Principals’ Sensemaking of Coaching for Ambitious Reading Instruction in a High-Stakes Accountability Policy Environment

In the present exploratory qualitative study we examine the contextual factors that influenced the implementation of a multi-year comprehensive literacy-coaching program (ContentFocused Coaching, CFC). We argue that principals’ sensemaking of the dialogic instructional strategies promoted by the program in light of high-stakes accountability policies influenced coaches’ work with teachers. Principals’ views of the efficacy of the teaching strategies promoted by CFC for meeting accountability targets influenced how principals socially positioned coaches in schools (i.e., the degree to which they promoted coaches as sources of expertise to teachers), and the extent to which the coaching received by teachers focused on implementing dialogic teaching epaa aape Education Policy Analysis Archives Vol. 22 No. 51 2 practices. Our results also suggest that principals’ sensemaking of the program was for the most part consistent across years, even in the face of shifting accountability status and changes in district leadership. Implications of our findings for improving the implementation of coaching programs are discussed.


Introduction
Results of national assessments consistently show that many students, especially from lowincome families, are unable to read with a high level of understanding (NCES, 2013).While there are many reasons for students' poor reading performance, research clearly shows that instructional quality is a major factor (Palincsar & Duke, 2004).Despite significant advances in the scientific knowledge of effective reading instruction (National Literacy Panel, 2000), many teachers are unprepared to teach higher-level comprehension skills and to meet the learning needs of struggling readers (Klingner, Vaughn, Hughes, & Schumm, 1998).
Literacy coaching is one of several policy tools that have emerged over the past three decades to increase the quality of reading instruction.Coaching has been included as a core component of education reform legislation at the state and federal levels (e.g., Reading First), district reform initiatives, and comprehension school reform models (e.g., America's Choice).The popularity of coach as a strategy to improve teaching quality is well warranted.Literacy coaching is aligned with research on effective professional development that emphasizes sustained, connected learning opportunities for teachers that is based in the curricula teachers are to teach (Guskey & Yoon, 2009;Hawley & Valli, 1999; studies reviewed in Wei et al., 2009).Coaching also is aligned with current learning theories that foreground social interaction and participation in authentic activities with more expert participants in the learning process (Collins, Brown, & Holum, 1991).When well implemented, coaches build teachers' knowledge of effective reading instruction by engaging them in ongoing, collaborative learning groups, and develop teachers' pedagogical skills by modeling new forms of practice and providing guidance and feedback to teachers as they enact these new practices in their classroom.
Despite the many reasons to believe that coaching would be an effective strategy for improving teaching and learning outcomes, however, the evidence in support of coaching initiatives is limited (Gamse et al., 2008;Garet et al., 2008;Marsh et al., 2008).Explanations for the disappointing outcomes of coaching reforms often have focused on problems of implementation.Like other educational reforms (McLaughlin, 1990), the enactment of coaching in schools often does not resemble the intentions of program developers.Significant variation exists within and among schools in the quality and amount of coaching received by teachers that complicates, or even undermines, efforts to discern the effects of coaching on teaching and learning (Blarney, Meyer, & Walpole, 2009;Marsh et al., 2008).Even under controlled study conditions, such as in experimental and quasi-experimental trials of specific coaching programs, fidelity of implementation is a significant challenge (Atteberry, Bryk, Walker, & Biancarosa, 2008;Matsumura, Garnier, & Spybrook, 2012;Matsumura, Garnier, & Spybrook, 2013).
While it is clear that coaching initiatives vary widely in their implementation, considerably less is known about why coaching programs are taken up differently in schools.We take up this question in the present study.Specifically, we draw on sensemaking theories (Coburn, 2004;2005;Spillane, Reiser, & Reimer, 2002a) to investigate principals' interpretation of the ambitious instructional practices promoted by their coaches and coaches' subsequent work with teachers.

Sensemaking and the Implementation of Instructional Reforms
Over the past decade, numerous studies have emerged that show that the implementation of instructional reforms is strongly influenced by individuals' understanding of these reforms and the context in which the reforms are implemented (Coburn, 2004;2005;Cohen, 1990;Cohen & Hill, 2001;Spillane, Reiser, & Reimer, 2002;Spillane & Zeuli, 1999).Drawing on cognitive and sociological theories, this approach to understanding policy implementation, termed sense-making, is predicated on the assumption that people act on the basis of what has meaning for them (Spillane, Reiser, & Reimer, 2002a;Spillane et al., 2002b).Specifically, this approach holds that actors interpret reform policy messages through the lens of their prior knowledge and experience.Because cognition is dependent on context (i.e., is not solely an individual matter) (Greeno, Collins & Resnick, 1986), how actors interpret and act on policy messages also is influenced by the social context and larger policy environment in which the reform is implemented, for example, the interactions among teachers to understand and operationalize ambiguous and sometimes conflicting policy messages (Coburn, 2001;2005;Coburn & Stein, 2006;Russell & Bray, 2013).
Most of the research on policy implementation that adopts a sensemaking frame has focused on teachers.This research suggests that teachers often interpret policy messages in ways that either reinforce preexisting practices or lead to surface-level adoption of new practices (Coburn, 2004;Cohen, 1990;Spillane & Zeuli, 1999).Coburn (2004), for example, examined teachers' interpretation of policy messages from various reading reforms (e.g., changing textbooks, widely publicized national reports, Language Arts frameworks).She found that teachers differed in their response to shifting visions of reading instruction promoted by these reforms in a process that was shaped by their preexisting knowledge and exposure to policy messages, as well as the nature of the policy messages themselves.Approximately half of the teachers transformed (i.e., assimilated) multiple policy messages to fit their underlying assumptions about learning and their own teaching practice.About a quarter of the teachers, in contrast, sifted through and actively rejected policy messages that they construed as running counter to their beliefs about reading instruction.Other smaller numbers of teachers engaged in 'symbolic' responses to reform messages, that is, a very limited adoption of a specific reform (e.g., posting writing rubrics on their walls, but not using the rubrics to grade student work), or created 'parallel structures' in their classroom, that is, activities and approaches that attempted to balance conflicting visions of reading instruction.Very small numbers of teachers responded to policy messages by restructuring their assumptions about teaching and learning.
A small, but growing body of research suggests that how teachers understand and implement new reform practices and policies also is influenced by other actors in the school system, including district leaders and principals (Coburn, 2005;Russell & Bray, 2013;Spillane, 2000).How leaders interpret and subsequently adapt and communicate with teachers about these instructional reforms is shaped, in part, by their prior understanding of subject-matter content and teacher learning (Coburn, 2005;Spillane, 2000).Coburn (2005), for example, showed that a superficial understanding of new reading practices on the part of principals led to confusing and contradictory messages about reading instruction communicated to teachers.
School leaders (e.g., district personnel, principals and coaches) also influence teachers' implementation of instructional reforms through their role as mediators of policy messages (Coburn, 2005;Woulfin, 2012;Russell & Bray, 2013;Spillane et al., 2002b).Principals mediate the link between "shifting logics in the environment and classroom change" by foregrounding some policy messages and deemphasizing others (Coburn, 2004, p. 214).Research conducted by Spillane and colleagues (2002), for example, showed that principals' interpretation of accountability policies was influenced by multiple social and organizational factors (e.g., teachers interpretations of test scores, student demographics) and influenced their school's curriculum priorities and classroom activities.Russell and Bray (2013) showed that educators' interpretation of the No Child Left Behind and the Individuals with Disabilities Act legislation was influenced by the structure of the policies themselves.In places where the policies were ambiguous, educators were more likely to "construct interpretations that strayed from explicit policy intent" (p.16).
Research that uses a sensemaking frame to understand how school leaders mediate the effect of policies on teaching is consistent with a larger body of research showing that principals play a key role in the leadership and implementation of educational reforms (see studies reviewed in Neumerski, 2013).As noted by Coburn (2005) and Spillane et al. (2002b), principals exert a significant influence on teaching and learning through their decisions of what, among a plethora of programs and sometimes conflicting policy mandates, to prioritize at their school.Specifically, research and theory suggest that students achieve at higher levels in schools where principals emphasize instructional improvement in their reform agendas and communicate a clear and ambitious vision for instruction and learning to teachers (Allensworth, Bryk, Easton, & Luppescu, 2006;Demoss, 2002;Dinham, 2005;Hallinger, 2003;Marks & Printy, 2003).A study of 8 lowperforming schools conducted by Demoss (2002), for example, showed that student achievement increased in schools where principals focused their efforts on increasing the rigor of the curriculum, in contrast to schools in which principals prioritized meeting accountability targets or noninstructional reforms such as the use of educational technology.
Teachers also are more likely to commit to and implement an instructional reform when principals actively promote new teaching practices and devote resources toward realizing this vision in classroom practice (Berman & McLaughlin, 1978;Marsh, Hamilton, & Gill, 2008;Quint et al., 2007).Through their management of the school budget and schedule, for example, principals influence the amount and type of professional development available to teachers, as well as the amount of time teachers have to collaborate and plan with colleagues (Goldenberg, 2004;Talbert & McLaughlin, 2006).Principals also influence the implementation of reforms through their efforts to promote collaboration and communication among teachers about instruction (Supovitz, Sirinedes, & May, 2010), and their contribution to creating a positive social climate that promotes trust and risktaking on the part of teachers to tryout new and challenging forms of instruction (Sebring & Bryk, 2000;studies reviewed in Brewster & Railsback, 2003).

The Role of Principals in Implementing Literacy Coaching
Research on the implementation of coaching reforms likewise foreground the critical role that principals and other school leaders play in the implementation of coaching reforms.Studies show that principals in specific vary significantly in their understanding of the coaching job (Camburn, Kimball, & Lowenhaupt, 2008;Coburn & Russell, 2008;Mangin, 2007;Matsumura, Sartoris, Bickel, & Garnier, 2009).In some schools, the coaching job is conceptualized as being to work primarily with teachers to improve their practice.In the majority of schools, however, coaches are expected by principals to perform a range of other tasks that include administrative duties, assessment coordination, and tutoring students.This interpretation of the coaches' role impedes teachers' professional learning opportunities by reducing the amount of time coaches have available to work with teachers (Bean, Draper, Hall, Vandermolen, & Zigmond, 2010;Matsumura et al., 2009).A survey of over 1,000 literacy coaches undertaken by the International Reading Association, for example, showed that 45% of respondents spent only 2-4 hours a week observing, modeling and talking to teachers about their instruction.Principals also vary in their beliefs about the skills and knowledge coaches need to perform their job effectively with consequences for their hiring practices and the quality of coaching available to teachers (Coburn & Russell, 2008;Mangin, 2008).
An emerging body of research suggests that principals also play a significant role in how coaches are situated within the school culture (Mangin, 2008;Matsumura, Garnier, & Resnick, 2010;Matsumura et al., 2009).For example, Matsumura et al., (2009) found that teachers participated more frequently in coaching in schools where the principals publically endorses their coach as a source of expertise, promoted the coaching program to teachers, encouraged teachers to work with their coaches, included their coach in school leadership activities and asked their coach to conduct whole school professional development sessions.Principals' willingness to share leadership responsibilities with teachers more generally also has been associated with increased frequency of teachers' participation in coaching and perception that the coaching they received was useful to them for improving their practice (Matsumura et al., 2010).

Present Study
While past research conducted by us and other researchers has established a critical link between principal sensemaking and the implementation of education reforms and policies, it is notable that very few studies have examined the principal's role in mediating the effect of accountability policies on coaching practice.An exception is Mangin (2009), who showed that districts were more receptive to literacy coaching when subgroups of students were in danger of not meeting accountability targets.We are not aware of any study, however, that has investigated how principals make sense of the instructional practices promoted by coaches in the face of accountability pressure, and the consequences of this for the content (i.e., instructional focus) of coaches' work with teachers.Understanding this dynamic is important given the widespread adoption of coaching as a means to build teachers' capacity to enact scientifically based reading instruction.Most coaching programs align with ambitious visions of instructional quality (Biancarosa et al., 2010;Matsumura et al., 2013;Sailors & Price, 2010).Research indicates, in contrast, that testbased accountability mandates can reinforce didactic teaching practices and content that run counter to research-based teaching practices that promote higher-level literacy skills (Diamond, 2007;Koretz & Hamilton, 2006;Spillane & Jennings, 1997).The coaching intervention may therefore be accepted as supporting a school's efforts to reach their accountability targets, or deemed extraneous to or a distraction from such efforts.As shown in other education research, we assert that the success of the implementation of a coaching program likely depends in part on this perceived compatibility of means and ends (i.e., the alignment of coaches' work and reform goals), and the negotiation thereof (Bryk, Sebring, Allensworth, Luppescu, & Easton, 2009).Investigating how the inescapable context of high-stakes accountability pressures may influence coaching to promote higher-level literacy skills would move us toward understanding how to better implement coaching programs.
Also left unanswered in the current research literature is the question of how principal support contributes to the implementation of coaching programs over multiple years.Despite the fact that coaching policies and programs generally are multi-year investments, research on the contextual factors influencing these reforms has primarily been based on cross-sectional designs.Longitudinal research is important because, as noted by McLaughlin (1990), contextual factors influencing educational reforms are fluid and can change over time as a result of local events that compete with the goals of the reform.The factors influencing the sustained implementation of coaching reforms, including different aspects of principal leadership may be fluid as well given the volatility of school and district contexts.
In the present exploratory study, we investigate how principal sensemaking contributes to the implementation of a literacy-coaching program (Content-Focused Coaching, CFC).We argue that principals' beliefs regarding the role of the dialogic teaching practices promoted by the coach in meeting accountability goals influences the implementation of CFC -the social positioning of coaches in schools and the content of coaches' work with teachers -in ways that are consequential to teachers' professional learning opportunities.

Content-Focused Coaching (CFC)
CFC, the literacy program targeted in our investigation, was developed at the University of Pittsburgh's Institute for Learning (IFL) (Staub & Bickel, 2003).The program draws on research showing that improvement in teaching and learning requires capacity building across the levels of the school system (Glennan & Resnick, 2004;Resnick & Spillane, 2006).Thus it provides intensive and ongoing professional development to coaches, principals, and district leaders to develop the pedagogical expertise and organizational conditions (i.e., leadership support) posited by the program to support the implementation of high-quality coaching.
Coaches met with IFL fellows 3 days a month (approximately 21 days per year) to develop coaches' knowledge of effective reading comprehension instruction and ability to work effectively with teachers individually and in small groups.The professional development focuses specifically on building coaches' ability to plan and enact rigorous and interactive text discussions based in authentic texts (referred to in this study as dialogic instruction).Building teachers capacity to plan for and enact dialogic instruction was chosen as the pedagogical focus of coaches' work based on a large body of research showing that interactive and rigorous text discussions are a high-leverage practice for increasing students' reading comprehension skills (Applebee, et al., 2003;Goldenberg, 1993;Kucan & Palinscar, 2013;Murphy, et al. 2009;Nystrand, 2006;Saunders & Goldenberg, 1999;Shanahan & Beck, 2006).Specifically, the substance of the professional development centered on Questioning the Author (QtA), a discussion-based approach to comprehension instruction in which teachers strategically pose high-level questions to students at key places in a text, and encourage students to share and challenge each other's ideas to grapple with these questions (Beck & McKeown, 2006;Beck, McKeown, Sandora, Kucan, & Worthy, 1996).
Based on research showing the critical role that principals play in the success of instructional reforms (studies in Neumerski, 2013), CFC includes a substantive focus on developing principals' knowledge of, and support for, coaches' work with teachers.Specifically, principals participate in the coach professional development one day a month to establish a shared understanding of the responsibilities of a CFC coach to help ensure that principals encourage their coach to spend their time working directly with teachers to improve their instruction (as opposed to performing administrative tasks or tutoring students) and to educate principals about the specific ways they could facilitate their coaches' work with teachers, for example, by helping coaches find time to meet with teachers.Based on research showing the importance of principals communicating a clear and ambitious vision of instruction to their teachers (Allensworth et al., 2006;Marks & Printy, 2003;Sebring & Bryk, 2000), including principals in the professional development also is intended to educate principals about the theory and research underlying effective reading comprehension instruction.The goal is to garner principals' support for the vision of instruction promoted by the coaches in their work with teachers, that is, that rigorous and interactive discussion about rich texts is key to improving students' reading achievement.The goal of CFC is to support a collaborative partnership between coaches and principals so that they work together to support instructional improvement.
District administrators are included in the training to create consensus among stakeholders in the vision for instructional improvement (see, for example, Aladjem & Borman, 2006) and support coaches and principals to work together as a team to plan for the professional needs of teachers at their school.District leaders also participate in the coach training in order to foster district-level understanding of effective reading comprehension instruction and responsibilities of a CFC coach and to encourage principals, through their supervisor's oversight, to support the work of their coach.
CFC coaches are expected to implement what they learn in their professional development with teachers at their school.With the support of their principal, CFC coaches are expected to work with fourth-and fifth-grade teachers in weekly grade-level teams to study the theory underlying dialogic reading instruction, and to plan Questioning the Author lessons.Coaches should also meet individually with teachers on a monthly basis to observe, model, or co-teach lessons as needed.

Study and School Context
Our study is situated within a three-year randomized controlled trial investigating the effectiveness of the CFC program for improving the quality of instruction and students' reading achievement in an urban district in the Southwestern United States.Twenty-nine Title 1 schools serving the lowest achieving students in the district were randomly assigned to receive a CFC coach (n=15) or to participate in the professional development (including literacy coaching) that was standard practice in the district (n=14) (Matsumura, et al., 2012;2013).Of the fifteen schools that implemented CFC, seven had the same principal for the three years of the studies.Because of our interest in the stability or change over time in principals' support for their CFC coaches' work with teachers, we focus on these seven schools for the present study.The seven schools are: Alder, Juniper, Laurel, Maple, Tulip Tree, White Pine, and Willow.
Approximately, 91% of the students in these schools at each study year were eligible for free or reduced price lunch; 80% of the students were Hispanic (15% African American), and approximately 40% of the sample was identified as English language learners.Teachers on average had approximately 10 years of teaching experience (SD=8 years, range=1-32 years), and about a third held master's degrees.Student and teacher characteristics did not differ across schools or study years.

Principals
Upon entry into the study, principals (N=7) had an average of 15 years experience as principals (range from 4 to 33 years), and had been serving as principals at their current school an average of 3 years (range from 1 to 7 years).All of the principals had a master's degree, and 1 held a doctorate.

Coaches
A total of 12 CFC coaches worked in our sample of 7 schools across the three years of the study.Coaches had an average of 2.2 years of literacy coaching experience (range from 1 to 7 years).Five of the coaches had a master's degree and one held a doctorate.Four coaches were hired to replace the coaches who left their school at the end of the first year of the study.The high rate of coach turnover between years 1 and 2 is due to the fact that when the CFC program was launched, four out of the seven coaches divided their time between two schools.In the second year, additional coaches were hired so that the coaches worked in a single school.At each year, half of the coaches' time was paid for by the district to work specifically as CFC coaches.The second half of their time was paid for by their principal out the school's professional development budget to perform other duties at the school (e.g., work with teachers in other grades, coordinate assessments, etc.).In the last year, a coach was hired to replace a coach who took a leave from work for personal reasons.Only two of the seven schools (Juniper and White Pine) had the same coach for all three years of the study.

Principal interviews
Each of the seven principals was interviewed at the end of each of the three years of the study using a structured protocol.In total then, there were 21 principal interviews.The interviews were conducted by the same member of the research team each year and were scheduled at a time and place that was convenient for the principals.The length of the interviews ranged from one to three hours.The protocol used for the principal interviews included (but were not limited to) questions related to principals' interpretation of the role of the CFC coach, their view of the success and importance of their CFC coaches' work with teachers, and particular actions they had taken to support the work of their literacy coach.For example, principals were asked to describe the duties they assigned and opportunities they offered to the CFC coach, how often they met with their coach, and what they discussed.Principals also were asked about the potential influence of state accountability targets on teaching and learning at the school, and more specifically, on coaches' work with teachers.

Coach interviews
Each coach was interviewed at the end of each of the three years of the study using a structured protocol.In total, there were 21 coach interviews.The interviews were conducted by the same member of the research team each year (who also conducted the principal interviews).Again, the length of the interviews ranged from one to three hours.The coach interview protocol focused on coaches' experience working with teachers and principals, and the factors related to principal leadership that supported or impeded their work.The questions coaches responded to included: How would you describe your role as a literacy coach at your school?What are your job responsibilities?How did the principal establish you in your role as coach?How, and in what ways, did your principal support your work with teachers?What, if anything, has the principal done to make your job harder than necessary?What was the focus of your work with teachers?Coaches were also asked about the potential influence of state standardized tests on instruction at the school.

Data Analysis
Principal and coach interviews were transcribed then analyzed, by a single researcher, using NVivo 9.2 (QSR International, 2011).The analysis involved multiple re-readings of the interviews and iterative coding (Miles & Huberman, 1994).Applied thematic analysis (Guest, MacQueen, & Namey, 2011), an inductive, content-driven approach to coding and identifying themes in qualitative data, was performed.First, broad codes (e.g., Principal Support, Coach's Duties) were applied to the major ideas contained within each response from each principal or coach to each question.Then, these codes and associated text passages were reexamined and compared, and more specific codes were generated (e.g., Principal Support for CFC, Principal Support of Coach).Following this, we completed further rounds of assigning codes to different ideas and refining codes as appropriate to characterize a particular action or idea (e.g., Principal's View of CFC as Insufficient for Meeting Accountability Targets, Coach Doing Work Related to Standardized Testing).Finally, we sought to identify themes that integrated sets of codes.The themes and codes generated and their definitions are presented in Table 1.The principal publicly endorses the coach as the expert, specifically in rigorous dialogic approach to literacy instruction.The principal may include the coach on the leadership team, actively encourage teachers to access the coach, give the coach opportunities to deliver in-service to staff related to CFC strategies.

Coach as Support for
Status Quo Literacy Instruction The principal may position the coach as an expert in literacy/ reading generally, but not in the CFC-advocated approach.The coach may be heavily involved in test-preparation type of literacy instruction.

Coach as Non-Expert
The principal gives the coach little public recognition and minimal public exposure in the school.The coach may be active or visible around the school, but is often engaged in non-CFC or non-literacy related activities.The coach's primary role as CFC coach and possessor of specialized knowledge is minimally recognized.The first theme, Principals' View of CFC as a Means to Achieve Literacy Goals, is concerned with what the principals believes the CFC can do for their teachers' teaching and student's literacy development.Specifically, it addresses how the principal conceives of the role of dialogic instruction as promoted by CFC in the context of accountability pressures and related mandates.Principals may be firmly committed to the idea that rigorous and interactive classroom text discussions based on authentic texts was the pathway to improving students' reading achievement.Alternatively, principals may only be considered a conditional proponent of CFC in that they might view CFC as valuable to a certain extent, but ultimately incompatible with or ineffective for helping students pass accountability targets.Finally, principals may be skeptical of the CFC as a necessary reform for their school altogether.
The second theme captures principal support in the form of their positioning of their coach in the school, among their faculty, as a source of literacy expertise.Strong and positive forms of principal support include introducing the coach to the faculty and actively encouraging teachers to work with the coach.Principals might also include the coach on a school-wide leadership team or arrange for the coach to lead an in-service session for the whole faculty.On the other hand, principals may neglect to endorse the coach as someone that possesses specialized knowledge; the principal may provide the coach exposure throughout the school, but not as an expert of rigorous instruction in literacy in the form intended by CFC; instead, coaches may be recruited to assist in improving the teaching of literacy as defined by standardized tests.Or principals may fail to provide opportunities for the coach to serve as a literacy expert for the school, or assign coaches duties unrelated to coaching and/or literacy.
Finally, we coded the nature of coach's work with teachers during the lead up to high-stakes testing season.The codes reflected whether work around high-quality dialogic instruction with authentic texts continued in its pure form or was adapted to support testing needs, whether it occurred in parallel with (i.e., not integrated with) standardized testing preparation, or whether CFC work was halted altogether in favor of intense and exclusive test preparation instruction.
After coding, using NVivo 9.2, we generated coding matrices (Miles & Huberman, 1994).The codes of interest were arrayed in rows.The columns reflected the three years of the study, and for each year, there were side-by-side columns for the principal's and the coach's interview.The cells of the matrix contained the text (or summary thereof) from the corresponding interviews that were coded to each node.Such a matrix effectively summarized the results of the coding for each school.It also helped to identify or assess potential changes over the three years.In addition, the matrix facilitated triangulation of data from the principal's and the coach's interviews.
To facilitate analysis, we reduced the data further.For each year, we examined each theme (e.g., Principal's View of CFC as Means to Achieve Literacy Goals, Principal's Positioning of Coach as Expert, and Coach's Work with Teachers During Test Season) and identified the node that best characterized that principal or the coach's work.For example, for a given year, principals were determined to be a Firm Supporter of CFC, a Conditional Proponent, or a Non-subscriber of the program.In several instances, applying the label/typology was straightforward, as text was coded to only one node.When text was coded to more than one node for a given theme, we considered both the number and the content of the passages coded to each node.Furthermore, in arriving at a typology, we considered both the principal's and the coach's interviews.More often than not, the two informants' views concurred.In instances where there were differences, we again examined the number and nature of the comments.(See Table 2 for a summary of the results.)

Results
As shown in Figure 1, our results suggest that principals' sensemaking of CFC was influenced by their understanding of dialogic instruction for improving reading outcomes and district policy messages regarding teachers' use of test-focused curricula and sanctions for failing to meet accountability targets.Specifically, according to the typology we have derived from data, four of the seven principals could be considered conditional proponents of CFC.Two other principals were firm supporters of CFC throughout the three years.One principal, a non-subscriber of CFC in year 1, renegotiated her view of CFC in year 2 only to return to the belief that CFC was largely unnecessary in Year 3. Principal's view of CFC, in turn, influenced their positioning of coaches as experts within their school.Principal that were firm supporters of CFC situated their coaches as sources of literacy expertise, particularly with regards to dialogic instruction, in their schools.At other schools, coaches were largely treated as support for status quo literacy instruction, which tended to be standardizedtest focused.In a few schools, the coaches were instructed to attend to administrative and other tasks unrelated to literacy improvement efforts.
Finally, our results suggest that variation in principal leadership -in terms of the vision for CFC and the action taken to position the coaches -was consequential to the school-level implementation of the program.On the whole, coaches' work with teachers focused on implementing dialogic reading instruction (Questioning the Author) in the first half of the school year (August through December) and again at the end of the school year (May onward); however, we saw significant variation in coaches' work with teachers from January through April -the months when the district expected schools to implement materials geared toward familiarizing students with the content and format of the state accountability test.In the schools with supportive principals, the coaches' work with teachers during these pivotal months (i.e., times of intense accountability pressure) continued to focus on dialogic teaching principles.In other schools, CFCrelated work was minimized or pre-empted altogether for test preparation.

Conditional Supporters of CFC: Promoting Parallel Visions of Reading Instruction
In the majority of schools (Alder, Maple, Tulip Tree, and Willow), principals expressed conditional support for dialogic teaching practices and CFC as the way to bring about reform goals for all three years of the study.These principals expressed the belief that while dialogic instruction was important for increasing students reading comprehension skills, it was in and of itself insufficient for meeting accountability targets.Dialogic instruction, and coaching to bring about dialogic instruction, would not get the 'dirty work' done of ensuring that their school's students would meet Adequate Yearly Progress goals.These principals endorsed parallel conflicting visions of instruction: One that emphasized dialogic instruction in the context of rich texts as a way to improve students' reading achievement, and the other that emphasized students spending their reading instruction time answering multiple choice questions about texts that were developed for the purpose of assessment (i.e., test-focused instruction).
For example, the coach at Alder made the following remark about her principal embracing the CFC initiative (in Year 1) until the principal sensed that different instructional practices were needed to help students pass the standardized test: We were doing really well until the results came back for the first round of the [state standardized] test…And then I saw something not typical for this principal, which was panic.And schedules were changed, things were more mandated than they've ever been before -a lot more structure, a lot more accountability…at the detriment of some of the things I could have done with a few of the teachers during that time.The principal at Tulip Tree (in Year 1) made statements conveying a similar view: …[We need to focus] on both immediate and long term goals for teachers and students.In terms of the long-term goals, developing effective readers, but also the short term or more immediate goal, in this case of having it be in alignment with the state accountability system.The fact is that they're taking the comprehension exams this next week and then again in April, so we have to be cognizant of how we can develop in the most global sense but also in a more specific sense, how are they going to be prepared for a particular format of testing.He went on to say, "…There are specific test-taking strategies that are not really addressed by this particular methodology [CFC]…I think [CFC] is useful for test preparation…It's necessary but not sufficient." As is evident, the principals' view of the CFC reform was to a large extent influenced by the state assessment and district mandates around the accountability test.The district levied significant sanctions against schools and principals that failed to meet accountability targets.Schools deemed Academically Unacceptable (AU) for two years in a row could be considered for closure.Principals whose school failed to meet accountability targets could be fired.Failing to meet accountability targets resulted in other problems as well, such as extreme and intrusive involvement of district personnel in schools (e.g., observing classrooms, taking over decision-making) that was demoralizing to teachers and principals alike.The principal at Maple recalled a "horrible" situation: …Our benchmark test didn't show that we were going to be successful, that we were going to be in trouble…That year we learned our lesson.It's like be careful with those benchmarks.They're very serious, and we took them seriously because it was like ok, we're going to come to the rooms.Where's your [testing] model?Where's your Thursday model?Where's your Friday model?…And so that made an impression on my teachers…You'd never want to be there again.So we're going to do whatever it took not to be under the microscope like that … Tulip Tree has felt the impact of district intervention too.For instance, because the school was assessed as Academically Unacceptable for one sub-population in fifth-grade Science in Year 3 of the study, there was, according to the coach, "an enormous change" for the entire school: What teachers had to teach became prescribed, and "weekly visitors" from the district office observed the classroom every week.Moreover, the district appointed an executive principal to supervise the school principal and a state education agency representative also visited frequently.
For the three principals that remained conditional proponents of CFC over the three years, while they saw CFC's dialogic instructional strategies (i.e., Questioning the Author), as important to developing strong literacy instruction and achieving student learning outcomes, they seemed deliberate about not over-investing in CFC because of the relative success the school has experienced in meeting accountability targets by engaging in test preparation activities.Related to this is a sense of distrust or at least uncertainty as to what a novel approach like CFC could contribute in the context of testing.The principal at Alder (in Year 2) expressed this sentiment: If you look at where in our MOY [middle-of-year] data were the least gains, it was at fourth and fifth grade…I am thinking it's because they are doing a lot of things that CFC is asking them to do with deeper comprehension strategies, but for some reason they are not showing up in progress on our [tests]…The big scary thing was that kids that did better on BOY [beginning-of-year assessments] went way down on MOY, like 20, 30 percentage points dropped…that was in fourth and fifth grade.The Alder coach's understanding of the principal's view aligned.She reported that "the administration is very bought into [test-taking] strategy and practice, separate from CFC.And there have even been statements made that CFC is taking people off track with [test] practice… The principal, that's her belief."The coach at Tulip Tree (Year 3) likewise observed that as the standardized testing dates drew near, "the fear sets in and we go back to what we know, which is doing intervention." We should note that these principals that are conditional proponents of CFC experience a very real internal conflict as to how to reconcile the two apparently different paths to achieving literacy goals.For example, the principal at Alder (in Year 2) said, "One of the things that [I] struggle with…is how to and when to emphasize test genre compared to the good reading strategies and comprehension things, that we are trying to improve comprehension in a more in-depth way…" Alder's coach also recognized the conflict, in saying, "[The principal] also wants the kids to be good strong readers and have a lifelong love of literature…[but] I think she wants them to love literature after March whatever [the testing date]."At Tulip Tree, the principal complied with the district directives to engage in test-strategies instruction and mock exams, yet expressed doubt that the preparing for state achievement tests can align with quality teaching: "I think that teaching reading and teaching the thinking of reading, and how to make those connections, is good teaching…" In all, despite some of the principals' inclination to support CFC and its dialogic instruction approach to improving literacy, the principals who were conditional supporters of CFC seemed unwilling or unable to become a true proponent.Policy messages from the district to adhere to test-preparation practices, along with the school accountability status, as inferred from benchmark tests and the like, factored heavily into the principals' sensemaking year after year, leading them to conclude that status quo should be upheld.

Principals' Positioning of Coaches as Experts
Principals who seemed to see the necessity of two separate and parallel types of literacy instructionone of dialogic instruction and one that is test-focused -were inclined to position their coaches throughout all three years as support for status quo instruction (i.e., generalists), which meant the coach might be heavily involved in test-preparation activities with teachers or directly with students, or provide the coach little public recognition and exposure altogether in the school.The coach may be active or visible around the school, but engaged often in non-CFC or non-literacy related activities, including administrative work.
Three of the four principals (Maple, Tulip Tree, and Willow) who are conditional proponents of CFC indeed failed to fully publicly endorse the coach's primary role as providing specialized support to teachers with regards to dialogic instruction, starting with not establishing the coach at the beginning of the year.At Tulip Tree, for example, the principal did not introduce the coach until about the fourth faculty meeting.At Willow, the coach claimed that the principal did not do much to establish her in her role at the school: "…Before I got here, I don't really know what he did.I mean the teachers knew I was coming, so he must have told them.But other than that, not really much." Also, the principals in the conditional proponent group did not provide opportunities for the coaches to be seen in leadership situations throughout the school.At Tulip Tree, for example, the principal did not seek to connect the coach with teachers beyond the designated 4 th and 5 th grades.And at Willow, the coach explicitly wished that the principal had allowed her to "do some professional development at staff meetings, even though it's supposed to be fourth and fifth [I'm working with]."She would have appreciated "introducing myself and explaining what I'm doing here, and the purpose of [CFC]." Coaches of principals that were conditional proponents were often expected to do tasks unrelated to CFC.For example, the coach at Willow was asked, among other things, to "study some different methods of teaching and their effectiveness…on test scores" and help identify and purchase leveled science readers, which were both beyond the scope of a CFC coach's work and tasks that did not access or help the staff recognize the coach's area of expertise.In general, the coach at Willow felt taken advantage of: The principal really just needs to understand that CFC has parameters because I think…any chance he gets, he tries to put things in that are not within that scope and I think…the main thing is just for him to really understand [that I am not] an all-around specialist…He doesn't want to understand it.I think he just sees that [teachers] have a lot to do, and they're stressed out and they think, oh, [the coach] is here today, she can do it."Likewise, at Maple, the coach was expected to provide support wherever the principal needed her: including tutoring students in math, taking care of honor roll certificates and assemblies, and administering the standardized test for special education.
In all, the principals at these schools used the coaches' time and expertise in a less than optimal or desired way.Instead of working to disseminate principles of dialogic instruction, coaches were expected to do work that directly helped the school meet its accountability targets using the status-quo methods of test-focused preparation.For example, at Tulip Tree, during the day as well as after school, the coach led intervention groups with students who were at risk of failing or had already failed the standardized test: "I…spent most of my time pulling and finding materials for [test] practice…for the intervention groups…for the after school tutoring, in running and copying and preparing and putting in teachers' boxes."In Year 3, when Tulip Tree was designated Academically Unacceptable (AU) in Science, the coach made a similar point, saying: A lot of my time and energy went into planning Saturday school intervention, planning after school interventions…so a lot of CFC things I would like to have done had to take a back seat to getting our school out of the AU label…[It's] difficult when your principal says you're doing this and you know that's not exactly how your time is supposed to be spent.Similarly, the coach at Willow spent half of her day coaching teachers and the other half working in direct contact with students in intervention groups.She also organized and taught Saturday [test preparation] camps.The coach at Maple endured the same experience, having to organize afterschool tutoring for test preparation.Willow's coach had approached her principal about the amount of duties unrelated to her CFC role: "I told him I needed to teach less.I needed to have less intervention groups.Because once these MOY scores kick in, they get so frantic about these lower kids that we're there teaching the groups, and then your coaching practice diminishes." Alder is an exception to the pattern that principals who are conditional proponents did little to position their coaches as experts with specialized knowledge about the dialogic approach to literacy instruction.In all three years, Alder's principal actively and publicly endorsed the coach.For example, she had the CFC coach run a staff training, "which said that I had a valid purpose here and I knew something …I think she wanted to put me out there and have people know who I was."Furthermore, the principal had the coach lead CFC professional learning community meetings, support vertical reading team meetings, deliver in-service on Questioning the Author and other CFC approaches to teachers of younger grades, which "really kind of helped…to make her more of a person that the whole faculty knew…" The principal said, "I think it's important for the entire faculty to see the big picture of the purpose of why we would be having a Content-Focused Coach…" Finally, in Year 2, the coach was responsible for being a point of contact for the reading specialist from the district.While the coach did some work directly with students on writing, she was not charged with administrative duties or other non-CFC-related work.The reason for Alder's departure from the expected pattern of minimally positioning the coach as expert is not entirely clear.Part of the explanation could be that unlike most other coaches, Alder's coach was fully funded by the study and only part time.Meanwhile, other coaches were paid part time by CFC and part time by school funds, so principals might have felt they could require the coach to do non-CFC or even non-literacy related activities.

Coaches' Work with Teachers
The principals' view of CFC as a means to achieve literacy goals appeared to influence how principals positioned coaches as sources of expertise to teachers.This in turn appeared consequential to the implementation of the program.Specifically, for the principals that promoted two parallel visions of instruction -dialogic and test-focused -the coach's work with teachers from December to April, when the district expected schools to engage in significant amount of test preparation activities, was such that CFC became rolled into test preparation activities.In other words, coaches approached teachers (and teachers approached coaches) with test-focused goals in mind.In certain years, particularly given a school's accountability status, CFC work appeared to have been halted altogether.As mentioned previously, often during this time, coaches themselves took on test-focused roles unrelated to working with teachers, such as tutoring students.
At Tulip Tree, in Year 1, for example, the coach admitted that to reach teachers, she "was selling [CFC] as…we can practice our Questioning the Author, get some good discussion, read some actually interesting stuff, and still be practicing our [state test] stuff…You can use these four [test preparation] questions for your assessment."Interestingly while in Year 1, the coach said that the test did have an influence at Tulip Tree, but not as big as in other schools, in Year 2, the coach noticed that from January onward, she did not have the opportunity to plan together with teachers because of the state standardized test: When January hit…I think a lot of [CFC] work kind of got derailed because [the test] trumps everything.Once the teachers are in intervention mode, and the school district sends us to training about how we're supposed to spend our time, and these are your directives, and this is…what language arts is going to look like, that doesn't lend itself to do this CFC work, to continue.In fact, the coach's work during this time revolved around "copying the stories out of the [test preparation] book and the questions that the district wrote, and those in and out was the reading plan.My [CFC] materials, everything was getting ready for [the test]."This trend continued.In Year 3, when the school was labeled Academically Unacceptable, the coach was asked how she married CFC with the prescribed test strategy curriculum.She replied, "There wasn't much because we were told to read [test preparation] passages."Also, "I didn't really meet with [teachers] much…because I was actually a teacher at that time," directly helping students through intervention groups.
At Willow, the coach's work with teachers was also halted.In Year 1, the principal reported: At certain points in the testing cycle…teachers really don't want to do reflective lesson study…We tend to go through a review of all of the objectives and do test taking strategies…Once those fundamental needs are met…then they're more willing to do the professional development that's related to lesson study with this model.Again, this sidelining of CFC continued in other years at Willow.The coach reported that once the MOY benchmark data rolled in, "principals go into panic mode and…want to see [test] strategies being taught in all the rooms.So the shift moved away from what we were trying to initially establish with Questioning the Author to the [district] method."In essence, once test season hits, "everything gets dominated by that." Similarly, at Maple, the principal revealed that during the gear up to test time, the coach abandoned CFC work: "Our coach knows [the state test] and knows how to get those babies to pass.So [the teachers] are going to be looking to her to help them help the kids pass.Not to help them do…the CFC."During this time, the teachers are willing to work with the coach, but only regarding test-taking strategies.Given this principal's (and other similar principals') positioning of the coach not as an expert of dialogic instruction, but more as a supporter of status-quo test-focused practices, this distortion of the CFC coach's role and the resulting infidelity of implementation are not surprising.
Finally, even at Alder, with the positioning of the coach as expert, it was difficult for the coach to accomplish CFC work during test-preparation period.In Year 2, the coach said, the CFC work "waxes and wanes.That's the truth of the matter…One of the things we did was started writing Questioning the Author-type lessons with [test] passages.I mean it's gotten to that point."More pointedly, in Year 3, the coach described that "the perception [at this school] is that CFC is this stuff in lala land, and [test preparation] is the stuff we have to do."To negotiate this, she "had to give up some territory.I had to be willing to be in it with the [standardized test] stuff too…I sort of negotiated a truce because it was clear to me that I wasn't going to change that system."

Firm Supporters of CFC: Promoting Dialogic Instruction
Only two principals (Juniper and Laurel) saw dialogic instruction and coaching as the 'answer' to improving students' reading achievement and meeting accountability targets.In contrast to others, the principals at Juniper and Laurel held fast to a pure vision of CFC for all three years.They believed that the forms of instruction promoted by CFC were necessary and sufficient for improving students' reading skills and performance on the state's accountability test.
There are great similarities between the sentiments of the principals at Juniper and Laurel.To begin, both principals saw a distinction between teaching to the standards versus teaching to the test, and were clear that instruction should not be test-focused.Laurel's principal said, We teach to what the kids need to know…You want kids to be critical thinkers.You want them to be problem solvers…I don't want them thinking of just a test.In Year 1, Laurel's principal expressed dissatisfaction with teaching to the test, saying that "The test is so limited -it doesn't fully gauge what a child can do…If I was teaching appropriately, I wouldn't care what you threw at my kids."Remarkably, and in contrast to the situation at Tulip Tree, where the principal reverted to test preparation in the face of district intervention, in Year 3, when Laurel was subject to district monitoring and prescriptions, the principal reiterated his view from Year 1, saying: My experiences as a teacher, as an educator, as administrator…if you're teaching good quality teaching, it doesn't matter what you throw at the kids… So it's not radically taking something out and you teach to a test.It's really going back to the good practices.At Juniper, the principal tried to help teachers understand that authentic literature can be used to help students grasp main idea and summarizing: "They don't have to do an isolated [test-like] passage."More notably, in great contrast to other principals, Juniper's principal intentionally did not institute test-focused after-school tutoring or other interventions, even at teachers' urging.She believed that, "Once I agree to that, that impacts their planning.And I believe that well planned lessons and core lessons will get more mileage than any intervention group or after school tutoring." In advocating quality instruction and backing away from test-preparation activities, the principals consciously disregarded some aspects of district policy.For example, when asked about how he was meeting the district's requirement that teachers prepare students for the state test, the principal at Laurel explained that while they administered Friday assessments as expected, they in fact moved further away from intense test preparation: "Where before it was let's take the hour, hour and a half to test you, [now] you don't need [that].The assessment should be all week as you're doing this [skill], as you're listening to kids read, as you're watching what they're doing…That's the check off.Those are your notes that you're taking."Asked whether it was acceptable to the district that the teachers did assessments that way, the principal replied, "It is ok by me," signaling that he has actively mediated or interpreted the accountability pressure for the teachers.Similarly, the coach at Juniper reported (in Year 1) that her principal essentially downplayed, even subverted the testpreparation directive from the district: [She] does not like the trainings that the district offers [on test preparation] because it's so much sort of gimmicky stuff that distracts from detracts from the process of reading and writing.[The district has the kids] do all kinds of bizarre [test-preparation] stuff that when the Principal sees that…she just tells the teacher to stop it.That crap is not helping your kids…[She calls that stuff] trash…She says it's awful.Stop doing it.In her own words, Juniper's principal said, "Matter of fact, I avoid buying a lot of [that] test prep stuff.I really hate it."In subsequent years, despite continued district pressure, the principal stayed on course in her dismissal of test preparation-focused instruction.In year 3, for example, the district purchased for the 4 th -grade teachers a set of test-preparation booklets and gave them a schedule that paced students to work through the books by the testing date.At the principal's (and coach's) recommendation, the materials were not implemented.According to the coach, the principal considered it "worth the risk" to disregard the district's prescribed program, even though the repercussions could greatly and adversely affect her career: She considers it worth the risk.The risk is if you don't do these things and your kids score badly, the district will say you should have done what we told you to do.If on the other hand your kids do really well, the district doesn't have anything to say to you.What can they say?Finally, the principals at Laurel and Juniper both perceived strengths in CFC's method and saw it as an alternative to the test-preparation approach to improving literacy skills.In Year 1, Laurel's principal expressed a preference for an instructional approach that did not teach skills in isolation, which is what he perceived test-taking strategies instruction was good for.Instead, he emphasized the need to "tie pieces together".In Year 2, he grew to view CFC as having the potential to tie parts of instruction together not only within a grade, but vertically throughout the school.In the interview, he discussed the need for Questioning the Author and other CFC-endorsed practices to be taken up in all grades at the school and to become "the way we do business….aneveryday routine [because] they are just good concepts."Similarly, Juniper's principal believed that CFC is "some of the best stuff we've ever done" and that teaching students higher-order comprehension skills, such as synthesizing, drawing inferences, making connections outside the text and across texts are essential to reading reform and should be taught more widely than in 4 th and 5 th grades.To fulfill this goal, she also sought to expand the reach of CFC.In fact, she instituted Questioning the Author strategies campus-wide in Year 2. For both Juniper's and Laurel's principals, for all three years, CFC was considered integral.

Principals' Positioning of Coaches as Experts
Principals that were unwavering in their support of CFC and their belief that dialogic reading instruction would achieve their school's literacy goals also clearly endorsed their coaches as sources of literacy expertise.They elevated the status of the coaches by signaling to teachers that the coaches possessed specialized knowledge that no one else in the school possessed.This was accomplished by including coaches in leadership teams, having coaches deliver in-service to faculty members, and not requiring that coaches do substantial (or any) work related to teaching to the accountability test or performing administrative duties.
A notable exception to the description above was Laurel in the first year of the project.According to the coach, the principal did introduce her to teachers saying, "She's here to support you"; however, this was done "way late" in the year.Furthermore, the coach reported feeling "left out of the loop when it came to professional development" because other instructional coaches at the school had a facilitating role, but she was not involved.In subsequent years, however, Laurel's principal significantly and actively took steps to position the CFC coach as a literacy and professional development expert.
In contrast, at Juniper the principal endorsed the CFC coach as expert and leader at each year of the project, even though the coach was only at the school half time.The principal set aside part of the time during a campus-wide faculty meeting time to formally introduce the CFC coach.She also insisted that by way of introduction, the teachers have a chance to see what coaching looked like.The result was a fishbowl-format presentation in which the CFC coach demonstrated a coaching session.This helped the teachers recognize the coach and the specific nature of her work.Throughout the year, the principal continued to put the coach in front of the whole campus.
On other efforts to position the coach, Juniper's and Laurel's principals were very similar.For example, Juniper's principal positioned the CFC coach on her leadership team.She attended leadership trainings and met with team leaders all day once a month.At one point, in recognizing that teachers needed support with lesson planning, Juniper's principal had the coach teach team leaders about lesson planning, using CFC materials.The coach subsequently led learning walks, asking teachers to look for evidence of a planned lesson.Likewise at Laurel, the principal sent "the message…that [the coach's] work is important" through trusting her to be a leader of professional development, to deliver training to teacher teams.The principal even instituted whole-campus afterschool planning sessions and specifically encouraged teachers to access the CFC and other coaches during this time.Moreover, in the face of district intervention in Year 3, Laurel's principal did not sideline the CFC coach; instead, he chose to have his coach liaise between the district's representative and his teachers.That is, the coach would meet with the teachers to learn what they wanted to accomplish, then work with the district coordinator to plan lessons that she then brought back to the teachers.Additionally, the principals actively invoked the coach's expertise in discussions with teachers.Juniper's principal would say, "I've been talking with [the coach] about this, and I think this is really good for our kids, and I want [her] to talk to you…" Or, she would reroute a teacher's question to the coach, saying, "She's the best person to ask that question to."In Year 2, when the CFC coach at Juniper was only being paid as the CFC coach half time, the principal ensured that the other half of her time was spent on similar work that made use of her CFC expertise.All these deliberate moves indicated that the two principals firmly regarded the coach as an expert and greatly desired that the faculty at large did too.
Finally, both Juniper's principal and Laurel's (in Years 2 and 3) also led by example and by positioning themselves as learners.When teachers tried to be excused from doing a book study with the coach, Juniper's principal "swooped in", declaring that she herself would participate in the book study.According to the coach, "Anytime a teacher says, 'I don't have time for this," she says, "Hey, I got two kids.I'm the principal…I'm still making time to meet with [the coach]."In doing so, the principal firmly established that the knowledge and expertise the coach can impart is worthwhile.More poignantly, according to Juniper's coach, her principal has stood up "in front of the whole faculty and said not a day goes by that I'm not asking my coaches for help and you should be too."Similarly, at Laurel, the principal recognized his own lack of expertise in reading and language arts and because of this, trusts the coach and accords her the freedom to be the reading expert on campus.In all, the two principals' efforts to position the coach as an expert are exemplary.

Coaches' Work with Teachers
At Juniper, throughout all three years, the coach's work with teachers even during the months of December to April -the time the schools are meant to invest significantly in test preparation, according to the district -continued to focus on CFC-endorsed activities.In fact, in order to explicitly and directly counter the inclination for coaching to fall by the wayside during this time, in Year 1, the principal intentionally assigned teachers a professional learning task.She explains: I gave them task 6…so…[the coach could] get with the teachers so they don't start getting, "Leave me alone.I want to do just [test preparation]."[The coach] is going to use Questioning the Author strategies to prepare new activity for the staff, so that it's research-based, they get to try it and they get to demonstrate it for the entire faculty.So that's about the best way we can keep the teachers engaged…at this point in time.As for the subsequent years, the coach recalls: We…really de-emphasized…the debugging strategies and the [test] model strategies that the district has been pushing…We did not follow the district's request [to spend time on test model strategies].We really tried to get into rigorous high quality reading instruction, and did not spend a lot of time practicing the test.CFC continued in its pure form.In fact, in both years, the binders full of test-preparation resources from the district went largely unused.The coach delivered the materials to the teachers, but "made it clear [she] was not going to do that…This school is not going to go down that path other schools are".
At Laurel, the other school with a principal that fully supported CFC, the implementation, as in the coach's work with the teachers during testing season, was less than exemplary.Recall in Year 1, while the principal seemed supportive of CFC, he had failed to introduce and actively position the coach as an expert in dialogic instruction.That year, the coach's work with the teachers focused on the district-mandated skim-and-scan strategy of reading and test taking.In Year 2, however, the coach folded test-related goals into CFC while working with teachers.In her words: I try to make a connection [between test preparation and CFC] by saying, while you are reading the text, I am going to teach them to stop along the way and ask questions, so therefore they are having some sort of interaction with the text.They're learning how to stop and think about what they are reading and learning a little.Laurel's coach continued discussing Questioning the Author with teachers, though the conversations often revolved around preparing students for the state assessment.In Year 3, Laurels' coach was "trying to push the independent reading and getting [students] excited for some read aloud and things that the teachers are doing…We give these kids a lot of practice with what we see is comprehension, but it's really like how to break down a piece of text and ask themselves questions," which is reflective of the dialogic approach advocated by CFC.Although the coach's work at Laurel does not quite exemplify CFC being continued in its pure form, we can argue that progression toward a CFC focus was clear over the course of the three years, and a condition that made this possible might be the principal's steady support for CFC, which over time, increasingly emboldened the coach and the teachers to follow the principal's vision and eschew policy messages related to test preparation activities.

Shifting Views of CFC and Dialogic Instruction
While all the other principals' views can be characterized by stability, White Pine's principal had a fluid interpretation of dialogic instruction; that is, her view of CFC as a means to achieve the school's literacy goals shifted notably from year to year in order to align with the district's policy message and her school's accountability status.Specifically, in Years 1 and 3, the principal at White Pine could be classified as a non-subscriber of CFC.In Year 1, she saw minimal need for CFC because math and science were the subjects where her school was most in danger of not meeting accountability targets (i.e., showing adequate yearly progress).This principal's attitude toward CFC is also betrayed by her attitude of mere compliance toward the CFC intervention.Her coach relates (in Year 1) that it is clear CFC "isn't the principal's agenda".After CFC training sessions, for example, instead of saying, "Let's partner on what to work on.Let's figure out how to take this happen in the school," the principal says, "This is a requirement that we have to take care of so we'll take care of that."The principal herself reiterated that she had not wanted her school to participate in the study, saying "I would have rather been included in a grant where my focus was in science and mathematics."And although CFC has brought some useful techniques such as better questioning techniques and lesson design, she had not seen those aspects as weaknesses prior to CFC.In other words, White Pine's principal did not perceive the CFC program as necessary in helping her achieve the school's reform goals.
The principal at White Pine, however, did soften her view of the value of CFC in Year 2, and could be considered a conditional proponent.Her coach recognized a "huge" attitude shift in the principal: "All of a sudden, she wants all of the coaches to coach this way…."This shift coincided with the principals' observation of a powerful coaching demonstration.It also appeared to have been influenced by actions on the part of the district that could be interpreted as support for the CFC.Specifically, district coordinators met with CFC professional developers to align district initiatives related to test preparation with the CFC initiative.The results were shared with the principals at a CFC session.It is not too surprising that White Pine's principal would be more apt to subscribe to CFC after this since she appears to be heavily influenced by policy messages from the district.Finally, as the coach observed, "we weren't a focus school [school with low assessment scores that required additional monitoring]… There weren't mandates on us…" The school's safe status likely influenced the principal to give CFC a chance.
By the same token, in Year 3, the principal retracted her endorsement of CFC.Significantly, the district hired a new principal supervisor for the schools in her geographic area.She was peripherally involved in the project but had made it very clear to principals, including White Pine's, that she did not support CFC.This resulted in a major disconnect between CFC and the testfocused curricula and practices promoted by the district.As the coach noted, "[If you were the principal], with that person being your supervisor, you just start to wonder how much do you push [CFC]"?In the meantime, the principal also felt added stress from the district directly related to standardized testing: We got the directive from our superintendent this year that there would be no academically unacceptable campuses…If you're a principal at a campus that goes AU two year in a row, you will not have a job anymore…It's extremely stressful.And you try as a principal to not bring that stress to your teachers.Moreover, after the MOY data, the principal was called in to meet with the district and was essentially told among other things, that "students should not be …reading novels between [then] and [testing time], not even independently," and that reading needed to be pushed aside for science instruction to improve the school's score in science.As a result of all this, in Year 3, White Pine's principal complied wholly with the district's directives to prioritize test preparation, as well as the district's preferred method for doing so -exposing students to repetitive test-like models.Indeed when asked in the interview how she dealt with the conflict between test practice preparation and the CFC model of teaching for reading comprehension, the principal replied, "Right now, I'm doing exactly what I'm told because I'm directed to do that.So if it's not aligned with CFC, it's not aligned with CFC…I have to do it."She also said of CFC, "I question it right now because our scores aren't all that great…", directly tying the effect of accountability to her belief about the CFC initiative.
White Pine's principal also rationalized not pushing for CFC because the situation in Year 3 at her school was such that too many demands were being placed on her teachers.The staff at the school seemed pulled in too many directions in terms of professional development, so much that when asked what the learning needs of her teachers are, the principal replied, "They don't need to learn anything.They need to stop learning so much and focus in on what really matters because we're losing teachers everyday to burn out."So in not subscribing to CFC, she was enacting a role as gatekeeper of sorts; she perceived herself as protecting her teachers from outside influences unnecessarily helping them to do a job they already knew how to do well.

Principals' Positioning of Coaches as Experts
The pattern of actions that White Pine's principal took to position the CFC coach in the school reflected her shifting view of CFC as a means to achieve literacy goals.Specifically, in Year 1, the principal did little to introduce or position the coach as a source of expertise in literacy at large, let alone CFC's vision of dialogic instruction.As the coach related, the principal "wants coaches.She wants three or four or five coaches if she could get them, but she doesn't want them to coach…[They're doing] administrative stuff."At best, the principal did not regard the CFC coach as possessing specialized knowledge.In fact, she situated the CFC coach as just another coach, so it was acceptable for teachers to not work with the CFC coach if they were accessing other coaches: "My other coaches do the exact same thing." In numerous other remarks, White Pine's principal downplayed the CFC coach as a potential expert and instead conveyed the belief that capacity existed within her staff to address the school's goals.For example, when asked whether the coach was meeting the learning needs of the teacher, she replied, "A great amount of discipline was going on prior to her getting here, so she's not bringing anything new…" She disagreed that the CFC coach had affected changes in teacher practice, rather attributing any changes to "the staff development we've been doing… prior to her getting here...So she's another cheerleader if you will, for best practices." As noted, in Year 2, for a number of possible reasons, the principal adopted a favorable view of CFC.Accordingly, she took actions to position her coach more prominently as an expert in the school.The coach herself perceived the "huge shift".For instance, when teachers approached the principal asking the CFC (and other coaches) to make photocopies, the principal would say, "No, they are busy.That is not their job.Bring them to me.I will make them."She also had the coach work with the third-grade teachers' learning group on lessons based on the Questioning the Author approach.Moreover, she set up teachers that needed support to work with the coach.In all of these ways, the principal adequately positioned the coach as an expert.
In spite of the principal's plans to launch the following year with a presentation to the staff of what CFC entailed, what the coach's role was, and how it would be a model for the campus, in Year 3, when she withdrew her support of CFC, she returned to a view of CFC coach as largely unnecessary, saying, "I think…we could have gotten the same learning without the help of [CFC]…I think we had the knowledge and skills within our own district to learn that."In Year 3, the CFC coach clearly felt that her schedule had changed.It was no longer "protected".There had been no faculty learning groups since the beginning of the year; instead, the coach, along with other instructional coaches, were required to conduct interventions to help students pass the standardized test.

Coaches' Work with Teachers
Predictably, at White Pine, given the principal's lack of support for CFC and for minimal effort to position the coach (at least in Years 1 and 3), the coaches' work with teachers essentially ceased altogether in favor of intensive and almost exclusive test-preparation activities.The coach said trying to work with the teachers on CFC-advocated strategies techniques was difficult: "It would not be difficult if it wasn't this time of year, but I mean, we have been testing all week long.And prep for that, you know it's creepy."During the months of December to April, White Pine's coach made minimal contact with teachers.In fact, she was occupied with running interventions, as with all other teachers at the school, to help students pass the assessment.Then when the standardized test was over, as the coach related, "You will be amazed at the attitude shift on campus.[Teachers] can breathe again.They can do things…" Clearly the principal's view of CFC as unessential for meeting the school's accountability goals adversely affected the implementation of the CFC coaching initiative at White Pine.

Discussion
Before discussing the implications of our study, we first want to draw attention to some of the limitations of our study that should be considered when interpreting our results.Our study was conducted in a very small number of schools, all of which were all drawn from a homogenous sample of Title 1 schools that served the poorest and lowest achieving students in the district.Because these schools often struggled to meet accountability targets, the pressure felt by the principals in our study might be greater than in other schools.We emphasize, however, that we do not intend our study results to be generalized to a broader population.Rather our goal is to use our data to contribute to emerging theories regarding the contexts that influence the implementation of coaching programs that may form the basis for future investigation.Future research would need to be conducted in larger and more diverse samples of schools to determine if the pattern of results we identified in our study holds true in other education contexts.
In addition, there were many factors that we did not measure that could affect principals' sensemaking of CFC and the implementation of the coaching program.For example, we did not directly measure the quality of coaches' interactions with teachers though this could conceivably influence principals' perceptions of their coaches' expertise and subsequent endorsement of their coach to teachers.We also did not assess the affective quality of the coach principal relationship in our study, though this too could have influenced principals' stance toward the program.Finally, we did not directly examine teachers' own sensemaking of CFC, though this again could have influenced principals' perceptions of the importance and relevance of their coaches' work.

Study Contributions
With those caveats in mind, our findings suggest that principals' negotiating of conflicting visions of reform teaching and social positioning of coaching are potentially important contextual factors that influence the implementation of coaching programs.Past research indicates that principals mediate the effect of reforms on teaching through their decisions about what, among conflicting policy messages, to promote to teachers and through the messages they send to teachers about how to implement new reading reforms (Coburn, 2005;Spillane et al., 2002b).Our findings regarding the influence of principals' sensemaking on teachers' professional development opportunities (i.e., coaching experiences) contributes a piece of the puzzle for understanding the mechanisms by which principals influence the success of instructional reforms (i.e., the extent to which teachers have the opportunity to learn how to implement new forms of instruction).
Our study also contributes to a broader understanding of the contextual factors that might impact teachers' school-based learning opportunities both within and across implementation years.As described earlier, the preponderance of research on teachers' professional development, and literacy coaching in specific, has been limited to cross-sectional designs.Our past research showed that how principals positioned coaches within the school's social culture (e.g., the actions principals took to signal the expertise of the coach and importance of coaching to teachers) was critical to launching the program in the first program year.Our current study extends our past research to show that principals continued active support for their coach influenced the program's implementation across all three years of the study.Our results also show that principals' views of, and support for the program, was mostly consistent across years with positive and negative consequences for coaches' work with teachers.

Designing and Implementing Coaching Programs
As described earlier, CFC is a complex multi-level intervention that emphasizes the importance of principals and district administrators in creating the organizational conditions in schools that support professional learning and instructional improvement in addition to also providing extensive professional development to coaches (Staub & Bickel, 2003).This emphasis on creating a 'learning organization' (Glennan & Resnick, 2004) by including principals and district leaders in extensive professional development with their coaches is unique among coaching programs.Specifically, the program aims to increase principals' content knowledge of effective teaching reading instruction, coaching practice, and the roles and responsibilities of literacy coaches (i.e., that they should work with teachers not perform administrative tasks or tutor students).The IFL designers of CFC hypothesize, based on other research, that increasing principals' subject matter content knowledge and best practices for instruction, as well as understanding of what good coaching should look like, would induce leadership support for coaches and their work with teachers.Our results suggest, however, that knowledge of best practices for reading instruction and coaching was insufficient for garnering principals' full support of the program and instructional practices promoted by coaches.Despite the fact that principals participated in a great deal of professional development (one day a month for three years), most principals were conditional in their support of CFC and the dialogic practices promoted by coaches.Moreover, as noted earlier, principals' stance with regard to CFC was mostly stable over time; that is, did not change as a function of increased exposure to the program over time, suggesting that increased professional development was not the answer to garnering principal support for the program.
Including district administrators in the CFC professional development also was intended to foster district support for CFC and in so doing, reinforce the importance of the program for principals.Many of the principals, however, appeared to view their support of CFC as a Hobson's choice -a perspective that was reinforced by the district rather than alleviated.On one hand, many of the principals expressed the belief that dialogic reading practices were important for developing students' reading skills.In this regard, CFC may have succeeded in their goal to increase principals' knowledge of high-quality reading instruction.Principals also were encouraged to support CFC by district administrators who attended the professional development.On the other hand, failure to meet accountability targets put principals' jobs in jeopardy and created conditions in schools that were difficult and demoralizing for teachers (e.g., frequent visits in teachers' classrooms by district administrators).Many principals held the view that test-focused reading instruction (reading of testlike passages and answering multiple-choice questions) would increase students' scores on the state accountability test.This vision of instruction also was reinforced by the district mandate that schools implement a test-focused curriculum after the winter holiday.District support for CFC, therefore, was complex and differed depending on the district administrators' responsibilities -if they were in charge of developing professional development or enforcing accountability policies.Garnering district support for CFC on the part of one constituency was not sufficient for creating the organizational conditions that support canonical implementation of the program.
Notably, lack of alignment (coherence) in policy messages communicated to teachers and school leaders is not a new problem (e.g., Russell & Bray, 2013), and national accountability policies requiring subgroups of students to show test score gains (NCLB) is not likely to end soon.While new state assessments aligned with the Common Core State Standards might communicate a vision of instruction that is more aligned with dialogic practices promoted by instructional reforms than traditional tests, it is notable that past research on state performance assessments have generally not shown positive effects on teaching (Firestone, Mayrowetz, & Fairman, 1998) or student learning (Koretz & Barron, 1999).In short, we believe that it is unrealistic to assume that the bedeviling influence of accountability policies on the implementation of instructional (coaching) reforms will disappear any time soon or be resolved by higher-quality state assessments.
Given these problems, we propose two recommendations for creating the organizational contexts in schools that might improve the implementation of coaching programs.First, and most importantly, we suggest that districts and/or program designers assess principals' readiness for change in advance of committing coaching resources to achieve particular instructional goals (see Matsumura et al., 2012).Assessing readiness in advance of implementing an intervention is common practice in the social sciences (Fixsen et al., 2005), but not in education reform where reforms often are top-down mandates -that is, are required for eligible schools by federal, state or district policymakers.Readiness for change emphasizes (among other things) the importance of full buy-in from stakeholders and a clear understanding on the part of stakeholders for how the interventions fits within the community's priorities for change.Given the wide range of reform programs and policies promoted to principals and the breadth of students' learning needs, it is likely that principals will not always hold improving reading comprehension as their primary reform goal or be willing to devote time to working with coaches to ensure teacher buy-in and participation.Notably, principals' readiness for the CFC was not assessed prior to their participation in our randomized cluster trial as we sought to assess the program's benefits in the most rigorous design possible by minimizing bias in our sample.Moving forward, however, we suggest that the reform priorities and time to participate in the intervention be an a priori consideration before attempting to implement a coaching program.
Second, we recommend that, as part of assessing readiness for a coaching program that principals and districts commit to limiting test-focused instruction to only a few weeks preceding the state's accountability test.Research on test preparation at the high school level indicates that a little timed test practice is beneficial to students, but that the benefit is small and diminishes quickly once students become familiar with a test's format and procedures (Allensworth et al., 2008).We suggest, therefore, that schools engage in some, but a very limited, amount of test preparation (i.e., test practice) and that coaches plan to suspend their regular coaching work during that period to fully support their principal and teachers during the testing period.We recognize that this recommendation might be very difficult to implement, as it would require support from multiple personnel in districts.This challenge must be balanced, however, against the (likely) possibility that multiple and conflicting visions of instruction communicated by districts to principals, and principals to coaches and teachers, will undermine the implementation of instructional reforms by reducing teachers' opportunities to learn the instructional practices they need to advance students' reading skills.

Figure 1 .
Figure 1.Relationship of principal sensemaking to the implementation of CFC

Table 1
Non-subscriber of CFC) literacy goals of the school.It does not supersede district directives to engage in test preparation.2. Principal's Positioning of Coach as Expert 2.1 Coach as Expert of Rigorous Dialogic Literacy Instruction The instructional approach advocated by CFC has value, but is incapable of or insufficient for meeting literacy goals as defined by accountability mandates.Separate standardized-testing focused instruction is needed in addition to dialogic instruction.1.3LargelyUnnecessary (Principal is aThe instructional approach advocated by CFC is largely unnecessary.CFC cannot address the accountability mandates and

Table 2
Summary of Typologies Derived for Each Coded Theme By Principal Over Time