the whole-class lesson to independent work time”) with precise and nonevaluative language (“At the end of the lesson, the teacher asked students what materials they needed to get for their upcoming independent work. She took a few responses and released students to go to their desks four at a time”).

Unlike many educators who call for “increased rigor” or “critical thinking skills,” with only a vague idea of what these terms mean, network members work together to develop detailed lists of what those abstract ideas should look like in real classrooms. They come to agreement on what teachers and students would be saying and doing if critical thinking skills were being demonstrated, or what students would be working on if their tasks were really rigorous. And when they don’t see these signs of critical thinking or rigor, they don’t blame teachers, students, parents, or other external factors. Instead they look within the school and district to suggest new and powerful ways educators can work together to achieve the student-learning outcomes they desire.

Getting Started with Rounds

This focused and purposeful work takes some getting used to. Our team at HGSE frames the rounds work in four steps (see table "A Four-Step Process").

A Four-Step Process
 Problem of Practice  Observation of Practice Observation Debrief Next Level of Work
School identifies a problem of practice that:

* focuses on the instructional core.
* is directly observable.
*is actionable (is within the school/district's control and can be improved in real time).
* connects to a broader strategy of improvement (school, system).
* is high-leverage (if you acted on it, it would make a significant difference for student learning).

Newtork adopts the problem of practice as the focus for the network's learning.
 Observation teams collect data that is:

* descriptive; not evaluative.
*about the instructional core.
*related to the problem of practice.
 Observation teams discuss the data in four steps:

* Describe what you saw.
* Analyze the descriptive evidence (What patterns do you see? How might you group the data?).
* Predict what students are learning. If you were a student in this class/school and you did everything the teacher told you to do, what would you know and be able to do?
 Brainstorm the next level of work:

* Share district-level theory of action.
* Share district context, including resources, professional development, current initiatives.
* Brainstorm suggestions for school level and for district level.
* Tie suggestions to the district's (and school's) theory of action.

Before hosting one of the network’s monthly visits, the host team identifies a problem of practice on which they ask members of the network to focus during classroom observations. The problem of practice is an instructional problem that the host team wants to solve in order to improve student learning. At the Jefferson School, Randall and his staff had spent a year’s worth of professional development trying to weave literacy strategies into their classrooms and were wondering why students didn’t seem to be benefiting from them.

The problem of practice is shared with the visitors at the start of the day and helps frame what is at the heart of any visit—observation of practice. Typically, groups of four or five visitors will observe in five or six classrooms for about 20 minutes each. The host site selects the classrooms to reflect the problem of ­practice. Because Jefferson’s literacy strategies were supposed to be embedded in all classes, the visits covered a wide range of classrooms and grades. In another setting, a focus on mathematics might bring visitors to a narrower swath of classes. The observers are guided by the host school’s problem of practice. They learn to take careful descriptive notes and to pay special attention to students and the tasks they are doing—not just what students are being asked to do, but what they are ­actually doing. At Jefferson, the observers were given a one-page summary of 14 literacy strategies that teachers had been trained to use and were asked to look for evidence and patterns of student use of these strategies.
The third step of the rounds process is the observation debrief, in which participants sift through the evidence they collected together. There are three stages in the debrief process: description, analysis, and prediction (see sidebar "Debriefing in Detail"). Debriefing in Detail

During one debrief session, a team of school visitors participating in instructional rounds discussed the following observations from a sixth-grade classroom:

Teacher referred to textbook and asked, “What were the branches of government in ancient Greece?” “What were the three social class groups in Greece?” “What was the main resource?”

Using Bloom’s Revised Taxonomy, which they had all read, the participants analyzed the tasks posed by the teacher and agreed that they could be categorized as “information recall.” But when they turned to the prediction stage of the discussion (“If you were a student at this school and you did everything you were expected to do, what would you know and be able to do?”), a fundamental—and unexpected—difference of understanding surfaced.

An experienced principal spoke first, “If I were a student in this class, I would have solid reading comprehension skills. I would be able to understand what I read.” To her surprise, she found that her colleagues disagreed. As they’d learned to do in the rounds process, they returned to the classroom evidence to examine it closely.

The group eventually agreed that, based on the interactions they had seen, they could predict that a student in the class they observed would know how to retrieve specific information from a text, to listen for what a question is asking and respond, and to read for factual information. Referring back to the Taxonomy, the network members then discussed the kinds of activities that predict reading comprehension—activities like summarizing, interpreting, inferring, or explaining. Since they had not observed these kinds of activities during their brief classroom visits, they agreed that the evidence they had gathered was not enough to determine whether students would be able to understand what they’d read.

By the end of the conversation, all members agreed that factual recall was not the same as reading comprehension. This was a new insight for the principal who had spoken first.

This anecdote highlights the importance of using detailed descriptive data as the basis for analysis and prediction. If the observations from this team had read, “Teacher asked questions about ancient Greece” and “Teacher asked questions about the book,” participants would not have been able to determine what students would know. It would have been harder to make the links between teaching and learning.

From Instructional Rounds in Education: A Network Approach to Improving Teaching and Learning, by Elizabeth A. City, Richard F. Elmore, Sarah E. Fiarman, and Lee Teitel (Cambridge, MA: Harvard Education Press, 2009). $24.95. To order, call 1-888-437-1437 or visit
The description stage keeps the focus on a factual description of what visitors actually saw—not their reactions, judgments, or inferences. Only after sharing their observations and agreeing on a fine-grained, detailed description of what they saw does the group go on to the analysis stage of the debrief, looking for patterns within and across the classrooms they observed.

Groups then build on these patterns to move to the predictive stage of the debrief, where the goal is to connect teaching and learning. Participants ask themselves, “If you were a student at this school and you did everything you were expected to do, what would you know and be able to do?” By linking the task and the teacher’s instruction directly to student learning, network members tackle the central question, “What causes the learning we want to see?” What specific teaching moves, what kinds of tasks, what forms of student engagement lead to powerful learning for students? This process ultimately helps identify potential areas for improvement and offers clues about how these areas could be improved, including the specific strategies and techniques that teachers could use and what the school or district could do to support them. Taken cumulatively, these debrief practices allow participants to describe the specific behaviors and structures they see that cause, enable, or at times constrain learning.

At Jefferson, the patterns that emerged in the analysis section of the debrief were clear and quite consistent across the dozens of classrooms visited. Visitors saw the teacher use one or more of the literacy strategies, but they saw almost no independent student use of the strategies. This led to the prediction that students in these classes would be able to follow directions in using specific literacy strategies when asked to do so by their teacher.

The final step of the rounds process is identifying the next level of work, when network members think together about what kinds of resources and supports teachers and administrators would need in order to move instruction to the next level. Here again, the more specific and precise the suggestions, the more helpful they are. At Jefferson, the visitors suggested that the school be more explicit with students about the goal of having them use these strategies in their own reading, writing, and thinking. Concrete suggestions included giving students a version of the one-page summary of 14 literacy strategies and having them track their own use of the strategies, combined with teaching students about metacognition and making explicit to students and teachers alike that the goal was that students, not just teachers, use the strategies.

Accelerating Instructional Improvement

Our goal in doing instructional rounds work is to help schools and districts develop effective and powerful teaching and learning on a large scale, not just isolated pockets of good teaching in the midst of mediocrity. Accordingly, the network’s suggestions for the next level of work are not about “fixing” any one teacher or group of teachers. They are about developing clarity, about good instructional practice, and about the leadership and organizational practices needed to support this kind of instruction at scale. Suggestions for the next level of work are intended more for administrators and other leaders than for individual teachers.

People often ask us, “Will doing rounds lead to an increase in student learning? Will it raise test scores?” The short answer is: by itself, no. Although the rounds process is not a silver bullet that will single-handedly lead to better test scores or increased learning for students, it is a powerful accelerant of school and district improvement efforts. Its focus on what goes on in classrooms anchors improvement efforts in the instructional core—the complex relationships among teachers, students, and content. The rounds process provides a key source of data and a powerful feedback loop to tell educators whether their systemic improvement efforts are actually reaching students. And the collaborative learning approach used in rounds networks creates norms that support adult learning and make organizational learning possible.

As one deputy superintendent from Ohio puts it: “The ‘next level of work’ has become a very common phrase now in our district conversation. We are all thinking more deeply about the supports. Are the supports in place to help [teachers and students] make the transition [to the next level]? Rounds is helping give us that firsthand data and getting us to think more deeply about it.”

Lee Teitel is lecturer on education and senior associate of the Executive Leadership Program for Educators at the Harvard Graduate School of Education. He is coauthor, with Elizabeth A. City, Richard F. Elmore, and Sarah E. Fiarman, of Instructional Rounds in Education: A Network Approach to Improving Teaching and Learning (Harvard Education Press, 2009).

Also by this Author