The influence of video clubs on teachers’ thinking and practice

Learning to notice framework

Several researchers argue that a key component of teaching expertise is the ability to notice and interpret what happens in one’s classroom. Berliner (1994) and Mason (2002) claim that expert teachers have heightened sensitivities to particular aspects of one’s practice, as well as techniques for analyzing, using, and inquiring into these aspects of their work. Ainley and Luntley (2007) refer to this as “attention-dependent knowledge” (p. 3), a range of attentional skills that expert teachers use to attend to the cognitive and affective aspect of classrooms.

We agree with these claims about the importance of noticing in expert teaching. Furthermore, as reported in van Es and Sherin (2002), we propose a framework that identifies three key aspects of teachers’ noticing. One aspect of noticing involves the ability to attend to what is significant in a complex situation, referred to as highlighting (Goodwin 1994) and marking (Mason 2002). In the context of a classroom, there are many things happening at one time. Teachers make choices about where to direct their attention and what events to pursue at any given moment.

Another characteristic of noticing involves using knowledge of one’s context to reason about noteworthy events. Prior research shows that as individuals become familiar with a particular type of situation, they are better able to analyze the same types of situations in the future (Lesgold et al. 1988). Teachers know a great deal about their students, curriculum, and school context, and they use this detailed knowledge to make sense of what they observe.

Noticing also involves the ability to make connections between specific events and the broader principles they represent (Copeland et al. 1994; Hughes et al. 2000). When viewing a class discussion, for example, expert teachers will describe what they see in terms of principles, such as, “This is an assessment issue.” When teachers extrapolate from the specific to the general, they form connections between the specific instances they see and the broader pedagogical issues such events represent.

Despite an increased focus on teacher noticing in recent years, little is known about the development of teachers’ ability to notice and how to effectively support this development. One approach that has received attention in research on teacher education and professional development involves the use of artifacts of practice, such as video or student work, to help teachers learn to observe and make sense of the complex practice of teaching (e.g., Kazemi and Franke 2004). We explore how video artifacts of practice might be productive for helping teachers learn to notice.

Video for teacher learning

Video has been used for decades to support teacher learning, and it appears to be a particularly useful tool for helping teachers learn to notice. Video can capture much of the complexity of classroom interactions, and it can be used in contexts that allow teachers time to reflect on these interactions (Brophy 2004; Tochon 2007). Video has been used to illustrate exemplary teaching practices and to represent the dilemmas teachers encounter in their day-to-day practice (Oonk et al. 2004; Sullivan and Mousley 2001). In either case, using video affords teachers’ opportunities to develop their professional judgment as they reason about the complex nature of teaching (Santagata et al. 2007). Furthermore, because video provides a permanent record of what took place, teachers have opportunities to see noteworthy events that they may not have noticed previously and to view video records several times, adopting different perspectives each time.

The video-club environment seems particularly well suited to helping teachers learn to notice. Teachers have few opportunities to observe their colleagues’ teaching, and video clubs provide teachers with access to each other’s classrooms (Little 2002). Research also suggests that there is value in teachers coming together to examine artifacts from their own classrooms (Lewis et al. 2006). Classroom artifacts become common referents on which teachers can focus their discussions, which can promote deep analyses of teaching and learning issues (Kazemi and Hubbard 2008). Moreover, analyzing video in a group context allows multiple perspectives on the same event to be explored, and this is much less likely to happen if teachers were to analyze video on their own (Tochon 2007).

While video can afford productive discussions of teaching and learning, there are also limitations to its use. First, what is captured on video is only a slice of what occurred in a classroom (Miller and Zhou 2007). Teachers who view records of practice may be quick to draw conclusions from watching a brief segment and not appreciate that it represents a particular moment in time. Second, watching and reflecting on video requires different practices than those teachers typically engage in teaching. Switching between these roles can be quite difficult (Lampert 2003). Finally, video is only a tool for learning. The tasks and structures for learning and the nature of facilitation with video need to be coordinated in such a way to capitalize on what video has to offer (Le Fevre 2004).

A focus on teacher interpretation of student mathematical thinking

When we talk of teachers’ noticing classroom events or interactions, a wide range of issues arise during a typical class session that might capture a teacher’s attention. In the context of mathematics education, a particular class of events is considered important—those relating to students’ thinking about mathematics. Reform efforts encourage teachers to attend to the detailed ideas that students raise, to student confusions and misconceptions, student errors, and the like (e.g., Schleppenbach et al. 2007). In this model, an important skill for teaching is “sizing up students’ ideas” (Ball et al. 2001, p. 453) and learning to carefully listen to what students actually say (Rodgers 2002). Research on effective professional development reveals the importance of focusing on student thinking of content (Kazemi and Hubbard 2008). Sowder (2007) reports that professional development focused on children’s thinking helps teachers design instruction that promotes learning mathematics for understanding and can improve student achievement. Additionally, teachers who attend to student ideas learn about their students and use what they learn while teaching, resulting in a cycle of teacher and student learning (Borko et al. 2008).

In line with these views, we adopt the perspective that teachers need to learn to focus their attention on the particulars of student ideas. Sherin (2007) suggests that this may be particularly important for veteran teachers, who are already skilled at interpreting classroom events. She proposes that they may need to learn to see different kinds of events and see them in different ways, given new views of mathematics teaching.

We also advocate that teachers adopt a particular discourse for analyzing teaching and learning. Nemirovsky et al. (2005) identify two types of discourse that teachers use when discussing case studies of classroom episodes, grounded narrative and evaluative discourse. The former highlights the sequential nature of teachers’ commentary, similar to following the plot of a story, while the latter invokes comments laden with values and judgments, such as what was good or bad. We suggest that a third discourse structure may be useful for teachers, one that is more tentative and interpretive in nature, in which the goal is to make sense of student thinking and use evidence to reason through worthwhile teaching issues (Borko et al. 2008; Seago 2004). The goal is to promote a discourse where teachers seek to ask and answer provocative questions, press on one another’s thinking, and critically analyze events they observe.

Research on changes in teachers’ practice

A key question for this research is the extent to which teacher learning in the video club—in which segments of mathematics lessons are viewed exclusively via video—might influence teachers outside of the video-club context, such as their beliefs about teaching and learning and their instruction. It can be quite difficult for teachers to make changes in their instructional practices (Franke et al. 2007). Teachers may hold well-established beliefs that are not in line with the vision of reform and resistant to revision (Forgasz and Leder 2008). Further, teachers may lack the necessary knowledge and skills to teach in ways envisioned by reform (Ma 1999). Teachers may believe they have adopted new teaching practices, but often their teaching remains relatively unchanged because they lack the knowledge to support students as they work through new types of mathematical problems (Perrin-Glorian et al. 2008). Finally, the school contexts in which teachers work are often resistant to change and may further constrain efforts to adopt new practices. Teachers may need substantial institutional support to change instruction (Correnti and Rowan 2007; Gamoran et al. 2003), including opportunities to participate in extended professional development and access to human and social resources to facilitate learning (Porter et al. 2000).

Thus, we believed it was critical to study not only what teachers learn in the video-club context, but also the extent to which participating in a year-long video club might influence teachers’ thinking and practice beyond the video club itself. On the one hand, the video-club design is based on key principles of effective professional development (Penuel et al. 2007). It is closely tied to teachers’ classroom instruction, is long-term and sustained around a specific focus, and encourages teacher inquiry into the particulars of their practice. Further, the facilitators modeled particular strategies, acting as social and intellectual resources to support teacher growth. At the same time, however, the discussions in the video club were not explicitly focused on changing teachers’ beliefs and the teachers were never asked to try out a new method in their classroom and to report back on their experience at the next meeting. Instead, we attempted to maintain a focus on analyzing student thinking as it appeared in the video excerpts that were viewed together.

Study design

The video-club group was formed as part of a university–school district partnership. The authors approached a district administrator and offered to facilitate a video club focused on student mathematical thinking. The administrator was enthusiastic about the idea and identified a school and the particular grade level teachers for this project, based on school-level student performance data and on previous mathematics professional development offerings in the school. Prior to the first meeting, the researchers met with the teachers to introduce the idea and to ask if they would be interested in participating in the club and the corresponding research. Teachers were paid a nominal fee for their participation.

We designed the video club to help teachers learn to notice the particulars of student mathematical thinking. Thus, we directed the teachers to examine student mathematical ideas and to use evidence from the video segment and transcript to support their claims about student understanding. To that end, we asked the following types of questions: “What method was Marisa using to solve that problem?”; “Where is that in the transcript?”; and “What does that tell us about her understanding of fractions?”

The teachers came to the meetings with other goals as well. Several of the teachers said they wanted to participate because they had opportunities to collaborate with their colleagues around the new curriculum. Further, the district required the teachers to participate in professional development, and some teachers planned to use the video club to fulfill that requirement. Finally, several teachers had prior experience using video or with other mathematics professional development programs that they believed were valuable. They thought they would benefit from additional opportunities of this sort. Nevertheless, the group agreed that the goal of the video club was to allow them to see images from each other’s classrooms and to learn how video might be a useful tool for reflecting on mathematics teaching and learning.

The video-club group met once or twice a month for a total of 10 video-club meetings, each one lasting about one hour. The meetings shared the same format. Two video segments from two teachers’ mathematics lessons were viewed and discussed at each meeting. Before each meeting, a member of the research team videotaped lessons, typically 50–60-min long, from two of the participating teachers’ classrooms. The video camera was placed in the back of the classroom. Three microphones were positioned throughout the classroom, and the teacher wore a lapel microphone. When the students worked together, the researcher used the audio mixer to listen to conversations and determine which discussions to capture on audio and video. These choices were made based on the extent to which students discussed the question or problem posed by the teacher and the substance and length of their conversation. The researcher zoomed the camera on individual or groups of students or moved around the classroom to capture up-close images of student work.

After the classroom observation, the same researcher viewed the tapes and selected a segment that had potential to foster productive discussions of student mathematical thinking. These segments came from whole-class discussions, student-to-student, and teacher-to-student conversations that highlighted student ideas. The clips showed students working through a mathematical problem or explaining how they solved a problem, or they highlighted teachers and students discussing a student’s solution method. The clips were 5–7 min in length. The researcher prepared a corresponding transcript to reference in the meetings. Clips from each of the teacher’s classrooms were viewed two or three times throughout the year.

Data collection

Several sources of data were used in this study. Each video-club meeting was videotaped. The facilitators arranged the desks in a semi-circle and placed a television monitor with a VCR in the center of the group so all members could view the segments clearly. The video camera was located to the side of the monitor, and a microphone was placed on a desk near the group.

Additional data sources include interviews and classroom observations. We conducted an Exit Interview with each teacher at the end of the series of meetings that lasted approximately 30 min. We asked the teachers to comment on what they thought were the most and least valuable aspects of the video club, and how participating in the video club influenced their thinking about mathematics teaching and student learning. We also asked them to discuss how their instructional practices had changed, if at all, as a result of participating in the video club. We constructed a list of questions around these issues to frame the discussion, however, we attempted to establish a conversational tone so the teachers were comfortable sharing additional thoughts and comments.

Finally, classroom observation data were used to study changes in teachers’ instruction from the beginning to the end of the series of meetings. The teachers were observed between three and nine times over the period of the video-club meetings. All of the classroom observations were videotaped, and field notes were collected for each observation. We focus on those observations that took place within 1 month of the first video-club meeting and those that took place within one month of the final video-club meeting. We refer to these as early and late observations.

Analysis

The research reported is largely qualitative in nature. In what follows we describe the specific methods used to analyze teacher learning in each of the three contexts that is a focus of this study. The goal of this analysis is not to make claims that participation in any video club would influence teachers in a particular way. Rather, the primary goal of this study is to investigate how teachers in this video club were influenced in order to explore the potential of video clubs as a context to support the development of teacher thinking and practice.

Analysis of video-club meetings

The first dimension examined whom the teachers commented on in the clip, the Actor of focus, including Student, Teacher, Self, Curriculum Developers, or Other. While the individuals in the clips were the students and teacher, in the meetings, the teachers often referred to their own teaching. Therefore, it seemed important to create a code to clearly note when the actor was not the teacher or students in the clip. They also often commented on the intended goals of the curriculum developers, so we chose to create a new category within the dimension of actor to capture this focus.

The second dimension examined the Topic of the teachers’ comments, which included Mathematical Thinking, Pedagogy, Climate, Management, or Other. Mathematical Thinking refers to mathematical ideas and understandings (e.g., “He was using his fingers to count the groups of ten.”); Pedagogy refers to techniques and strategies for teaching the subject matter (e.g., “What method do you use to teach two-digit addition?”); Climate refers to the social environment of the classroom (e.g., “That was a fun lesson.”); and Management refers to statements about the mechanics of the classroom (e.g., “The teacher handled that disruption really well.”).

The third dimension focused on how the teachers analyzed practice (Describe, Interpret, or Evaluate), or their Stance. Describe refers to statements that recounted events that occurred (e.g., “The students had their hands up. The teacher then called on a girl to come to the board.”); Evaluate refers to statements in which the teachers commented on what was good or bad or should have been done differently (e.g., “I really like how the lesson was set up. That was great.”); and Interpret refers to statements in which the teachers tried to make inferences about what they observed. In other words, they used their observations of what was happening in the video to make hypotheses about why these events were taking place (e.g., “The way he was tapping on the cards and counting the groups means he understands multiplication.”).

The fourth dimension focused on the level of Specificity teachers used to discuss events they noticed, and this category consisted of two codes, General or Specific. General refers to statements that examine a whole class or that are broad generalizations, such as “They all have trouble with fractions.” Specific statements are those that are specific to particular events, ideas, individuals, or issues, such as one student’s idea or a particular teacher move (e.g., “Juan has trouble adding improper fractions.”).

Finally, the fifth dimension refers to the sources of evidence teachers used and examined whether their evidence was based on the video segment they viewed or on events outside of these segments (video-based or non video-based).

To begin, the video-club transcripts were segmented based on when a new topic was raised for discussion in this context (Grant and Kline 2004). Then, each teacher’s participation within each segment was coded along the same five dimensions. Because of the dynamic nature of conversations in the video-club meetings, each teacher may have participated differently within a given segment. To characterize how the individual teachers analyzed video, however, each teacher received one code per dimension for each segment based on his or her primary focus, which was determined by looking at the context in which his or her comments were made. One teacher, Drew, the first-year teacher, was eliminated from this analysis because he made few comments in the second and final meetings, making it difficult to determine the nature of his noticing in this context. Research on quantifying analysis of verbal data (Chi 1997) highlights the value of conceptualizing chunks at different grain sizes, while research on discourse analysis points to the importance of considering the broader context of the conversation in which individual utterances are made (Goffman 1981; Hymes 1974).

One researcher, the first author, coded all 10 of the video-club meetings in this way, while a second researcher coded 5 of the 10 meetings. These five meetings were randomly selected but represented those from the beginning, middle, and end of the series of meetings. Inter-rater reliability was initially 88%. Any differences between the two coders were discussed and resolved through consensus. Once all of the segments were coded, percentages were calculated for each category within the five dimensions. This was done for each teacher, for each of the ten meetings. Based on this analysis, a table was created indicating these percentages.

We then used quantitative methods to investigate significant differences between the ways the teachers analyzed video early and late in the series of meetings. First, to examine whether they talked differently about classroom interactions over time in the video-club context, the percentages for each dimension, per teacher, for the second meeting in which the teachers were present and for the final video-club meeting were compared. We chose not to use the first video club meeting because the discussion focused on introductory and management issues so it was not representative of the kind of discussion that took place in the other meetings. Then, statistical methods were used to test the significance of any changes in teachers’ analyses from the early to the late video-club meeting. Specifically, a two-tailed z-test for dependent samples was conducted to examine the significance of the differences in the percentages of comments the teachers made in the areas of Student, Mathematical Thinking, Interpret, Specific, and Video-Based, from the early to the late meetings. We used a z-test because it is the standard statistical test for analyzing proportions of known populations (Kirk 1990). Z-values greater than 1.99 indicate statistically significant difference in comments between early and late meetings.

Analysis of teacher exit interviews

Analysis of the exit interviews began with a member of the research team transcribing each of the interviews. Next, two researchers, again the first author and another member of the research team, created summaries of each teacher’s comments for each question, following methods similar to Grant et al. (1998). Each summary was then reviewed and any instances in which the teachers indicated they had learned through participating in the video club were highlighted. Following this, the highlighted segments within an individual teacher’s interview were compared in order to identify those areas in which each teacher perceived that he or she had learned. The final phase of analysis involved looking for patterns across the teachers’ perceptions of their learning. This iterative process resulted in the identification of three common areas in which the majority of the teachers reported to have learned: (a) the importance of attending to student thinking, (b) attending to student thinking during instruction, and (c) the school’s mathematics curriculum.

We recognize the limitations of self-report data. However, we also recognize the strong link between teachers’ beliefs and practice (Hoy and Davis 2006). Thus, we wanted to examine whether participation in the video club influenced the teachers’ perceptions of mathematics teaching and learning, particularly around the issue of student thinking. Further, triangulating the data provides additional insight into the ways the teachers may have been influenced by the video-club activities.

Analysis of the classroom observations

Finally, classroom observation data from early and late in the year was used to investigate changes in practice. This analysis provides an existence proof that participation in a video club can have an impact on teachers’ instruction (Eisner 1998). Additionally, we propose that this analysis will provide a useful framework for conducting future investigations into the influence of participation in video-based professional development on teachers’ practice.

Our analysis of the teachers’ classroom practices focused on whole-class and small-group discussions because we thought these would be places where we could see the extent to which teachers focused on student ideas (Hufferd-Ackles et al. 2004). To begin, all portions of the lessons consisting of whole-class or large-group discussion were identified. Two researchers, the first author and another member of the research team, then reviewed these segments of video and the corresponding field notes and created analytic memos (Miles and Huberman 1994) concerning the ways in which the teachers promoted and engaged with student thinking during each lesson observed. This analysis was guided by research that identified teaching practices related to exploring student thinking including the nature of teachers’ questions and explanations and how teachers solicit and respond to student ideas, strategies, and confusions (Cohen 2004; Empson and Jacobs 2008; Hufferd-Ackles et al. 2004; Franke et al. 2007). We used these themes as a starting point for reviewing the data, but we also accounted for additional issues observed in the data set (Strauss and Corbin 1998).

The memos from both researchers were then reviewed, and we identified several common areas in which the teachers appeared to shift from the early to late observations. The memos pointed to six teaching practices that developed over time. Three related to Making Space for Student Thinking: (a) publicly recognizing students have ideas to contribute, (b) providing extended opportunities for student thinking, and (c) eliciting multiple methods or solutions from students. Two additional practices related to Probing Student Thinking: (a) asking for explanations and (b) probing explanations. Finally, review of the memos suggested that the teachers more often took a stance of Teacher as Learner in the later classroom observations.

To systematically investigate these changes, we returned once more to the videotapes of the early and late classroom observations. All portions of the lessons identified as whole-class or large-group discussion were divided into 2-min intervals, similar to the methods employed by Borko et al. (2008). To be clear, both early and late in the year, the length of the lesson across the seven teachers remained the same, 50–60 min-long. Across the classroom observations, there were a total of 125 2-min segments early in the year and 109 late in the year. In the early observations, the teachers spent 22 min on average in whole-class discussions, and they spent 17 min on average in whole-class discussions late in the year. A researcher then coded each interval for evidence of each of the six teaching practices. In addition, a second researcher coded all lessons from three of the teachers’ classroom observations. Inter-rater reliability across each of the six practices was at least 88%. We then created a table that indicated the numbers and percentage of 2-min intervals in which teachers exhibited evidence for each of the six teaching practices that were the focus of analysis.

Next, we used quantitative methods to investigate significant differences between the teachers’ practices early and late in the year. In particular, we compared the percentages of 2-min intervals in which evidence of each of the six practices was identified for all teachers. Again, we used a two-tailed z-test for dependent samples to test the significance of any changes in teachers’ practices from the early to the late classroom observations.