DISSEMINATING SUCCESS FOR ALL:

Lessons for Policy and Practice

Robert E. Slavin, Johns Hopkins University

Nancy A. Madden, Success for All Foundation

Report No. 30 (Revised) 

Published by the Center for Research on the Education of Students Placed At Risk (CRESPAR), supported as a national research and development center by funds from the Office of Educational Research and Improvement, U. S. Department of Education (R-117-D40005). The opinions expressed in this publication do not necessarily reflect the position or policy of OERI, and no official endorsement should be inferred.

The Center

Every child has the capacity to succeed in school and in life. Yet far too many children, especially those from poor and minority families, are placed at risk by school practices that are based on a sorting paradigm in which some students receive high-expectations instruction while the rest are relegated to lower quality education and lower quality futures. The sorting perspective must be replaced by a "talent development" model that asserts that all children are capable of succeeding in a rich and demanding curriculum with appropriate assistance and support.

The mission of the Center for Research on the Education of Students Placed at Risk (CRESPAR) is to conduct the research, development, evaluation, and dissemination needed to transform schooling for students placed at risk. The work of the Center is guided by three central themes — ensuring the success of all students at key development points, building on students’ personal and cultural assets, and scaling up effective programs — and conducted through seven research and development programs and a program of institutional activities.

CRESPAR is a partnership of Johns Hopkins University and Howard University.

Abstract

Success for All, a comprehensive schoolwide reform program for elementary schools serving many children placed at risk of school failure, was first piloted in one Baltimore elementary school in the 1987-88 school year. Since then, the program has expanded rapidly; as of 1999-2000, it is in more than 1,500 schools in 400 districts in 46 states throughout the United States.

Success for All requires substantial change in many aspects of curriculum and instruction and extensive professional development to help schools start children with success and build on that foundation through the elementary grades. It takes time for teachers to learn and perfect new forms of instruction, and for other school personnel to learn new roles. Therefore, the program requires a great deal of professional development done over an extended period of time.

This report describes our experience — the strategies pursued, the relative success of various dissemination routes, and the implications of those experiences for public policies. This experience has led us to conclude that successful dissemination requires a combination of talented, dedicated trainers and a local and national network of schools willing and able to provide technical and emotional support. Employing full-time, regionally-based trainers recruited from outstanding Success for All schools is the most effective way of building our staff. In contrast, strategies depending on involvement of other organizations or on part-time certified trainers have rarely been successful.

To offer research-based, replicable school designs to large numbers of schools, national and state policymakers need to help existing and new reform networks build up their training capacity, and provide funds to schools to help them with start-up costs.

Introduction

Never in the history of American education has the potential for fundamental reform been as great. The bipartisan embrace of ambitious national goals, the reauthorization and substantial restructuring of Title I, the availability of new designs for school change supported by the New American Schools (NAS), and the growing strength of such national professional development networks as Sizer’s Coalition for Essential Schools, Comer’s School Development Project, Reading Recovery, and the National Writing Project, are all developments that create unprecedented possibilities for change. New federal legislation that provides competitive funding to schools to adopt proven, comprehensive reform designs adds both resources and attention to the movement toward school-by-school, standards-based reform.

However, it is by no means certain that the potential for reform will be realized. Changes will take place, but will these changes actually make a difference in the school success of large numbers of children? For this to happen, the nearly three million teachers in America’s schools will have to learn and regularly apply very different and far more effective instructional methods than those they use now. School organization, assessment, grouping, and many other aspects of schooling will have to change. The systemic changes happening at many levels of government are creating a fast-rising demand for high quality, sustained professional development, particularly the professional development needed for schools to adopt proven models of school change. Yet the national infrastructure for professional development of this kind is quite limited.

If reform is to produce results, major changes in the structure of professional development are needed. This report is intended to shed light on the question of how a national approach to professional development might enable professional development networks to bring proven school change models to scale by describing the lessons we have learned in disseminating Success for All, a comprehensive reform program for high-poverty elementary schools. In the course of disseminating Success for All we have learned a great deal about the process of change, about factors that support and inhibit school-level reform, and about ways of enlisting others in support of our efforts. This report describes our experience with dissemination, the strategies we are pursuing, the relative success of various dissemination routes, and the implications of our experiences for public policies.

Success for All

Success for All (Slavin et al., 1992, 1994, 1996; Slavin & Madden, 1999a; Madden et al., 1993) is a program designed to comprehensively restructure elementary schools serving many children placed at risk of school failure. It emphasizes prevention, early intervention, use of innovative reading, writing and language arts curriculum, and extensive professional development to help schools start children with success and then build on that foundation throughout the elementary grades. The box below summarizes the main elements of the program.

Research comparing Success for All to control schools in 23 schools in nine districts has consistently shown that Success for All has substantial positive effects on student reading achievement throughout the elementary grades (Slavin et al., 1994, 1996; Madden et al., 1993) as well as reducing special education placements and retentions and improving attendance (Slavin et al., 1992, 1996; Slavin & Madden, 1999).

In recent years we have developed and evaluated programs in mathematics (MathWings) and in social studies/science (WorldLab). In general, schools implement Success for All first, and then add MathWings and/or WorldLab in subsequent years. Research also shows positive effects of MathWings (Madden, Slavin, & Simons, 1998) and WorldLab (Slavin & Madden, 1999b).

Success for All was first piloted in one Baltimore elementary school in the 1987-1988 school year. In 1988-89 it was expanded to a total of five schools in Baltimore and one in Philadelphia. Since then the number of schools has grown by 40 - 100% each year. In 1998-99, Success for All is being implemented in approximately 1,500 schools in 400 districts in 46 states throughout the United States, and has been adapted for use in five other countries. Of these, approximately 150 are using MathWings and/or WorldLab.

Table 1 shows the growth in numbers of schools implementing Success for All from 1987 to the present. Figure 1 shows the same data for the years 1990-1999. Note that the annual percentage gain has never fallen below 40%, and every year the number of schools added is more than those added in the previous year. This is a very rapid rate of growth; for example, in three years (1995-1998), the total number of schools has almost quadrupled, after sextupling in the three years before that. Over the past two years (1997-99), the number doubled again. At this growth rate we have to continually focus on maintaining the quality of our implementations as it becomes impossible for our central staff to know every school. This paper discusses the successful as well as unsuccessful strategies we have pursued to enable us to scale up our operation without compromising on quality and effectiveness.

Table 1

Approximate Numbers of Success for All/ Roots & Wings Schools, 1987-1999

Program Characteristics Affecting Dissemination

There are several unique characteristics of Success for All that have an important bearing on the strategies we use in disseminating the program. First, while Success for All is always adapted to the needs and resources of each school using it, there are definite elements common to all. A fully functional Success for All school will always implement our kindergarten program and reading program in grades 1-5 or 1-6, will have at least one tutor for first-graders, and will have a full-time facilitator and a family support team. Other elements, such as preschool and full-day kindergarten, are optional, and schools vary in the number of tutors, the staff time devoted to family support, and other features. Yet despite this variation, we believe that the integrity of the program must be maintained if schools are to produce the results we have found so consistently in our research. The whole school must make a free and informed choice to adopt Success for All; we require a vote by secret ballot of at least 80%. But when schools make this choice they are choosing a particular model of reading instruction, a particular use of Title I and special education resources, a particular within-school support structure, and so on. Unlike many alternative schoolwide change models, Success for All is not reinvented from scratch for each school staff.

Success for All requires substantial change in many aspects of curriculum and instruction. It takes time for teachers to learn and perfect new forms of instruction, and for facilitators, tutors, family support team members, and principals to learn new roles. Therefore, the program requires a great deal of professional development done over an extended period of time. While the initial training period is only three days for classroom teachers, many follow-up visits from Success for All trainers take place each year. Schools budget for 23 person-days of training in the first implementation year, 12 in the second, and 5-8 in each subsequent year.

Success for All requires that schools invest in tutors, a facilitator, materials and extensive professional development. Because of the focus of the program and its cost, the program is primarily used in high-poverty schools with substantial Title I resources. As of 1999-2000, the cost of the program for a school of 500 students averages $73,000 in the first year for materials and training, plus salaries for a facilitator, tutors, and other staff (usually reallocated from other functions). Success for All schools rarely receive funds beyond their usual Title I allocations, so in one sense the program has no incremental costs, but there are many schools that could not afford a credible version of the model. While the cost of the program does restrict its use, it also has an important benefit: it increases the likelihood that the school and district will take it seriously and work to see that their investment pays off.

The comprehensiveness, complexity, and cost of Success for All have important consequences for dissemination. First, they mean that the program cannot be mandated en masse. Instead, districts usually start with a few schools and gradually add more. Second, they mean that the commitment to the program must be long-term, and we must be prepared to be engaged with schools for many years, perhaps forever. Third, they mean that we must maintain a large, very highly skilled staff of trainers to work with schools. While we do occasionally use teachers and facilitators from successful schools in our training programs, the program does not lend itself to an easy "trainer-of-trainers" strategy in which a small staff trains local trainers to work with schools.

Dissemination Staff

Until July, 1998, the dissemination of Success for All was primarily carried out by our staff at the Center for Research on the Education of Students Placed at Risk, Johns Hopkins University. Since that time, our dissemination has been moved to a separate not-for-profit organization, the Success for All Foundation, or SFAF (see below). In fall, 1999, our training staff consists of approximately 250 full-time trainers, plus 30 in two partner organizations (see below). Almost all of our trainers are teachers; most have been building facilitators or teachers in Success for All schools. The only trainers who are not former teachers are those who focus on family support. Their backgrounds are usually in social work or counseling.

The trainers who work for SFAF are organized in 16 regions of the U.S., each with a very experienced trainer as a regional manager. In addition, we have a small number of part-time trainers (some of whom have formerly worked for us) located in various parts of the country, and we will often ask an especially talented teacher or facilitator to help us with training and follow-up in their own area.

In addition to SFAF staff, there are also two regional training programs for Success for All. One is at Education Partners, a small for-profit organization that has developed marketing, training, and follow-up capacity for Success for All. We negotiated an agreement with Education Partners that gives them rights to disseminate Success for All in a territory consisting of the San Francisco Bay Area, Washington, Oregon, Idaho, Colorado, and New Mexico. Education Partners agreed to a set of performance standards that specify the quality of implementation we expect to see in schools they work with. In 1999-2000, Education Partners is working with approximately 180 schools.

The second currently functional regional training site is at the University of Memphis. This group, led by Steven Ross and Lana Smith, has had a long-standing involvement with Success for All schools in Memphis, and has conducted research on Success for All in districts around the United States. The University of Memphis group has taken responsibility for implementing Success for All in Tennessee, Arkansas, Georgia, Mississippi, Louisiana, and Missouri.

Formerly, WestEd, a federally-funded educational laboratory, maintained a regional training program for Success for All in most of California, Arizona, Utah, and Nevada. However, problems with this arrangement led to its cancellation in April, 1998. Most of the SFA trainers working for WestEd moved to the Success for All Foundation, which now serves schools in the former WestEd region.

Dissemination Strategies

Schools first become aware of Success for All in a variety of ways. Many articles have been written about the program in educational journals, and our staff has made many presentations at conferences. We have an awareness video and materials, including a book describing the program and its outcomes. Educators may write for information, call members of our dissemination staff, or otherwise make contact with us. School or district staff may then invite our staff to make awareness presentations. These often take place as part of "effective methods fairs" in which large districts or states invite principals or school teams to learn about many promising models. We encourage schools to send delegations to visit other Success for All schools in their region if at all possible. If there is interest in schools after these awareness presentations, school staff will send us a "Preliminary Data Form," which enables us to calculate a price for training and materials. We will negotiate a contract specifying what we and the school and the district promise to do. The contract makes our intentions and requirements clear. At some point a presentation will be made to the whole staff of each interested school. Following opportunities to examine materials, visit other schools, and discuss among themselves, school staffs vote by secret ballot. As noted earlier, we require a positive vote of at least 80% of the professional staff. It is rare that we would go through the entire process and then have a vote of less than 80%; more often votes are closer to 100% positive. However, the exercise is essential in that it assures teachers that they had a free choice and that the program is supported by the great majority of their colleagues.

This integrity of this vote is so important to us that in some districts in which we suspect there may be pressure to implement Success for All, we have asked the local teacher's unions to monitor the voting. We have found that it is very difficult to work in schools in which the staff did not make an overwhelming, informed, and unfettered choice to work with us.

As soon as a school has decided to adopt the program, planning for implementation begins. A member of the SFAF staff or one of our regional training sites is appointed to serve as the school’s lead contact.

A facilitator is then chosen, usually an experienced and respected teacher from within the school’s own staff. The facilitator and the principal will attend a week-long training session in one of a few central locations in different parts of the U.S. held well in advance of training for the school staff. For example, we hold our main facilitator/principal "new sites" trainings in May, June, and July for schools planning their training for teachers in August. This gives the facilitators and principals time to work out issues of staffing, space, finances, ordering and storing materials, and so on. Facilitators may also visit other schools to see the program in action and to get a first-hand view of what facilitators are expected to do.

If a school is planning to begin Success for All in September, training will generally take place over a three-day period in August. Additional training is provided later for tutors and for family support staff.

Some schools start at mid-year, implementing portions of the program that can be started then, but holding off the first-grade reading program until the following September.

The initial training is typically done by the school’s "point trainer," other staff from SFAF or regional training organizations, and (occasionally) adjunct trainers who are facilitators or teachers in existing Success for All schools. After initial training, follow-up visits will be conducted by these same staff. A first-year school will receive 6-8 person-days of followup.

Our main objective during follow-up visits is to strengthen the skills of the building facilitators and principals. We cannot hope to adequately monitor and refine implementations from a great distance; instead, we must rely on the facilitator, who is the change agent within the school, as well as the principal and teachers. Our staff members jointly conduct an implementation review, visiting classes, interviewing teachers, family support members, tutors, and others, and looking together at the data on student performance, pacing, attendance, special education placements, and so on. Our trainers model ways of giving feedback to teachers, give the building facilitators advice on solving their problems, share perspectives on strengths and weaknesses of the program, and plan with the building facilitator and principal the goals for individual teachers and for general program implementation that the facilitator will follow up on. Trainers meet with teachers to provide additional training on such issues as writing, pacing, or classroom management. They respond to questions and discuss issues needing further attention. Later, trainers complete implementation check forms and write up site reports summarizing what they have seen, noting promises made, issues to be followed up on, and ratings of the quality of implementation of each program element.

In general, we are very satisfied with the dissemination model we are using. In regular implementation checks that are part of our follow-up visits, we find more than 90% of teachers to be doing an adequate job of implementing the programs, and many teachers are doing inspired teaching, using our materials and methods as a jumping-off point for innovative and exciting instruction. The relative prescriptiveness of the model and the training and follow-up that support it are sometimes perceived to be problematic before implementation begins, but are rarely a long-term problem, as teachers and other staff come to see the flexibility within the program and to see the outcomes for children. In fact, for teachers used to inadequate professional development without the material or human supports necessary to change their teaching on a day-to-day basis, the completeness of Success for All, from materials to training to follow-up, is a major plus. The consistent positive findings in evaluations of Success for All in its dissemination sites tell us that our model of dissemination is working.

However, while we are confident that the Success for All program can be successfully adapted to local circumstances and replicated nationally using the model of dissemination we have evolved, the problem we face is how to provide such an intensive level of service on a broad scale. America has more than 50,000 Title I elementary schools. We are currently working in more than 1,500 schools, about 3%. Our network of schools has been expanding by about 40-50% each year, an enormous rate of growth. We have had to continually restructure ourselves to accommodate this growth without compromising on quality, and will need to continue to do so for the foreseeable future. We still have a long way to go.

Obey-Porter Comprehensive School Reform Demonstration (CSRD)

In 1997, the U.S. Congress allocated $150 million for a new program designed to assist schools in adopting "proven, comprehensive" reform designs. These are programs that provide external assistance to schools to upgrade their curricula, parent involvement approaches, assessments, professional development approaches, and other features. Success for All and Roots & Wings were named in the legislation among 17 examples of such comprehensive designs. Schools can apply for three-year grants of at least $50,000 per year to pay for the start-up costs of adopting comprehensive designs. Funds were also allocated for labs and state departments of education to help in the awareness and review process (see Slavin, 1998).

The CSRD program could have a profound impact on scaling up of Success for All and Roots & Wings. Obviously, it provides funding for schools that might not have been able to afford adequate implementations. More importantly, perhaps, it raises the profile of programs like SFA/R&W at the policy level. State departments of education, laboratories, and districts are all engaged in disseminating information about comprehensive reform models, and this may have an impact on their own involvement in and knowledge about whole-school reform.

One difficulty inherent to CSRD, however, is that it is likely to greatly increase the demand for research-based reform models without increasing the supply. All existing programs will be struggling to add training capacity without reducing quality. Another potential difficulty with CSRD is the possibility that some schools will apply for funds because they want money, not because they are eager for reform. With this danger in mind, we have tried to ensure that our rigorous standards for schools to adopt SFA are followed in spirit as well as process. We are issuing certificates to schools that fully meet our standards, and are asking state departments to make sure that any application to implement Success for All or Roots & Wings encloses a certificate from us, which commits us to provide training and materials to the school if it is funded.

In CSRD grants made to date (fall, 1999), 212 out of 1431 grants in 40 states have been made to schools to implement SFA/R&W. This is the largest number of grants to any program, more than to the next three programs combined. However, the surprise in the CSRD process is that grants are being made to support an enormous array of programs that are neither well-researched nor even comprehensive. Collectively, the 17 programs listed in the legislation are receiving only about half of the grants.

Extending Our Reach

As Success for All has become a national program, we have had to confront the problem of providing adequate training and follow-up in many widely dispersed locations with very different needs, resources, and circumstances. Early on, we began searching for ways of engaging regionally based educators in training or support roles, to extend our training capacity, to reduce travel costs for schools, and to provide schools with trainers who are more familiar with the local scene. For a program as complex as Success for All, with such extensive requirements for training and follow-up, it is not a simple matter to train trainers to work in their own areas. As we disseminate Success for All we do not want to compromise on the quality or integrity of the model we have developed and researched. It is difficult to train educators who have not been teachers or facilitators in Success for All schools, and the need for lengthy follow-up makes it difficult to have part-time trainers with other jobs play a major role in training. With these concerns in mind, however, we have pursued a variety of strategies for building a local and regional capacity for training, follow-up, and support. The following sections discuss our experiences with each.

Regional Training Sites

As noted earlier, we have two regional training sites for Success for All managed by other organizations. One of these is at the University of Memphis and one at Education Partners. The stories of how these sites were established and how other attempts to create regional training sites in other organizations have failed provide an interesting perspective on the possibilities and difficulties of regional training strategies.

Universities

One obvious candidate for regional training sites is universities. This is the route taken by several other national school reform networks, such as Reading Recovery (Pinnell, DeFord, & Lyons, 1988) and Accelerated Schools (Levin, 1987). However, Success for All does not lend itself as easily to dissemination from universities. Reading Recovery is a tutoring program for at-risk first graders that provides its training as courses with graduate credit. It therefore fits easily into established structures. Accelerated Schools emphasizes an organizational development consulting approach that is also familiar to university faculty members (see McCarthy, 1991). In contrast, working with whole schools over extended time periods is an unusual activity for university faculty, who are typically too involved with courses, committees, and research to put much time into such activities. With the sole exception of the University of Memphis, no university has attempted to establish a regional training program for Success for All.

Our successful training site at the University of Memphis exists because of some unusual circumstances. The University of Memphis has a Center for Research in Education Policy funded primarily by the State of Tennessee. Two researchers in that Center, Steven Ross and Lana Smith, developed a close relationship with the Memphis City Schools in the course of their research there. In 1989, they began working with one Memphis Success for All school and assessing outcomes in that school. Later, we contracted with them to conduct independent evaluations of Success for All in several additional districts. Initially, their main involvement with Success for All was as researchers, not trainers. However, in 1992, a new superintendent in Memphis, Gerry House, became interested in Success for All and asked their help in scaling it up. Since then, the University of Memphis center has begun work in additional districts in Tennessee, Georgia, Arkansas, Mississippi, Louisiana, and Missouri. They have built up their own training staff, and also use part-time trainers who are from Memphis City Schools.

The success of the University of Memphis regional training site depends on several relatively unique characteristics. One is the existence of a research center at the university. Another is the unusual motivation and skill of the researchers, and their close relationships both with our center and with the Memphis City Schools. However, it is important to note that the University of Memphis training site came into being through a traditional university activity, research, and not training per se. In fact, the emphasis of this center is still as much on research as on training. Other attempts to recruit universities to house regional-training programs have not worked out.

Education Partners

Education Partners (EP), a for-profit company, is the largest regional training program for Success for All housed in a different organization. Headquartered in San Francisco, EP currently serves approximately 180 schools in the San Francisco Bay Area, Oregon, Washington, Idaho, Colorado, and New Mexico. In 1996, EP's President approached us about taking on a training role. At the time, EP was very small, and had only been in operation for less than two years, contracting to help reconstituted schools in San Francisco with professional development. EP proposed to carry out dissemination of SFA in a defined region under a stringent set of performance standards, to be monitored by us, that required them to maintain a high quality of training and implementation at each school they served. They agreed to pay JHU (currently SFAF) a set fee on all revenues. We agreed to EP's proposal as an experiment, to see if a for-profit organization could do a better job than we could as a not-for-profit. Later, when a contract with the Xerox Corporation to do our printing and distribution fell apart in fall, 1996, EP bid for and won a contract to take on this function as well.

Because EP started out so small, it was able to design itself solely for the purpose of serving as a training program for SFA. With a few exceptions (such as a more proactive marketing plan), EP operates much the same way as SFAF does. However, while there was an initial plan to gradually expand EP's training territory, this is no longer contemplated. The problem is in the for-profit nature of EP. While any operating surpluses in SFAF go into further development and research, those in EP go to investors. The involvement of SFAF with a for-profit organization has created some political and legal problems. For example, it led to concerns at the IRS in terms of our not-for-profit status, and concerns in teacher's unions and policy circles about our motives. EP will continue to add schools within its region, and continues to play an important role in printing and distribution for our entire network of schools. However, our experience does not support the idea of subcontracting to for-profit organizations for training services.

Educational Laboratories

The regional laboratories would appear to be ideal organizations to become regional training sites for Success for All. They are responsible for helping districts in their regions learn about and implement effective programs. In fact, when they were first established in the 1960s, labs were meant to complement the work of national research centers, such as the one at Johns Hopkins in which Success for All was developed and researched. As originally conceived, research centers were designed to do basic and applied research on important educational problems, while labs were supposed to interpret this research for their regions and help local schools apply the findings. Labs and centers are funded by the same agency and are still expected to work together.

In fact, lab-center collaboration is relatively uncommon, and it is almost unheard of for labs to actively disseminate programs developed and researched in research centers. Labs often develop (and disseminate) their own programs and see little advantage in disseminating others’ work. Also, labs must serve an enormous region, and it is difficult for them to provide services to individual schools.

Despite these problems, we attempted to engage labs in support of Success for All dissemination. We spoke to lab directors and lab communication directors, and had various communications with individual labs. However, WestEd, in California, was the only lab to establish a regional training program for Success for All. The North Central Lab (NCREL) was once helping us work with a few Chicago schools, but other labs have expressed little or no interest.

The arrangement with WestEd was initially successful, based in part on a history of prior collaboration between Johns Hopkins and the individual who started the regional training program at WestEd. However, ultimately this arrangement did not work out. Part of the problem was in maintaining consistency between JHU/SFAF and WestEd; in many cases, WestEd reinterpreted SFA policies, failed to implement various program elements, or otherwise insisted on its own approaches. These and other problems led to a schism within the WestEd SFA staff, with more than half resigning or threatening to resign during the 1997-98 school year. SFAF took back the region in April, 1998.

While our experiences with WestEd and with labs in general does not support the idea of having labs establish their own training programs, labs can be helpful in an awareness and brokering role. In particular, the Obey-Porter Comprehensive School Reform Demonstration, described earlier, provided $500,000 grants to each lab to help schools and districts in their regions learn about and adopt effective whole-school reform designs. The labs have been helping states set up awareness activities, such as effective methods fairs, to help schools and districts apply for funding to implement proven designs, including Success for All and Roots & Wings.

State Departments of Education

One potential source of assistance in providing regionally-based training and assistance to Success for All schools is state departments of education. We have worked with a few state departments and have found them to be helpful in some ways, but have not yet found a way to have state departments play a major role in dissemination of our program.

The state department with which we have had the greatest involvement is in New Jersey. There, starting in 1994, we worked with an Urban Education Initiative related to a statewide court-ordered plan to increase state funding to the 28 lowest-wealth districts. The state department staff involved with this initiative coordinated several meetings with high-poverty districts and intended to build its own training and support capacity for Success for All. However, the political turmoil resulting in part from the tax increases necessitated by the funding equity decision caused a change of control in the legislature and then of the governorship, dooming the Urban Education Initiative. Still, our involvement with the New Jersey State Department resulted in our building a small network of schools in that continued on its own with minimal involvement by the state. In 1998, the approval by the New Jersey Supreme Court of a plan to focus substantial new resources for these 28 districts on proven programs, with Success for All prominently highlighted as the preferred model for elementary schools, has re-established a close collaboration with the New Jersey Department of Education, and an expansion in the number of SFA/R&W schools in the state.

Other state departments of education have highlighted Success for All in state conferences and in other forums. The New York State Department is recommending Success for All to schools identified as low-achieving in state assessments, and is helping to fund the program in several New York City schools.

State departments of education have taken a central role in the Obey-Porter Comprehensive School Reform Demonstration. They are responsible for writing overall state plans, making schools and districts aware of options available to them, reviewing proposals, and making awards. In this process many state departments are becoming aware of the capacity problem among existing research-based programs, and are beginning to think about ways to support quality adoptions and thereby attract national programs to their states and increase their ability to work with large numbers of schools.

In general, state departments of education have been important in providing appropriate policy environments, awareness, and (in New Jersey, Ohio, and New York), money for implementing schools, but have not taken a serious direct role in extending or supporting Success for All implementations.

School Districts

School districts themselves are logical sources of training and follow-up for Success for All and other reform models. Many school districts with several schools implementing Success for All designate a district coordinator for the program. The district coordinator is intended to serve as a liaison between our staff, the schools, and the central administration. In some districts this person is expected to learn the program and provide direct support to teachers, facilitators, and other staff, much like that which our staff gives to schools in follow-up visits.

Our experience with district coordinators is that they can be very useful in their liaison function, but are rarely effective in training or follow-up with schools. The need for a liaison is great, especially in large districts. District coordinators can and do help make sure that schools get the resources they need and that district policies are interpreted for the Success for All schools. For example, if the district adopts a new reading curriculum, the liaison can help figure out whether Success for All schools should simply be exempted from it, or whether some attempt should be made to adapt the Success for All curriculum to the new guidelines. The district coordinator can advocate for the program within the central office and see that it remains on the district’s broader agenda. He or she can provide a single point of contact for our program staff on all issues that go beyond individual schools, from arranging for ordering, duplication, and delivery of materials, to helping with assessments, to keeping our staff aware of changes in district policies.

As important as the liaison role is, our experiences with district coordinators have been mixed. In some districts, district coordinators have been people who already have many other responsibilities, and Success for All is added to their list with nothing else being removed. Further, there is often an institutional imperative to contain, isolate, and marginalize innovations, even ones valued by the administration and which are clearly, demonstrably effective. Assigning a program to a relatively low ranking central office official can be one way to ensure that a project remains at the periphery of the district’s operations (even if it was no one’s intention that this take place). We have found that it is important to maintain close relationships with someone in the district who has line authority (such as the superintendent, assistant superintendent for instruction, or Title I director) and not to let the project be seen as "belonging" to a lower-level district coordinator.

We have rarely found district coordinators to be helpful in training or implementation monitoring. Most often, district coordinators do not have the time to devote many whole days to detailed, classroom-level follow-up. More importantly, most district coordinators have no experience in Success for All schools, so after a short time every teacher and administrator in the schools knows more about the program than they do. The district coordinators who have been most useful in training and follow-up have been former principals or facilitators in Success for All schools. As time goes on we expect to see district coordinators recruited from within Success for All schools able to take on more of the training and follow-up role for Success for All than they have previously. This is working moderately well in Memphis and in Houston, two of our largest implementations, where the districts have designated staff with responsibility for training and follow-up.

Regionally-Based Project Staff

As the Success for All network has expanded and matured, another means of establishing regional training sites has become dominant. This is the establishment of regional training programs staffed by trainers who are full-time employees of SFAF but remain in their home areas. This arrangement solves several problems. First, we often find staff (usually facilitators) in Success for All schools who are outstanding educators, excellent trainers, willing to leave the security of their school district jobs, are eager to travel and work with schools all over the country, but are not willing or able to move to Baltimore. In our early years we did require most new trainers to relocate to Baltimore, but found that a requirement that such unusually able and exceptional people also must be willing to move to Baltimore put a severe constraint on our hiring qualified staff. Having regionally-based full-time staff allows us to hire the very best experienced trainers regardless of where they happen to be located.

Second, hiring trainers to serve a region gives us far more control and assurance of fidelity to our program’s goals than does engaging regional training sites in universities or other existing agencies, which may have their own agendas and constraints. Otherwise, regionally-based SFAF trainers have the same advantages as institutionally-based regional training sites. They reduce travel costs to local districts and increase the probability that our training staff will know about and can adapt to local circumstances and needs.

On the other hand, regionally-based SFAF trainers also have several drawbacks. One is that they are often isolated, working from their own homes without the informal collegial supports that are so important to our Baltimore staff. Operating far from our center, these trainers cannot routinely attend meetings or keep up easily with the latest information or developments. To deal with this, we hold regular staff retreats at least three times a year, plus meetings around other functions, to keep everyone on the same wavelength. These meetings have major costs, but are essential in a widely distributed organization. Monitoring the performance of regionally-based trainers is also problematic, although we have recently put in place a system of regional managers to help with training, mentoring, and supervision of new and junior trainers in their areas in 14 regions, plus the University of Memphis and Education Partners. These regions are shown in the chart on page 8. Although regionally-based trainers can decrease costs for the districts they serve, they increase costs for us, such as unreimbursed travel costs to bring these trainers to Baltimore for meetings, telephone costs for coordination, costs of setting up home offices, and costs of supervision.

Despite the problems of coordination and the additional costs involved, in the future we expect to see a continuing increase in regionally-based SFAF trainers. In fact, this is the only dissemination model we expect to expand in the coming years.

For Success for All, our Baltimore training center has become just one of 16 regional offices, serving the Mid-Atlantic region. It is interesting to note, however, that in our much smaller and newer training programs for MathWings and WorldLab (social studies/science), we have also started off relying on Baltimore-based trainers until training in these areas becomes more routine and well-specified.

Schools Helping Schools

One important way Success for All is extending its reach to a larger number of schools without sacrificing quality is by increasingly engaging staffs of successful, experienced schools in local support networks, particularly in mentoring new schools (Cooper, Slavin, & Madden, 1998). School staffs never take full responsibility for training and follow-up without the active involvement of trainers from SFAF or regional training sites, but they can take over some of the follow-up. The advantages of this are significant. Local schools that have successfully implemented Success for All have staffs that know every detail of the model and how to make it work in an environment very similar to that of the new school. Experienced and new schools can establish "mentoring" relationships in which staff exchange visits, materials, and ideas. When the new school runs into troubles large or small, the mentor school is nearby and ready to help. Schools often establish local support networks to help each other implement Success for All over the long run.

Success for All schools are increasingly joining forces with each other to create local support networks. For example, experienced schools are becoming mentors for new schools just starting out in Success for All. In some places schools are helping schools within their own districts, opening their doors to repeated visits, having school staffs visit new schools, sharing materials, coordinating training, and providing ideas, advice, and emotional support. Many schools are mentoring schools outside of their districts. In addition to experienced schools mentoring new schools, there are several areas in which staff from Success for All schools of all levels of experience meet on a regular basis to share ideas and discuss common problems.

School-to-school mentoring lets the real experts — the teachers, facilitators, and principals in successful Success for All schools — share their wisdom of practice and hard-won experience. It gives new schools an attainable vision of what Success for All should be like. It gives staffs of new schools support when they run into problems or opposition. Mentor schools know the local situation, so they can help new schools adapt to local realities more readily than our trainers can.

Perhaps most importantly, building support networks enables school staffs to provide emotional support to one another. In an isolated Success for All school, staff members are always wondering if they’re doing the right thing, if problems they perceive are temporary setbacks or serious roadblocks. Others in the district or even in the same school are often skeptical, either wondering what madness has come over the school or feeling envious about the attention and resources the school is getting. A local support network or mentoring arrangement helps school staffs withstand the inevitable pressures that come with innovation. Teachers, facilitators, and principals in Success for All schools speak a common language and share a common vision and purpose.

The following sections discuss ways in which schools help schools implement Success for All.

Mentoring.
Staffs of mentor schools can play a critical role in providing local models of effective implementation that others can visit, in participating in training and follow-up, and in giving advice and support to new schools. Mentoring arrangements can significantly reduce the amount of time SFAF or regional training center facilitators need to spend at a school, and can therefore reduce the costs of implementing Success for All. However, mentoring is not intended to completely replace training and follow-up from SFAF. To maintain the integrity and effectiveness of the program we need to know that its main elements exist in high quality form in each school that calls itself a Success for All school. It is important that we achieve an appropriate balance between local mentoring and continued training and follow-up from SFAF or regional training centers. Working out this balance is done on a case-by-case basis, but we do not anticipate having staff of new Success for All schools completely trained by local mentor schools without our involvement. To do this would risk our most important resource, our reputation for providing consistent, first-rate professional development to enable high-poverty schools to substantially improve the school success for all their children. The ideal arrangement for bringing new schools into the Success for All family is a combination of initial training from SFAF or regional training program staff in collaboration with local mentor schools followed up by the mentor school staff, with occasional visits from SFAF to add additional elements and check on the schools’ progress.

Relationships between mentors and new schools can be established either formally or informally. In the Success for All network today, many mentoring relationships were established informally by the schools themselves. Usually, a delegation from a new school visited an existing school, and when the new school adopted Success for All, the relationship between the new and existing school has continued. This happens all the time within districts that are expanding their number of Success for All schools, but often happens between districts as well. In other situations mentoring arrangements have been established as an explicit part of an implementation plan. New schools have sometimes been assigned existing schools, or at least designated staff from existing schools, to serve as their mentors. The participation as mentors in each of these cases was voluntary, but was strongly encouraged and supported by the districts.

The best way for new schools to obtain the "attainable vision" of what a Success for All school should look like and be able to accomplish is to arrange to have staff members from new schools visit experienced schools. Such visits are a typical and important part of the decision-making process that leads to a school’s adopting Success for All, and continue to play an important role as experienced schools become mentors for new schools. In initial visits, likely visitors are those who will be most involved in the decision: principal, members of a site-based management team, representatives of teachers and parents, and district staff. After the decision has been made, the designated facilitator is often the most frequent visitor, along with teachers who visit their counterparts in the mentor school to see what various program elements look like in practice.

Apprenticeship
Just as one-to-one tutoring is the most effective form of instruction for children, one-to-one apprenticeship is the most effective way to learn a complex task. One of the best services a mentor school can provide to a new Success for All school is to provide an opportunity for key staff from the new school to work side-by-side with their experienced counterparts.

Apprenticeship opportunities often involve individuals designated to be facilitators in new schools working with experienced facilitators. This is a particularly effective use of apprenticeship. The facilitator’s role is extremely complex and demanding; no training program or manual could begin to describe in adequate detail the thousand tasks every facilitator must do well.

The more time a new facilitator can spend in an experienced school, the better. The new facilitator can be of concrete value to the experienced one in helping with eight-week assessments and regrouping, organizing materials and supplies, and so on, while gaining concrete information about the facilitator’s role. The new facilitator can attend all the meetings the experienced facilitator attends, including those of grade-level teams, family support teams, and building advisory teams. The new facilitator can also visit classrooms and tutoring sessions with the experienced facilitator so that they can compare notes afterwards on strengths and weaknesses of lessons and constructive strategies for helping teachers improve the teachers’ lessons, classroom management strategies, and other practices. The new facilitator might spend time with family support staff, going on home visits with them or participating in programs for parents. After a week or more in a mentor school, a new facilitator will be far more confident about how to make his or her own school equally effective, and will have built relationships with the mentor school staff that will be important to both schools for a long time.

After the new school is up and running, the facilitator from the mentor school might arrange to spend a day or two with the new facilitator to "shadow" him or her and provide feedback and ideas.

Apprenticeships can be equally useful for other key staff. It is rarely possible for a principal to spend a whole week in another principal’s school, but several day-long visits can often be arranged and can be an excellent way to give the new principal a better vision of the whole program, knowledge of his or her own role, and a strong relationship with the mentor principal and staff. Family support team members can spend significant amounts of time in their mentor school working with its family support team, facilitator, and principal to learn how to make this program component work well. Individual teachers frequently visit teachers in the mentor school but less often apprentice with them. However, this does happen on occasion, especially when teachers are new to a Success for All school that does not yet have fully proficient teachers of its own to serve as mentors at the appropriate grade level.

Networking

Building a national network of Success for All schools is one of the most important things we’re trying to do (Cooper, Slavin, & Madden, 1998). An isolated school out on the frontier of innovation can sometimes hang on for a few years, but systemic and lasting change is far more likely when schools work together as part of a network in which school staff share a common vision and a common language, share ideas and technical assistance, and create an emotional connection and support system. This is the main reason we have an annual conference for experienced and new sites. At the annual conference we provide valuable information on new developments and new ideas (most of which we have gotten directly from the schools we work with). We are also trying to build connections between the experienced schools, so that they can share ideas on issues of common interest and build significant relationships with other schools pursuing similar objectives. We are also trying to create an esprit de corps, a pride in what we are all trying to do together, an understanding and acceptance of the struggle needed to achieve the goal of success for every child. We have "t-shirt days" and team-building activities that can be as important as the formal sessions. The breaks, when staff from different schools get to know each other and exchange information and telephone numbers, may be even more important.

In addition to the national conferences, there are many other things we try to do to build an effective support network. Our newsletter, Success Story, is one example. Our training sessions and the manuals and materials we produce invariably use contributions from experienced Success for All schools and reflect them back to all schools. In particular, our family support and facilitator’s manuals are primarily composed of ideas we’ve gotten from extraordinary Success for All schools, and we keep revising these and other materials as we learn more from the schools. For example, school staff often modify various materials, forms, and assessments for their own use. We pay attention to these modifications and if they seem broadly applicable, we use them to revise our materials. Further, in our conversations with schools we are constantly putting schools in touch with other schools to help them with specific issues, such as bilingual education, year-round schedules, use of Title I funds in non-schoolwide circumstances, use of special education funds to support tutoring, and so on.

Many local support networks for Success for All arise in informal ways, where individual principals or facilitators happen to be friends or get to know one another at Hopkins conferences or other activities related to Success for All. Mentoring relationships between experienced and new schools often grow into local support networks. In other cases, local support networks have been established by school districts or (in the case of New Jersey) by the state department of education.

Local Meetings
One of the most common activities of local support networks for Success for All is regular meetings among key staff. Most often it is facilitators or facilitators and principals together who meet about once a month to discuss common problems and explore ways to help each other. Sometimes principals meet separately from time to time to discuss issues of particular concern to them. The benefits of these meetings are like the benefits of mentoring, discussed earlier. Principals, facilitators, and family support team members can learn a great deal from others who are facing similar problems in similar environments under similar circumstances. Further, regular meetings among the leaders of Success for All schools provide routine opportunities for these staff to build positive relationships and to establish opportunities for other types of mutual assistance.

Some local support networks schedule some sort of demonstration at the host school for the visiting staff from other schools. For example, the host school may have developed a new computer system to help with regrouping, a new thematic unit for preschool or kindergarten, or a family involvement or parent volunteer program they want to show off. The demonstration might take place before or after the meeting.

Local Conferences. One of the problems with our national Success for All conferences is that since most school staff must travel great distances to attend, few schools send more than one or two people, usually the facilitator and/or the principal. Because of funding limitations, some schools cannot send anyone. Yet a similar purpose is sometimes served by holding local conferences. These can be scheduled on designated staff development days so that all staff can attend. The activities are like those of the national conference, with various inservices, updates, and other sessions, and with opportunities for schools to show off their accomplishments in a variety of ways. SFAF staff participate, but center stage is reserved for the schools themselves. This provides a basis for local networking among the whole staffs of the schools that has remained long after the conference itself.

Sharing Resources. There are many ways in which Success for All schools make effective use of limited resources by working together in a local support network. For example, schools often schedule combined training sessions. This does not work if numbers get too large, but if, for example, there is a training session just for kindergarten teachers or just for tutors, then it makes sense for this training to be offered to all members of a local support network.

Many schools share materials with each other. For example, schools making extensive use of novels find that they can make resources go much further if they share their novels with other Success for All schools. Two schools may agree on a common list of novels and each can then order half of the list, expecting to exchange novels at mid-year.

Schools involved in local support networks also collaborate on development of new materials. This happens frequently when a district adopts a basal or literature series for which supportive materials do not exist. The district or several individual schools often designate a development team of teachers from several schools to create these materials (using our format and standards) and then share them with all schools. We often contract to pay some of the costs of this development if we can then share the new materials with all districts using the same books.

Organization and Capital

Scaling up a successful school reform model is not only a question of building a strong training corps capable of working nationally. It also involves creating an organization capable of supporting trainers, developing materials and strategies, carrying out research and awareness activities, and so on. This, in turn, requires capital. Any reform organization needs to spend large amounts of money each year on recruiting and training new trainers, developing and printing materials, and other activities, many months before school districts pay their bills. This means that a line of credit is needed indefinitely, over and above whatever funding was necessary to develop and evaluate the program in the first place.

These issues, both creating an efficient organization and securing operating capital, having consumed enormous amounts of our energies in recent years. For ten years, Success for All existed as part of Johns Hopkins University. When the program was small, this worked very well. Johns Hopkins took care of most routine business functions, such as payroll, benefits, insurance, and some legal services. It allowed us to run a deficit each spring as long as we had accounts receivable to cover the deficit when schools paid us in the fall.

However, by summer of 1997, it became apparent that this arrangement could no longer work. On our side, the university's salary scales, policies, and practices were constant impediments to growth. We could not hire trainers in the Northeast, for example, because Johns Hopkins salary scales were much lower than those of Northeastern school districts. Similarly, we had difficulty hiring business-related staff, such as accountants, human resources staff, and a finance director, because the University's rates for such staff were half of what commercial businesses were paying. On the University's side, the size and complexity of our operation were very difficult and time-consuming to manage, and the University was understandably uncomfortable advancing us ever-larger amounts of capital each spring.

As a result, we decided to separate from the University. We reached agreement with University officials by February, 1998, and completed the separation by July 1, 1998, establishing the Success for All Foundation as a not-for-profit entity to be responsible for the development and dissemination of Success for All and Roots & Wings.

One of the key issues we had to resolve early on was whether to remain as a not-for-profit organization. This was a difficult decision. On one hand, it was clear that as a for-profit we would have no problem raising capital; many venture capital firms and individuals courted us heavily. However, there were several factors that led us to strongly favor staying in the not-for-profit world if we could. One was a desire to maintain an institutional ethos that focused on what is best for children, not what is best for profits or investors. Our staff is deeply committed to children and to school reform, and we did not want to undermine this spirit in any way. Another factor related to the public perception of our efforts. Watching the hostile reception in many quarters to the Edison Project and other for-profit education reform groups, we wanted to be sure that our program was seen as having unmixed motivations. The American Federation of Teachers and, to a lesser extent, the National Education Association, have strongly supported us (and opposed Edison). We did not want to endanger support of that kind. Finally, as a practical matter, we wanted to be certain that any operating profits would go back into development, research, and quality control, not into investors or taxes.

The decision to remain as a not-for-profit organization did have serious costs, however. We found that banks were unwilling to make loans to us unless we had substantial assets. We were able to secure approximately $5 million in grants and loans from two family foundations, the MacArthur and Ford Foundations, and New American Schools; on the basis of these assets, we obtained a line of credit from a commercial bank. Even with this we remain seriously undercapitalized for an organization of our size and rate of growth. For example, we have an annual printing bill of about $20 million, which we must pay many months before school districts begin to pay us. Had we had investors rather than loans, these problems would not have existed. On balance, we are sure we made the right decision, and that, in the long run, we will be much stronger as a not-for-profit organization.

In addition to capital needs, we have had to recruit a large corps of people to duplicate all of the functions the University had previously fulfilled: finance, accounting, payroll, benefits, insurance, legal services, information technology, space, and so on. All of these new people had to be recruited and trained at the same time that we were increasing our number of schools by about 50% and our total institutional budget by almost double. As these people have settled in, it has become apparent that we can do a much better job outside of the University, creating an organization completely tailored to our needs.

Lessons Learned

Our experience with the national dissemination of Success for All has led us to several conclusions. These are as follows.

• Successful dissemination of a program as comprehensive and complex as Success for All requires a combination of two types of assistance to schools. One is a core of talented, dedicated trainers operating from the project’s home and/or regional training sites closely coordinated with the project headquarters. The second is a local and national network of schools willing and able to provide technical and emotional support to schools entering the network.

• While other institutions can be helpful in dissemination, we are finding greater success in employing staff from outstanding Success for All schools to be full- or part-time trainers. Regional laboratories, other universities, and state departments of education have been helpful in our dissemination efforts, but with the exception of the University of Memphis they have not taken major responsibility for disseminating Success for All in their regions. District coordinators are very helpful as liaisons between our project, Success for All schools, and their central offices, but have been less helpful in training or follow-up. Regionally-based trainers on our payroll and staff in Success for All schools who are willing to do some training and follow-up for us are usually much more effective.

• Quality control is a constant concern. Whatever dissemination strategy we use, constantly checking on the quality of training, implementation, and outcomes is essential. Without it, all programs fade into nothingness.

• To maintain over a long period of time, schools implementing innovations must be part of a national network of like-minded schools. To survive the inevitable changes of superintendents, principals, teachers, and district policies, school staffs need to feel that there is a valued and important group beyond the confines of their district that cares about and supports what they are doing.

Success for All is only one of many national models of school reform, and it has unique characteristics that may make some dissemination strategies effective and others difficult or ineffective. Other types of programs may find very different strategies to be more effective. However, to the extent that other programs emphasize a strong research base, a well-specified set of materials and procedures, and a comprehensive approach to reform, we believe that our experiences will be a useful guide and will inform policies regarding technical assistance and reform at the local, state, and federal levels.

Policy Implications

Our experiences with the dissemination of Success for All have given us some degree of insight into the ways that systemic issues, such as federal, state, and local policies, can promote or inhibit school-by-school reform, and have given us some ideas about how these policies might change to support what we and other school change networks are trying to do.

Substantial positive change in student learning can only come about on a broad scale when major changes take place in the daily interactions of teachers and students. Ideally, we would have a variety of curricula, instructional methods, professional development methods, and school organizational forms for each subject and grade level, each of which has been rigorously researched and evaluated in comparison to traditional practices and found to be effective on valid measures of student achievement. School staffs would be made aware of these effective alternatives and would have the time and resources to learn about them, visit schools using them, see video tapes on them, and ultimately make an informed choice among them. Their exploration of alternatives might be assisted by local "brokers" who are knowledgeable about effective programs, organizational development, and the change process, and are aware of local needs, circumstances, and resources (see Slavin, 1997, 1999).

School staffs would control significant resources for materials and professional development and would be able to invest them in the exploration process and in well developed models supported by national training staffs and local support networks. These national programs would themselves be primarily supported by revenues from schools, but would also have seed money for developing materials and awareness and training materials, establishing national networks and regional training sites, and building qualified staffs of trainers and support personnel. Federal and state policies would support the process of school-by-school change by developing and promulgating standards, assessments, and accountability mechanisms likely to encourage school staffs to explore alternative models for change and to invest in professional development. They would push existing resources (such as Title I, Title VI, and Goals 2000 funds) to the school level, with a clearly stated expectation that these funds are intended for whole-school reform, not for maintaining current operations or patching around the edges. Some portion of school change funds would be provided on a competitive basis to schools, based on their willingness to engage in whole-school reform and allocate their own resources (especially Title I) to this purpose. Further, funds would be allocated to outstanding exemplars of school reform methods to compensate them for the costs of serving as demonstration sites, mentoring other schools in their local networks, and participating in local training and follow-up.

The remainder of this paper discusses the current state of policy support for school-by-school changes and the policy reforms needed to provide this support on a broad scale.

1. Increase support for research and development of school change models.

One of the most important deficiencies in the current structure of professional development is a shortage of whole school reform programs proven in rigorous research to be markedly more effective than traditional instruction, and thus ready for national dissemination (see Slavin & Fashola, 1998). Besides Success for All, only a few, such as the Comer project (see Becker & Hedges, 1994) and Direct Instruction (Adams & Engelmann, 1997) have conducted and reported comparisons with traditional methods. Even at the classroom level, there are only a few proven, well-evaluated methods capable of national dissemination, such as Reading Recovery (Pinnell et al., 1994) and cooperative learning (Slavin, 1995). There is progress on the development of new school change models; the New American Schools (NAS) funded seven design teams to develop such models. However, there is no current plan to formally evaluate the outcomes of these new designs in comparison to traditional methods.

It is interesting to note that the federal involvement in the development, evaluation, and dissemination of these models has been minimal. Private foundation and corporate funding has almost entirely been responsible for the development and dissemination of the Comer, Levin, and Sizer projects. Success for All has benefited from federal funding (its development and evaluation have been part of the work of the Center for Research on the Education of Students Placed At Risk at Johns Hopkins University), but it could not have been successfully developed and evaluated without funding from private foundations, especially the Carnegie and Pew Foundations and New American Schools.

There is a need for federal investment in the development of schoolwide change models, in evaluation of these models by their developers, and in third-party evaluations that compare the effects of the models to the effects of traditional methods (see Slavin, 1997, for more on this). Only when we have many successful models with clear and widely accepted evidence of effectiveness will we be able to confidently offer schools an array of choices, each of which may be quite different in philosophy or main elements but each of which is known to be effective under well-specified and replicable conditions of implementation.

2. Help proven professional development networks build capacity.

The most important limitation on the broad dissemination of Success for All is our own capacity to provide high-quality professional development services to a very large number of schools. Our model requires a great deal of training and follow-up, and any equally ambitious restructuring program that intends to change the daily instructional practices of all teachers would require equally intense training. We can only add so many schools each year without overtaxing our staff’s considerable energies, hiring more trainers than we can train and mentor, or seeing the quality of professional development decline. As a result, we must decline to work with further schools whenever our training calendars are full.

Our professional development organization is self-funding; our trainers’ salaries are supported by fees we charge schools for their time. However, rapid scale-up has costs. While we are training new trainers, we must pay their salaries, fly them to observe schools or training sessions, and so on. Costs for establishing trainers in sites other than the project’s home site may be particularly great, as these trainers must travel frequently to the home site. There is no source of funding for these costs. By the time a trainer is fully operative and bringing in enough revenue to cover his or her salary, we may have spent more than $50,000. As noted earlier, an even larger problem of scale-up is obtaining a line of credit to cover printing and other cyclical costs.

There is a need to provide training organizations like ours with funds to scale up their operations. Ultimately such organizations must be self-funding, but they need capitalization as they begin their work and as they engage in significant expansion of their national capacity. As noted earlier, private foundations have largely fulfilled this capitalization function for some projects, including Success for All, but if training organizations are to remain in the not-for-profit sector and to operate at significant scale, there must be much larger sources of capital for this purpose from government or donors.

3. Provide resources to schools earmarked for adoption of effective programs.

Serious reform at the school level takes serious funding at the school level. School staffs must have control of resources they can spend only on professional development, especially on adoption of demonstrably effective programs. For example, the Commission on Title I (1992), led by David Hornbeck, proposed a 20% set-aside of Title I funds for professional development. However, this did not prevail in the Congress.

School staffs should control professional development funds so that they can choose the development that they feel will meet their needs. When they freely select a given program or service provider, they will feel a commitment to that choice, in contrast to the more common case in which teachers resist inservice presentations that they feel do not respond to their needs. A school should be able to purchase services from any provider, including universities, regional laboratories, federal, state, or local technical assistance centers, professional development networks (such as the National Writing Project), or even their own district’s staff development office. Funds for this purpose may be awarded on a competitive basis; our own experience with the Venture Capital program in Ohio suggests one means of providing resources to schools with coherent plans for implementing major changes in their school and classroom programs.

The Obey-Porter Comprehensive School Reform Demonstration is making an excellent start in this area, at least as far as whole-school, comprehensive designs are concerned. This initiative is providing modest funding on a competitive basis both to help schools adopt research-based programs and to give them an incentive to use their existing resources (especially Title I) on programs likely to make a difference in all aspects of school functioning and in student achievement (see Slavin, 1998).

4. Provide awareness and brokering services to schools so they can choose professional development services wisely.

Individual school staffs are poorly placed to select promising or effective programs, as they may not be aware of what is available or how to go about obtaining the programs and materials they need.

Providing awareness (and some brokering) of promising programs is one area in which the federal government has played a significant role. As noted earlier, the National Diffusion Network (NDN) provided small Developer/Demonstrator grants to developers of programs that met an evaluation standard. NDN state facilitators organized awareness conferences and helped schools adopt these "validated" programs. However, the evaluation standards were low, and NDN funding was never adequate to provide much more than a clearinghouse, informational function (although, even with its limitations, NDN efforts led to thousands of successful adoptions of research-based programs in every state). In 1996, funding for the NDN was eliminated.

There is a need for far more ambitious outreach to school and district staffs to help them assess their needs and make them aware of a range of alternative programs and services available to them. Schools might invest their own professional development resources in such brokering services, or might pool resources with other schools to bring in awareness presentations on a variety of available programs and practices. State or federal support might be important in helping establish brokering agencies or individuals, but in a system in which professional development resources are focused at the school level, agencies or individuals providing any professional development services to schools would ultimately have to support themselves on fees from schools. Existing agencies, such as the regional laboratories and the new regional comprehensive assistance centers, could also play an important role in helping schools make wise choices of professional development services and programs. A process of this kind has been set in motion by the Obey-Porter Comprehensive School Reform Demonstration, which provides funds to labs and state departments to increase awareness of proven, comprehensive models.

5. Provide funds to successful exemplars of proven programs to serve as demonstration/training sites.

One thing we have learned in the dissemination of Success for All is how important it is to have schools successfully implementing the program whose staffs are willing to receive visitors and assist neighboring schools in the process of adopting the program. Many of our outstanding schools have put hundreds or thousands of person-hours into helping other schools start and maintain the program.

However, all this help comes at a price. Many schools can provide only minimal assistance to other schools without overly taxing their own staff resources. Some principals are concerned that if they let their best staff members work to help other schools, they will be hired away. More often, school staffs find that while their efforts to help other schools bring them recognition and satisfaction, they must put a limit on this activity.

It is unfair and unrealistic to expect that outstanding exemplars of proven programs will work indefinitely as demonstration and training sites without any outside compensation. There is a need to provide resources to these schools for the real costs of serving as demonstration sites (such as hiring substitutes when staff are elsewhere helping other schools) and to help them see aiding other schools as a part of their responsibilities. A model for this is professional development schools associated with schools of education.

Schools willing to serve as demonstration/training sites might receive funds amounting to half of their facilitators’ salary (about $25,000). In return, they would be expected to be open to a specified number of formal visits (e.g., one visit per month) and to provide some number of person-days of training and follow-up to other schools (e.g., 40 person-days per year). This would create a situation in which schools would be motivated to serve as demonstration/training sites, and would receive special recognition (as well as funds) for agreeing to serve in this role.

Conclusion

Our experience in the national dissemination of Success for All is instructive in many ways. We have discovered that there are far more schools eager to make thoroughgoing changes in their instructional programs than we or other national training networks can possibly serve. Policy changes, such as those contained in the reauthorized Title I, the Obey-Porter Comprehensive School Reform Demonstration, and state and local systemic reforms, are further motivating schools to seek high-quality, intensive, and extensive professional development services to fundamentally transform themselves. The key limitation in making this change take place is the limited national capacity to provide schools with well-researched models backed by networks of trainers, demonstration schools, materials, and other requirements.

The focus of this paper is on the ways we have tried to expand the capacity of our Success for All program to serve a rapidly expanding network of schools across the United States, and on the policy changes that would be needed to support our network and others in building our nation’s capacity for quality professional development. In brief, we have found that our network of schools and our own dedicated staff are the bedrock of a national dissemination strategy, and that building on the strengths of this network is the most promising approach to scale-up. Federal, state, and other support to help establish and maintain professional development networks like ours, along with providing money to schools earmarked for professional development, are most likely to create conditions in which schools throughout the United States will focus their energy on exploring alternatives, seeking professional development appropriate to their needs, and then engaging in a long-term, thoughtful process of change that results in measurably improved achievement for all children.

References

Adams, G., & Engelmann, S. (1997). Research on Direct Instruction: 25 years beyond DISTAR. Seattle, WA: Educational Achievement Systems.

Becker, B.J. & Hedges, L.V. (1992). A review of the literature of effectiveness of Comer’s school development program. Unpublished manuscript, Michigan State University.

Commission on Title I (1992). Making schools work for children in poverty. Washington, DC: Council of Chief State School Officers.

Cooper, R., Slavin, R.E., & Madden, N.A. (1998). Success for All: Improving the quality of implementation of whole-school change through the use of a national reform network. Education and Urban Society, 30 (3), 385-408.

Dianda, M. & Flaherty, J. (1994). Annual report on promising practices and program adaptations and successes. Los Alamitos, CA: Southwest Regional Laboratory.

Haynes, N.M. (Ed.) (1994). School development program: Research monograph. New Haven, CT: Yale University.

Levin, H.M. (1987). Accelerated schools for disadvantaged students. Educational Leadership, 44(6), 19-21.

Madden, N.A., Slavin, R.E., Karweit, N.L., Dolan, L.J., & Wasik, B.A. (1993). Success for All: Longitudinal effects of a restructuring program for inner-city elementary schools. American Educational Research Journal, 30, 123-148.

McCarthy, J. (1991, April). Accelerated schools — The Satellite Center Project. Paper presented at the annual meeting of the American Educational Research Association, Chicago, IL.

Pinnell, G.S., Lyons, C.A., DeFord, D.E., & Seltzer, M. (1994). Comparing instructional models for the literacy education of high risk first graders. Reading Research Quarterly, 29, 8-38.

Pinnell, G.S., DeFord, D.E., & Lyons, C.A. (1988). Reading Recovery: Early intervention for at-risk first graders. Arlington, VA; Educational Research Service.

Slavin, R.E. (1990). On making a difference. Educational Researcher, 19 (3). 30-34.

Slavin, R.E. (1995). Cooperative learning: Theory, research, and practice (2nd ed.). Boston: Allyn & Bacon.

Slavin, R.E. (1997). Design competitions: A proposal for a new federal role in educational research and development. Educational Researcher, 26 (1), 22-28.

Slavin, R.E. (1998). Far and wide: Developing and disseminating research-based programs. American Educator, 22, (3), 8-11, 45.

Slavin, R.E., & Madden, N.A. (1999a). Success for All/Roots & Wings: 1999 summary of research on achievement outcomes. Baltimore: Johns Hopkins University, Center for Research on the Education of Students Placed at Risk.

Slavin, R.E., & Madden, N.A. (in press b). Roots & Wings: Effects of whole-school reform on student achievement. Journal of Education for Students Placed at Risk.

Slavin, R.E., Madden, N.A., Dolan, L.J., & Wasik, B.A. (1996). Every child, every school: Success for All. Newbury Park, CA: Corwin.

Slavin, R.E., Madden, N.A., Dolan, L.J., Wasik, B.A., Ross, S.M., & Smith, L.J. (1994). "Whenever and wherever we choose...": The replication of Success for All. Phi Delta Kappan, 75(8), 639-647.

Slavin, R.E., Madden, N.A., Dolan, L.J., Wasik, B.A., Ross, S.M., & Smith, L.J. (1994, April). Success for All: Longitudinal effects of systemic school-by-school reform in seven districts. Paper presented at the annual meeting of the American Educational Research Association, New Orleans, MO.

Slavin, R.E., Madden, N.A., Karweit, N.L., Dolan, L., & Wasik, B.A. (1992). Success for All: A relentless approach to prevention and early intervention in elementary schools. Arlington, VA: Educational Research Service.