Introduction

Recent reviews of research on professional learning suggest that structured opportunities for teachers to meet and collaborate around content-focused practice and artifacts of teaching support instructional improvement (Hill & Papay, 2022; Weddle, 2022). Moreover, there has been increasing attention to the importance of teacher agency and the need for professional development that is responsive to teachers’ needs and contexts (Borko et al., 2015; Imants & Van der Wal, 2020). There has been little research, however, into what is involved in facilitating collaborative discussion that both supports teacher agency and leads to productive learning or how school-based leaders can learn effective facilitation practices (Even, 2008; Krainer et al., 2021). This paper adds to the literature by exploring the development of responsive facilitation moves in teacher learning communities focused on collaborative analysis of artifacts of K-8 mathematics instruction. Using artifacts that are selected by teachers themselves allows them to generate and explore problems of practice that are specific to the needs of their students and unique classroom and school contexts, but also creates challenges for facilitation, as the direction that discussion takes cannot always be predetermined. As Richards (2022) argues, while much has been written about what it means for classroom teachers to be responsive to student thinking, there are very few studies that focus on professional learning facilitation that is responsive to teachers’ questions about their own practice.

In this study, as part of a larger research–practice partnership around mathematics instructional improvement, teacher leaders were learning to facilitate Collaborative Lesson Design (CLD) inquiry cycles, where grade-level groups of teachers and teacher leaders from different schools: (a) met online to plan a lesson around a rich mathematical task, (b) enacted the lesson in their classrooms, and then (c) met again online to debrief, selecting and bringing artifacts of their instruction for discussion and reflection. Each inquiry cycle focused on implementing one of the following responsive teaching practices that participants had explored, analyzed, and practiced in prior professional development sessions: launching a challenging task, facilitating students’ productive struggle, or orchestrating discussion of students’ strategies (Responsive Math Teaching Project, 2021b). The inquiry cycles therefore reflected a combination of both structure and agency. The CLD cycles were designed to provide the opportunity for teachers to learn “in and from practice” supported by “professional discourse and engagement in communities of practice” (Ball & Cohen, 1999, p. 12). Furthermore, by engaging teachers in a modified form of lesson study, the inquiry cycles provided structured opportunities to adjust and tailor their developing instructional practices in response to their own students, thereby supporting transfer of the learning of new practices into the classroom context in ways that are responsive to local needs (Lewis, 2016). Rather than observing each other’s lessons, however, in the current study, both the teacher participants and the teacher leader (novice facilitator) brought video clips of their enactment as an artifact of the instructional practice that was the focus of the inquiry cycle. A university-based mentor provided tailored support depending on the needs of the novice facilitator, sometimes modeling facilitation moves, particularly in the areas of building community and supporting teacher participants as learners, and gradually stepping back as the novice facilitators were ready to take more of these moves.

This study focused on five CLD groups, each of which was composed of a mentor, 1–2 teacher leadersFootnote 1 who were learning to facilitate the sessions, and 4–6 K-8 teachers from participating schools, as they met online to engage in six cycles over the course of one school year. The novice facilitators gradually took on more facilitation responsibility as the year progressed, providing a rich opportunity to study the developing practices of both mentors and novice facilitators over time. Drawing on a situated perspective on professional learning, we explored the following questions: What facilitation moves do mentors and novice facilitators use to orchestrate collaborative discussion of teachers’ artifacts of practice in relation to professional learning goals? How do novice facilitators’ practices and teacher participation change or develop over time?

In the sections that follow, we first situate our study in the existing literature on the facilitation of artifact-based professional development and collaborative inquiry into practice. We then present the conceptual framework that guided our approach to understanding teacher and facilitator learning.

Facilitation of artifact-based professional learning

As described above, in this study, professional learning was situated in a context where inquiry was oriented around video records of practice and pedagogical goals, teachers had agency over the questions and problems of practice that were explored, and the facilitator was positioned as a co-learner by implementing and reflecting on the collaboratively designed lessons along with the participants. We, therefore, looked to the existing literature on the facilitation of video records of practice for professional learning as well as the facilitation of lesson study and teacher inquiry into practice.

In 1999, Ball and Cohen called for professional learning grounded in the use of artifacts of classroom practice to generate opportunities for educators to engage in ongoing, collaborative inquiry into student learning and instructional practice. Since that time, much of the research on the use of video records has focused on how to help teachers notice students’ mathematical thinking (for reviews, see Baecher et al., 2018 and Gaudin & Chaliès, 2015). More recently, there has been some attention to the role of the facilitator in video-based professional development where the discussion is oriented around noticing students’ mathematical thinking. Several studies of the facilitation of “video clubs” focus on how teachers can be supported to move from description or evaluation of video episodes toward interpretation and analysis of student thinking. Coles (2013) highlights the importance of setting up norms for discussion and argues that metacommenting, or drawing together participants’ descriptions to “identify and label a purpose,” is a key facilitation practice for supporting the move to interpretation. van Es et al. (2014) present a framework for facilitation moves that includes orienting the group to the video analysis task, sustaining an inquiry stance, maintaining a focus on the video and the mathematics, and supporting group collaboration. Gonzalez et al. (2016) apply this framework to the facilitation of discussions of teaching animations as well as video, with only minor modifications. In these studies, the facilitator’s role is focused on guiding and steering teachers’ attention to notice student thinking. As van Es (2009) notes, “If teachers are to use student thinking to inform pedagogical decisions, they need to ‘learn to notice’ student mathematical thinking” (p.101).

Borko et al. (2015) studied the facilitation of professional development focused on Problem Solving Cycles, where participants first solve a mathematical task and develop lesson plans, and then teach and videorecord the lesson in their own classrooms. Facilitators then review the videos and select clips for participants to discuss in subsequent workshops, again with the goal of developing teacher knowledge through the analysis of student thinking. They show that strong facilitators demonstrated two contrasting styles, both of which led to productive, high-level discussions of the videos: They either used questioning and probing to elicit connections from the participants, or they more directly proposed those connections themselves and then engaged the participants in conversation around those ideas.

Richards (2022) uses the term “focused responsiveness” to characterize the facilitation of professional learning that centers teachers’ “assets, aims, and needs while also bringing them into conversation with research-based instructional approaches that support student learning (p. 985).” This approach to facilitation is fundamentally asset-based, starting with the premise that teachers bring valuable knowledge and experience and allowing for teacher agency while also being focused on professional learning goals. We hypothesized that in our context, the facilitator’s role may be less about focusing teachers’ attention, as it is in video clubs, and more about being responsive to teachers’ problems of practice while supporting inquiry into practice and group collaboration.

In a study of teacher study groups, Jenlink and Kinnucan-Welsch (2001) highlight the tension that emerged when facilitators, who positioned themselves as co-learners with participants, needed to assume a role of “directive collegiality” or “responsibility for guiding, and sometimes directing, behavior” of the group (p. 720). Similarly, Lewis (2016) notes that novice facilitators of lesson study groups needed to learn to manage tensions in supporting teachers’ autonomy while at times being more directive. Gibbons et al. (2021) propose some additional dimensions that are important in the facilitation of professional learning where “teachers are asked to make their practice public for collective inquiry.” Their framework involves three components: (a) moves that support learning through collaboration, including publicly valuing participants’ expertise and experiences and nurturing a shared purpose, (b) promoting experimentation and analysis of instructional practice, and (c) shaping the emotional space.

In sum, existing studies of facilitation around video artifacts and inquiry into teaching practice highlight that there are multiple dimensions to the work of the professional development facilitator, including managing logistics, focusing teacher attention, establishing trust and a supportive environment, and modeling an inquiry stance.

A situated perspective on learning to engage in productive discourse around instructional practice

Our research draws on a situative perspective to define teacher learning as changes in their participation as they moved from the periphery to more active and central roles in the learning community and made sense of the shared social practices (Borko, 2004). From this perspective, we view the mentors, teacher leaders, and participants as situated within communities of practice (Lave & Wenger, 1991), where members were mutually engaged in an activity (reflecting on instructional practice in enacted lessons) and focused on a joint enterprise (improving math instruction) with a shared repertoire (a shared instructional model, planning template, session structure, and common vocabulary) (Wenger, 1998). Viewing learning as situated within the setting within which it takes place means that it cannot be separated from the context (Brown et al., 1989). Moreover, a situative perspective views learning as a process of development and transformation rather than one of acquisition of knowledge or transfer from the professional development context to the classroom context (Rogoff et al., 1995). For both novice facilitators and participants in CLD groups, this development might involve an increase or change in their contributions to the discussions, offering their own ideas or adaptations to the joint enterprise of collaborative inquiry into practice, or shaping the focus or direction of the group’s inquiry into practice. We therefore attended to these kind of changes in participation for mentors, teacher leaders, and teachers as evidence of their learning over time.

In the CLD groups, after collaboratively planning a lesson around a challenging mathematical task, novice facilitators and participants recorded and selected their own video clips of their enactment of the lesson to bring to the debrief session for discussion. Sherin and Dyer (2017) and Richards et al. (2021) argue that the activities of video capture and selection are themselves an important part of the professional learning process. Participants also posed their own questions about their practice and goals for instructional improvement, further shaping the learning opportunities that occurred in debrief sessions. The planning and debriefing sessions were structured around a particular responsive teaching practice, and though analysis of student thinking was a component of the reflection, the overall focus was on instructional decision-making. In addition, the participants, having all taught the same lesson in their classroom, were positioned as having expertise, and knowledge was co-constructed by the group. A particular challenge of facilitation in this context is the fact that it is contingent on what participants bring to the conversation. Although teachers were asked to submit videos from their lesson ahead of time, this did not always happen, and facilitators most often had little time to prepare.

To conceptualize productive discussions around instructional practice in this context, we turned to the existing work on teacher learning through collaborative discourse. In a review of research on teacher learning communities, Lefstein et al. (2020) propose that pedagogically productive talk is collaborative discourse that (a) is focused on problems of practice, (b) involves pedagogical reasoning, (c) is anchored in rich representations of practice, (d) is multi-voiced, (e) includes generative orientations toward students, learning, content, teaching, and problems of practice, and (f) reflects a combination of support and critique (pp. 363–4). In their analysis of dialogue in teacher workgroups, Horn et al. (2017) argue that instructional learning opportunities are generated when participants have conceptual resources to interpret their teaching experiences and are mobilized to reimagine their instructional practice through future action. In their analysis of over 100 teacher workgroup meetings, they found that “the richest learning opportunities occurred when groups developed concepts in conversation—discussing the why’s of instructional strategies, not just what or how” (Horn & Garner, 2022, pp. 103–4). Similarly, Borko et al. (2015) describe high-level discussions of Problem Solving Cycles as those where teachers “connected events in the video with general principles of teaching and learning, or they proposed alternative pedagogical solutions” (p. 107). We draw on these frameworks to conceptualize productive discourse around instructional practice in the CLD inquiry cycles as instances where facilitators or participants draw on principles of teaching and learning to make sense of artifacts of teaching practice and/or propose alternative pedagogical decisions or instructional moves in relation to the learning goals in the design of the professional development.

Research design

Study context

This research was part of the Responsive Math Teaching (RMT) Project, a research–practice partnership with a network of 14 under-funded public K-8 schools serving predominantly Black and low-income communities in a large urban city in the Northeastern US. The study took place during the 2020–2021 school year when the district returned to hybrid learning after COVID shutdowns but shifted to virtual learning on a school-by-school basis due to continuing high rates of infections. In all but one school, 100% of students were classified as low-income, and educators faced ongoing challenges around human and capital resources, along with COVID-19 disruptions and the effects of systemic racism in urban communities that had experienced decades of disinvestment.

In this context, attention to improving mathematics instruction through professional development was not a sustained or coherent focus of school and district improvement efforts. At the beginning of the partnership, the district had adopted a new vision for mathematics with commitments to “equitable discourse, rich and meaningful tasks, purpose-driven work, questioning and curiosity, and valuing diverse thinking” along with new curricular materials, but most of the focus of district-led efforts was on accountability, through quarterly interim assessments and checklists for administrators to use for classroom observations. There were few opportunities for teachers to learn new instructional practices, and most teachers were using a direct instruction or didactic model rather than an approach that valued and was responsive to student thinking.

The partnership focused on developing the instructional expertise of school-based teacher leaders in the network through five cohorts that went through a three-phase multi-year professional development program designed to help them experience, understand, and implement a dialogic model for mathematics instruction that was responsive to student thinking (RMT Project, 2021b). This model included launching an open-ended challenging task to elicit students’ prior knowledge and develop collective understanding, facilitating students’ productive struggle without showing or leading them to a predetermined strategy or solution path, and facilitating a discussion of student strategies to build on their thinking and connect to important mathematical ideas and concepts (Ebby et al., in press; RMT Project, 2021a). For each cohort, the first phase of professional development focused on experiencing responsive mathematics teaching as learners through monthly problem-solving communities. The second phase focused on classroom implementation through six practice-based professional development (PD) sessions where participants analyzed and rehearsed component parts of the responsive instructional practices in concert with Collaborative Lesson Design (CLD) cycles designed to support implementation and inquiry into practice. Some teachers continued to participate in these CLD groups for multiple years, while others went on to a third phase, where they learned to facilitate practice-based PD or CLD cycles.

During the year of this study, the teacher leaders were in the third phase, where they were learning to facilitate CLD cycles over the year with a university-based mentor and a group of 4–6 teachers from different schools, some of whom had participated in CLD during the previous year, and some who were participating for the first time while also participating in the practice-based PD sessions. In the practice-based PD sessions, the teachers came together to learn about, unpack, and participate in rehearsals and simulations around specific responsive and inclusive instructional practices related to launching a task, facilitating productive struggle, or discussing learning thinking, after which they met with their CLD groups to incorporate those practices into lesson plans. The CLD participants selected and solved the task beforehand, shared and analyzed potential strategies, and then worked together to plan the lesson to maintain the cognitive demand while supporting all students. For example, they considered how to launch the task (help students make sense of the task, elicit prior knowledge, and prepare to solve), how to facilitate productive struggle (how to support students who cannot get started or get stuck), how to prepare for the whole-class discussion (what to look for in student work and how to help students make sense of the work that is shared), how to connect the whole-class discussion of student strategies toward the mathematical goal, and how to collect evidence of student learning to inform decisions about possible next instructional steps (RMT, 2021b). After teaching the collaboratively designed lessons in their classrooms, participants came together again to debrief and reflect on the enactment, each teacher choosing a short clip from their lesson that highlighted an instructional moment for discussion. The teacher leaders—who also taught the lesson, either in their own or in another teacher’s classroom—met to plan with their university-based mentor before each CLD session and took increasing responsibility for facilitation over time.

In a prior study, Valerio (2023) traced CLD participants’ integration of responsive teaching practices into their classroom instruction while they engaged in the inquiry cycles and found that teachers took up an array of responsive moves related to launching, facilitating productive struggle, and discussing student thinking. This study also highlighted structural features of the practice-based PD and CLD sessions that leveraged teacher agency and identified supports that helped teachers adjust their instruction to become more inclusive and responsive to student thinking.

Participants and data collection

The analysis reported in this paper focused on six teacher leaders (see Table 1) and 20 teachers who participated in five grade-level CLD groups as they engaged in six cycles over the school year. The teacher leaders had between two and four prior years of experience in the RMT project and worked closely with a university-based project mentor to plan and learn to facilitate a grade-level CLD group. Meredith was the only teacher leader who was released from a classroom to support other teachers in her school full time; the others were all grade-level teachers who had been selected by their principals as having the potential to be mathematics instructional leaders in their schools. Notably, Alicia, Wanda, and Hannah all went on to become full-time released teacher leaders in their schools in subsequent years.

Table 1 Teacher leader participants
Full size table

Each CLD cycle included a preparation session between the mentor and the teacher leader, a planning session for participants, where they collaboratively planned the lesson, and then a debrief session after the participants had taught the lesson in their school contexts. We focused our analysis on the 30 debrief sessions around artifacts of practice because we felt that this is where we were most likely to see more of the core characteristics that Lefstein et al. (2020) identify in pedagogically productive talk (focused on problems of practice, pedagogical reasoning, and rich representations of practice) and where facilitators had to be responsive to what participants brought to the discussion. Data sources for this study included over 180 Zoom recordings and transcripts of the online CLD sessions, Zoom recordings and transcripts of the planning and debriefing sessions between the teacher leaders and their mentors, mid-year surveys and end-of-year interviews with the teacher leaders, written journal reflections, and bi-weekly written reflections from mentors on their ongoing work with teacher leaders.

The second and fourth authors served as mentors for the teacher leaders as well as members of the research team, while the first and third authors did not participate in the CLDs. This provided both insider and outsider perspectives on the CLD groups. As a team of White heterosexual women situated at the university, we recognize that our own identities, beliefs, values, and commitments shape the lens through which we approached and engaged with this research. As members of a dominant racial group, we were cognizant of the importance of centering the voices and experiences of our diverse group of participants in our analysis. As we explore in more detail in the next section, this meant continuously examining our own biases and assumptions and making a commitment to recognize and value different leadership styles, facilitation moves, and perspectives even when they differed from our expectations around what constituted effective PD facilitation.

Data analysis

The first stage of analysis involved developing a coding scheme for facilitation moves in the debrief sessions made by both the mentors and teacher leaders. We began with a subset of transcripts, separating the transcripts into turns taken by different speakers. Through an inductive and deductive process, we developed a set of emergent codes for the moves made by facilitators, refined them through constant comparison and grounded analysis (Charmaz, 2008; Glaser & Strauss, 1967), and eventually grouped the codes into larger categories, based on the function of the move in supporting productive discussion around instructional practice. We used this scheme to code each turn made by mentors or teacher leaders, at times applying multiple codes when the turn reflected more than one facilitation move. During this process, two researchers reviewed and coded each transcript and then met with a third researcher to discuss and refine the codes as needed. The two researchers continued to code the remainder of the debrief transcripts and met regularly to resolve disagreements. The process was iterative, involving edits to the codes and categories to incorporate new learning and then going back to apply those changes to previously coded transcripts.

In the second stage, we used the resulting framework, shown in Fig. 1, to code all facilitation moves made by the teacher leaders and mentors in debriefs that took place at the beginning, middle, and end of each year (Cycle 1, 3, and 6). All transcripts were coded in Dedoose and were either double-coded and reconciled or coded by two researchers together. Once coding was complete, we pulled excerpts from the transcript that corresponded to each code and subcode and constructed analytic memos to capture the nature and purpose of each type of facilitation move.

Fig. 1
figure 1

Coding framework for facilitation of discussion of artifacts of teaching

Full size image

We then downloaded the codes into a spreadsheet and looked for changes in the types and frequency of moves made by each facilitator (teacher leaders and mentors) over time. We used our framework to guide the review of the videos of the debriefs in all six cycles, including those that had not been coded, and wrote analytic memos to describe the overall focus of each session, the nature of the teacher leader’s and mentor’s facilitation, and relationships between moves. We also noted instances in each cycle where participants were engaged in productive talk around instructional practice, characterized by reference to (a) connections to general principles of teaching and learning, (b) explicit purposes behind an instructional move or decision (pedagogical reasoning), or (c) alternatives for future instruction (Borko et al., 2015; Horn & Kane, 2015; Lefstein et al., 2020).

Next we triangulated our emerging findings by reviewing the additional data sources for each CLD, including video recordings of the planning and debriefing sessions between the teacher leaders and their mentors, the teacher leader interviews and journal reflections, and the mentor reflections. We categorized each teacher leader on a continuum of focused responsiveness, based in large part on the extent to which we saw them using eliciting and connecting moves to press on and shape teacher contributions toward the goals of the CLD inquiry cycle. Through further reflection, however, we came to recognize ways that our analysis may have been influenced by our own biases about what good facilitation looked and sounded like. More careful examination of our coding and the transcripts showed that many teacher leaders were positioning themselves as fellow learners rather than experts, frequently using cultivating moves to affirm participant contributions and draw connections to their own teaching and learning, and that these moves also generated productive discourse around instructional practice in the group discussions. By abandoning the idea of a continuum and looking more closely at what was happening in the debriefs for the participants, we broadened our lens to learn about different styles of facilitation and the functions of the different types of moves. It also prompted us to focus on what was happening for participants during the debriefs as a way to characterize effective facilitation, rather than only the presence or absence of particular types of moves.

As a result, in the next stage, we explored the interaction between facilitation moves and changes in teacher participation and productive discourse around instructional practice by applying our framework to code participant turns during the discussion of the artifacts. We found that participants were increasingly building off each other’s contributions and that participants as well as facilitators were utilizing cultivating moves that served to build community and connecting moves that generated productive discourse around instruction in relation to the learning goals. The other two categories of moves, structuring and eliciting, were used almost exclusively by facilitators.

In the final stage, we conducted a member check by sharing our framework of facilitation moves with the teacher leaders, who were continuing to lead CLDs in a new school year with less support from their mentors but still coming together with the mentors after every cycle to debrief. Asking them to talk about whether the framework resonated with their experiences as facilitators helped to reinforce the credibility of our findings as well as provide additional insights. We also asked them to use the framework to choose video artifacts of their developing leadership practice to share in group debriefs, which generated additional examples of these practices across the different CLD groups and contexts. Their discussions around these artifacts both validated the framework and helped us to better understand the growth in their facilitation practices as well as the questions and tensions that arose as they continued to develop as leaders.

A framework for focused responsiveness in discussion of artifacts of teaching

Before turning to the findings from the analysis of the coding of facilitation and participation, we first explain each category of discussion moves in the framework in more detail to illustrate how each type of move functioned to support productive discourse around instructional practice, which we defined as discussions that were connected to general principles, pedagogical reasoning, or future instruction in relation to the responsive teaching practices that were the focus of the CLD cycle.

Structuring

The first category, structuring, was almost exclusively used by teacher leaders and mentors who were responsible for facilitating the CLD sessions. Structuring ensures that the group members understand and stick to the goals of the session and move through the activities on the agenda, guided by the professional learning goals of the session. As K-1 teacher leader Alicia described it:

You have to remember, “This is what I want them to work on today” and keep going back to that. It really is just focusing and really understanding what you are looking for from the teachers (Interview, 07/06/2021).

Presenting a task lays out to the group members what they will be doing and, ideally, how it relates to the goals for the session. Transitioning occurs when the facilitator moves from one artifact and/or sharer to the next. This move involves balancing the timing of the session with the content of the discussion, making sure everyone gets the opportunity to share, and ensuring that important conversations have space to develop. Redirecting the discussion allows the facilitator to shift the conversation when it strays from the goals of the inquiry cycle (i.e., responsive instructional practices), sometimes involving gently pausing a participant in the middle of a turn. Together, these moves function to keep the focus of discussion on the professional learning goals of the CLD inquiry cycle.

Eliciting

Eliciting moves allow the facilitator to create space for participants to contribute to the discussion, share ideas, and make connections. These moves can also be used to make sure ideas on the table are clear and understood by the group. Eliciting moves were also exclusively made by facilitators. Participants sometimes posed genuine questions, but not as a mechanism to elicit new ideas or generate participation. Drawing out participant ideas invites group members to say more or think more deeply about a topic of discussion. These moves, typically questions posed by the facilitator, could be as simple as asking the group members what they notice about an artifact or asking them more specific questions about topics brought up in the discussion around an artifact. Asking a clarifying question helps the group make sense of what is happening in a lesson artifact. Frequently, this type of question was asked to the person sharing to gather more information about what was happening before or directly after their video to better contextualize what they were watching. Pressing for pedagogical reasoning occurs when the facilitator asks for the rationale behind a particular instructional decision, such as asking why a teacher may have used a specific talk move in the lesson. Facilitators were able to use eliciting moves to press on participants’ thinking, deepen their understanding of responsive teaching practices, and lead the participants to make their own connections to principles of teaching and learning. In this way, they made space for participants to engage in productive discourse around their own and each other’s instructional practice.

Cultivating

Cultivating moves functioned to build the community of the group, establishing feelings of trust and connection between teachers who were sharing their practice and learning together. This category includes two types of moves that were made only by facilitators. When establishing or reinforcing norms, facilitators encouraged teachers to be vulnerable with each other by reminding them of the learning potential of the activities in which they were engaging. Often this facilitation move involved supporting teachers to work through feelings of discomfort around videorecording, watching, and sharing their practice publicly. Encouraging group expertise moves invited other group members to contribute to the conversation. Sometimes facilitators directed these moves at specific group members, for example, asking a teacher to share something specific about their own lesson. In other cases, facilitators opened conversations up to the whole group, encouraging members to contribute their ideas to enrich the discussion.

Although cultivating moves were initially more frequently used by facilitators, most participants quickly picked up this way of interacting with their peers. Part of the artifact sharing involved receiving feedback from the rest of the group. Sometimes this took the form of participants noticing instructional moves, and other times this entailed answering a question posed by the artifact sharer. During this time, leaders and teachers alike often used affirming a participant moves to compliment something about the sharer’s instruction and affirming the process moves to articulate and share the benefits of engaging in the cycle of planning, enacting, and sharing lesson artifacts. Reference to own or other’s teaching practice was also used during artifact discussions to make connections between participants’ artifacts or to draw on a previously tried instructional practice to give advice or support. Participants and facilitators used references to their own learning experience to bring something they had learned or were in the process of learning to the group. They sometimes also made references to their own learning around the mathematical problem that was the focus of the lesson for that cycle. Cultivating moves were frequently used by both novice facilitators and participants, growing over the course of the year as trust built within the community. As one teacher leader, Meredith, reflected, this trust allowed teachers to be vulnerable and share their practice, supporting deeper learning about areas of need and strengths.

It's very humbling to [video] tape yourself, and the fact that they were willing to do that and then look at it and make changes, I think that's huge as a teacher. I mean most teachers don't videotape themselves teaching and that's such a valuable practice, so you can really see where changes need to be made. And also, be proud of yourself, like see what you're doing and also have other people look at it. If you're not able to see the positives, others can point them out for you (Interview, 06/07/2021).

Cultivating moves served as an important way for the group to collaborate and function as a community, to support each other in learning from their own and each other’s practice, and to engage in what Gibbons et al. (2021), drawing on the literature on school-reform, describe as reciprocity, where “each of the parties involved demonstrates equality and humility in the face of the tasks’ complexity and the limits of their own understanding” (p. 6). These moves also functioned to position all members as having expertise to contribute to the collective understanding of responsive teaching practices and kept the talk grounded in classroom instructional practice.

Connecting

Connecting moves were critical for supporting pedagogically productive discussions around instruction. These moves help teachers to generalize instructional decisions, understand the reasoning behind them, and refine their instruction in the future. When connecting to general principles of teaching and learning, group members used examples of instruction that they saw in the artifacts to make broader statements and conjectures about pedagogy or important components of instructional practice. Connecting to student understanding occurred when group members analyzed student thinking represented in the student work or videos that were shared to reflect on teaching practice. Teachers and leaders connected to pedagogical reasoning when they interrogated the purpose behind an instructional decision, often articulating the potential effects of a specific teacher move. Connecting to future instruction allowed participants to consider bringing something new into their classroom instruction and invited them to think about making changes. Typically, this move occurred in one of two ways: Either facilitators or participants presented an alternative action to one that was presented, or they discussed some potential next instructional steps building on the current lesson. What ties together the connecting moves is the way that they extend teachers’ ideas and initial understandings about applying responsive teaching practices in their classrooms to more general principles of teaching and learning, a central characteristic of productive discourse around instructional practice (Borko et al., 2015). Importantly, connecting moves were not just made by facilitators but also, as we show in the next section, increasingly by participants themselves.

The development of responsive facilitation moves over time

An important feature of a community of practice is that individuals gradually take on the social activities of the more experienced members of the group as they move from peripheral to full participation (Lave & Wenger, 1991). In the CLD groups, productive discourse around instruction, or making connections between instructional practices, student learning, and pedagogical principles, was increasingly a shared and collaborative endeavor.

Table 2 illustrates the results of our coding of the mentors’, teacher leaders’, and participants’ use of these categories of responsive facilitation moves over time. As they took over more facilitation responsibility, all teacher leaders increased their use of structuring and cultivating moves; some also increased in either connecting or eliciting moves. Moreover, these increases were often concurrent with decreases of those moves by the mentors as shown by the shaded cells in Table 2. This suggests that the mentor was modeling facilitation moves at the beginning of the year that were then taken up by teacher leaders in subsequent sessions, as they apprenticed into the role. We expected to see this pattern, given that mentors were gradually releasing responsibility for facilitation and both encouraging and supporting teacher leaders during planning sessions. The patterns in this table also show that teacher leaders were learning different types of practices over the year. For example, although Wanda and Whitney grew in the use of cultivating moves over time, Hannah and Sasha started the year with more frequent use of those moves.

Table 2 Frequency of facilitation moves in CLD cycles over time
Full size table

Figure 2 illustrates this pattern when the moves are combined across all five CLDs. The points on each graph illustrate the relative proportion of each type of move made by mentors, leaders, and participants. In general, there was an increase in cultivating moves made by both teacher leaders and participants as those moves decreased by mentors. We also noticed a more dramatic increase in connecting moves made by participants in four of the five groups, accompanied by a decrease in those moves by both teacher leaders and mentors. In contrast, in the K-1 group, participants were already making connecting moves in the first cycle, but the teacher leaders increasingly made them over time.

Fig. 2
figure 2

Changes in the proportion of different types of facilitation moves by mentors, teacher leaders, and participants across groups

Full size image

Co-construction of facilitation in the grade 3 CLD group

To understand these patterns and the dynamics of the relationships between these moves, we looked more closely at the videos across all six cycles to see how the different types of moves made by both facilitator and teacher participants were functioning to support productive discourse around the implementation of responsive instructional practices. In this section, we present a narrative analysis from the third-grade CLD group as an illustrative case of these dynamics between facilitator moves, teacher participation, and growth over time of productive discussion, characterized by participants making more connections to general principles, pedagogical reasoning, and inquiry into their instructional practice. As illustrated by the code frequencies in Table 2 and in line with the patterns across groups shown in Fig. 2, Wanda, the teacher leader, increasingly used structuring and eliciting moves, while decreasing the use of connecting moves. However, at the same time, the participants substantially increased their use of connecting moves that reflected inquiry into responsive teaching practices.

For the first cycle, as in many of the other groups, the university-based mentor did most of the facilitation. The mentor began by structuring the debrief to focus on student learning, explaining that each person would get five minutes to share, first by stating their mathematical goal for the lesson and then sharing artifacts to illustrate the progress students were making toward that goal. She explained that the other participants would then each give feedback in the form of a compliment and question, followed by open feedback. The mentor then asked Wanda, the teacher leader who was apprenticing as a facilitator, to share first. Wanda began by sharing examples of student work from her teaching of the lesson, while also reflecting on things she might have done differently (e.g., “What I should have done is I should have pushed [the student] to recognize that this was repeated addition”). At times, the mentor prompted Wanda to pause and then offered her own reflections or posed questions to the group about the understanding represented in the student solutions. The discussion that ensued was a back-and-forth between Wanda and the mentor, the mentor making connections to student understanding and Wanda reflecting on things she might have done differently as a teacher. The other participants then shared their artifacts in a similar manner. Although there was a lot of productive discourse around student understanding of the math and alternative pedagogical decisions, in Cycle 1, it was the mentor who was doing much of the structuring, cultivating, and prompting others to make connections to general principles of teaching and learning.

In the second cycle, Wanda took on more of the structuring by posing the focusing question for the debrief around the launch of the task, focusing on the professional learning goal by asking “What goals did you want to accomplish in the launch and how did you determine that your students were ready to be let go?” She also utilized more cultivating moves, mostly in the form of giving the other teachers positive feedback about the artifacts that they shared. At one point, Wanda asked the other teachers to make recommendations for the potential next steps—an attempt to elicit pedagogical alternatives from participants—but she then quickly responded with her own suggestion rather than waiting for or encouraging contributions from other participants. Although Wanda structured the session and helped participants feel comfortable sharing their practice, the connecting moves were made by the mentor and sometimes by other participants. Overall, these first two sessions functioned more like a “show and tell” with each participant presenting and reflecting on their artifacts than a true dialogue.

By mid-year, as shown by the coding of Cycle 3 in Fig. 2, Wanda began using more eliciting moves, a trend that we noted in the next three cycles as participants also increasingly made connecting moves and the mentor drew back. For example, in the Cycle 3 debrief, Becky shared a video clip where she was working with a student who was having trouble getting started with the task. Rather than just asking the others for compliments, Wanda posed a direct question to draw out Becky’s thinking and elicit connections to pedagogical alternatives: “What questions could you have asked or how?” When Becky responded by considering more directive questions she could have asked to help the student see the important mathematical relationships, Carrie countered with an affirmation: “I don’t know, I thought you did well,” and continued to explain why she liked the way Becky was continually asking the student questions to help them work productively through the struggle. This prompted Becky to articulate the tension she was feeling around allowing students to struggle and knowing when to “give a little push:”

I think that's where I struggle the most is, when is it time to just kind of give him that little push, right? To kind of “here, do this, try this.” But then I think it would've been okay because [the task asks] how many different ways can you find, so that would've got him to one way, right? And then he could move forward from that.

Wanda responded by restating the tension Becky named and acknowledging that it was something she also struggled with in her teaching, a move that cultivated a sense of community and inquiry by making a connection between their instructional practices and acknowledging an essential tension in responsive teaching. Wanda’s facilitation moves maintained the group’s focus on pedagogical decisions and brought other group members into the conversation to consider the challenges Becky was presenting, highlighting the fact that they were all learning how to implement these responsive practices together.

In Cycles 4 and 5, Wanda continued to pose questions that elicited contributions from participants and, after careful planning with her mentor, more strategically structured the debrief around instructional practices that were being introduced in the practice-based PD sessions. For example, in Cycle 4, when participants shared video clips of “facilitating productive struggle,” she asked the other participants to make a list of questions they heard the teacher ask and to identify those questions that helped the teacher understand the learner’s ideas and those that helped the learner move forward in their thinking. This was language and a distinction about responsive questioning practices that had been introduced in the recent practice-based PD session that some of the participants had attended, and Wanda’s prompting primed participants to connect the instructional practices they observed to general principles about teaching and learning. By narrowing the scope of what could be noticed in the artifacts to focus on a particular aspect of responsive teaching, Wanda was able to facilitate a discussion that connected to prior learning of the group, while still maintaining the very real connection to experiences and questions faced by the participants as they enacted this lesson in their various school contexts.

In Cycle 6, there was a dramatic increase in the presence of connecting moves made by participants. Teachers were asked to bring video clips from the portion of their lesson where they were leading a whole-group discussion of student solutions. After Becky shared her clip and started to explain her reasoning for selecting student work to share, Wanda paused her and said, “Instead of you telling us why you chose it, let’s see if we can figure out why you chose it.” This prompted the participants to engage in productive discourse where they both analyzed the examples of student thinking and connected their interpretation to alternative instructional choices and rationales. Wanda was the next to share her video clip, again asking the group for their comments on her selection and sequencing of solutions. Becky commented by complimenting and naming the progression of strategies in order of complexity, from a student who had modeled the task concretely, to one who used an equation, and then one that had an equation and an explanation,” noting, “That was something that mine was lacking… I didn’t really have one that had both.” In this turn, Becky was both complimenting Wanda’s strategic sequencing of strategies to provide students access to each other’s thinking and critically reflecting on her own practice.

Susan then complimented another instructional practice that she saw in Wanda’s video, noting Wanda’s use of mathematical language during the discussion and offering her reasoning for why this was important for supporting deeper student understanding.

I noticed that you talked about the math vocabulary or the math language about the denominators. So not just that the denominator is this, but what is it, what does it mean? What does it tell us? So I felt like that was really critical just in general math understanding, but also to reiterate to kids, what does this mean in terms of the problem? Why is this important for this particular problem?

This led to the participant engagement in a dialogue that drew connections between Wanda’s use of mathematical language to help the student make sense of the important concepts in the task and Becky’s point about students not always being able to explain their solutions. Carrie then connected the idea to student learning by remarking that it reminded her of the work they had all done to solve a different task in a recent PD, where she recalled having to go back to think about the meaning of a solution in the context of the problem. Wanda picked up on this shared experience to connect their ideas to the role of the teacher, noting that when planning this task, she also thought about how the facilitator of the PD they had all recently experienced had sequenced the solutions to give learners access and help them gradually build understanding of more complex mathematical ideas.

Wanda: I was like, I'm going to follow [the PD facilitator], I'm going to start with a picture, with a more visual one, and then [my students] can understand what the other students did.

Carrie: I see it! Yep, Wanda, you were in my head!

Carrie and Susan continued to reflect on how they made sense of the challenging task in the PD and how “showing your work” was important for the learning process as well as for orchestrating a discussion to help learners understand the important math.

Throughout this session, the combination of Wanda’s structuring and eliciting moves supported productive discussions as participants increasingly made connections between their noticing of teaching practices in each other’s videos to pedagogical reasoning, student understanding, their own learning experiences, and future instruction. The conversations were multi-voiced and dialogic, anchored in the artifacts of their practice, and oriented toward making adjustments and improvements to increase student access to mathematical reasoning and sensemaking.

Over time, the combination of more intentional structuring of the debriefs and Wanda’s increasing comfort in her use of both cultivating and eliciting moves (while the mentor pulled back) supported more varied and multi-voiced productive discourse from participants, characterized by connections to pedagogical reasoning, student understanding, or future instruction. Moreover, these connections were generated by the participants themselves rather than only by the mentor and teacher leader and were taken up collectively through dialogue.

At the end of the year, Wanda reflected on how she had grown as a facilitator and how challenging it was to know how to ask the right questions:

You're the one, you know, asking those thoughtful questions, that get everybody to the point where you need them to be…. I think the biggest and most important thing with [Project Name] is really facilitating the productive struggle, and that's through questioning.

Wanda’s reflection on the facilitator’s role of “asking thoughtful questions that get everybody to the point where you need them to be” reflects her growth in using eliciting moves, which can also be seen in Table 2, as well as the value of staying focused on learning goals for participants. In other words, she was being responsive to both the participants’ contributions and the professional learning goals for each cycle. She also went on to draw a parallel between the facilitation practice of asking good questions to the teacher’s role in facilitating students’ productive struggle:

And that's the hardest part because normally, I mean it happened to me when I was in school, the teacher would just come and give you the answer, and you go on about it having not learned it. But you want it to be meaningful to the kids. So being able to ask the questions. . . and that's hard, that's very hard (Interview, 06/24/2021).

Teacher leader perspectives on focused responsiveness

Our analysis of the interviews, debrief sessions, and the member check with teacher leaders illuminated how they were making sense of and conceptualizing their roles as facilitators as well as some of the challenges and tensions associated with learning to use these responsive facilitation moves to support teacher learning. Meredith, who was the most experienced leader in the group, described it as “true facilitation” that was different from her experience planning and delivering more traditional professional development because it involved adapting on the fly in response to unanticipated ideas that teachers brought up and disagreements that emerged. She also articulated how it involved being responsive to teacher contributions and questions in the moment instead of simply moving forward with a predetermined plan:

I've also learned that no matter how much planning you do, they're always going to think of things and bring up things I didn't think of. So just being prepared and ready to address those and think on my feet and being able to make sense of what they're thinking (Interview, 06/07/2021).

At the same time, however, Meredith acknowledged the importance of being focused: “It really forced me to facilitate conversations and guide them in a way I wanted them to go.” Focused responsiveness, then, involves navigating the tensions between eliciting and responding to emergent ideas while ensuring that conversations do not stray too far from the goals of the session. This is similar to the tension encountered by teachers as they make instructional decisions in response to both their students’ ideas and the mathematical goals of their lesson.

Providing adequate wait time and refraining from answering questions themselves was an important part of eliciting contributions from the group, and several teacher leaders mentioned the need to pull back, not do too much of the talking, and not always provide answers. As Whitney remarked, “Sometimes it seems like you know a question that was asked is unanswerable. But if you give them enough time [the participants] will gather their thoughts” (Interview, 06/15/2021). Others focused on balancing their own contributions to the group. Reflecting on her growth as a facilitator after two years, Sasha commented on her improvement in “knowing when to add to the conversation and offer my opinion or just listening and following the train of thought” (Debrief session, 06/01/2022).

Although some of the teacher leaders were initially uncomfortable with the idea of being positioned as a leader in this role, they eventually recognized that the group was all learning together and supporting each other. As Alicia explained:

I'm thinking everybody knows more than I know . . . so I always went in just very nervous and like, "Oh God. They're going to see through this. I don't know what the hell I'm doing." That's always how I felt, but it did get better. I think teachers are always looking for ways to get better. The best teachers are always like, "Oh, that's a great idea. I'm going to try that." So, I think it ended up being so collaborative and helpful for everybody. (Interview, 07/06/2021)

The collegial and collaborative nature of the CLD sessions enabled teacher leaders to position themselves less as experts and more as colleagues and co-learners, offering guidance and support without making teachers feel like they were doing something wrong or needed to correct their practice. Sasha appreciated being able to share ideas about different approaches in the debrief sessions in these collegial ways:

The debriefing sessions were really cool because I felt like, especially with being able to share out my thoughts with the participants there and saying “You did it that way, maybe I would have done it a little differently,” and then I might share how I would have done it differently, without having that person think that they did it wrong.

Similarly, Alicia, who felt uncomfortable positioning herself as an expert, reflected that she could focus on the strengths in teachers’ practice to support their ongoing learning and growth:

I really understand now how to encourage someone and point out maybe somewhere or areas that might need work, without saying, "Oh, you didn't do that right." You know what I mean? You don't have to do that. You really can focus on all of the positives, on the strengths, and have them pull out what they need to work on, and then go from there (Interview, 07/06/2021).

One important way, therefore, that teacher leaders managed the tension around what Coles (2013) calls “directive collegiality” (p. 720) was to cultivate a community where participants felt safe and supported to name the strengths they saw in each other’s practice as well as offer alternative suggestions and ideas that could generate future growth. In other words, they recognized and positioned the participants as having important expertise to share, while also helping them feel safe to take risks and experiment to improve their own practice (Gibbons et al., 2021).

The reflections of the teacher leaders helped us understand that focused responsiveness, in the context of the facilitation of a teacher learning community, entails striking a balance between following participant contributions and professional learning goals, making space for participants’ ideas and sharing one’s own, providing enough wait time to stimulate meaningful contributions while also being mindful of keeping the conversation moving, and finding a comfortable balance between being an expert and being a colleague when offering feedback.

When we discussed the framework of facilitation moves with the teacher leaders during a debrief in the year following this study, they used it to both reflect on their own practice and support their continued growth. Four teacher leaders set goals to incorporate more cultivating moves, three decided to work on incorporating more connecting moves, and one prioritized using more moves of both types. Focusing on facilitation moves, and not on innate ability or performance, enabled the leaders to see specific ways they could improve and grow. Moreover, it expanded their views, and ours, of what it meant to lead professional learning, as they shifted from a belief that they had to be experts to focus on the idea that they were continually learning along with the participants.

Discussion and conclusions

Our framework of facilitation moves adds to the existing work on core practices of leaders around PD focused on mathematical tasks, video, and practice (Borko et al., 2014; Even, 2014; Gibbons et al., 2021; van Es et al., 2014) by specifying the work of focused responsiveness in the context of facilitating teacher learning communities around artifacts of instructional practice. The CLD communities reflected important differences from more traditional PD settings: Instructional expertise was distributed across participants rather than residing in the facilitator, participants made their practice public and available for inquiry, and learning occurred through collaborative dialogue and reflection on practice in relation to specific instructional improvement goals. Although each session had an overall focus on an element of instructional practice, participants helped to shape the direction and focus of the dialogue through their selection of video clips, raising questions about their own practice, and offering feedback and suggestions to others. In this way, there was a balance between structure and teacher agency.

Specifying the work of this kind of facilitation is important so that others can learn to see, enact, and improve the component parts of complex practice (Grossman et al., 2009). We identified four categories of facilitation moves that supported productive discourse around instructional practice: (a) structuring moves, which function to keep the group focused on the professional learning goals; (b) eliciting moves, which create space for participants to contribute; (c) cultivating moves, which build trust and a sense of community; and (d) connecting moves, which serve as conceptual resources (Horn et al., 2017), by bridging teachers reflections on their experiences to more formal theories of teaching and learning. Moreover, teacher leaders found this framework to be helpful in thinking about their roles as facilitators and improving their own facilitation practices.

Taking a situative perspective prompted us to attend to the interactions between teachers, teacher leaders, and mentors as well as changes in participation over time as they engaged in collaborative lesson debriefs. This allowed us to see how productive discourse around instructional practice was co-constructed and sustained by the group, as teachers began to take up some of those facilitation moves themselves. Connecting moves, which were made by both facilitators and increasingly by participants, were supported by structuring, eliciting, and cultivating moves. As the teacher leaders gradually took on these practices which had been modeled by their mentors, more space was created for participants, who were also becoming apprenticed into the practices of the community to engage in productive discourse and inquiry into practice.

Because we did not just study experienced facilitators but also teacher leaders as they apprenticed with a mentor, our findings offer insight into how to prepare facilitators—both in terms of what they need to learn and how they might be mentored. The six teacher leaders in our study followed different trajectories of growth in relation to the types of moves that they took on, but in each case, there was an increase in connecting moves made by the group. For example, as Table 2 shows, novice teacher leaders Meredith, Wanda, and Whitney increased in the use of cultivating moves over the year, while Hannah, a teacher who gradually took on leadership later in the year, grew most in her use of structuring moves. Moreover, the novice facilitators found that the framework helped them reflect on and improve specific aspects of their facilitation practice as they moved into independent roles.

The current approaches to mathematics education call for instruction that is responsive to learners as they engage in productive struggle, reasoning, and communication around challenging tasks while developing deep conceptual understanding (Hiebert & Grouws, 2007; NCTM, 2014). In this research–practice partnership, participating teachers first learned what this kind of instruction looks and feels like, then developed the necessary teaching skills and practices through enactment, reflection, and giving and receiving feedback in a community of peers, teacher leaders, and mentors. This approach to professional learning positions teachers’ experience and knowledge as important assets while supporting them to try out and reflect on new classroom practices.

Unlike more traditional professional development, in the CLD communities, participants had some agency over the focus of the discussions. Although the mentor and teacher leader planned each session around a particular element of responsive mathematics instruction (e.g., the launch of the task, facilitating students’ productive struggle on the task, or facilitating discussion of student solutions), the teacher participants chose their own artifacts to bring and posed questions about their own practice. The facilitators, therefore, had to respond in the moment to build on participant contributions in a way that was responsive to both the teachers’ emerging ideas and the learning goals for the session. This is, in turn, parallel to what the teachers were working on in their own classroom practice, building on student contributions in a way that was responsive to both students’ emerging ideas and the mathematical goals of a lesson.

The power of professional learning through collaboration is well-established (e.g., Grossman et al., 2001; Horn & Kane, 2015; Little, 2003; McLaughlin & Talbert, 2001). Our study shows how facilitators and participants can learn to cultivate productive discourse that supports teacher learning around artifacts of instructional practice through mentored apprenticeship. The framework for focused responsiveness in the discussion of artifacts of teaching contributes to the understanding of the complexity of facilitation and informs our understanding of how to prepare mathematics leaders to support productive discourse around instructional practice in professional learning communities.

Our findings are limited, of course, by the fact that we studied a small group of teacher leaders in one specific professional learning context where they were supported by university-based mentors. An important direction for future inquiry will be to explore whether these facilitation moves for focused responsiveness and patterns in their co-construction and use hold true in other contexts where teachers have opportunities to collaborate and have agency over their professional learning.