Home Articles Reader Opinion Editorial Book Reviews Discussion Writers Guide About TCRecord
transparent 13

School Turnaround Through Scaffolded Craftsmanship

by Charles L. Thompson, Gary Henry & Courtney Preston - 2016

Between 2006 and 2010, the North Carolina Department of Public Instruction intervened in 128 low-performing schools, combining approaches consistent with school restructuring and transformation. In improved schools, local educators reconstructed key school functions, a distinctly nonlinear process more like the work of skilled craftsmen than that of design engineers that we refer to as “scaffolded craftsmanship.” We interviewed key stakeholders in 12 high schools to learn about the dynamics accounting for the improvement or stalemate at each school. In sum, in the improved schools we studied, the turnaround process was not a matter of initial external design and subsequent implementation, but a non-linear process of planning, inventing, adjusting, and re-planning as well as a process of learning, doing, and learning from doing. The improvement generally began with the installation of new leadership and involved four main components: new commitment, climate, and culture; improved knowledge and skills; strategically organized and managed structures and supports for instruction; and strengthened external support. Our findings suggest that judicious personnel replacement followed by professional development and coaching targeted to key functions may be a more effective method for implementing school turnaround than the structural approaches promoted via NCLB sanctions and Race to the Top.

In response to increasingly urgent calls to “turn around” America’s low-performing schools, policy makers and researchers have proposed a variety of approaches to the problem, emphasizing either “transformation” or “restructuring.” Transformation refers to improvements in organizational and instructional practice within the context of schools as currently governed and staffed, largely through professional development and technical assistance. In its current usage, restructuring refers to changes in school governance and personnel replacement, including closing a school entirely and dispersing students to other schools, implementing state or district takeover, restarting a school as a charter school, or replacing the principal and some or all of the teachers in a school.

Combining approaches consistent with both restructuring and transformation, between 2006 and 2010 the North Carolina Department of Public Instruction (NCDPI) intervened in 66 low-performing high schools, 37 middle schools, and 25 elementary schools (n.d.). Local districts replaced principals and substantial numbers of teachers, but equally important were professional development and sustained school- and classroom-level coaching keyed to an NCDPI-mandated planning framework. In schools that improved, the combination of elements supported a process we refer to as “scaffolded craftsmanship” (Thompson, Brown, Townsend, Henry, & Fortner, 2011). Together, the planning framework, professional development, and coaching scaffolded a distinctly non-linear process through which local educators reconstructed key school functions, a process more like the work of skilled craftsmen and their apprentices than that of design engineers and installers.

On average, in high schools assisted by the NCDPI’s Turnaround Schools program performance composites (the percentage of students passing state tests out of tests taken) improved by about 11 percentage points over the four year period, while the 67 next-higher-performing, un-served high schools improved by an average of about three percentage points (Thompson et al., 2011). Comparison based on a broader index of improvement, gains in learning, adjusted for school and student characteristics, reveals smaller gains for the treated schools but supports the efficacy of NCDPI interventions (Thompson et al., 2011).

Although on average the treated schools outgained similar untreated schools, the range of improvement among treated schools was wide. Through a study of 30 strategically sampled schools, 10 that improved sharply, 10 that improved moderately, and 10 that made little progress or even worsened, we sought to isolate the factors that promoted improvement in some schools and those that frustrated change in others (Thompson et al., 2011). Here, we focus on the 12 high schools we studied. The balance of this chapter describes the main features of the North Carolina Turnaround Schools program, outlines our research methods, and details the process of scaffolded craftsmanship in improved schools, and concludes by comparing our findings with prior research. In improved schools, the turnaround process echoes some of the main findings from prior research on school turnaround, with some significant additions. More importantly, the NCDPI’s scaffolding enabled the state to implement the turnaround process on a statewide scale.


NCDPI initiated its program to turn around low-achieving schools in 2005 in response to judicial and gubernatorial mandates (Fiscal Research Division, 2007; SERVE Center, Friday Institute, & Carolina Institute for Public Policy, 2010). The judge in the long-running Leandro v. State of North Carolina school finance suit, Judge Howard Manning, Jr., held that North Carolina’s constitution obligates the state to provide every child a “sound basic education.” Accordingly, he determined that whether students were making adequate progress toward a sound basic education could be measured by whether they achieved proficiency on the state’s End-of-Grade and End-of-Course tests. By Judge Manning’s standards, a high school where fewer than 55% (later raised to 60%) of its students achieved proficiency should be closed unless urgent steps were taken to turn it around. Thus, in 2006, he ordered the NCDPI to assess such schools to determine why they were performing so poorly and how they could be improved.

In response, during 2005–2006, NCDPI sent assessment teams to 35 high schools whose performance composites had fallen below 60% for two consecutive years. From Summer 2006 through Spring 2007, UNC-Chapel Hill’s Kenan-Flagler Business School collaborated with the UNC system-based Principal’s Executive Program (PEP) to provide a 24-day program of professional development to leadership teams for about half of the 35 high schools. Kenan-Flagler’s contribution focused on general management skills plus the development of an entrepreneurial fund-raising plan while PEP focused on instruction and instructional monitoring.

Participants found the former a distraction from the central tasks of school turnaround and the latter far more helpful, but complained that the program removed them from their schools for far too many days. As such, for the second half of the 35 high schools, professional development was cut to 13 days, focusing more squarely on instruction, instructional supervision, teacher recruitment and retention, data-driven decision-making, and other topics closely related to school operation and performance. Rather than developing business plans for entrepreneurial initiatives, second-year participants worked on plans responsive to NCDPI’s Framework for Action. During the 2007–2008 school year, the program focused even more intensely on school functions, with even closer links to NCDPI’s planning framework for turnaround schools.

The Framework for Action required all 66 high schools to develop formal plans to address key issues, including ninth grade transition, formative assessment, support for struggling students, literacy needs, professional development based on student achievement data, professional learning community, community involvement, and alignment of all school processes and procedures to support student achievement. After NCDPI reshaped the PD program, NCDPI-contracted “Leadership Facilitators” (school coaches) helped the turnaround schools develop their plans. They followed up with weekly visits to the schools, providing advice and technical assistance throughout the turnaround process. Leadership Facilitators were former administrators with successful experience in schools serving similar student populations. Leadership Facilitators who served as school-level coaches, were joined by Instructional Facilitators, accomplished teachers who worked with teachers at the classroom level. Instructional Facilitators were spread more thinly, assisting each teacher in a turnaround school about once a month.


To assess the impact of the interventions during the 2006–2010 period, we used a difference-in-differences design comparing the estimated test score gains in the 66 turnaround high schools with those in the 64 high schools most similar to them, controlling for numerous school and student variables. As indicated earlier, we also compared the average improvement in the two groups’ performance composites.

The Turnaround Schools Program had a 0.08 standard deviation unit gain (approximately 1 point) on End-of-Course tests after four years in the program, but the margin of improvement varied widely (Thompson et al., 2011), We undertook the present qualitative study to investigate sources of the variation. At the high school level, we selected 12 schools that had made contrasting levels of progress. We ranked schools by their 2009–2010 performance composite gains since 2007–2008, identifying those with the highest levels of total improvement from 2007–2008 to 2009-10 (25 points or more). We then selected a group that had made significant but more moderate levels of progress, averaging about 15 points, and a group that had either worsened or had improved by less than five points. Additionally, we chose schools so that the sample reflected variation in urbanicity, region of the state, school size, ethnic composition, and poverty levels. We chose nine high schools using the transformation approach, emphasizing change within the existing school structure, and three that had undergone restructuring, which involved breaking up the schools into smaller theme-based academies.

To learn about the dynamics accounting for the improvement or stalemate at each school, we interviewed principals, assistant principals, five to seven teachers, and any other personnel whom the principals identified as especially knowledgeable about the school’s experience during the turnaround process. In addition, we interviewed the central office administrators who had worked most closely with the school during turnaround, as well as each school’s Leadership Facilitator and, when possible, one or more of the Instructional Facilitators. We supplemented our facilitator interviews by reviewing a sample of the reports they filed with NCDPI.

Parallel semi-structured interview protocols for each of these categories asked about the reasons for the school’s initial low performance; the steps the school had taken to improve and which of these were particularly effective or ineffective; what assistance they had received along the way and the degree to which the assistance was genuinely helpful; what obstacles to improvement they had encountered and how they had surmounted the obstacles, if indeed they had; and whether the school now had the capacity to continue to improve and perform at higher levels. Most interviews were recorded and transcribed. When interviewees declined to be recorded, we relied on notes taken during the interviews.

On the basis of the transcriptions and notes, we wrote field notes on each school in a common format, capturing the main themes across responses to each of our questions, and including quotations that expressed the themes in striking or economical ways and anecdotes that would help us illustrate them. We then reduced the field notes on each school to a one- or two-page summary table. Using these summary tables, we identified similarities and contrasts across schools and organized these themes into a graphic model to portray the dynamics of the turnaround process in schools that made significant progress (Figure 1).


Although the low-achieving high schools were encouraged to choose and adopt a comprehensive school reform (CSR) model and many did so, implementation of externally designed models was not a central feature of the turnaround process in these schools. Indeed, after a year or two of struggling to implement a CSR model, all but one of the schools in our sample that adopted a CSR model abandoned it and substituted their own locally developed plans. They treated the parts of a CSR model as building blocks, adapting them to fit into locally developed designs rather than carrying out model developers’ blueprints comprehensively or faithfully. The CSR models contributed components, ideas, and skills to these schools, but anything approaching full implementation of the models was rare.

If model implementation is not an accurate way to characterize the turnaround process, what is? The director of the Turnaround Schools Program put it this way:

It is very complicated. It is a lot of what I call craft work. It’s really using processes and procedures [thoughtfully]. Where you put kids, how you hire and develop your teaching population, then your other systems, like how you manage student behavior, how you manage time, how you manage instructional practice—all those systems. In a low-achieving school, generally you find none of them are functioning very well. So you have to rebuild them. But if you get the right routines flowing in a school … each piece kind of fits together.

Our research confirmed this characterization of the process. Rather than model implementation, a more accurate term for the turnaround process is scaffolded craftsmanship. That is, improvement came through painstaking, piece-by-piece reconstruction, scaffolded by NCDPI and/or the NC New Schools Project facilitators but depending at least equally on the energy, commitment, and inventiveness of local educators. As our interviewees described the turnaround process, transformation did not proceed through a pre-specified, linear series of steps. Instead, external facilitators, school leaders, and teachers worked on one part, shifted their attention to another, recognized that there was a piece missing between the two and worked on that, circled back to rework the first piece so that it dovetailed better with the middle one, and so on until the pieces began to take shape and work together as a functioning whole.

If it is illuminating to view the turnaround process as scaffolded craftsmanship, it is also helpful to see it as a process of learning. That is, principals and teachers in the improved schools in our sample were learning how to construct and operate a well-functioning school as they were constructing it. Their learning process is captured in learning theorists’ notion of cognitive apprenticeship (See, for example, Collins, Brown, & Newman, 1987; Brown, Collins, & Duguid, 1989; Collins, Brown, & Holum, 1991). Consistent with the concept of cognitive apprenticeship, principals and teachers in turnaround schools learned largely from accomplished practitioners, Leadership and Instructional Facilitators, who were modeling and coaching good practice in their schools and classrooms. Thus, school turnaround entailed a scaffolded process of professional learning supporting a process of transformation.

We cannot be certain how much credit for improvement to attribute to school personnel themselves and how much to attribute to the professional development providers, Leadership Facilitators, Instructional Facilitators, and others that supported them. Principals and teachers themselves were center stage, while NCDPI and other external support organizations were in the background. Teachers often credited their principals with kick-starting the process, putting on the pressure for change, establishing discipline, raising morale, and providing ongoing support, while largely mentioning help from external organizations or facilitators only later in the interviews. The Turnaround Schools Program Director acknowledged, “When all is said and done, they are the ones running these schools and teaching the kids, and what they do is what actually makes the difference.” Local educators have to own the process, she explained. Accordingly, facilitators were trained to take a background role, suggesting, modeling, and coaching, not taking charge, directing, or starring. Yet it was clear that in the improved high schools, the turnaround process was clearly shaped to substantial degrees by the Framework for Action and the professional development and coaching that NCDPI and its partners provided.

In sum, in the improved schools we studied, the turnaround process was not a matter of initial external design and subsequent implementation, but a non-linear process of planning, inventing, adjusting, and re-planning as well as a process of learning, doing, and learning from doing. The improvement generally began with the installation of new leadership and involved four main components: (1) new commitment, climate, and culture; (2) improved knowledge and skills; (3) strategically organized and managed structures and supports for instruction; and (4) strengthened external support. Figure 1 provides an overview of the dynamics of improvement.

Figure 1. The school turnaround process



At the start, most of the improved high schools were challenged by serious problems of discipline and low expectations for student achievement, and the early actions that successful leaders took combined attention to both issues. Action on both issues generally involved a combination of tough discipline on one hand and active efforts to forge relationships and mobilize engagement on the other. Over time, these processes led to a more orderly and caring school climate, with greater academic press and increased expectations for student learning.

Safe, Orderly, and Caring Environment

The clearest illustration of creating a more orderly and caring environment comes from a small rural high school that was plagued for more than 20 years by conflict between students from two communities that had been consolidated into this county-wide school. The school was overwhelmingly African-American, so the conflict was based not on racial differences, but on long-smoldering enmity between the two communities. There were small-scale conflicts almost daily, and periodically, large group fights. An atmosphere of conflict and disorder permeated the school and contributed to the rapid turnover of principals and teachers.

Leadership on this issue came from an unexpected quarter. A man who had grown up in the county, now a member of the local police force, took note of the problems in the high school. Believing that his military and police experience gave him a special perspective on the issue, he approached the chairman of the school board with his ideas for addressing it. After creating a specific plan, the chair hired him as chief of security.

The plan had two sides, implemented in cooperation with a responsive principal. First, the new chief of security added two more officers to the two already in place, deployed all four to walk the halls, and instituted a zero-tolerance policy against fights. Offenders would not simply be disciplined by the school, but arrested, taken to jail, charged with crimes, and prosecuted. Simultaneously, the security officers were instructed to chat with students, get to know them personally, eat lunch with them, attend their sporting events, and ask the students to come to security officers, teachers, or the principal with information about developing conflicts or planned fights. Teachers were also asked to show more of a presence in the halls and to listen for signs of trouble in their classrooms. The combination of a get-tough policy and relationship building worked. After an arrest or two, students began to approach the staff to head off fights and talk through the conflicts that would have previously sparked fights.

The leadership provided by the school’s chief of security was unusual, but the serious discipline problems stemming from inter-community conflict were not. They were common in the consolidated rural high schools in our sample, and when they were addressed successfully, it was through similar approaches that combined hard-nosed enforcement with relationship building.

Academic Press and Increased Expectations for Student Learning

Displayed on the office whiteboard of the principal of a small rural high school whose composite score had increased 28 points in 2 years were teachers’ End-of-Course (EOC) examination passing rates and average scores, together with the goals the principal and each teacher had jointly set for the number of students in each class who would pass the exam this year. He explained that early in a semester, he sat down with each teacher of an EOC course to review students’ prior test scores and the prediction for each student’s likely score in the course. They discussed what the teacher and others would have to do to help the student make a passing score, focusing especially on students who were not predicted to pass, and set a goal for the number of students who should pass. Periodically during the year, they met to review benchmark and formative assessment results to adjust the program of extra support required to meet the goals. Because the goals set for every teacher are displayed on this public whiteboard, each can compare her own goals—and, at the end of the semester, her own students’ performance relative to the goals—with those of other teachers in the same and other EOC courses. This spurred competition among teachers, but it also prompted teachers to seek help from colleagues with better success rates.

Teachers loved this principal and spoke of how hard he worked, how well he knew the students, how often he was in their classrooms, and how well he listened and responded to their problems and needs, which led to his success in commanding their loyalty and mobilizing their support. He treated teachers with respect, cared about them as people and not solely as cogs in the test score machine, was motivated primarily by a concern for student learning rather than his own advancement, followed through in a competent way on the decisions and plans they made together, and evaluated teachers evenhandedly rather than playing favorites. He was in their classrooms on a regular basis, observing and making useful suggestions for improvement. On several occasions, the principal accepted decisions of his leadership team even when they went against his personal preferences. As the principal recalled:

Leaving that leadership retreat, those teachers felt pretty much empowered—that it wasn’t the principal’s ideas that they have to carry through. They were … their ideas. So … they were able to come back and go to the departments and be able to sell it because they had input throughout the process. And we set up our goal. We said at the end of the year, we’re going to be at 70% with students passing the EOCs.

A major challenge in establishing higher expectations for academic performance was overcoming the ingrained belief that poor or mediocre performance was the best that could be expected of students. Tying improved student achievement directly to teachers was pivotal in breaking through this barrier. The account that one teacher gave us was echoed in other improved schools:

But when the new administration came in and put just as much pressure on the teachers as the teachers put on the students, that’s really when the change started taking place. And yes, it was pressure. But after the end of last year, we saw, “Okay, wow, this is possible! We did raise scores.” So at the end of the first semester, now we’re like, “Okay, well, we can go higher.” And now the test score expectation for this end of year is way above anything that any of us would have even imagined three years ago. [Before the change, there was] … a mindset that excuses would be taken. And from my first year to the second, [my attitude shifted] … from why this isn’t my fault that my students aren’t doing well to “I’m taking responsibility for my students’ success and their failures.”

Intensified demands on teachers led directly to more demands on students, which led to unanticipated levels of improvement in test score performance, leading in turn to a sense that far more was possible than teachers, students, and others had imagined. That is, in the formerly low-achieving schools in our sample, change came in waves, with the initial focus of accountability and engagement leading to changes in teacher and student behavior, leading in turn to improved outcomes that inspired still higher expectations.

Evidence from our interviews across schools suggests that this is how a culture of higher expectations was created. Here, “culture,” refers to beliefs, expectations, and norms that shape teachers’ and students’ ongoing behavior, distinct from external pressures. Accountability pressures within the context of strong relationships generated commitment to new goals and standards for student behavior and learning. Similarly, strong and consistently enforced discipline policies together with energetic efforts to cultivate caring relationships with students combined to create safer and more orderly environments. While this initial commitment was crucial, it was not sufficient to complete the culture-building process. Only when teachers told themselves, “Wow, this is possible” and “Okay, well, we can go higher,” as was reported across the improved schools, had a school’s culture been transformed. These findings may provide concrete examples of the types of changes that the turnaround literature refers to as “quick wins” (Herman, Dawson, Dee, Greene, Maynard, Redding, & Darwin, 2008).


Personnel Replacement

The single most important thing that districts did to support the high schools that made significant progress was to select and install a new principal. The days of choosing principals primarily for their ability to keep order and keep parents happy are long gone, they said. The new principals’ mandate was to raise test scores, and quickly. A common image of the “turnaround principal” is of an energetic dynamo who transforms a lagging school by force of personality. But in only one case did a principal conform closely to this image. In general, the principals of improved high schools were distinguished more by their ability to develop rapport with teachers and students, by their knowledge of instruction, and by an unshowy determination to improve academic performance.

They arrived early, worked late, knew every student’s name and many details about them, worked the halls talking with students and teachers, observed classrooms daily, held teachers personally responsible for helping to meet school goals, and made tough decisions about teachers who failed to respond to suggestions and pressure for improvement. The district administrators who hired them emphasized knowledge of and experience in managing instruction as their primary reasons for choosing these principals.

Frequently, substantial numbers of teachers were replaced soon after a new principal was installed, as a means for principals to demonstrate their commitment to their turnaround agenda. The principals of the improved schools hired as many as half of the teachers in the school, and in one case, all of them. Vacancies may have stemmed from high turnover rates in these disorderly, low-achieving schools, but in some cases, principals deliberately created vacancies by encouraging poor performers to “retire or move on.” One central office intervention was initially unwelcome from the principal’s point of view but was unusually productive: the decision to replace approximately half of a struggling school’s faculty with TFA teachers. At the end of the first school year thereafter, the school’s performance composite rose by some 20 points, and by the end of the second school year, by another 10. In one small urban school created by “redesigning” a comprehensive high school, the entire faculty was new, all hand-selected by the principal. In other cases, the number of teachers replaced was smaller, but by all accounts, the new teachers brought new energy for transformation into the schools.

Many of the new teachers in these mass replacements were new to teaching as well as new to the school. Particularly during their first two or three years, inexperienced teachers produce lower test score results than their more experienced counterparts (Henry, Bastian, & Fortner, 2011). Yet the sharp rise in performance composites at schools with many new teachers suggests that with strong professional development and coaching, new teachers can make a positive contribution to a school’s performance. Across the schools in our sample, teachers credited extensive professional development from multiple sources for bringing bright, energetic, but inexperienced teachers quickly up to speed, sometimes enabling them to outperform more experienced teachers.

The replacement of teachers in improving schools did not end with the initial wave but continued with the deliberate discharge of underperforming teachers who failed to respond to pressure and assistance to improve. A prominent feature of the improvement process in these schools was frequent classroom observation by principals, assistant principals, and other administrators, Leadership Facilitators, and Instructional Facilitators. These underperforming teachers were provided with feedback and suggestions, but when they showed no improvement or outright resistance, discussions turned toward the need for principals to put recalcitrant teachers on action plans and eventually to encourage them to find new jobs or retire.

In some cases, support organizations pressed principals to replace teachers. An NCDPI official told one principal, “You need to get rid of these teachers. They are killing your scores.” The principal acknowledged the problem: “I knew she was right. I had known for some time that I should do it. I guess the pressure from her made me do what I knew all along I should do.” When he told the teachers to retire or move on, they did so largely without protest. Although the conventional wisdom says that it is very difficult for an administrator to fire a tenured teacher, in school after school, teachers with lagging scores, poor classroom observation results, and a reluctance to change were reported to depart in response to low evaluations, action plans, and pressure to retire or move on.

Creating a vacancy was only step one in replacing a low-performing teacher: the harder problem was recruiting and hiring a replacement with stronger commitment and skills. Asked how he did so, one principal said, “I can’t compete on money, so I sell the mission and a chance to work in a school that is on the move.” By “mission,” he meant the opportunity to give low-income and minority students a good education and a better chance in life. In several schools, other incentives were provided, including an increase in the local supplement, the chance to teach relatively small classes, and a district-sponsored apartment complex for teachers. But across schools, the most compelling draw seemed to be the opportunity to pursue the mission in a school that was improving.

Coaching and School-Specific Professional Development

Leadership Facilitators were recruited for their experience as successful principals and trained to work in a facilitative rather than a directive way. They generally visited each of their assigned schools once a week and performed a range of functions, often beginning by carrying out their own needs assessments by reviewing data on the school, interviewing principals and teachers, observing in classrooms, and moving about the school informally. A typical Leadership Facilitator visit might involve a brief orienting conversation with the principal, several classroom observations, and participation in a School Improvement Team meeting or a meeting with a small group of teachers and an assistant principal working on a specific problem, such as difficulties in the in-school suspension program or how to improve tutoring arrangements for struggling students. At the end of a day, Leadership Facilitators usually met again with principals to discuss what they had learned.

Leadership Facilitators sometimes served as neutral discussion leaders during leadership team meetings, School Improvement Team meetings, and planning retreats as well as taking initiative to organize special meetings to address problems they had identified. Additionally, they provided tools such as classroom observation protocols and common lesson plan formats to principals and teachers, modeled the use of these tools in joint instructional monitoring and feedback sessions, and then followed up by observing and coaching principals and teachers as they used the tools. They suggested ways of handling important tasks, such as reviewing the data on incoming ninth graders and developing a master schedule that would assign them to appropriate courses and teachers while also providing their teachers with common planning time. They worked closely with testing coordinators to ensure that NC Wise, the state’s new student information system, would come online properly in the school, and to determine how the data from End-of-Course, benchmark, and formative assessments might be reported to and interpreted for teachers. Facilitators’ focus on the master schedule, NC Wise, and formative assessment demonstrates the role they played in organizing systems and processes within the schools.

Another function of Leadership Facilitators was supporting follow-through on schools’ Framework for Action plans. As one NCDPI manager put it, “You need to remind them on a regular basis … to keep people on track in really low-capacity schools,” but there is little evidence that facilitators tried to dictate actions to either principals or others. In contrast, once facilitators had discussed a problem several times with a principal, NCDPI officials sometimes stepped in and urged actions in a very pointed way, as indicated in the example described above about the supervisor who told a principal point-blank, “You need to get rid of these teachers. They are killing your scores.”

Instructional Facilitators were selected for recent experience as successful teachers and provided assistance to individual teachers and groups of teachers in their assigned subject areas. Because resource constraints limited the number of Instructional Facilitators on staff, they visited schools less frequently than leadership facilitators, once or twice a month at most. There was also more variation in the frequency of visits across facilitators, schools, and time. Most teachers’ comments about Instructional Facilitators were general but positive and complaints from teachers about Instructional Facilitators were about seeing them too seldom. One NCDPI supervisor conceded that resources were too limited to provide the depth and frequency of instructional facilitation that she thought necessary in the lowest capacity schools.

Particularly when working with new teachers, Instructional Facilitators often focused on the NC Standard Course of Study, breaking it down objective by objective to clarify exactly what teachers should be focusing on. Instructional Facilitators taught demonstration lessons, observed as teachers gave new techniques or material a try, and provided a combination of encouragement and corrective feedback. In one school, an Instructional Facilitator team-taught with the chair of the science department, leading to major improvements in science instruction. According to the principal of an improved school, the demonstration lessons and the fact that the Instructional Facilitators were teachers only recently out of the classroom gave them credibility and leverage. Instructional Facilitators also brought in classroom materials and lesson plans to shore up observed weaknesses. They helped teachers understand End-of-Course, benchmark, and formative assessment data and suggested strategies to address objectives on which many students scored poorly. On occasion, Instructional Facilitators offered targeted professional development, “I noticed that one of the things we needed help with was differentiation. All of the coaches came in together and did that workshop for us.” Like the Leadership Facilitators, Instructional Facilitators often met toward the end of a day with principals or assistant principals to discuss their observations.


Coordinating Curriculum and Assigning Students and Teachers Strategically

The improved schools in our sample used a variety of strategies to shepherd individual students through curricular paths matched to their evolving skills and to ensure that students encountered solid teaching and re-teaching along the path to proficiency. This sounds simple, but it required many distinct components, each carefully crafted to perform its function within a coordinated whole.

One key to improvement was to break the curriculum down into course-sized chunks, then route individual students through the right courses in the right order to best prepare them for graduation. Specific attention was paid to what aspects of the Standard Course of Study were taught in which courses. As one principal explained, crafting the details of the ninth grade transition courses was crucial:

What is so essential is exactly what you teach. In Algebra I, [the NC Standard Course of Study specifies] four goals. When you look at the End of Course exam … you may have 60% of the test come from Goal 3 and Goal 4. So we design our curriculum in a way that the Foundations of Algebra [students] will get Goal 3 and Goal 4. [Further,] the pacing guide is crucial. You may have a bridge course, but if it does not have a good pacing guide, it’s a failure. I’ve seen students go into Algebra 1A and 1B still struggling because of the way the pacing guide was designed.

Yet detailed planning of the courses comprising the various pathways only prepared the way for another essential step: strategic assignment of students and teachers. In each of the substantially improved schools, principals, counselors, and assistant principals worked together to choose an appropriate series of courses for individual students, using test data, as well as personal knowledge of teachers and students to make the best set of matches for each course, what one principal referred to as “hand-scheduling.” Constructing the master schedule and student assignment rosters were complex tasks that required thinking not just semester by semester but over full academic years, anticipating the courses that students would need in future years. Principals and Leadership Facilitators consistently pointed to the master schedule as a key instrument for improved academic performance.

Even when students were hand-scheduled into strategic reading and other courses designed to bolster weak skills, many still needed additional preparation to understand material in EOC-tested courses. The Framework for Action called for “a plan for identifying and addressing literacy issues and needs,” requiring additional efforts to develop students’ content-specific reading skills, but many teachers lacked the knowledge to teach such skills.  One English teacher told us, “I had never been trained in how to teach reading. We had just assumed that kids would come to us with reading skills.” To address this, the district’s central office, NCDPI, and its facilitators provided training in teaching literacy in the content areas. The Instructional Facilitators in particular provided concrete, subject-specific suggestions and materials. One teacher recalled, it was “literacy in math, literacy in science, literacy in history, literacy in shop, literacy in Phys Ed. We all got involved in teaching literacy.”

Supervising Instruction, Building Professional Community, and Using Assessments

The improved schools took additional steps to ensure that the NC Standard Course of Study for each course was actually taught, taught well, and re-taught when necessary. With the support of Leadership and Instructional Facilitators, principals supervised instruction closely, organized teachers into professional learning communities, and promoted the use of benchmark and formative assessment to check students’ learning regularly, guide assistance for struggling students, and shore up weak spots in teaching.

Interviews with central office staff, principals, and Leadership Facilitators indicate that when the turnaround process began, little real teaching was going on in many classrooms. A leadership facilitator’s report illustrates a common occurrence:

My first observation today was in the classroom of a science teacher [...] I saw no teaching. A quiz lasted for half the period and for the remaining 45 minutes, the teacher instructed the students to read the next chapter. He did point out several things they should remember. When students became a little chatty, he had them answer questions at the end of the chapter.

This was not an isolated case. As a central office administrator observed about another school, “We always got the sense that teachers were not really teaching the Standard Course of Study. If they were teaching at all, they were teaching whatever they enjoyed teaching.”

One step toward assuring that the Standard Course of Study was taught was simply to stress its importance and help teachers, especially new teachers, understand it. In the rural high school with large numbers of TFA teachers, one interviewee reported,

We got the content area people, the coaches or facilitators, and they were very good coming in and working with our teachers. Because we had all new teachers, and I liked the fact that they were really, really dedicated to making sure that our teachers understood the curriculum.

Other teachers gave similar reports of Instructional Facilitators breaking down the Standard Course of Study, objective by objective to “make sure that our teachers understood the curriculum.”

Another step in ensuring that the Standard Course of Study was taught was the development of pacing guides to distribute objectives effectively over time, coordinated with benchmark assessments to check students’ progress at regular intervals. Even with a good understanding of the curriculum and a pacing guide, one Leadership Facilitator stressed that many teachers had trouble constructing lesson plans that worked well in a 90-minute block period. He created a common lesson plan format based on Madeline Hunter’s principles of effective instruction and his own experiences. The format could be modified to fit a specific class, but included required elements like explicit statement of the objectives, bell-to-bell teaching, a mixture of presentation with progressively more independent student work, multiple transitions from one mode of activity to another, and a closing summary of learning.

This Leadership Facilitator shared it with the principal and teachers, and showed the principal how it could be used to guide classroom observations. He modeled its use as a format for making notes during the lesson and for providing feedback to teachers afterwards. Over time, as teachers used the format to guide planning, and administrators used it for observation and feedback, the lesson format has become routine in the school.

Common lesson plan formats and frequent classroom observations took different forms in different schools but were regular practices in the improved schools in our sample. Principals, assistant principals, and Leadership Facilitators focused primarily on the Standard Course of Study objectives being taught, if lessons were well planned, and whether students seemed actively engaged during a lesson.

As frequent as classroom observations were, in improved high schools this type of administrative supervision was not the sole means of ensuring that the Standard Course of Study was taught well. A strong complement to administrative supervision came professional learning communities (PLCs), in which teachers worked together to develop pacing guides and lesson plans, observed and gave each other feedback, created formative assessments, and used their results to improve their teaching and pinpoint which of their students needed further instruction on which objectives.

One step in the creation of PLCs was to schedule common planning times for the teachers of a subject or an EOC-tested course. But principals and assistant principals took additional steps to jumpstart collaboration. In one moderately improved high school, the principal and an assistant principal led required weekly department meetings, focusing discussions on curriculum, teaching, specific students’ problems, and on how some teachers were able to succeed with particular students whom others could not reach. The sessions were designed primarily to help lagging teachers learn from their peers. Many teachers testified that these collaborative sessions represented more powerful contributors to their professional development than any formal workshops. There was some initial resentment of and resistance to the sessions, but according to the principal, the meetings have now become routine, and the administrators only participate when invited by teachers or when they need to address some problem.

In one of the most improved schools in our sample, the chair of the science department recalled, “Our significant change began when [the Instructional Facilitator] started working with us.” With support from the Instructional Facilitator, the science chair began team teaching one large group of students with two younger teachers. The science chair took the lead, but all three planned and taught the class together. In classes that the two younger teachers taught on their own, they used the same lesson plan and patterned their teaching after the approach that the chair had modeled. Over time, teachers throughout the department began team planning and teaching some of their classes. They used the school’s common lesson format but put special emphasis on hands-on approaches, including physical models (e.g., of atoms, molecules, cells) along with regular vocabulary drills and review, attention to test-taking strategies, and common formative assessments. Through team teaching, the chair spread best practices for test-taking strategies throughout the department. Teachers in the department even developed a practice they referred to as “rotations,” in which the teacher who was best at teaching a given set of objectives would teach it to all of the students enrolled in an EOC-tested subject rather than keeping students in fixed class groupings. This level of team teaching and student exchange was uncommon, but teachers in improved schools often reported observing each other to pick up ideas and make suggestions.

The use of multiple layers of assessment data by teachers in these PLCs seems to have been particularly powerful, ranging from using Educational Value-Added Assessment Software (EVAAS) to analyze EOC results to benchmark tests to ongoing formative assessments. An Instructional Facilitator who was particularly knowledgeable about EVAAS became “a kind of EVAAS guru” for teachers in one improved school. With guidance from the “EVAAS guru,” the teachers responsible for each EOC-tested subject examined the predicted score for each student in each of their classes and brainstormed ways to beat the predictions. One teacher noted that EVAAS helped them “identify which students are right on the borderline, which ones you need to push a little more, and which ones are probably stronger and can work with other students if you want to pair them up.” In addition, district-developed benchmark tests administered at nine-week intervals kept the teachers themselves on track with the scope and sequence as well as helping them track student progress.

According to the teachers, the combined effect of examining all of this assessment was to focus them on what students were actually learning, on needed changes in their own curriculum and teaching, and on common errors that students make. As one teacher explained,

All of the Algebra I teachers give the same assessment, and we can look at the objectives and see, “Which objectives is my class weak in? This class over there was not weak in it, so let me talk to that teacher and find out what I can do better to improve my teaching of that particular objective.” Or, “Why were my students weaker here versus there?” And it just lets you know what you maybe need to go back and focus on what the students are not getting. And this processing has helped build teamwork. The teachers teaching the same EOC [course] are really working together… far more than they ever did before. They’re developing lessons together, they share ideas, they share notes, and see what works best.

Teachers in improved schools often reported using 20-question assessments on a weekly basis, with 5 of the 20 questions focusing on material taught in previous weeks. They stressed that the weekly assessments not only served the obvious functions of generating information to guide improvement of teaching as well as tutoring for students who missed certain items, but they also prompted students to review the week’s lessons and to refresh their memory of material learned earlier in the semester.

Organizing Assistance for Struggling Students

The NCDPI’s Framework for Action also required the turnaround schools to submit plans for assistance for struggling students. In improved schools, principals, assistant principals, and teachers provided extra help to struggling students before, during, and after school, using information from the benchmark tests or formative assessments. In the highest performing high school in our sample, teachers went to extraordinary lengths to work with students who needed help. One math teacher arrived at school at 6:30 each morning to tutor students before school, often stayed until 5:30 or 6:00 p.m., and sometimes met students after church on Sundays. While weekend hours were unusual, before- and after-school tutoring by teachers and even principals was common in the improved schools. In one rural school, the central office also provided extra funds for after-school tutoring, strengthening support for struggling students.

Many students, particularly in rural areas, found it difficult to get to school early or stay late for extra help, so the improved high schools scheduled periods during the school day for this purpose. One school called this Great Expectations: to make time in the day for these sessions, they eliminated a ten-minute break from the schedule and shaved five minutes off of each class period. When some teachers were not using the time well, the administration laid down some ground rules for the Great Expectations periods:

They can’t introduce new material. They can’t just provide free time for students to work. So after benchmark assessments, we sit down with teachers and we look at [what objectives the students in each class seem to be having trouble with] and we say very clearly, “This is what you need to reteach during Great Expectations time.”

Great Expectations also served as a time for pullouts for students who need intensive help in any classes. The focus of pullouts changed over time. For example, there was an intensive focus on writing right before the writing test. To allow for more intensive and targeted intervention, virtually all faculty helped with pullout sessions, including the principal and the schools’ instructional coordinator.

In addition to tutoring during the school day, some schools created special programs to prepare for EOC exams. In one improved school, English teachers wore camouflage and combat boots to stir up interest in 2 weeks of “boot camp” sessions held after school that about half of the sophomores participated in at least once. Several teachers of other subjects joined the English teachers to staff the boot camp sessions. EOC preparation sessions took different forms across the schools, but some form of special sessions, often with participation by teachers of non-tested subjects, were a regular feature at the improved schools.


District officials often followed up on the installation of a new principal by providing a variety of continuing supports. Thus in some improved schools, the principals complained about the lack of support they were getting and expressed fears that district decisions would undermine the progress the school was making. It appears that continued central office support was helpful but not absolutely essential to a turnaround effort, as some schools made progress without it. But district intervention was essential at the point of installing the right new principal.  

Building community support for schools was another essential element of turnaround for improved schools. At a small rural high school, the new principal lined up a series of appearances at churches throughout the largely African-American community. At each, he explained what he and his colleagues were undertaking and how they were going about it. These visits paid off later when he instituted new policies requiring a higher GPA to participate in sports, thus threatening the participation of some football players. Grumbling arose among parents and athletic boosters, “but some important people in the community told them that I knew what I was doing, so they should leave me alone,” he recalled. Despite any opposition that may have been aroused by the new GPA requirement, the county commissioners were also persuaded to raise the local teacher supplement in this low-wealth community by $1,000.

In at least two other schools, the relationship between the school and community was slow to turn around. One school’s identity in the community was shaped in contrast to the other high school in the district that was seen as “the good high school,” and further damaged by footage of a student being helicoptered out after being shot as he was leaving school. The new principal appointed soon thereafter took several steps to improve the school’s image, including a Saturday event devoted to cleaning up the school and painting the entranceway and the atrium where assemblies and other events were held. According to the principal, “We got 400 parents and students to work with us that day.” Yet the principal continued to worry that the school’s identity as second best might persist in the minds of the school board and superintendent. He was anxiously awaiting the effects of budget cuts, fearing personnel cuts would leave him without the handful of new people he had managed to bring in to help him lead the turnaround effort.


Rowan and colleagues have argued that high academic standards, assessment-based accountability, and enlarged choice are necessary but not sufficient conditions for the improvement of low-achieving schools. While these conditions may motivate teachers to work harder, stronger motivation on its own cannot bring about large improvements in troubled schools. Such macro-level approaches as recruiting more highly qualified people into teaching produced have produced only “discouragingly small” average change, and locally-driven school-by-school approaches have typically produced improvement in only a small fraction of low-achieving schools (Rowman, Correnti, Miller, & Camburn, 2009, p. 10). Alternatively, in a study of 115 elementary schools that implemented CSR models, they provide evidence that “school improvement by design” can be implemented faithfully and produce substantial improvements in student performance.

Our study suggests a second scalable alternative to macro-level and locally driven approaches: scaffolded craftsmanship. On the premise that chronically low-performing schools were incapable of self-improvement, the high schools we studied were urged to choose from a menu of CSR models, and most did so. Yet only one of the 12 high schools we studied implemented the chosen model successfully, and that school was virtually a new start, with its principal, entire faculty, and student body recruited from scratch. The other 11, where the effort was to turnaround a largely intact school, abandoned their chosen models in favor of designs largely structured by the NCDPI’s Framework for Action, a general pattern across the entire set of 66 high schools served by the program.

NCDPI’s approach enjoyed several advantages over the CSR models. First, it focused on function rather than form. For example, the High School Framework for Action called on the schools to develop a “plan for 9th grade transition” rather than to implement a single organizational form such as a Freshman Academy, a prominent feature of one CSR model. The balance of the Framework was similarly oriented to functions rather than forms, calling on the schools to develop plans for using formative assessments, assisting struggling students, improving students’ literacy skills, creating professional development based on student achievement data, building professional learning community, involving the school community in addressing school needs, and reviewing all school processes and procedures to ensure consistency with a priority on school-wide academic proficiency. The emphasis on function over form focused principals and teachers on central influences on student learning while permitting them to develop approaches that fit their circumstances and commanded their commitment. Other advantages included the PD providers’ and coaches’ reputations as principals and teachers fresh from successful experiences in similar schools and the frequency and long-term continuity of their visits to the schools.

On first examination, the adoption and subsequent abandonment of most components of the CSR models appear to represent an instance of Berman, Greenwood, McLaughlin, and Pincus’ “mutual adaptation” (1975), where in successful implementations of innovations, not only do innovations change schools, but schools also changed the innovations themselves, dropping some features and adjusting others to fit their circumstances. But in the schools we studied, CSR models played a very minor role, and to conceive the change process in terms of model implementation, even via mutual adaptation, would be like mistaking details in one corner of a painting for the central subject and theme of the painting. In these schools, change was about the craft of reconstructing the schools’ major functions with judicious external guidance, not implementing the CSR models. The importance of NCDPI’s guidance also differentiates the change process in these schools from purely locally-driven change.

If scaffolded craftsmanship characterizes much of the turnaround process, substantial personnel replacement also played a key role in successful turnaround efforts. Although NCDPI did not order personnel replacement, in all of the schools that did improve, local districts installed new principals near the beginning of the intervention. One school built a new faculty from scratch, and the other improved schools replaced from a third to half of their teachers. The personnel changes brought new energy and talent into the schools and eased the process of mobilizing the faculty behind a reform agenda. But several cautions concerning personnel replacement are in order. First, as bitter experiences in the unimproved schools in our sample showed, without assured district support, new principals could not assert tough accountability and discipline policies, or take other steps necessary to stabilize the school’s climate and reduce uncontrolled turnover. Absent stability, personnel replacement was just more turnover. Second, after the personnel changes, new principals had to work hard to build both teacher-principal and teacher-teacher trust. Due in part to the pressures and negative publicity affecting these low-performing schools, distrust was already pervasive, and personnel replacement tended to exacerbate the problem until deliberate trust-building efforts could overcome it. Finally, both principals and teachers emphasized that professional development and continued coaching were necessary to convert the newcomers’ talent and energy into skilled practice.

Although our findings extend those of prior research, it is more important to emphasize that the NCDPI’s Turnaround Schools program made it possible to scale up from isolated changes in individual schools to improvement on a statewide scale in a total of 66 low-performing high schools, 37 middle schools, and 25 elementary schools. Improvement across the set of high schools was not uniform, but the average improvement in the percentage of their students that were proficient was more than 10 points, consistent with the standard set by Herman et al. (2008) for “a turnaround school.” While we cannot be certain that it was scaffolded craftsmanship that brought about the improvements, over the 2006–2010 period, the 66 treated schools improved proficiency rates and learning gains substantially more than the next 64 higher performing schools.

Nationally, growing frustration with the failure of many low-performing schools to improve has fueled interest in more “robust” or radical interventions, such as (a) replacing all faculty in a school, (b) closing them and redistributing students to other schools, and (c) re-starting them as charters. Yet research does not provide strong support for the effectiveness of such interventions. (De la Torre, Allensworth, Jagesic, Sebastian, & Salmonowicz, 2012; Hess, 2003; Rowan et al., 2009). Our findings suggest that judicious personnel replacement followed with professional development and coaching targeted to key functions may be a more effective method for implementing school turnaround than the structural approaches promoted via NCLB sanctions and Race to the Top.


Berman, P., Greenwood, P. W., McLaughlin, M., & Pincus, J. A. (1977). Federal programs supporting educational change, Volume IV: A summary of the findings in review. Santa Monica, CA: RAND Corporation.

Brown, J. S., Collins, A., & Duguid, P. (1989). Situated cognition and the culture of learning. Educational Researcher, 18(1), 32–42.

Collins, A., Brown, J. S., & Holum, A. (1991). Cognitive apprenticeship: Making thinking visible. American Educator, 15(3), 6–46.

Collins, A., Brown, J. S., & Newman, S. E. (1987). Cognitive apprenticeship: Teaching the craft of reading, writing and mathematics (Technical Report No. 403). Cambridge, MA: BBN Laboratories; Champaign, IL: Center for the Study of Reading, University of Illinois.

De la Torre, M., Allensworth, E., Jagesic, S., Sebastian, J., & Salmonowicz, M. (2012). Turning around low-performing schools in Chicago: Summary report. Chicago, IL: The University of Chicago Consortium on Chicago School Research.

Fiscal Research Division. (2007). North Carolina high school reform. [Fiscal brief]. North Carolina General Assembly, Fiscal Research Division. Retrieved from http://www.ncleg.net/fiscalresearch/fiscal_briefs/Fiscal_Briefs_PDFs/High%20School%20Reform%20Fiscal%20Brief_Final_12%2021%2007.pdf

Henry, G. T., Bastian, K. C. and Fortner, C. K. (2011). Stayers and leavers: Early-career teacher effectiveness and attrition. Educational Researcher, 40(6), 271–280.

Herman, R., Dawson, P., Dee, T., Greene, J., Maynard, R., Redding, S., & Darwin, M. (2008). Turning around chronically low-performing Schools: A practice guide (NCEE #2008–4020). Washington, DC: National Center for Education Evaluation and Regional Assistance, Institute of Education Services, U.S. Department of Education. Retrieved from http://ies.ed.gov/ncee/wwc/pdf/practice_guides/Turnaround_pg_04181.pdf

Hess, G. A., Jr. (2003). Reconstitution—three years later: Monitoring the effect of sanctions on Chicago high schools. Education and Urban Society, 35(3), 300–327.

North Carolina Department of Public Instruction. (n.d.). High school framework for action [Handout]. Raleigh, NC: http://www.ncpublicschools.org/schooltransformation/turnaround/archives/framework/

Rowan, B., Correnti, R., Miller, R. J., & Camburn, E. (2009). School improvement by design: Lessons from a study of comprehensive school reform programs. Philadelphia, PA: Consortium for Policy Research in Education.

SERVE Center, Friday Institute, & Carolina Institute for Public Policy. (2010). Race to the Top proposal [Proposal submitted to the US Department of Education]. Greensboro, NC: Author.

Thompson, C., Brown, K. M., Townsend, L. W., Henry, G. T. & Fortner, C. K. (2011). Turning around North Carolina’s lowest achieving schools: 2006–2010. North Carolina: Consortium for Educational Research and Evaluation.

Cite This Article as: Teachers College Record Volume 118 Number 13, 2016, p. 1-26
http://www.tcrecord.org ID Number: 20554, Date Accessed: 3/23/2019 1:10:29 AM

Purchase Reprint Rights for this article or review
Article Tools
Related Articles

Related Discussion
Post a Comment | Read All

About the Author
  • Charles Thompson
    East Carolina University
    E-mail Author
    CHARLES L. THOMPSON retired in 2013 after serving as Director of the Teacher Quality Research at the Carolina Institute for Public Policy and Research Professor in the Department of Public Policy at the University of North Carolina at Chapel Hill.
  • Gary Henry
    Vanderbilt University
    E-mail Author
    GARY T. HENRY holds the Patricia and H. Rodes Hart Chair and serves as Professor of Public Policy and Education in the Department of Leadership, Policy and Organization, Peabody College, Vanderbilt University. Professor Henry’s research focuses on teacher quality, teacher preparation, school reform, quantitative research methods, and educational evaluation and has recently published in Educational Evaluation and Policy Analysis, Journal of Teacher Education, Education Finance, and Policy and Science.
  • Courtney Preston
    Florida State University
    E-mail Author
    COURTNEY PRESTON is an Assistant Professor of Educational Policy Studies at Florida State University. Dr. Preston research focuses on policies and practices to improve school personnel effectiveness, including teacher quality, teacher preparation, and educator labor markets. Her research has recently appeared in Teachers College Record, Economics of Education Review, and Educational Management, Administration, and Leadership.
Member Center
In Print
This Month's Issue