Transforming High Poverty Urban Middle Schools
into Strong Learning Institutions: Lessons From the
First Five Years of the Talent Development Middle School

Robert Balfanz and Douglas Mac Iver
Johns Hopkins University

This article appeared in the Winter of 2000 in: Journal of Education for Students Placed At Risk, 5(1 & 2 [special double issue])

Abstract

The Johns Hopkins Talent Development Middle School Model (TDMS) aims to transform high poverty urban middle schools into strong learning institutions that reliably provide every student with a standards-based education and every teacher with the training, support, and materials she/he needs to deliver it. Two of the model's developers discuss ten lessons learned from implementing, refining, and evaluating this model in five high poverty middle schools in Philadelphia and discuss obstacles they have encountered and breakthroughs that they made in developing the knowledge base, materials, and infrastructure needed. Taken together, the lessons suggest that "improvement now" in curriculum and instruction and in student achievement is a realistic goal even in high poverty urban middle schools, but such improvement requires multiple layers of sustained technical assistance and implementation support and also requires local partners who can help the model to become integrated into the fabric of the school district.

Transforming High Poverty Urban Middle Schools

In many respects, it is during the middle grades that the battle of urban education is lost. It is here that the absence of strong curricula and the lack of well prepared teachers are most severe (Cooney, 1998; Schmidt, McKnight, & Raizen, 1997). It is also during the early years of adolescence that students become disengaged from school and fail to receive the academic preparation that they need to succeed in high school. Alienated, unsure, and having received a very uneven and substandard middle grades education, up to half the students in our nation's largest cities are unable to make a successful transition to high school (Neild, & Weiss, 1999; Wilson & Corbett, 1999).

During the past five years, researchers and practitioners at Johns Hopkins University as part of the Center for Research on Educating Students Placed at Risk (CRESPAR) have worked in collaboration with the Philadelphia Education Fund and the School District of Philadelphia to develop a solution to the problem of middle grades education in our nation's cities. Our efforts are two-pronged. First, we are working to develop a powerful but flexible whole-school reform model for urban middle schools that will reliably provide every student with a standards-based education and every teacher with the training, support, and materials she/he needs to deliver it. Secondly, we are working to develop the knowledge base, materials, and infrastructure needed to help urban school districts systematically create middle schools that work for all of their students.

Currently, the Johns Hopkins Talent Development Middle School (TDMS) model (Mac Iver & Plank, 1997; Mac Iver, Mac Iver, Balfanz, Plank, & Ruby, in press) is being implemented, refined, and evaluated in five high poverty middle schools in Philadelphia. In addition, we have begun a national field test of the TDMS model which will involve 3 additional middle schools in Philadelphia and 6 to 9 middle schools in three school districts (Mobile, Detroit, and Memphis) chosen to represent the diversity of urban middle schooling in the U.S.(Mac Iver, Balfanz, Plank, & Ruby, 1998). In this article, we will summarize the major lessons we have learned over the past five years, highlight the major obstacles we have faced, and discuss our next steps.

Building Upon Prior Middle Grades Reform Efforts

The Talent Development Middle School Program builds on the foundation of prior major middle school reform efforts begun in the 1980s (California State Department of Education, 1987; Carnegie Task Force on the Education of Young Adolescents, 1989; Children's Defense Fund, 1988; Dorman, 1984; Lipsitz, 1984; Maryland Task Force on the Middle Learning Years, 1989). These efforts have helped create a consensus concerning the kinds of structural and organizational changes needed in middle grades schools to make them more caring and more supportive institutions.

These efforts have been successful in helping many middle schools change their climates and structures to become "warmer, happier, and more peaceful places for students and adults... [However, most of these schools] have not moved off this plateau and taken the critical next step to develop students who perform well academically, with the intellectual wherewithal to improve their life conditions" (Lipsitz, Mizell, Jackson, & Austin, 1997, p. 535). That is, despite evidence that a comprehensive implementation of the recommended reforms results in encouraging gains in student achievement and well-being (Felner, Jackson, Kasak, Mulhal, Brand, & Flowers, 1997), high performing middle schools are still quite rare (Cooney, 1988; Killion & Hirsh, 1998; Johnston & Williamson, 1998) because few schools move far enough "along the developmental continuum from changing climate and structure toward changing curriculum and instruction" (Lipsitz et al., 1997, p. 540).

In short, although structures and practices that are in keeping with the best of the middle grades reform documents are an essential foundation for middle grades reform, dramatic improvements in student performance result only if teachers also provide all students with markedly better learning opportunities every day. In other words, the structural foundation provided by the "signature practices" of middle grades reform must be combined with an instructional edifice of a demanding core curriculum for all students accompanied by instructional strategies which maximize "teaching for meaning," motivation to learn, and peer support for learning. This instructional edifice must be supported by ongoing high-caliber, curriculum-specific professional development that is accompanied by extensive follow-up support for teachers in their classrooms by experienced curriculum coaches.

Five Critical Components of the Talent Development Middle School Model

The Talent Development Middle School (TDMS) is based on the philosophy that students are "at promise" not "at risk" (Boykin 1994). It brings together five critical elements that are needed to turn low performing high poverty middle schools into strong learning institutions which take a "no excuses" approach toward student success (Wilson & Corbet 1999). In a Talent Development Middle School:
* A communal organization of schooling is combined with research-recommended forms of instruction, standards-based curriculum, and student assessment in each of the major subject areas.

* Teachers receive four tiers of intensive and continuous support. This includes on-going grade and subject specific staff development (38 hours a year for at least two years), sustained in-classroom implementation support from respected peers, and opportunities to network with teachers in other schools using TDMS instructional programs.

* Students receive a wide array of learning supports and extra-help opportunities which increase their motivation to learn and enable them to succeed in high-level standards based courses.

* Research, evaluation, and refinement are on going and multi-leveled.

* There is a deep commitment to develop the local capacity needed at the school, district, and community level to sustain and spread strong implementations of the TDMS model.

The Talent Development team works intensively with a school for three to five years to implement these elements and build the school's and district's capacity to sustain them. Our aim is to create schools in which every student has a world class education which includes the study of high quality literature, Algebra, hands-on science, and the use of primary sources in history. In working alongside teachers, administrators, and our local partners in Philadelphia and elsewhere to reach this aim, we have learned a number of lessons and confronted several major obstacles. This has greatly increased our understanding of what it takes to successfully turn high poverty middle schools into strong learning institutions and provided us what we believe are important insights into the comprehensive whole school reform process.

Lessons Learned

Lesson 1. Low performing urban middle schools which serve high poverty populations can realize substantial and systematic improvements in student learning and academic achievement even during the first year of comprehensive whole school reform.

In some circles, it has almost become an article of faith that it can take up to three to five years before comprehensive whole school reform results in significant achievement gains. We have found that this does not have to be the case.

We have worked closely with five urban middle schools to develop and implement the Talent Development Middle School model. Each of these schools serves high poverty populations (their poverty rates range from 75 to 90%) and are broadly representative of the middle schools found in the large urban and northeastern school district of Philadelphia. On a district wide performance index which measures the achievement of eighth graders in mathematics, English, and science (based on Stanford-9 results) and takes into account student and faculty attendance, and promotion rates, three of the schools (Central East, Cooke, & Clemente) ranked slightly below the mid-point of the district's 42 middle schools prior to the implementation of TDMS, one (Shoemaker) ranked near the bottom, and one (Beeber) ranked in the top quartile.

Taken together, these schools face most of the strains associated with urban schooling including large size (the schools range from 750 to 1500 students), high rates of teacher and administrator turnover, large numbers of apprentice teachers, uncertified teachers and long-term substitutes, substantial ESL and/or bi-lingual populations, low faculty morale, limited student expectations and poor academic performance. Prior to the implementation of TDMS, the typical student entered four of our five schools testing approximately two years below grade level in math and reading and completed the eighth grade testing three years below grade level.

Yet despite these strains, the three schools for which post-test data is available experienced significant and systematic gains in student learning during their first implementation year of the Talent Development Middle School Model.

Achievement Gains at Central East Middle School During its First Year as a TD school. Since becoming a Talent Development Middle School, Central East has experienced strong broad-based, and sustained achievement gains in all achievement areas assessed by the district's testing program. Gains at Central East during its first year as a TD school were particularly impressive in Reading, English and Language Arts. This was not surprising because Central East did not begin phasing in TDMS's math and science programs until its second year as a TDMS school.

All Talent Development Middle Schools use Student Team Literature, a middle school language arts curriculum and instructional program that is designed to improve students' skills in reading vocabulary, literary analysis, and student collaboration by using outstanding literature, higher-level questioning, and working with other students. It includes (1) curricular materials (partner discussion guides) to assist students' study of high-quality fiction and non-fiction books, (2) recommended instructional practices, peer assistance processes, and assessments, and (3) staff development, mentoring, and advising for teachers to support the curricular and instructional reforms. The National Staff Development Council recently selected the Student Team Literature Program for inclusion in their Consumer's Guide of effective staff development programs that meet national standards in language arts (Killion, 1999).

After one year of implementation of the Student Team Literature program, a significant advantage in reading comprehension improvement was observed in 21 classes at Central East Middle School relative to 25 classes in a closely matched comparison site selected by the school district. The improvements in reading comprehension at Central East were substantial (Mac Iver, Plank, and Balfanz, 1997). In analyses that controlled for prior achievement and current grade level, the typical Central East student outperformed his or her counterpart at the comparison school on the Stanford 9 multiple choice test of reading comprehension by almost 12 scale score points. The observed effect size of .51 is large compared to the impact of other educational reforms and is of the magnitude needed to achieve serious academic gains (Mosteller, Light, & Sachs, 1996).

Achievement Gains at Cooke Middle School During Its First Year as a TD school. Stanford 9 achievement test data indicate that large and systematic achievement gains were made at Cooke Middle School during its first year of implementing the TDMS model. These gains were significantly higher than at a district-selected, demographically-similar control school. Effect size estimates from simple HLM models indicate that Cooke outperformed its comparison school by .24 standard deviations in reading (Plank & Young, 1999) and .52 standard deviations in math on the Stanford 9 achievement tests (Balfanz, Mac Iver, & Ryan, 1999).

In total math achievement on the Stanford 9, the average national gain between fall and spring is twelve scaled score points, and nearly all the class sections at Cooke (97%) met or exceeded this average. In contrast, only 77% of the class sections in the comparison school gained at least twelve scaled score points. The gains are even more dramatic when considering the percentage of class sections that gained twenty-four points - double the national average, or roughly equivalent to two years of learning. At Cooke, 59% of the class sections gained over twenty-four scaled score points as compared to only 13% of the class sections at the comparison school (Balfanz, Mac Iver, & Ryan, 1999).

Achievement Gains at Beeber Middle School During its First Year as a TD school. The only data available to gauge Beeber Middle School's accomplishments during its first year as a TDMS school were summarized in the school district's official performance index for Beeber 8th-graders before and immediately after beginning implementation of the TDMS model. This data is more limited than that available for Central East and Cooke Middle School. However, as shown in Table 1, it suggests that ,during its first year as a TD school, Beeber Middle School also showed substantial gains in all subject areas (reading, math, and science) assessed by the school district of Philadelphia and met its performance index "growth target." (An official target established by the district's accountability system.)

[Table 1. Performance Index Scores (based on the Stanford 9 achievement of 8th-graders) at Beeber Middle School before and immediately after initial implementation of TDMS]

In sum, results from the nation's first three Talent Development Middle Schools suggest that "improvement now" is a realistic goal even in high poverty urban middle schools.

Lesson 2. Inattention to the technical core of schooling is a major source of poor student performance in high poverty, urban middle schools.

Many prior attempts to improve urban schools have operated under the assumption that the root causes of low performance are found in economic, social, cultural, and bureaucratic factors which are difficult to change and usually beyond the control of an individual school. There is considerable truth in this assumption, but it leaves a critical source of poor performance unrecognized. In high poverty secondary schools, low student achievement is actively manufactured. This occurs, often unacknowledged or unknowingly, when inattention to the technical core of schooling (curriculum, instructional materials, academic learning time, professional development etc.) severely limits students' opportunity to learn (Balfanz 1997, in press, Balfanz & MacIver, 1998).

Achievement suffers when students are not provided a coherent, consistent, and increasingly complex standards-based curriculum in each major subject area which builds year upon year in a systematic and thoughtful manner (Balfanz, MacIver, & Ryan 1999). Achievement suffers when students are not provided with organized and sustained extra-help opportunities that are linked to their classroom experiences (MacIver, Balfanz, & Plank 1998).

Teachers' ability to teach is limited when they are not provided on-going subject and grade specific professional development which gives them the content knowledge, instructional strategies, classroom management advice, and hands-on experience they need to successfully implement new standards based instructional programs or to teach well a subject with which they have limited experience (Balfanz & MacIver, 1998). Teachers' ability to teach is also limited when they do not know what their teaching assignment is going to be until the start of the school year and/or their teaching assignment is switched from a subject for which they have been recently trained to one for which they have not (Ruby, 1999). It also suffers when they are not provided with instructional materials they need until well into the school year.

Students' and teachers' ability to function at a high level is undermined when poor planning and lack of attention helps create an unruly and disrespectful school climate (Balfanz, in press). All of these problems commonly occur in high poverty middle schools and work together to manufacture low levels of student achievement. In short, in trying to address the myriad of problems and larger social context that confront high poverty middle schools, it is often the fundamentals which get overlooked.

Lesson 3. An initial focus on the technical core combined with effort to create a communal organization of schooling is therefore both essential and strategic.

Low performing, high poverty middle schools need to be thoroughly transformed. The organization of teaching and learning, the nature of classroom practice, the roles and responsibilities of students and teachers, and belief systems regarding what constitutes success often need to be fundamentally altered (Wilson & Corbett, 1999). This requires hard and sustained work by the school's faculty, administration, and students. To summon the energy needed to achieve this, to let down protective layers of cynicism about reform, and to risk changing established work patterns, teachers and students need to see an immediate benefit from reform efforts.

We have found that this can be achieved by placing an initial emphasis on the teaching and learning tools that students and teachers have to work with (curriculum, instructional materials, academic learning time, and professional development) and creating conditions which personalize, improve the quality, and increase the depth of student-teacher interactions.

Teachers notice when, in some cases for the first time in their careers, they are provided with a (1) coherent curriculum that is coordinated and builds grade by grade, (2) the essential supplies and learning materials they need to teach, (3) staff development that is immediately useful because it is linked to the instructional program they are using and (4) the grade they are teaching, and in-classroom implementation assistance from a respected peer who is there to support rather than to evaluate (Useem 1998, Darling-Hammond 1998).

Students notice when, often for the first time in their educational experience, they are provided with organized extra-help opportunities during the school day, and a curriculum which challenges and engages them (Wilson and Corbett 1999). Students and teachers notice when a variety of organizational approaches (such as small learning communities, semi-departmentalization, looping) are used to create a communal organization of schooling which enables teachers to be both more caring and daring (MacIver & Prioleau, 1999), students to receive the individual attention they need (McGrath, 1998), and students and teachers to develop respectful and supportive attitudes toward each other (Wilson & Corbett, 1999).

In the middle schools we have been working with, we have strong evidence that both students (Wilson & Corbett, 1999) and teachers (Useem 1998, 1999) draw strength and inspiration from these improvements. They recognize that it leads to a better learning environment and--in their words-- to becoming a "real" school. It also provides the energy, hope, and belief needed to deepen and sustain the implementation of comprehensive school reform.

Wilson and Corbett (1999) interviewed 210 8th-graders from six high poverty middle schools in Philadelphia to document these students' perceptions of their educational experiences in the middle grades. They found that students at Central East Middle School (School # 6 in their report) were much more likely to report that quality instruction and engaging pedagogy occurred regularly in the major subject areas than were students in the five comparison schools. Further, the students at Central East reported that their teachers consistently a) "stayed on" students to complete assignments, b) went out of their way to provide help, c) explained things until the "light bulb went on" for the whole class, d) provided students with a variety of challenging hands-on activities through which to learn, e) gave students frequent and meaningful opportunities to write, revise based upon feedback, and then "publish", f) were able to control student behavior without ignoring the lesson, and g) understood students' personal situations and factored them into their lessons.

Wilson and Corbett credit the communal organization of the TD school and the "outside support" of professional development and technical assistance provided by the TD program for helping Central East "scale up" higher quality instruction and more engaging pedagogy to encompass virtually the whole school. This is not to say that visitors to School #6 are always impressed by every lesson they observe in every classroom , but rather to say that serious learning opportunities are provided in every classroom. The students are spared the watered-down and below-grade-level learning opportunities so common in high poverty urban schools.

Useem (1998, 1999) of the Philadelphia Education Fund has completed several rounds of focus groups with teachers in TD middle schools to learn about teachers' evaluations of TDMS's training, materials, and professional development. She reports that TDMS's content-specific training and classroom support was commended by all focus group participants. The participants noted the initiative's openness to and encouragement of teachers' refinements, adaptations, and suggestions. The multiple levels of support -- from TDMS instructional facilitators, teacher leaders in the school, and curriculum coaches (teachers on special assignment) -- appear to provide a highly effective package of professional development services.

Teachers' assessments confirmed findings from the research community that teacher learning is effectively nourished when multi-layered opportunities embedded in the school and classroom are readily available. Useem's reports (1998, 1999) are encouraging because they stress that most aspects of the TD model are being well-received in the schools and that teachers are committed to moving forward with the effort.

Lesson 4. The most effective way to systematically achieve high standards is to implement and support high-quality research-based instructional programs in each major subject area on a schoolwide basis.

The eminent physicist, Freeman Dyson, in a recent series of lectures (Dyson, 1999) argues that the driving force behind scientific revolutions are not concepts but tools. From our experience, we have come to believe that the same is true for educational revolutions. The notion that all students can and should achieve at high levels in multiple subjects is a revolutionary idea, particularly for high poverty middle schools. This revolution, however, will not be achieved by exhortation, the publication of district-wide standards, or even the study of student work alone. In order for high standards for all to become a reality, the tools teachers and students need are research-based, high quality, instructional programs which have embedded within them standards-based work and the means to overcome poor prior student preparation. By instructional programs, we mean a coordinated and comprehensive set of student and teacher materials that provide students and teachers with the resources they need to engage in standards-based lessons everyday. These include daily classroom activities, longer term projects, and practice exercises, as well as, extra-help and academic recovery components.

These instructional programs need to be implemented schoolwide and across all grades in each major subject area. This does not have to occur in a rigid and lockstep fashion. There must be room for teacher innovation and individuality. But what is needed are the tools which will enable a baseline of good instruction to happen in every classroom. Establishing benchmarks, providing examples, and even identifying best practices are simply not strong enough interventions to systematically overcome both the realities and associated beliefs (and misbeliefs) that exist regarding student achievement in high poverty middle schools.

These interventions leave the hard work of creating a coherent, consistent and increasingly complex set of standards-based activities to individual teachers working mostly in isolation (Bol et al., 1998; Glennan, 1998; Newmann, Lopez, & Bryk 1998). This can result in some islands of excellence but it does not typically lead to the consistent, day -to-day standards-based instruction in all grades and classrooms that students need to overcome poor prior preparations and perform at high levels (Balfanz 1997, in press, Balfanz & MacIver, 1999; Ruby, 1999).

We have also observed that when standards-based, high quality instructional programs are implemented schoolwide they have positive effects on teacher's self-conceptions and the expectations they hold for their students. We have seen this particularly with the Talent Development Mathematics program whose goal is to prepare all students to succeed in a standards-based Algebra class in eighth grade.

Many of the teachers we have worked with have expressed initial skepticism about their students' ability to do high level work in mathematics. The experience of implementing a high quality, research based instructional program and working with us to customize and localize it to their students needs, however, has changed many of their minds. This can be seen in the following statement made by a teacher in a reflection paper written for a graduate course she took as part of the professional development ffered through the TDMS.
The Everyday Math series (the standards based mathematics curriculum used in 5th and 6th grade in TD middle schools) has restructured my approach to teaching mathematics. I am more aware of the thinking process students use to solve problems than I was before. My job takes more time now since I am looking at the entire process and not just a final answer. Since I've seen such incredible results, I'm not complaining! . . .
If someone asked me last year if they though my students would be able to conquer a math program such as Everyday Mathematics, I would have laughed and said "Yeah, right!" I am astonished at how far my students have come since September. They actually look forward to math class! Who would have thought?

Her sentiments were echoed by a number of teachers in the course.

Having said all this, our experience has also taught us that it is easier said then done. Complete high quality, standards-based instructional programs in the middle grades are rare. In subjects such as science, standards-based modules exist that often require additional materials, professional development, or adaptation to make them implementable and to construct a comprehensive curriculum that is systematic and coherent across the middle grades (Ruby, 1999). In other subjects, like social studies, there are high quality student materials (e.g. Hakim's A History of Us) but not complete instructional programs. Even in mathematics, where there are several new standards-based mathematics programs available, we have found that they still need to supplemented to provide bridges and supports for students with poor prior preparations.

Over the past five years, we have (and for at least the next five years we will) put a significant effort into developing examples of the standards-based instructional programs that are needed (for example, see Dangel & Garriott, 1999). But this is an area that is too important to be left to the developers of whole school reform models alone. School districts, foundations, and governments also need to become involved in helping to provide teachers and students with the tools (instructional programs) they need to move the standards movement from rhetoric to reality in high poverty schools.

Lesson 5. Implementing a comprehensive whole school reform model in high poverty middle schools requires multiple layers of sustained technical assistance and implementation support.

We have learned that few significant reforms or interventions are effectively implemented and take hold in low performing, high poverty secondary schools unless they are accompanied by on-site and often intensive technical assistance and implementation support. Successful and sustained whole school reform in high poverty middle schools is labor intensive. It requires continual work and attention. Providing low performing, high poverty secondary schools with a vision, a planning process, and even a blueprint accompanied by initial training and assessment tools is typically not enough.

The "managed" chaos, day-to-day stress, and high faculty and administrator mobility rates that are characteristic of low performing secondary schools continually saps the energy, divert the focus, and undermines the follow through needed to implement and sustain reforms. Through our on-going work, we have learned that each significant intervention requires multiple layers of implementation support. When we have been able to provide this support, we have seen significant success even in the most difficult schools. When we have not, the interventions have faltered (Useem 1998).

To achieve systematic improvements in teaching and learning, we have learned it is necessary to provide teachers with four layers of support. The first layer is on-going subject and grade specific staff development that is intimately linked to the curriculum they are enacting. This professional development needs to have three primary focuses. First, on a monthly basis in a very "hands on" and concrete fashion it models upcoming instructional activities for teachers. Second, it provides both the content knowledge required by these activities and demonstrates effective instructional strategies tied to the activities. Third, it provides teachers with the opportunity to network and learn from each other.

The second layer of support is non-evaluatory in-classroom implementation assistance provided by a respected peer. This curriculum coach, who is often a school district teacher on special assignment to the Talent Development model, performs a wide range of support functions including modeling, troubleshooting, helping the teacher customize the curriculum to his or her classroom, and making sure that the teacher has all the materials he or she needs.

The third layer of support is provided by lead teachers in the school who receive intensive training in the instructional programs being implemented. The final and fourth layer of support is provided by TDMS instructional facilitators who work closely with both the curriculum coaches, lead teachers, and principals to design the on-going staff development, customize and localize the instructional programs, and keep the instructional intervention on track.

At one time or another, each of these layers of support has proven critical and together they have proven robust enough to enable substantial and systematic achievement gains in high poverty middle schools (Balfanz, MacIver, & Ryan 1999, Plank & Young, 1999).

Intensive instructional support, however, is not enough. We have also learned that high poverty, low performing middle schools often need organizational assistance as well. This may involve helping schools re-organize their scheduling, staffing, and budgeting in order to support comprehensive whole school reform. This assistance can involve ordering and delivering materials to teachers or helping schools to diagnose the source of school climate problems (i.e. students in the halls, lack of respect between students and teacher, tardiness, and misbehavior) and then working with them to develop and implement effective interventions.

Lastly, schools may need assistance in how to re-conceptualize leadership roles and responsibilities to provide teacher and students the support and guidance they need to achieve at high levels. The greatest challenge is when schools need high levels of support in all areas. When this occurs, as it often does in the most troubled and low performing schools, it can be necessary to simaltaneously provide technical assistance, implementation support, and the organizing energy required to begin and initially sustain significant reforms. We have learned that, in these cases, the only way to make a significant impact is to gather the resources necessary to help schools in these multiple areas.

Any less of an intervention, or even a phased approach, is not enough to overcome the chaos and malaise.

Lesson 6. Faculty buy-in and ownership are essential, but can be achieved through the customization and localization of proven whole school reform models.

Significant and transformative reforms will not be fully implemented and sustained without the support of a large segment of the faculty and administration. It is very difficult to reform schools by fiat. At the same time, many low performing, high poverty urban secondary schools do not have the wherewithal to invent and follow through with their own reform agenda, even if they are provided with a facilitated planning process.

We believe that middle ground can be found. Our strategy is to present faculties with a proven comprehensive model designed specifically for high poverty middle schools and then achieve buy-in and ownership by inviting them to customize and localize it. Customization occurs when schools incorporate existing success and strengths into the model. Localization occurs when the model is adapted to support on-going school, district, and community initiatives and requirements. This process begins in an up to one-year long information sharing and planning process which culminates with faculty approval to move forward, and then continues throughout the implementation years.

Lesson 7. The pervasive mobility of teachers, administrators and students continually threatens the sustainability and institutionalization of even proven reforms.

The biggest obstacle to implementing and sustaining the whole school reform in high poverty urban schools is the pervasive mobility of administrators, teachers, and students. Central East Middle School, the initial pilot site for the Talent Development Middle School has had five principals in five years. This year, there is a strong likelihood that up to three of the seven middle schools we will be working with in Philadelphia will have new principals. Administrative turnover is not limited to principals. Next year, we will be working with at least two new Cluster leaders and multiple new assistant principals.

Each new principal is a challenge. Each brings his or her own vision and was not part of the initial buy-in process which brought the whole school reform model to the school. As a result, working and supportive relationships need to be built anew and each new principal has to be educated about the reform model and convinced of its merits.

We have found that a key to surviving high rates of principal turnover is to build a strong relationship with the faculty. For example, one new principal was considering fundamentally altering the Talent Development approach to extra help in mathematics. The principal listened to our arguments about its proven success. It was the passionate defense of the extra-help program and general support of the school's partnership with Talent Development mounted by several key faculty members, however, that convinced the principal to maintain its current form.

Consequently, a critical lesson we have learned is that it is a mistake to put to much faith in the ability of the principal to carry out and sustain whole school reform initiatives. In low performing high poverty middle schools, principals have a short tenure. This is especially true for strong and effective principals who often use their success in a middle school as an immediate stepping stone to a better-paying, less stressful job in the suburbs, a high school principalship, or leadership of a cluster of schools. Deep and lasting reform in high poverty middle schools will take hold only if significant numbers of the faculty are committed to them.

The problem here, of course, is that low performing, high poverty urban middle schools tend to have high rates of teacher mobility as well. The impact and sustainability of reforms are undermined by both high rates of teachers leaving the school and constant shifting of teaching assignments within the school (Useem, Christman, Gold, & Simon, 1997). Low performing, high poverty, urban middle schools are among the toughest places for teachers to work. This leads to two types of mobility. Since it is often not possible to fill all of the teaching positions with fully certified and experienced teachers, these schools employ significant numbers of novice and provisionally-certified teachers, as well as long-term substitutes. Many of these teachers do not stay in teaching. Some leave on their own accord, others are asked to leave. The ranks of these teachers are also effected by shifting enrollment patterns.

The second type of mobility affects experienced and certified teachers. Here there is almost a "culture of transfer". As soon as teachers acquire enough seniority to transfer to a stronger school, many seize the opportunity. There is also recruitment both within and from without the school district. In many respects, low performing schools serve as a farm system for stronger schools. As soon as teachers gather the experience and know-how to be effective teachers, they are sought out by other, often higher performing schools. This process is abetted in school districts which use a cost-averaging system, in which all middle schools are charged a district-wide average salary for each teacher they employ. In practice, this means that lower performing schools with young and less educated staff, provide a subsidy to strong schools with veteran and highly trained staffs.

A less visible but still significant threat to successful reform is the high rate at which teachers have yearly changes in their teaching assignments. This is in part a reaction to the high rate of teacher mobility out of the school. Teachers with seniority are often given more say in their assignments and as a result tend to pick the subjects and grades with which they are most comfortable. Because this phenomenon is greatest in districts that have primarily elementary certified teachers teaching in middle schools, it tends to have its greatest impact on mathematics and science instruction. These are the subjects that elementary certified teachers tend to feel least prepared to teach in middle school. As a result these assignments are often given to the newest and least experienced teachers. For example, at two of our TD schools, half of the science faculty were new last year. A second source of constant shifting of job assignments is that principals are juggling many competing staffing priorities. Developing strong mathematics and science staffs historically have not been given high priority (Balfanz 1997, Ruby 1999).

This turnover due to transfers and to changing teaching assignments even affects those teachers who have been designated as science resource leaders and have received special training. By Spring 1999, half of the teachers who went through an intensive three-year Science Resource Leader training program ending in June 1995 were no longer teaching science at four of the TD schools in Philadelphia (Ruby, 1999).

We have found that the only practical antidote to high rates of teacher mobility in Talent Development Middle Schools is for us to work with the school to a) create better a environment in which teachers feel a strong sense of community and receive the support they need, b) establish a permanent professional development and in-classroom implementation support infrastructure, and c) develop a relatively stable corps of teachers who are comfortable, skilled, and trained at teaching standards-based mathematics and science.

Lesson 8. You can get the technical aspects of school reform right and still be undone by getting the relationships wrong.

The simple fact is that in large multi-layered school districts, there are more people in a position to block reforms than provide active support and assistance. This occurs in large part because each layer is thinly staffed (often by very capable people). It is not atypical, for example, for the Curriculum and Instruction Department to have as few as two staff people for each major subject area, one person for the secondary grades and one person for the elementary grades. As a result, except for a few, generally small initiatives, the central office departments are pushed into a compliance role. All they can do with the resources at hand is monitor the extent to which a reform is in accordance with district mandates and goals.

This becomes particularly apparent when attempts are made to scale-up a whole school reform model to multiple schools. Since we believe that reforms need to be substantial, designed to have immediate impact, robust enough to survive in the real-world environment of urban schools, and focused on the technical core of schooling, there a few divisions of the school district that the TDMS model does not eventually intersect with. In Philadelphia, for example, we have had to coordinate our efforts with Office of Curriculum and Instruction, the Office of Assessment, the Office of Best Practices, the research department, the Urban Systematic Initiative, the Teaching and Learning Network, Office of Equity, the Chief of Staff, three cluster leaders and one school support team.

Moreover, since coordination and compliance are often in the eye of the beholder, acquiring school district support requires a lot of face time. Positive relationships must be built with many individuals, at the same time that being pulled into school district politics must be avoided. This is one reason why a local partner with deep knowledge of the school district can be essential (See Lesson 9). It is also why comprehensive whole school reform is closer to a craft than a mass production enterprise. Each school and district is to some extent unique and requires constant attention (Balfanz & MacIver, 1998).

Lesson 9. To succeed in large complex school districts, you need local friends and partners both in and out of the school district.

In our view it will be difficult for externally developed whole school reform models to go it alone on a significant scale in large urban school districts (Balfanz & MacIver 1998). Localism is very much alive at the school district level. Over time the model needs to become integrated into the fabric of the school district. This combined with the high rate of administrator turnover and the multiple reform initiatives which are occurring simultaneously (Hess 1999) means that there is a constant need to introduce and re-introduce the reform to multiple audiences, convince them of its merits, demonstrate its accordance with local initiatives, and show that you are willing and able to work with and in the best interests of the school district.

This requires a lot of work and navigation. Without friends and supporters both inside the school district and in local reform organizations who are well versed in the model and its accomplishments, it is difficult to accomplish and maintain the necessary level interaction with the school district at multiple levels. Local friends both inside and outside the school district also can play an essential role in the long term localization and institutionalization of the reform model (Balfanz & MacIver, 1998). We have found that local education funds can be excellent partners (Useem 1998b).

Lesson 10. A key to scaling up will be getting schools and districts to understand and fund the true costs of transforming low performing, high poverty urban middle schools into strong learning institutions.

The five core research and development Talent Development Middle Schools we are currently working with in Philadelphia, along with the six to nine national field test schools we will be adding this year, are not bearing the full cost of implementing the model. To varying degrees, each school is receiving subsidized support and materials through grants provided to CRESPAR and/or directly to the schools from the federal government and private foundations. This is appropriate for a research and development effort. In order for the Talent Development Middle School Model or any other successful comprehensive whole school reform model to become widespread, however, schools and districts will have to be willing to pay the true cost of reform. In the case of the TDMS model, these costs are not overbearing but in many cases will necessitate the schools and districts re-conceptualize their funding habits and priorities.

The major costs of the TDMS model can be broken down into three categories: instructional materials, professional development, and instructional facilitators/curriculum coaches. The instructional material costs are no more or less than schools traditionally spend. In each major subject area, the TDMS model uses research- and standards-based materials that are commercially available, as well as materials produced by TDMS.
Schools can stagger their purchase of these materials (buying one subject a year) to spread their costs over time (much as they would do in the absence of adopting the TDMS model.)

Professional development costs will vary by school district depending on the extent to which teachers must be paid to attend professional development, the local rate of pay, and the amount of professional development time built into the school's calendar. Philadelphia has a relatively high rate of a little more than 20 dollars per hour. This means it costs approximately $700 per teacher per subject to provide the recommend 35 hours of professional development per year (for two years). In Philadelphia, however, we have been able to dramatically reduce these costs by forming a partnership with a local university to provide teachers three graduate elective credits for 35 hours of professional development for a reduced rate of 175 dollars.

The biggest, and perhaps most essential cost of implementing the TDMS model is the expense of providing sustained in-classroom implementation support. We have learned that, at a minimum each school, needs to receive one day a week of support in each core academic subject they are implementing. We have also learned that the instructional facilitators/curriculum coaches providing this support can successfully work with no more than three schools at a time. From this we have developed a basic formula for providing in-classroom implementation support. It requires each of three schools to fund one position. These three schools then share the services of three subject specific instructional facilitators/curriculum coaches (e.g. one English, one Math, and one Science coach). In this way, the schools can be implementing up to three core subjects at one time and each instructional facilitator/curriculum coach can spend one day a week at each school, have flex day to spend at the school in most need during a given week, and a day to plan, develop materials, and organize.

In addition to this basic support, we have also learned that schools also often require supplemental support to launch effective extra-help programs in math and English and/or to help organize the school for change (i.e. scheduling, improving school climate etc.). Taken together, this implies that schools will need to budget from 60 to 90 thousand dollars per year to fund the in-school support we have found is needed to bring about and sustain transformative change and improvement.

All in all, we estimate that it will cost approximately 100 to 150 thousand dollars a year for about five years to successfully turn low performing, high poverty urban middle schools into strong learning institutions. Part of the challenge in the near term will be convincing schools and districts that this is not a large sum.

If a 1,000 student middle school spends about 5,000 dollars per student per year, then 100 thousand dollars represents 1% of the school's budget. Finding 1% to 2% for effective reform is an achievable goal. But it will take work. Schools and districts will need help and encouragement to find ways to re-allocate existing funds or raise additional money.

For some schools this will be relatively painless. Many large, high poverty urban middle schools receive one hundred to one hundred and fifty thousand in Title 1 funds. For other schools, making student achievement and teacher learning the overriding goal may necessitate changes in the school's staffing model. In many of the Philadelphia schools we work with, for example, there is almost a 1 to 1 ratio between teaching and non-teaching positions. Part of the challenge will be in convincing schools which are used to supporting multiple piecemeal initiatives that they need to husband their resources and apply them all to one proven comprehensive reform efforts.

Another challenge will be pointing out that some of the obvious cost-cutting measures do not work. When initially presented with the true costs of reforms, several schools have proposed giving their own teachers partial release time toserve as in-school instructional facilitators. The problem with this is two-fold. First, these teachers are often pulled off task to work on short term administrative chores such as administering Title 1 tests, providing extra discipline support, and most commonly class coverage (when subs can not be found). As a result, the are unable to provide a regular schedule of teacher support. Second, these teachers are often veterans who are used to doing things in certain way, and as a result, do not become strong advocates or experts in the TDMS research- and standards-based curriculum programs. Rather they see their role as providing teachers with their own wisdom born of experience. Sometimes this wisdom is good, sometimes it is not, but in either case it often does not support the systematic implementation of the TDMS instructional programs.

Where does this leave us?

After five years of working to develop and implement a comprehensive whole school reform model that has been specifically designed for low performing, high poverty schools, we are both hopeful and sobered. Hopeful in that we have seen schools achieve substantial and systematic improvements in both student learning and achievement and teacher support and performance. This tells us that it can be done and much more quickly than is commonly assumed.

At the same time, we are sobered by how much constant attention, energy, and hard work this progress has required from the developers, the instructional facilitators and curriculum coaches, the administrators, the teachers, and our local partners. We are hopeful because many of the key reforms are broadly reproducible. We are sobered however that few school districts are creating the infrastructure needed to support such reforms. In closing, we agree with Wilson and Corbett (1999) that in order for high poverty middle schools to become strong learning institutions, they will need to move from an ethos that seeks to do the best that can be done under difficult circumstances to a "No Excuses" mentality which relentlessly finds ways to help all students succeed. What remains unknown is how this can be accomplished on a large scale in multiple settings and sustained over time. These are the questions to which will turn our attention over the next five years.


References

Balfanz, R. (1997, March). Mathematics for all in two urban schools: A view from the trenches. Presented at the Annual Meeting of the American Educational Research Association, Chicago, Illinois.

Balfanz, R. (in press). Why do so many urban public school students demonstrate so little academic achievement?: The underappreciated importance of time and place. In M.G. Sanders (Ed.) Schooling at risk: Research, policy, and practice in the education of poor and minority adolescents, Erlbaum Associates, Inc.

Balfanz, R. & Mac Iver, D.J. (1998) The school district's role in creating high performing middle schools. Baltimore, MD: Johns Hopkins University Center for Research on the Education of Students Placed at Risk.

Balfanz, R., Mac Iver, D.J., & Ryan, D. (1999). Achieving algebra for all with a facilitated instructional program: First year results of the Talent Development Middle School Mathematics Program. Paper presented at the annual meeting of the American Educational Research Association, Montreal.

Bol, L., Nunnery, J. A., Lowther, D. L., Dietrich, A. P., Pace, J. B., Anderson, R. S., Bassoppo-Moyo, T. C., & Phillipsen, L.C. (1998). Inside-in and outside-in support for restructuring: The effects of internal and external support on change in the New American Schools. Education and Urban Society, 30(3), 358-384.

Boykin, A W. (1994). Harvesting talent and culture: African-American children and educational reform. In R. J. Rossi (Ed.), Schools and students at risk (pp 116-138). New York: Teachers College Press.

California State Department of Education. (1987). Caught in the middle: Educational reform for young adolescents in California public schools. Sacramento: California State Department of Education.

Carnegie Task Force on the Education of Young Adolescents. (1989). Turning points: Preparing American youth for the 21st century. New York:

Carnegie Council on Adolescent Development of the Carnegie Corporation. Children's Defense Fund. (1988). Making the middle grades work. Washington, DC: CDF.

Cooney, S. (1998). Education's weak link: Student performance in the middle grades. Atlanta, GA: Southern Regional Education Board.

Dangel, S. & Garriott, M. (1999). Teaching Guides for Books 6-10 of "A History of US" by Joy Hakim. Baltimore, MD: Talent Development Middle School American History Curriculum, Center for the Social Organization of Schools, Johns Hopkins University.

Darling-Hammond, L. (1998). Teacher learning that supports student learning. The Best of Educational Leadership 1998. (pp.6-10) Alexandria, VA: ASCD.

Dorman, G. (1984). Middle grades assessment program. Carrboro, NC: Center for Early Adolescence, University of North Carolina, Chapel Hill.

Dyson, F. J. (1999). The sun, the genome, and the internet: Tools of scientific revolutions. New York: Oxford University Press.

Felner, R.D., Jackson, A.W., Kasak, D., Mulhall, P., Brand, S., & Flowers, N. (1997, March). The impact of school reform for the middle years: Longitudinal study of a network engage in Turning Points-based comprehensive school transformation, Phi Delta Kappan, 528-532, 541-550.

Neild, R.C., & Weiss, C. C. (1999). The Philadelphia Education Longitudinal Study (PELS): Report on the transition to high school in the School District of Philadelphia. Philadelphia Education Fund.

Glennan, T. K. (1998). New American Schools after six years. Washington, D.C.: RAND.

Hess, F.M. (1999). Spinning wheels: The politics of urban school reform. Washington, DC: Brookings.

Johnston, J.H. & Williamson, R.D. (1998, April). Listening to four communities: Parent and public concerns about middle level schools, NASSP Bulletin.

Killion, J. (1999). What works in the middle: Results-based staff development. Oxford, OH: National Staff Development Council.

Killion, J & Hirsh, S. (1998, March 18). A crack in the middle, Education Week on the Web.

Lipsitz, J., Mizell, M.H., Jackson, A. W. & Austin, L.M. (1997, March). Speaking with one voice: A manifesto for middle grades reform, Phi Delta Kappan, 533-540.

Lipsitz, J. (1984). Successful schools for young adolescents. New Brunswick, NJ: Transaction Books.

Mac Iver, D.J. Balfanz, R., & Plank, S. B. (1998). An 'elective replacement' approach to providing extra help in math: The Talent Development Middle Schools' Computer- and Team-Assisted Mathematics Acceleration (CATAMA) Program. Research in Middle Level Education Quarterly, 22(2), 1-23.

Mac Iver, D. J., Balfanz, R., Plank, S. B., & Ruby, A. (1998). The national field test of the talent development middle school. Baltimore: Johns Hopkins University Center for the Social Organization of Schools.

Mac Iver, D. Mac Iver, M., Balfanz, R. Plank, S.B., & Ruby, A. (in press). Talent Development Middle Schools: Blueprint and results for a comprehensive whole- school reform model. In M.G. Sanders (Ed.), Schooling students placed at risk: Research, policy, and practice in the education of poor and minority adolescents. Mahwah, NJ: Erlbaum.

Mac Iver, D. J., & Plank, S. B. (1997). Improving urban schools: Developing the talents of students place at risk. In J. L. Irvin (Ed.), What current research says to the middle level practitioner. (pp. 243-256). Columbus, OH: National Middle School Association.

Mac Iver, D. J., Plank, S. B., & Balfanz, R. (1997). Working together to become proficient readers: Early impact of the Talent Development Middle School's Student Team Literature Program. (Report 15). Baltimore, MD & Washington, DC: Center for Research on the Education of Students Placed at Risk

Mac Iver, D.J. & Prioleau, A.D. (1999). Looping: Helping middle school teachers to be caring, daring, and effective. Paper presented at the annual meeting of the American Educational Research Association, Montreal.

Marlyand Task Force on the Middle Learning Years. (1989). What matters in the middle grades? Baltimore: Maryland State Department of Education.

McGrath, A. (1998, April 20). Algebra and sympathy. U.S. News & World Report, 57-58.

Mosteller, F. D., Light, R. J., & Sachs, J. A. (1996). Sustained inquiry in education: Lessons from skill grouping and class size. Harvard Education Review, 66(4), 797-842.

Newmann, F. M., Lopez, Gudelia, L. & Bryk, A.S. (1998). The quality of intellectual work in Chicago schools: A baseline report. Chicago: Consortium of Chicago School Research.

Plank, S.B. & Mac Iver, D.J. (1998, January). The Talent Development Middle School: A blueprint for reform. Principal, 77(3), 54-55.

Plank, S. & Young, E. (1999). In the long run: Longitudinal assessments of the Student Team Literature Program. Paper presented at the annual meeting of the American Educational Research Association, Montreal.

Ruby, A. (1999). An implementable curriculum approach to improving science instruction in urban middle schools. Paper presented at the annual meeting of the American Educational Research Association, Montreal.

Schmidt, W.H., McKnight, C.C., & Raizen, S.A. (1997). A splintered vision: An investigation of U.S. science and mathematics education. Dordrecht: Kluwer Academic Publishers.

Useem, E.L. (1998). Teachers' appraisals of Talent Development Middle School training, materials, and student progress: Results from six focus groups at Central East Middle School and Cook Middle School. Baltimore: Center for Research on the Education of Students Placed at Risk.

Useem, E. L. (1998b) Local education funds: What they do, how they do It, and the difference they make: A report on the Public Education Network. Public Education Network, Washington, D.C.

Useem, E.L. (1999, March). Year two talent development at Cooke middle school: A report from two focus groups. Philadelphia: Philadelphia Education Fund

Useem, E., Christman, J.B., Gold, E., & Simon, E. (1997). Reforming alone: Barriers to organizational learning in urban school change initiatives. Journal of Education for Students Placed At Risk. 2(1), 55-78.

Wilson, B.L. and Corbett, H.D. (1999). "No excuses": The eighth grade year in six Philadelphia middle schools. Philadelphia Education Fund.