Lessons Learned Coaching Teachers in Behavior Management: The PBISplus Coaching Model

Abstract

There is growing interest in coaching as a means of promoting professional development and the use of evidence-based practices in schools. This paper describes the PBISplus coaching model used to provide technical assistance for classroom- and school-wide behavior management to elementary schools over the course of three years. This tier-two coaching model was implemented within the context of school-wide Positive Behavioral Interventions and Supports (PBIS) and tested in a 42-school randomized controlled trial. We summarize some of the lessons learned by coaches regarding their efforts to gain access to the administrators, teachers, and student support staff in order to effect change and improve student outcomes. We conclude with a discussion of ways to successfully collaborate with teachers to promote effective classroom- and school-wide behavior management.

In response to concerns regarding the effectiveness of training for teachers implementing classroom-based curricula, there has been increased interest in the use of coaching and consultation models (Denton & Hasbrouck, 2009). Some of this momentum stems from legislation (e.g., the Reading Excellence Act and Reading First, as enacted under the No Child Left Behind Act) which required schools to deliver evidence-based instruction and allowed for funding of professional development. More recently, there has been interest in the application of coaching models to behavior management – both at the school- and classroom-levels (e.g., Kratochwill & Bergan, 1990; Sheridan & Kratochwill, 2007; Sprick, Knight, Reinke, & McKale, 2006). Research on professional development models suggests that situated learning (i.e., professional development and learning that takes place in its natural context) promotes greater outcomes than discrete training that occurs outside of the day-to-day context (Lave & Wenger, 1991). Collaborative approaches to professional development not only instill learning in a more effective manner, but also create professional networks that serve to sustain newly acquired skill implementation (Dunlap et al., 2000; Joyce & Showers, 1980). Providing teachers with opportunities to reflect on newly acquired skills and work collaboratively with other teachers renders professional development far more effective and sustainable than traditional models (Garet, Porter, Desimone, Birman, & Yoon, 2001).

Despite the growing interest in coaching, there have been few rigorous outcome studies designed to examine the effectiveness of coaching models on student outcomes (Pas, Bradshaw, & Cash, in press). The available studies have focused mostly on academic content coaching and curriculum implementation (e.g., reading, math, or science; e.g., American Institutes for Research [AIR], 2004; Deussen, Coskie, Robinson, & Autio, 2007). Furthermore, there has been limited research specifying coaching models (Pas et al., in press). The current paper addresses this gap in the literature by focusing on a coaching model and its implementation by three expert coaches. The coaching model was used in the PBISplus randomized trial of Positive Behavioral Interventions and Supports (PBIS; Sugai & Horner, 2006), which is a non-curricular universal prevention strategy that aims to alter the school environment by creating improved systems (e.g., discipline, reinforcement, and data management) and procedures (e.g., office referral, reinforcement, training, and leadership) that promote positive change in staff and student behaviors. The whole-school PBIS strategy aims to prevent disruptive behavior and enhance the school's organizational climate by implementing a three-tiered prevention model (Bradshaw, Koth, Thornton, & Leaf, 2009; Mrazek & Haggerty, 1994; O'Connell, Boat, & Warner, 2009), where more selected interventions complement the universal school-wide components of the model (Sugai & Horner, 2006; Walker et al., 1996).

The PBISplus trial had a particular focus on tier-two interventions, whereby coaching supports were provided to schools already trained in and implementing the universal elements of school-wide PBIS to address the needs of students not responding adequately to the universal model; a core focus of the program was on assisting teachers in the implementation of classroom-based interventions. Findings from the three-year, 42-elementary school trial indicated that the PBISplus model is effective at increasing both teacher efficacy and students' academic achievement (see Bradshaw, Pas, Goldweber, Rosenberg, & Leaf, in press).

This paper aims to describe lessons learned from the PBISplus trial with regard to helping coaches gain access to the building, classroom, and teachers, and the efforts necessary to improve student behavior. There are a number of insights gained from these coaching experiences that may guide other researchers and practitioners in the use of coaching and consulting supports in schools and to improve future implementation of such models. We begin with a review of the extant literature on coaching models and then apply this work to the coaching model developed for use in the PBISplus trial. Case examples from this trial are used to illustrate the implementation of best practices in coaching and provide practical guidance to those interested in implementing coaching models or designing studies to test the impact of coaching.

Definitions of Coaching

One challenge in the coaching literature is the lack of consensus over what coaching is and what it entails. Without a clear operational definition of coaching, it is difficult to determine the effectiveness of coaching. We draw on the work of Denton and Hasbrouck (2009), which focused on reading coaches, to define coaching as a role in which an individual provides another with direction and support to accomplish their goals. In the case of coaching within the educational context, a coach is one who works with teachers to provide support in the implementation of their duties as a teacher; this could include providing instruction, engaging in effective classroom management, or addressing the needs of a specific student.

A related concern is how the coaching is done. This includes who the coach works with, the techniques used in coaching, and the model applied. Furthermore, there is limited work specifying the core features of coaching, including (a) how coaches spend their time (or how they should), (b) the techniques used to improve teacher practices, (c) the model used (e.g., expert, peer, or collaborative), and (d) the training coaches need to be effective (Denton & Hasbrouck, 2009).

Approaches to Coaching

Coaching approaches have been outlined and subsequently categorized by a number of researchers (see Denton & Hasbrouck, 2009 for a full review). Each of these categorizations is based on who the coach works with and the approach taken. Among those reviewed, the most complete and comprehensive categorization system was put forth by the American Institutes for Research (AIR, 2004) and modified by Denton and Hasbrouck (2009). This categorization not only takes into account how coaches spend their time (as in Deussen et al., 2007), but also includes the style or approach used. The categories identified were: (a) technical coaching which helps improve teachers' instruction by focusing on professional development provided to teachers and restructuring of classrooms using an “expert” (Gutkin, 1999) model (i.e., the coach is an expert teacher providing technical assistance to a novice teacher [e.g., Poglinco et al., 2003]); (b) collaborative problem solving, in which a coach helps teachers address the needs of students through facilitation of problem-solving stages (i.e., problem identification, identifying and prioritizing goals, developing an action plan, and evaluating the outcomes [e.g., Hasbrouck & Denton, 2007; Rosenfield, 1987]) to enable teachers to implement a plan; (c) reflective coaching, in which the coach prompts teachers to think reflectively and critically about their teaching practices in order to change behavior (e.g., Garmston, Linder, & Whitaker, 1993); (d) team-building coaching, which utilizes reflective coaching techniques within a group context to create “learning communities” in schools (e.g., Showers & Joyce, 1996), and (e) reform or change coaching, which targets whole-school (rather than individual teacher) improvement through engaging principals and teachers in leadership development and helping the school to better allocate and utilize its resources. While most coaching or consultation approaches (i.e., named models) utilize multiple approaches, we believe this is a helpful heuristic for identifying and categorizing the key components commonly used in coaching. It should be noted that this heuristic explicitly includes some consultation models (e.g., Instructional Consultation as an example of collaborative problem solving). The PBISplus coaching model has some similarities to existing consultation models in terms of using an indirect service model to assist teachers to effect change in students, as well as the specific coaching techniques used (e.g., building rapport, the focus on data and problem solving; Denton & Hasbrouck, 2009),

The PBISplus Coaching Model

This PBISplus coaching model was developed and tested within the context of a randomized controlled trial of PBIS (Bradshaw et al., 2011). The trial focused on supporting elementary classroom teachers in their acquisition and implementation of evidence-based classroom management practices in the classroom, as well as the implementation of tier-two targeted, or selected, preventive interventions. Three PBISplus “Liaisons” served as coaches to provide technical assistance to teachers and student support teams regarding the use of evidence-based practices in the prevention and interventions for problem behaviors. This was accomplished through consulting with teachers, providing support in the use of evidence-based practices, providing support in problem-solving student issues through a behavioral approach (with a focus on the function of behavior), and attending meetings which addressed student needs (e.g., student support team meetings). Using the categorization approach provided by Denton and Hasbrouck (2009), the PBISplus coaching model integrates the technical, collaborative, and reflective coaching approaches to assist schools in better addressing the needs of students who did not respond to the universal elements of PBIS.

When consulting with teachers in the classroom, the coaches utilized techniques including observing and providing feedback to teachers, modeling the use of evidence-based tools and processes, and delivering formal, didactic professional development sessions. This coaching model is considered by Hasbrouck and Denton (2005, 2007) to be a “student-focused” model, where the intended outcome was to guide teachers in the development of plans to support student success. In the trial, we broadened this model to include additional technical assistance at the child, classroom, and school levels. The coaching supports were intended to help teachers develop new skills, which would become routine, and ultimately sustained, over time.

The theoretical foundation for the coaching model was grounded in socio-cultural learning theory, which maintains that individuals embrace new information not simply through learning but also through practice (Vygotsky, 1978). In addition, learning in contextual settings promotes a stronger connection and commitment to the implementation of newly acquired skills. In this way, social interaction may be a vital component of the learning process as teachers first acquired new information via social interactions and contexts and then incorporated the new information into their own individual cognitive structure (Vygotsky, 1978). The sociocultural learning theory has implications in assisting teachers in developing and implementing new ideas in schools and classrooms.

Social interaction also elucidates the critical importance and value of the cultural awareness that teachers cultivate for their students using tools such as “Double Check,” a professional development framework for culturally responsive teaching (see Bottiani et al., in press). Finding the culturally appropriate “voice,” or manner in which communication is facilitated and new information is offered, can assist in teacher acceptance and enhance the use of new information in schools. Additionally, when teachers learn new strategies, scaffolding (Vygotsky, 1978) is a powerful tool to further develop skills, apply them, and sustain them over time.

Coach Training and Assignments

The three coaches were doctorally-trained individuals whose backgrounds included advanced training in functional behavioral assessment, consultation, teaching, and facilitation. They were each assigned a caseload of between five and eight elementary schools enrolled in the trial. The coaches were external to the school, and thus itinerant in nature. Each school received 16 hours per month (approximately four hours weekly) of coaching support in the first year of the project and 8 hours of coaching support per month in the subsequent two years, for a total of three years of support. The principal was informed that the technical assistance focused on decreasing referrals to the office by increasing teacher capacity to manage student problem behaviors in the classroom implementing the PBIS framework. All schools had been trained in the universal, school-wide PBIS model prior to enrollment in the project, as the trial intended to build on the universal model by providing training and coaching on classroom-based management strategies and tier-two supports.

Lessons Learned

Throughout the PBISplus trial, the three coaches met twice a month for supervisory meetings with the Project Directors. During this time, the practices used in the schools, progress noted, and obstacles faced were discussed and documented. Based on the information compiled during these meetings, common themes were noted and delineated for the purpose of this paper. As this trial focused on the implementation of tier-two supports, and specifically the use of evidence-based practices, coaches also utilized best practices defined in the coaching and consultation literature to achieve the desired changes at the school and classroom levels; therefore, many of the lessons learned presented are explicitly tied to the literature base. In this section, we present the lessons learned from the PBISplus trial, with regard to providing coaching supports. We align each lesson with existing literature in order to illustrate the connection between the literature and practice-based examples. We also highlight examples of some of the strategies for overcoming common challenges implementing evidence-based approaches. Evidence for the effectiveness of these approaches comes from the overall impact of the trial (Bradshaw et al., in press), as well as quantitative and qualitative data collected by the coaches and teachers.

Getting support from the top

The principal plays an instrumental role in the change process (Sarason, 1996). Thus, a coach providing technical assistance in today's schools will succeed only with both the approval and active support of the principal. This support can be displayed through the principal's actions, such as individual and group meetings with the coach, helping connect the coach with other school staff, involving the coach in relevant school meetings, and ensuring that information is shared with the coach. To get the support needed, coaches asserted clarity of intentions and outcomes, so that there was full transparency as to purpose and impact of the role the coach will play. In the PBISplus trial, all of the schools were voluntary participants (as opposed to mandated), so in many schools, the principals willingly provided the coaches with access to key school leaders such as assistant principals and teachers, to the classrooms, and to existing school resources. This may not always be the case for others engaging in coaching, consulting, or professional development capacities.

While many schools made overtures of welcome, there were still instances of resistance from several principals to get direct access to classrooms and teachers. Reasons for this may have included issues of power, trust, protectiveness, or reluctance to change. An example of this was when principals protected their classroom teachers, insisting that any intervention focus exclusively on the student's behavior (i.e., rather than teachers' skills). In another example of resistance, the principal at one of the project schools never granted the coach access to the classrooms. Rather, the coach was restricted to supporting the school through ancillary roles, such as serving on the student services team or on afterschool committees. This was in contrast to other schools, where the principal granted the coach full access to classrooms, resulting in greater willingness on the part of teachers to request assistance. Though the reasons for principal resistance are difficult to determine, such resistance is unfortunately common (Sarason, 1996). It should be noted that the importance of the principal likely varies across the different school levels (e.g., elementary vs. secondary schools); in the case of the PBISplus trial, the schools were all elementary schools and therefore the principal was the key instructional and administrative leader. This may differ in very large schools where there are multiple administrators and decision-makers with different responsibilities (e.g., instruction vs. discipline); in such instances, the support may need to come from a different individual or group of individuals. Nevertheless, the issue of gaining administrative support is still applicable.

Gaining access

Having permission to be in the building does not necessarily mean that each teacher allows (or accepts) a coach access to his/her classroom. This can be alleviated when a principal communicates the importance of the coaching process to the faculty. It is also helpful to have the opportunity to meet the faculty and staff prior to working in the school to ensure access to classrooms and teachers. In the trial schools, when this meeting occurred and the principal took the time to introduce and explain the purpose for the coach's visits, teachers were more inquisitive about the coach's role and the services provided; this also helped to ‘start the conversation’ and served as a positive, initial encounter. In one school, a principal toured the school with the coach and introduced all faculty and staff to the coach, while emphasizing that the coach was a resource to help with challenging student behaviors. Once introduced to staff, it was beneficial for the coach to invite conversations that further clarified the coach's role. The coaches found that teachers responded most favorably when they felt that their needs were heard and the responses were personalized as much as possible.

It was also important for the coach to be connected via communication tools to the staff members. One principal provided a mailbox for coach correspondence to achieve this goal. Getting on school mailing lists (e.g., an email listserv) was another way to stay informed and engaged. These introductions and actions sent a clear message to staff members that the coach was integrated into the school. In schools where these steps were not taken, the coaches observed faculty members to be more reserved and less engaged. In addition, without clarity on the coach's role, teachers perceived that the coach's presence in their classroom was to evaluate, rather than collaborate. In the case of external coaches or consultants, gaining access starts with basic connections (e.g., being invited to staff meetings and included on communication tools), which may not be necessary for internal coaches to obtain; however, oftentimes support staff may be assigned to multiple schools and also need to engage in these activities to gain access. Even when placed within a school full-time, those providing coaching services still may face the obstacle of gaining access to individual classrooms.

Getting your foot in the door

Schools can be complex, difficult to navigate, and frequently closed to outsiders. A school in effect is a collection of subsystems (Curtis & Stollar, 2002), which are delineated by grade level, experience level, administrator and faculty characteristics, cultural differences, positions as a supporter or detractor of the existing administration, and differences in levels of receptiveness to change. Fostering working relationships in these types of situations required patience, perseverance, and at times, tactics to increase acceptance into the school. Despite the best possible “selling” or presentation of a program or innovation, several factors can lead the school staff to perceive the coach as an outsider and therefore resist their influence. This resistance had a number of negative consequences; three of the most consequential were reduced program traction, resistance to the coaching process, and lowered coach morale.

Techniques from the field of social psychology can be employed to gain further entry into schools which present as “closed systems” (Curtis & Stollar, 2002; von Bertalanffy, 1968). One method is referred to as the “foot in the door technique” (Freedman & Fraser, 1966). The logic is that if someone consents to a small, almost nominal request, they are more likely to consent to a later, larger request. Applied to coaching, in the case where reluctance seems to be the result of an unwillingness to commit time and resources to engage in the coaching, a coach's initial request of a principal may be to meet very briefly, simply to introduce him/herself. The next request may increase the response cost on the school's part. For example, the PBISplus coaches' work with the schools' Student Support Team (SST) was a small request for working with teachers and served as an avenue for gaining access to individual teachers. After forming connections and gaining trust with the SST, the coaches were able to gain classroom access and were permitted to schedule meetings with individual teachers. This allowed for the school personnel to get a sense of the coach and process without making a large commitment at the outset. This approach employs similar principles to behavioral momentum (Mace et al., 1988), a classroom technique where positive behavior is initiated through shaping behaviors until the ultimate desired level of behavior is reached.

One critical step for a coach to take in order to “get one's foot in the door” is learning the etiquette and culture of the building (Sarason, 1996). Often times, that begins with signing in and displaying proper identification. The coaches needed to learn what methods of communication were preferred by school staff (e.g., telephone, handwritten correspondence, email) and how to keep the administrators engaged and informed. Meeting key support staff was also crucial, as they knew the school operations and culture, and provided support and assistance. Bringing bread, plants, and other small treats to the schools showed that the PBISplus coaches appreciated the schools' hospitality and effort, and was also consistent with the underlying PBIS framework.

Although the coaches in this case came into the schools with credentials (i.e., doctoral degrees), experience (i.e., all had at least 15 years of professional experience working in an educational setting), and represented a credible local university, there were still occasions when distrust and reluctance to engage were apparent. To overcome this, coaches persisted, showed up as scheduled, and offered to do what the school needed done, even if it was only tangentially related to the PBIS goals of the trial. This was done in part to prove the coaches' commitment to the schools' and teachers' success. These tangential tasks involved activities like working in the school's PBIS store (i.e., where students turned in incentives for tangible rewards), sitting and talking informally to students and staff in the main office, or doing additional research on behaviors in the Media Center (a task that could have otherwise been done off-site). Over time, trust with the teachers and staff was built; they shared more freely, which is a necessary component of the coaching relationship. Through presence and persistence, the coaches were able to get their foot into the door and gain acceptance among staff.

Understanding school culture

Ultimately, schools are unique when compared to other professional environments and require a different level of understanding and technical assistance (Hoy & Feldman, 1987; Hoy, Hannup, & Tschannen-Moran, 1998). First, in a typical school environment, teachers work in isolation from one another; there is a culture that when the door shuts, the teacher has full control over the classroom. This “professional isolation” leads to a culture where teachers do not receive clear or constructive feedback on their professional performance (Fullan & Hargreaves, 1991). This is often exacerbated by the absence of schedules and assignments that promote collaborative relationships between teachers and other professionals (Fullan & Hargreaves, 1991). On the other hand, the prevailing culture in a school has a way of dictating practice as well. While all schools have the mission to improve outcomes for students, schools achieve this goal in different ways. This relates to the culture of the school and the school leadership. Priorities are set, both explicitly and implicitly, which relate to the school culture (Curtis & Stollar, 2002; Sarason, 1996).

As an outsider seeking to coach teachers, the PBISplus coaches needed to become familiar with the overarching school culture as presented through the faculty's explicit and implicit goals, practices, and priorities. Understanding goals, priorities, and practices within the specific context was important to understanding school outcomes. Research shows that teachers in both high- and low-performing schools academically report progress monitoring as a priority (Hobby, 2004). In addition, high-performing schools also report prioritizing building capacity of all learners and pushing the limits of excellence, while lower-performing schools prioritize warmth and effort-centered (as opposed to achievement-oriented) constructs (Hobby, 2004). Therefore, the coaches looked at building-specific priorities and tried to link coaching priorities to these school priorities, with the ultimate goal of building capacity for all learners. To establish that culture, PBISplus coaches asked to see the building's School Improvement Plan. One building's School Improvement Plan listed improvement in student behavior (or a reduction in suspensions) as the sixth out of six goals. This school ranked academics much higher; therefore, the coach emphasized the link between behavior and academic performance and tried to reinforce this in all school interactions.

In addition to understanding the school's goals and priorities, coaches must have the ability to navigate the “hidden curriculum,” or implicit priorities and practices. Within every school, there are unwritten rules and invisible nuances that are critical for a coach to understand (Eisner, 2002). While student behaviors contribute to the hidden curriculum, so do the behaviors of the adults. Examples of student behaviors that relate to the hidden curriculum might include preventing younger students from entering spaces “reserved” for older ones; these spaces may be on the playground, on the bus, and in the cafeteria. Other “hidden curricula” may involve membership on athletic teams, band, clubs, and other school activities.

Examples of adult behaviors that contribute to the hidden curriculum might include: honoring senior members of the faculty, lunching in the teacher's lounge rather than the classroom, participating in school-wide events (e.g., fundraisers, sports), enlisting the support of the front office staff, and determining who on the faculty can help generate buy-in for a new initiative or strategy or even for a new principal. The coaches observed the different school settings and student and staff behaviors in order to identify the hidden curricula in each of the schools and then worked with this unwritten protocol in order to promote change within the building.

A final factor to consider is the level to which teachers are included in decision making. In a building where decision making is entirely the responsibility of the administrator, the PBISplus coach applied a different approach than in a building where consensus decision making was practiced. This factor determined when and how often to engage the administrator in coaching actions. The decision-making hierarchy also relates to teachers' ability, and therefore willingness, to engage in certain activities (e.g., coaching as a whole) or interventions (i.e., for a specific student). In summary, a coach first needs to understand the culture of the school and then try working with, rather than against, that culture where possible. When this is not possible, understanding the barriers allows for the coach to set realistic expectations and employ the most effective strategies.

Identifying and promoting buy-in at the school level

A significant challenge that faced the PBISplus coaches was dealing with reluctant, and sometimes resistant, teachers. Often those who are vocally resistant to new ideas can sway widespread opinion in a negative direction, derail momentum, and limit change. More senior teachers were often very effective at stopping or resisting change if they did not buy in. On the other hand, these more vocal or senior teachers were also able to move an initiative forward, if the coach could garner their buy-in. Senior teachers often serve as key opinion leaders, and thus can be critical to the success of a new initiative or innovation (Atkins et al., 2008; Neal et al., 2008).

Another way of navigating the situation of reluctance is to use what is referred to as the psychology of inevitability (Brehm, 1959). Though it is usually more helpful to invite voluntary participation and activity, in some situations and for some individuals, it may be helpful to simply inform them of the change and then discuss how to make the transition instead of initially asking if they want to make the transition. It is important to determine the cause of resistance when deciding how to proceed as the coach. In the case of the PBISplus trial, the decision to participate and randomization to treatment occurred before the coaches entered the schools. Despite this, in most schools, there were several faculty and staff members who were reluctant to consider the benefits of change. For many, knowing the change was inevitable helped them move beyond their reluctance to it and changed their mindset from one of “should I?” to “how do I?” The PBISplus coaches then took two steps: they sought the expertise and opinions of those who moved from resistance to acceptance, and they used small group meetings to integrate these “converted” teachers with the more resistant staff. By engaging in these small groups, coaches used the same phenomenon as did the resistant teachers; they maximized acceptance and utilization of coaching by taking advantage of networks among teachers and by creating and utilizing champions for promoting change (Fixsen et al., 2005; Rogers 2002; Schoenwald & Hoagwood, 2001). This has been found to be most effective as a means for shaping attitudes regarding a new initiative or concept (Rogers, 2002).

Building rapport, trust, and collaboration at the classroom level

Teachers, in many cases, were initially guarded about having a coach in the classroom, especially someone from outside the district. In general, teachers still were not accustomed to individualized support coming from outside “guests,” which made the PBISplus coaches recognize the importance of establishing a deep and personal level of trust with individual teachers. The experience of teacher wariness to allow coaches and other professionals into the classroom is a common experience cited in the literature (e.g., Ancess, Barnett, & Allen, 2007; Sarason, 1996), and can be seen as a normal reaction that people have to attempts at behavior change (Miller & Rollnick, 2002). The issue of trust is an ongoing concern when trying to effect change in schools (Bryk & Schneider, 2002); this is also true in the case when university personnel partner with schools. Continuous dialogue and responsiveness among the partners, working alongside practitioners to seek and craft solutions, and expressing sensitivity to teacher needs help to build trust between the coach and teachers (Ancess et al., 2007). The development of supportive relationships through non-directive communication, such as motivational interviewing strategies, often results in effective collaboration and behavior change (Miller & Rollnick, 2002).

When setting out to provide technical assistance to classroom teachers, the coaches were sure to explain that their role was to provide resources and support. It was important that the teacher understood that information and data would be gathered to develop solutions, not to critically evaluate or measure teachers against performance criteria, and that the teacher agreed to the steps to be taken (e.g., classroom observations). This shared understanding promoted a trusting relationship which assisted the coaches in providing feedback to teachers. In addition, teachers were more likely to embrace the recommended actions when the PBISplus coach was able to observe in a classroom. They first provided praise for the teachers' strengths, followed by constructive feedback. They then engaged in solution-focused conversations in order to assist the teachers in developing a series of realistic solutions or strategies that could be implemented to meet the identified changes.

An effective working relationship was built upon a relationship of trust that was itself constructed by mutual and equal exchange among participants (Bryk & Schneider, 2002). Framing the work with teachers as asset-based and focused on skill building and skill mastery was important; it needed to be clear that it was not to document skill deficits. This approach also modeled a strength- (rather than deficit-) based approach for teachers to use with students. This also reduced the potential for the coaches being perceived as judgmental or evaluative.

Power sharing is another critical element of establishing an effective coach-teacher relationship (Erchul & Raven, 1997). This is in contrast to mandated support provided to troubled teachers or in support of troubled students. When the PBISplus coach-teacher relationship was first being developed, it was important to create a situation or opportunity to be ‘invited’ back to the classroom and school. For example, it was much easier for a teacher to simply accept the summary of a visit with recommendations and escort the coach to the door. If the feedback was seen as the final product, the teacher may not actually implement the recommendations. To create the situation to be invited (or welcomed) back into the classroom, PBISplus coaches offered assistance with the implementation of a particular recommendation. This not only communicated that there was indeed shared power but also that the coach would be returning to follow up about the progress of the recommendation. When a teacher was agreeable to both a specific recommendation and scheduled a follow-up meeting with the coach, the teacher was more likely to implement the recommendations discussed.

It was important as part of the trust-building process to schedule future appointments with the teacher and follow up soon after leaving the classroom. By communicating persistence to the teacher, the PBISplus coaches found the teachers' efforts tended to mirror theirs. When the teachers were more engaged in the process, they followed through with recommendations, followed up with the coach, and shared successes with other faculty.

In order for the coach-teacher relationship to remain productive, the principal needed to understand and honor the confidential nature of the work in progress. In addition, due to the potential exposure a teacher may feel when working with a coach, it was helpful to be explicit about and uphold the terms of confidentiality between the teacher and the coach. Sometimes this meant keeping all statements confidential, while other times it meant crafting a message to the principal with the teacher's input. A similar approach was taken regarding the collection and sharing of data (e.g., observations) about specific teachers; the expectation was that teachers would be the sole recipients of this information. Administrators varied in their comfort with this, as some requested that any data collected on teachers be shared with the administration. However, the PBISplus coaches made and kept a promise to teachers that there would be “no surprises” in their communication with administration.

Scheduling classroom visits

Scheduling a time to visit with a teacher can be challenging, as teachers have very little flexibility within their schedules; they are expected to be in the classroom and actively engaged in the instructional process throughout the day. The teachers in the trial schools often sought assistance in handling the behavioral concerns of a specific student and thus wanted the PBISplus coaches to visit when the child with challenging behaviors was behaving the worst. However, this was not always possible, given that the coaches' time in the schools was limited. Arriving during the scheduled class time often did not allow for coaches to meet individually with the teacher, as the students were in the classroom and it was not an appropriate time to have such discussions. It was insightful, however, to observe the classroom over time, in order to understand when and how the behavior occurred, and what preceded and followed it.

The coaches also found that conducting classroom observations at various times, and not only when a student was misbehaving, was more informative. It was then possible to identify potential triggers for both the students and teachers, factors in the setting or around events that affected behavior both positively and negatively, and variability in interaction patterns. Understanding these variables, as opposed to relying solely on the teacher's self report of the student behavior, provided greater insight into the broader classroom context. As a result, the PBISplus coaches were better prepared to support the teacher and apply a problem-solving framework.

Using data to inform practice

One of the greatest challenges in coaching teachers was to help teachers collect data with fidelity over time. Data collection can be interpreted by teachers as overly formal, burdensome, and complex. As mentioned earlier, a potential way to increase teachers' adherence to data collection is to first work with a small and easily completed task (e.g., tallying; Freedman & Fraser, 1966) and for the coach to share in some responsibility for data collection (Erchul & Raven, 1997). One way the PBISplus coaches overcame this barrier was to present the teacher with simple strategies, such as counting frequency and duration of problem behaviors (e.g., using tick marks on a paper or placing items in a jar). By showing teachers the simplicity of the process, they were more amenable to this tallying type of data collection.

On the other hand, conceptualizing the antecedents (A) and consequences (C) of a student's behavior (B), or the ABCs of behavior, in an objective way was more difficult for teachers. Data on specific behaviors, their frequency, duration, and intensity, as well as data on context (i.e., information about what else was happening in the classroom), were the most helpful in developing a function-based action plan. Once the teacher experienced the value of collecting data, it was easier to understand and collect more complex data, such as the ABC data.

It was also challenging to help teachers address the consequences of students' behavior as a means of altering that behavior. Teachers often lacked consistent responses to the same behavior over time, because of the challenges of being consistent within a dynamic classroom. On many occasions, it was necessary to help the teacher learn to diffuse the situation first, rather than confront the behavior with an automatic, and often inflammatory, response. With this approach, the teacher could regain or retain control when the behavior occurred. After mastering the skill of diffusing the behavior, the teacher was better prepared to observe the situation, collect meaningful data, and address the student with authority. This challenge was better addressed when the teacher understood the data collection process, could assess the consequences which helped sustain student behavior, and in turn better understood the importance of responding consistently.

Conclusion

The current paper highlighted some key lessons learned through providing coaching support services over a three-year randomized trial in elementary schools across six counties in Maryland. We provide practitioners of coaching with practical advice on how to gain access to and work effectively with teachers. We aimed to contribute case example knowledge to the obstacles, and possible solutions to overcome these, when coaching teachers in classroom management. Given the increased interest in the implementation and maintenance of classroom-based programs, school practitioners and researchers alike are interested in coaching and consultation as a means for enhancing the teacher practice. Of particular importance is how to effectively implement coaching or consultation models.

The experiences summarized illustrate the importance of engaging teachers, administrators, and other school personnel effectively to be successful at promoting adult and child behavior change. These lessons also emphasize the importance of gaining school access and teacher trust (Bryk & Schneider, 2002; Sarason, 1996); clarifying multiple roles, including those of the teachers, coaches, and administrators; as well as the mechanics of scheduling and engaging in accurate data collection.

Although the primary role for the coaches in these schools was to provide technical assistance, it required the use of many skills (e.g., listening, reflection, collaborative problem-solving) to promote teacher change. This illustrates that, to be most effective, coaches must be skilled as listeners and facilitators with grounding in educational principles. Teachers who are overwhelmed by challenging student behaviors are often in need of someone to listen. There appeared to be great power in storytelling and reflective questioning, as teachers often needed an opportunity to vent frustration before engaging in the work of problem solving. The PBISplus coaches found that when given the opportunity to voice their emotions about an issue, teachers could then think more clearly and often could communicate the solutions to their problems with their student(s). When frustrated and experiencing emotional stress, teachers were not ready to engage in the work of problem solving, and therefore the coach would be forcing a conversation that the teacher was not quite ready for.

On the other hand, a skilled coach is able to balance the discussion and help the teacher move past the frustration in an efficient manner; otherwise the emotions could overtake the problem-solving process. Therefore, a coach must also be strong at facilitation. Once a teacher has ‘vented’ about their frustration, an effective coach will facilitate the process by asking questions that lead the teacher in the direction of a solution. A coach can accomplish this by asking questions, listening critically to the response, reflecting the teacher's ideas back to them, and guiding choice making (Rosenfield, 1987). By approaching a problem as a facilitator, the teacher is provided opportunities to develop the capacity to problem solve collaboratively and, with time, more independently.

It is important to note some potential limitations of the current study. These elementary schools were in primarily located in urban-fringe and suburban settings, and all were trained to implement the PBIS framework. As a result, it is unclear whether these lessons learned will be applicable to all schools. Due to the group randomized controlled trial design of the PBISplus Project, the data collected on student outcomes were aggregated at the school-level and are summarized elsewhere (see Bradshaw et al., in press). Data on the effectiveness of particular coaching strategies are not available, as additional research is needed to identify which components of the coaching process are associated with significant changes for teachers as well as students.

Despite these limitations, the lessons learned from the coaching supports provided through the PBISplus trial are informative for advancing the field of implementation science (Pas et al., in press). Teachers today are presented with increased responsibilities; the idea of adding additional work is rarely well received. Therefore, it is important for the coach to change the prevailing perception of their role as a change agent. The coach must project him/herself as an asset to the teacher, not a liability or an additional responsibility - as a problem solver, not a problem maker. This can be achieved by explaining to the teacher that the coaching goal is to decrease the workload, not increase it. Recommendations from the coaching process that are considered ‘low maintenance' and ‘high yield’ by the teacher have the greatest chance of being carried out. Any action that is easy to implement has the greatest potential for success. This is consistent with the foot in the door technique (Freedman & Fraser, 1966), in that the teacher can participate in the process initially in a way requiring a smaller commitment and contribution. Once the teacher has experienced success, he or she will be more likely to make a larger commitment in future interactions. From these lessons learned, personnel assigned to coach in the school setting can learn to navigate this setting more efficiently and effectively, thereby increasing the fidelity with which interventions are implemented and enhancing the outcomes achieved. These recommendations are a first step in helping ensure implementation of a coaching model within schools.