“Collaboration” includes a broad category of practices. It is not one specific intervention whose effects can be neatly isolated. Researchers have therefore defined it in many different ways. In addition, schools, districts or individual departments do not necessarily institute only one collaborative practice but may foster collaboration in various ways. Researchers have not yet determined which approaches to collaboration or elements of those approaches are most effective. Effective collaboration likely takes shape in a variety of ways across contexts.
A typology of collaborative activities and structures:
Vangrieken and colleagues proposed a typology of collaborative activities and structures:
Describing the specific ways in which teachers collaborate is surprisingly difficult. The terms used to define different forms of collaboration are not universally understood to mean the same thing. For example, what one researcher may call “peer coaching” might fall under another’s definition of “mentoring” or “induction.” The most recent comprehensive literature review of studies on teacher collaboration by Katrien Vangrieken and colleagues concluded there was a lack of coherence in the research about terms used to describe types of collaboration.70
Furthermore, as with other educational practices, a particular approach to collaboration can be implemented differently in terms of purpose, scope, quality and depth. It is therefore important to be cautious when drawing conclusions based on research about the impact of any specific approach to collaboration if that research does not make clear how that type of collaboration was implemented. Vangrieken and colleagues concluded that “superficial collaboration”—such as planning teacher activities or the nature and content of testing—was much more common than “deep collaboration”—such as participating in peer observations or discussing common problems and their causes.71 For example, an induction program for teachers new to a school might consist merely of a single week’s orientation at the beginning of the school year, followed by infrequent and unstructured meetings with a colleague. But a more comprehensive induction program might continue over multiple years and incorporate frequent peer mentoring, regular collaborative planning, quarterly feedback following observations of instruction and repeated opportunities to observe master teachers’ instruction.
Moreover, no educational practice is used in isolation. A school that encourages a collaborative approach to induction may also be characterized by more collegial relationships between teachers and principals or by greater coordination of curricula across grade levels. Therefore, it can be hard, though not necessarily impossible, for researchers to isolate any one approach to collaboration from the broader context and character of a school.
The approaches to collaboration described below are not the only ways in which teachers work together. Rather, they are the ways of working together that thus far have been reasonably well documented by researchers.
Even in schools where most of teachers’ work occurs in isolation, teams of teachers may nonetheless already be meeting in subject-area or grade-level teams to coordinate curricula and assessment, plan events or discuss community engagement activities, or for a variety of reasons as part of their union contracts. Working in teams does not automatically imply that any collaborative activities take place.72 Yet existing teams could become opportunities for collaboration aimed at goals such as school improvement or enhancing teacher learning. Research has focused both on naturally occurring, routine types of teamwork73 and on collaboration that is arranged or developed in the context of a specific improvement-oriented intervention from school leaders, district or state policymakers or other administrative actors.74
Establishing shared goals and effective leadership can help teacher teams collaborate, leading to improvements. Researchers conducted an exploratory interview study with 142 teachers, administrators and staff in six high-poverty, high-minority, high-achieving elementary and middle schools in one city. Five of the six schools used teacher teams, devoting large blocks of time for team meetings. These meetings were used to discuss specific lesson plans, the curriculum more broadly and students’ achievement and behavior, both as individuals and in groups. Teams’ effectiveness was related to regularly scheduled meeting times, facilitation by trained teacher leaders and support from administrators. Teachers who agreed with the teams’ purpose were more motivated to work with their teams. Teachers generally valued their teams, saying they felt less isolated as a result and that the teams helped create “coherence across classrooms” and “shared responsibility for students.” The researchers found that both teachers and administrators credited teachers’ work in teams for individual teachers’ improvement in instruction, for students’ improvement on state achievement measures and for schoolwide improvement more generally.75
Professional communities, or professional learning communities (PLCs), are an approach to school improvement that includes teamwork. Professional communities emerged in the 1960s and are one of the collaborative models that some policymakers have recently called for in schools.76 Professional communities are usually understood to constitute a group of people across a school who are engaged in common work; share a set of values, norms and orientations toward teaching, students and schooling; and “operate collaboratively with structures that foster interdependence.”77
This approach to school improvement is undergirded by two broad ideas. The first is that if teachers share knowledge and learn actively with one another, this will benefit their instruction and their students’ achievement. The second is that structure and support are required for these exchanges to occur at the frequency and depth that are necessary for them to be beneficial.78
Research by the University of Chicago Consortium on School Research, led by Anthony Bryk, on school organizational features and student achievement outcomes in Chicago public elementary schools has particularly moved the concept of “professional community” forward. Bryk and colleagues acknowledge that work in a professional community represents a very new arrangement for teachers, one dependent on collaboration. It makes their work public to their colleagues, requires critical questions and relies on a normative commitment to student improvement.79 Yet their findings suggest that a strong professional community is an integral component of what they deem a school’s “professional capacity,” which helps account for a school’s trajectory of improvement.80 (For more on this topic, see the section on student achievement.)
The term “professional learning community” (PLC) is sometimes used as a variant of or even in place of the term “professional community.”81 PLCs can be polarizing: A survey funded by the Bill & Melinda Gates Foundation found substantial dissatisfaction among teachers with PLCs as a form of professional development in their schools, but it found substantial support among educational leaders for devoting more resources to PLCs.82 The survey—which did not specify what types of teachers or leaders were surveyed or how representative those samples were—did not explore why teachers expressed dissatisfaction with PLCs. But even Rick DuFour, whose company consults on the establishment of PLCs and other educational reforms, acknowledged that the term “PLC” has been inaccurately applied to poorly structured meetings rather than to “authentic” PLC processes.83
A 2008 review of 11 studies suggests that PLCs can increase collaboration among teachers and that a focus on student learning in PLCs is key to their potential to improve student achievement.84 (For more on this topic, see the section on student achievement.) But PLCs can be instituted in ways that vary in their effectiveness at improving student achievement. For example, a study compared nine elementary schools that received an intervention meant to foster PLCs with six control schools in the same large urban school district that did not receive that intervention. Teachers in both the intervention and the control schools met a few times a month in grade-level groups to work together.85 The study found that teachers in intervention schools canceled fewer meetings and tended to have structured agendas for their meetings.86 But student achievement in the intervention schools did not improve relative to control schools until teacher leaders were trained on using protocols for meetings and other ways of making teamwork work.87 Ultimately, whether a school creates PLCs or other collaborative teams may be less important than how the practices used in those collaborative configurations are implemented.
Teacher-to-teacher mentoring can be hierarchical—with a teacher regarded as highly skilled mentoring a newer or struggling teacher. But mentoring need not be reserved for teachers construed to be in need of help. It may be carried out for the sake of continuous improvement. Mentors may observe their mentees teaching and provide feedback, or mentors may invite their mentees to watch them teach and then discuss what each observed.88 Some schools make mentoring a part of induction for faculty who are either new to teaching entirely or who are new to that school.89
Evidence for the effectiveness of mentoring at improving student outcomes is mixed. Of two randomized controlled studies, one showed mentoring to be effective at improving student achievement and the other did not. The first study, a randomized controlled study led by Steven Glazerman at Mathematica Policy Research and in partnership with the U.S. Department of Education Institute of Education Sciences, found that the two mentorship programs it examined did not improve student achievement.90 The study was conducted with 418 elementary schools in 17 urban school districts assigned by lottery to implement a comprehensive induction and mentorship program for new teachers or to implement their district’s standard, less comprehensive induction program. Based on student achievement data and surveys with teachers and mentors, the researchers found that while teachers in the comprehensive program reported receiving more support than teachers in the standard program for their first year, there was no significant difference in teacher instruction, teacher retention or student achievement after either the first or second year. They did find that among teachers in the comprehensive program for two years there was a small but significant increase in student math and reading scores in the third year.91
By contrast, John Papay and colleagues explored the impact of mentoring in a study of 16 schools—elementary, middle and high—in a low-income, high-minority Tennessee district.92 State-mandated performance evaluations had rated some teachers as more accomplished than others in specific instructional skills. The researchers believed that the more skilled teachers might be able to help their less skilled colleagues build their instructional capacities. The researchers therefore randomly selected eight “intervention” schools in which teachers who had been rated as more skilled in a specific instructional area (such as “managing student behavior,” “teacher content knowledge” and many others) mentored their colleagues who had been rated as less skilled in that same area. The researchers compared student achievement in these eight “intervention” schools with eight “control” schools in which teachers were not paired with mentors. Achievement increased for students of the less skilled teachers who had been mentored by their more highly skilled colleagues. In fact, student achievement in the intervention schools increased overall compared with achievement in the control schools overall.93 Because the schools were randomly selected to get the intervention, it is more likely that the improvements in achievement were due to the collaboration rather than to some other factor.94
While evidence for mentoring’s effectiveness at improving student achievement is mixed, mentoring may reduce teacher turnover. One nationally representative study suggests that a more comprehensive approach to induction, one that includes mentoring, can reduce turnover.95 In this study, the combination of mentoring and support from an administrator, principal or department chair had no effect on teacher turnover. However, a seminar with other new teachers and shared planning with teachers in their subject area as well as having a mentor and support from an administrator, principal or department chair did reduce the chances of teachers leaving their school or the profession.96
Research from New York City public schools also showed that mentoring can reduce turnover. In 2004, in compliance with a new state law requiring mentoring for all new teachers, the city’s Department of Education tried to assign mentors for all teachers in elementary, middle and high schools with less than one year’s teaching experience. Teachers whose mentors had worked in the same school were less likely to transfer schools or to leave teaching with the city’s Department of Education entirely.97 Teachers who received more hours of mentoring also had higher student achievement in both math and reading.98
It is possible that a school organized with significant levels of coordination in curriculum and teaching methods, within and across grades, may have a better chance at making mentoring work for new teachers. Since few schools in the United States are organized with such significant levels of coordination, there is little empirical evidence on the effects of collaboration in such contexts. Research from countries such as Canada and Finland suggests mentoring can be valuable as part of broader approaches to induction.99
Lesson study has been defined as “classroom-based collaborative research” that is designed and undertaken by teachers.100 Lesson study originated in Japan at the turn of the 20th century and has since been a key feature of their education system. Japan saw sustained high levels of student achievement during the 20th century. The institutionalized use of lesson study in Japanese schools suggests a relationship between lesson study and improvements to student achievement.101 Lesson study is intentionally designed to lead to slow but steady change.102
In Japan, the process typically works as follows: A group of teachers reviews a curriculum and works collaboratively to identify goals for student learning and to design a lesson. They conduct a live classroom lesson led by one teacher and observed by the rest, who collect data and make observations on teaching and learning during the lesson. Teachers then meet to discuss and reflect on the data to evaluate the lesson on whether and how it achieved the student learning goals. Finally, this reflection is documented and carried forward in an iterative process to continue to refine the lesson and teaching methods. In addition to or instead of these steps, teachers may observe a highly accomplished teacher talk through the planning of a lesson and then observe that teacher teach it and reflect on it. New teachers may be asked to do the same, with guidance from peers and from more accomplished teachers.103
Lesson study has spread to other countries, coming to the attention of educators in the United States in the late 1990s.104 The practice was adopted by individual schools and some entire districts, in some cases aided by inviting teachers from Japan to coach or mentor. But lesson study has not been widely adopted in the United States. Where it does happen, the practice is most often used at the school or district level and especially for mathematics. Research on lesson study for mathematics instruction in particular indicates that it has potential to help teachers develop their content knowledge and instructional practices.105
Research on lesson study in the United States has primarily taken the form of descriptive case studies.106 This research has documented and described how lesson study is being adopted, implemented and improved upon, in part to understand what aspects of it work and do not work in carrying the practice into American schools. A descriptive study of how a group of teachers in a California district undertook lesson study highlighted the importance of helping teachers understand both the practice and the theory of lesson study.107
There are few experimental studies designed to determine by what mechanism lesson study improves instruction.108 But researchers Catherine C. Lewis and Rebecca R. Perry conducted a randomized control trial study in which 39 already existing collaborative groups of elementary school math teachers were assigned to three different conditions: implementing a lesson study practice; implementing lesson study with a specifically developed guide modeled from Japanese curriculum guides that explained the curriculum as well as common student thought processes and misconceptions; and continuing collaborative practices as usual but with a stipend equivalent to the other two groups, to serve as the control group. Groups were matched on school demographics and socioeconomic status and location within the same district, if possible, and were tracked and measured across one semester. The study found that in groups that implemented lesson study with the guide, teachers self-reported more awareness of student thought processes and higher expectations of their students.109 Those teachers also scored higher on a measure of their belief that they were helped by learning from other teachers.110
Professional development typically entails teachers attending classes or conferences led by experts from outside their school. Teachers are generally then left to implement what they learn on their own in their classrooms. Professional development opportunities and requirements have seen significant investment from school districts and states in recent decades.111 Research suggests the majority of professional development opportunities are not ongoing.112 Yet professional development activities that are not ongoing do not lead to dramatic changes in instruction or student achievement.
Elements of effective professional development:
Based on systematic review of studies on professional development, researchers posit that there are key elements or characteristics that professional development activities must include in order to be able to meet improvement expectations, including:
A case study by Elham Kazemi and Megan Loef Franke provides an example of a professional development program of sufficient duration that enabled teacher collaboration. The study examined a professional development program designed to bring together a team of 10 math teachers in one elementary school in a small urban school district for monthly team meetings. The meetings focused on students’ math work and on observations of other teachers’ classrooms in order to spur “collective inquiry,” or collaborative teacher learning. School administrators, support teachers and the principal offered support for the program. Based on systematic analysis of meeting and classroom observations, the researchers documented two shifts in teachers’ participation as a result of the program. Teachers became more aware of students’ thought processes in solving math problems, and because of this focus, they worked to develop shared goals and instructional techniques for their students’ mathematical learning.114
In schools that do more to coordinate the work of teachers in different classrooms, even professional development that is not ongoing may have potential to lead to lasting improvements in instruction. In their multiyear study of hundreds of elementary schools in Chicago, Bryk and colleagues found that quality professional development led to improved academic outcomes, but only in schools characterized with a strong “work orientation”—support for innovation and a collective sense of responsibility for improving the school.115
Researchers have only just begun to develop an understanding of what makes for effective collaboration. But many researchers have concluded that teacher collaboration tends to be more successful when teachers have goals and shared values.116 In their 2015 review of the literature on teacher collaboration, Vangrieken and colleagues note that shared goals, among other conditions, is mentioned across many studies as a facilitating factor for teacher collaboration and that, on the other hand, a lack of clarity around goals is an oft cited factor hindering collaboration.117
Collaboration centered on shared goals can in some cases be sufficient for spurring improvement. A team of researchers studied a “continuous improvement” intervention structured entirely around goals in Title I elementary schools, which serve substantial proportions of low-income students. Teachers in “intervention” schools met a few times a month in grade-level groups to work together. They set goals, monitored indicators of progress toward them and got help in achieving them, including help from principals. Teachers in comparison schools also met a few times a month in grade-level groups to work together but followed other reform models that did not center around shared goals for student learning.118 The researchers described teachers in intervention schools shifting toward a focus on student learning and away from thinking it was out of their hands if they planned a lesson and their students “didn’t get it.”119 The researchers observed teachers in intervention schools adapting their teaching practices to meet students’ needs; assigning more responsibility for student learning to their own work as teachers; and ascribing less responsibility to students’ previously demonstrated engagement, family resources or parents’ involvement.120
In fields outside education, members of successful organizations “interact regularly to share their ideas and expertise and develop common understanding of organizational goals and the means to their attainment.”121 In addition to developing common understandings of school and district goals as well as the means to achieve them, the process of collaborating can lead teachers to develop shared goals.122
Of course, goals can sometimes be too narrowly focused. For instance, teachers might collaborate to focus a disproportionate amount of their attention on supporting only those students performing just below a proficiency level on a state-mandated exam in hopes of raising a school’s status, instead of supporting all students.
Besides having goals, researchers have only just begun to develop an understanding of what makes for effective collaboration.
Within their own study design, researchers can determine whether or not a studied collaborative activity or intervention was “effective” to the degree that it did or did not achieve a certain outcome being measured, such as allowing teachers to establish a shared vision or leading to improvements in students’ test scores. However, it is more difficult to ascertain from the field of research on teacher collaboration overall elements of “effective” teacher collaboration. The most recent review of the literature, published in 2015 by Vangrieken and colleagues, attempts to contribute an understanding of what makes for “effective” teacher collaboration.
Despite different outcomes being preferred and different interpretations of effectiveness being considered important depending upon the goal and context of collaboration, Vangrieken and colleagues proposed some criteria for “effectiveness.”
Process-level criteria for effective collaboration included:
Outcome-level criteria for effective collaboration included:
Next: PRINCIPALS' ROLES