Achieving High Educational Standards for All: Conference Summary (2002)

A major objective of the conference was to examine the role of research in improving the quality of education, especially for students from those racial/ethnic minority groups that historically have been poorly served by schools. In discussing this topic, Samuel Stringfield of the Center for the Social Organization of Schools at Johns Hopkins University noted that although there have been substantial increases in the number of years of schooling completed and in degree attainment for all segments of the population (U.S. Department of Education, 2000b), improvements in academic outcomes, as measured by the National Assessment of Education Progress (U.S. Department of Education, 2000a), have been modest, particularly since the late 1980s. Nevertheless, he expressed his opinion that substantial achievement gains reflecting improvements in the quality of instruction will be soon forthcoming. His rationale for this optimistic assessment is twofold.

First, the cost of not improving educational outcomes for the country, as for individual students, has become unacceptably high. As discussed in Chapter 2, today’s technology-intensive economy requires a highly educated workforce. Much more than in the past, the prosperity of the country, like the prosperity of individuals, rests on a solid educational

foundation. In short, Stringfield argued that, as a country, we no longer can afford to fail to provide students with a quality education.

Second, Stringfield cited as a rationale the accumulated lessons from nearly 60 years of research on education reform. “We can say a great deal about what is necessary to make reform work at the classroom and the school levels…yet we haven’t been able to stabilize our gains.” He noted that even the most valid, research-based reform models are not yet fully reliable. Six large-scale studies cited by Stringfield are: Aikin (1942); Berman and McLaughlin (1977); McLaughlin (1990); Stallings and Kaskowitz (1974); Crandall and Louks (1983); Stringfield et.al. (1997); and Ross et al. (1999).

In discussing the challenge of developing and implementing reliable models for improving student achievement, he cited the work of Lewis Thomas (1979) on the history of medicine. Thomas found that medicine did not start advancing “on the day that it turned toward science.” Rather, it took decades of work laying the scientific foundation before major advances in medicine began to materialize. As Stringfield put it, “people realized that an awful lot of what was being passed off as medicine was just bunk, and that opened the door to what [Thomas] called 100 years of science.”

Stringfield argued that the path of scientific research in education is not unlike that described by Thomas for medicine. Beginning with the Eight Year Study conducted during the 1930s (Aikin, 1942), many lessons have been learned in seven decades of education research. For example, he noted that Ron Edmunds’ (1979) findings about the characteristics of effective schools recently were supported and elaborated on by Charles Tedlie and David Reynolds (2000): “We (now) know tremendously more than Ronald Edmunds did, but recent research reaffirms (Edmunds’) statement that we can fix a school if we want to, wherever that school is.” Stringfield also cited work by Jerry Brophy (1988) and E. Bridges (1986) on teacher effects and J. Millman and Linda Darling-Hammond’s work (1990) on teacher professional development as providing a scientific basis to “conduct meaningful evaluations of teachers.” He also cited Preventing Reading Difficulties in Young Children (National Research Council, 1998b) as “providing an extremely valuable summary of decades of research on beginning reading.”

Stringfield further elaborated on the reasons for his optimism about the future of education reform by referring to the work of Peter Senge

(1990). Senge wrote about organizations that deliberately set out to learn how to improve their performance. He argued that lasting improvements in the performance of organizational goals occur only when good ideas and inventions are transformed and systematized into innovations that work consistently and reliably.

Drawing parallels to the history of both medical and education research, Stringfield used Senge’s illustration from aviation history on how the Douglas Corporation incorporated inventions developed for the ultimately unsuccessful Boeing 257 aircraft into its design of the innovative DC-3. The Boeing 257, introduced in 1934, incorporated four important new inventions: a variable pitch propeller; retractable landing gear; light-weight, molded body construction; and a radial cooled engine. Unfortunately for Boeing, the plane still was unstable on takeoffs and landings despite these important advances. The following year, Douglas Corporation introduced the aerodynamically and commercially successful DC-3, which incorporated the four important advances developed by Boeing but added an additional feature to address the stability problem—wing flaps. Senge concluded that the scientific and engineering research done by Boeing was necessary but not sufficient to produce a reliable, high-capacity aircraft. Douglas Corporation brought all of the pieces together to produce a truly innovative airplane.

For Stringfield, the current state of education is analogous to that of aviation in 1934, just prior to the introduction of the DC-3. The field has benefited from much solid research, but the findings have not yet been systematically brought together in a way that reliably produces the kind of outcomes that are desired. Stringfield argued that there are quite a few examples of schools using reform models that are achieving good results. Continuing with Senge’s terminology, he considers these to be examples of educators’ good ideas having been transformed into inventions. So far, none of these “inventive” models has proven that it can be reliably and economically replicated. In other words, none can yet qualify as a true innovation by Senge’s definition. Nevertheless, Stringfield argued, researchers are identifying more of the conditions under which specific reform models work and do not work. Furthermore, in his judgment, some reform models are getting close to the point at which they can be reliably and economically replicated and are not far from fitting Senge’s definition of an innovation. Overall, in his view, research is getting closer to providing a solid foundation for “differential diagnosis and prescription” that will better match reform strategies to specific needs. Referring to two prominent whole-school reform models, he offered the following opinion:

He went on to add that, although some reform models are more solidly based in research than are others, any number of national and locally developed models have produced good learning outcomes in at least one site (Herman et al., 1999; Northwest Regional Education Laboratory, 1999; Stringfield, 2001). For Stringfield, what is most important is not the particular theory of reform, but how the model is implemented. No matter what the model, better outcomes depend on better professional skills of teachers, better curriculum, better instruction, and better professional development. “An interesting theory is no good to a superintendent. They get fired on interesting theories.” Instead, Stringfield sees the primary challenge in education in the coming years as more akin to engineering— learning how to implement on a large scale the important lessons that have been learned from research so as to reliably produce in a cost-effective way the desired learning outcomes. That is, to move from good ideas and educational inventions to reliable innovations.

Stringfield cautioned, however, that even the best reform models depend on good implementation and that, invariably, every national reform model is recreated or cocreated at the school level by local educators:

In other words, Stringfield sees no magic formulae or shortcuts to success that can avoid dealing with these fundamental issues.

Stringfield argued that research being done on highly reliable organizations is highly relevant for the education reform process:

Stringfield identified some of the principles of highly reliable organizations:

• They require that the public and employees hold the perception that failure to achieve the organization’s core goals would be disastrous.

• They require clarity regarding goals. Stringfield noted that in his observation of schools that were successfully implementing various types of reform, the staffs invariably were united regarding a finite set of goals.

• They are alert to surprises or lapses. Small failures can cascade into major failures. Stringfield noted that “all of us make dozens of small mistakes a day. In highly reliable organizations, areas in which mistakes can cascade are monitored very closely.” He added that various reform models agree that success in basic reading and math skills is critical to students’ long-term success in school. As he put it, “it takes a young child years to learn that he or she is not a skilled reader. During that time, several adults over literally hundreds of occasions will have observed small failures in the student’s learning. It is not critical to catch any one error; however, it is critical to avoid a cascade of reading failures and derailed self-confidence.”

• They build powerful databases on dimensions highly relevant to the organization’s ability to achieve its core goals.

• They extend formal, logical decision analysis, based on standard operating procedures, as far as knowledge allows.

• They have initiatives that identify flaws in standard operating procedures, and then nominate and validate changes in those procedures that have proven inadequate.

Stringfield mentioned other important principles: highly reliable organizations recruit extensively and then train and retrain staff constantly. They take performance evaluation seriously and allow for mutual monitoring of job performance. These organizations are hierarchically structured but incorporate “a second layer of behavior that focuses on collegial decision making regardless of rank” during times of peak loads. Highly

reliable organizations invariably are valued by their supervising organization. Finally, they prioritize high reliability over short-term efficiency.

Stringfield noted that, as a member of the Baltimore City School Board, he has worked to implement the principles of highly reliable organizations in the city’s school districts. One still would not need to look very far to find serious problems in Baltimore’s schools, yet he presented data documenting substantial improvement in the reading skills of Baltimore’s elementary school students over a period of three years as evidence that the district’s efforts to implement the principles of highly reliable organizations were beginning to produce results.

Technical assistance, in the form of teacher professional development and various forms of collaboration between university and school-based professionals, was the primary focus of a preconference workshop. Teacher professional development, in particular, is a topic that was discussed by several speakers. Michael Klentschy, superintendent of the El Centro Unified School District in California, articulated a point made by many: the way that in-service professional development typically is done should be reexamined. Often, in-service professional development is seen as an opportunity for university-based researchers and experts to deliver information to teachers. Louis Gomez, Phyllis Hunter, Barbara Foorman, and Diana Lam articulated a different model of teacher professional development—one that emphasizes a two-way flow of information between providers of technical assistance and the teachers who have real-world experience in classroom settings. All emphasized the importance of partnership and mutual respect as preconditions to effective collaboration.

Diana Lam highlighted several other factors that she saw as important to effective professional development from her perspective as a superintendent:

• Time: School districts must allocate sufficient time for teachers to participate in professional development. She noted that this often is a problem.

• Reallocation of Resources: Since most school districts have limited resources for noninstructional purposes, superintendents may need to take away resources from other areas of the budget to support professional development activities.

• Consistency with District Policies: Professional development activities should be consistent and well aligned with the district’s curriculum.

• Public Engagement: Efforts to engage parents, the business community, and others in support of reform efforts are important to their success. Professional development activities in support of a program of education reform should be supported by the community.

• On-Site Availability: Professional development is more effective when provided on-site, either in person or via technology. For example, many school districts have hired instructional coaches or guides as a strategy to provide teachers with ongoing professional development.

Louis Gomez discussed his work as a provider of technical assistance, although he would prefer to be considered a partner rather than a provider. From his base at Northwestern University, he codirects the Center for Learning Technologies in Urban Schools—a collaboration involving the Chicago Public Schools, Northwestern University, the Detroit Public Schools, and the University of Michigan. (For more information, see the center’s web site: < http://www.letus.org/welcome.htm>. Accessed October 22, 2001.) Gomez stated that the center’s initial goal was to help to make technology available to increase the engagement of urban students in learning, particularly in science. It soon became apparent that spending money on technology would be ineffective if the curriculum was boring. So, in collaboration with teachers from 60 schools in Detroit and Chicago, the university-based experts began working to design curricula that would be appealing to students in the participating schools and that used technology to help visualize and model the scientific principles being taught. A common vision of how to do this was developed—something that could only be accomplished by combining the knowledge and perspectives of the university and school-based professionals. As Gomez put it, “For us, learning, including learning with technology, is the social construction of knowledge in a community with distributed resources.”

Like other presenters, Gomez underscored the importance of aligning the curriculum, the technology and assessments to build instructional capacity. The center’s work has concentrated on aligning technology with curricula to facilitate “adventurous teaching and learning.” He added, however, that alignment with district assessments has been a problem, especially in Chicago. As has occurred with other inquiry-based curricula, “[teachers] say that work on our stuff stops because the tests are being done.” Teachers are not confident that the information and skills emphasized in the technology-enhanced curricula developed with the center are covered by the standardized promotion test administered by the school system.

While acknowledging that standardized testing can play a useful role helpful in promoting achievement, Gomez noted that he and other educators committed to inquiry-based learning need to work more closely with district accountability offices so that the assessments appropriately measure the knowledge and skills that students develop through inquiry-based learning.

Barbara Foorman and Phyllis Hunter described a model for delivering technical assistance to improve the reading of kindergarten and early grade schoolchildren in Texas. Foorman is a researcher who has conducted studies of the effectiveness of different strategies for teaching reading to very young children enrolled in Title I schools (Foorman et al., 1998). She also provides technical assistance to the Texas Reading Initiative. (To learn more about the Texas Reading Initiative, see <http://www.tea.state.tx.us/reading/>. Accessed October 22, 2001.)

She argued that research-based instructional practices can improve the reading skills of at-risk children, providing many with the literacy skills they need to avoid academic difficulties in later grades. Among the lessons that Foorman said she had learned is that more explicit instruction in the alphabetic principle can be effective in helping 1st and 2nd graders who are served by Title I to read better (Foorman et al., 1997). Knowledge of the alphabetic principle is awareness that written words are composed of letters that are intentionally and conventionally related to phonemic segments of the words of oral language (National Research Council, 1999b:147). Her research also has looked at the effects of teacher competencies on student reading skills. This research has informed the efforts of the Texas Reading Initiative’s strategy to develop the capacity of elementary schools to teach reading more effectively in the early grades— especially in those schools with many children who are academically at risk. The initiative’s efforts center on teacher professional development.

Phyllis Hunter is a consultant to the Texas Reading Initiative. She also had worked for several years as reading manager for the Houston Independent School District (HISD). Rodney Paige, then superintendent of HISD, had placed her in charge of the reading programs of 185 schools serving 225,000 students. She reported that her work in Houston taught her that caring and knowledgeable teachers are indispensable for effective reading instruction.

Referring to the statewide Texas Reading Initiative, Hunter stated: “[In 1999] Teacher Reading Academies trained 20,000 teachers. They were trained by the State of Texas—not the university, but the State of Texas—

for four days on what we think is critical to kindergarten reading, oral language development, print awareness, scaffolding expansion. . . . [In 2000] we are training all of our 1st grade teachers for five days—four days in class and one follow-up day.”

Hunter noted that the follow-up training provided to teachers at their own schools is supported by several major grants to the state from the business community, and that there are other components to the reading initiative that are supported by multiple sources. One of these programs provides an incentive of $7,000 for teachers to acquire specific skills needed to be designated as a master reading teacher in a Title I school. “We are aiming to put an expert reading teacher in every school in Texas.” Hunter and Foorman emphasized that the Texas program is a rare example of a research-based professional development program that has been scaled up for implementation statewide. It is supported by the state government, the private sector, and the policies of school districts statewide.

In Chapter 3, Catherine Snow noted that early reading is one of several areas in education in which the research base is strong enough that it can be expected to provide a solid foundation for improving students’ reading skills if properly implemented. The Texas Reading Initiative is providing teachers with that information as well as various forms of technical assistance to help teachers to use it effectively.

Robert Slavin, chairman of the Success for All Foundation and codirector of Johns Hopkins University’s Center for the Social Organization of Schools, observed in his conference presentation:

Slavin’s message about “sweating the details” and finding ways to bring about school change on a large scale is strikingly similar to comments made by Samuel Stringfield. Success for All, the most widely disseminated whole-school reform model, is comprehensive in scope, highly scripted, and specifically designed to “sweat the details.” Success for All began in a single, high-poverty, inner-city school in Baltimore in 1987. By 2000, approximately 1,800 elementary schools serving 1 million children

in 48 states were implementing Success for All. Approximately two-thirds of the students were black or Hispanic, and the great majority of the sites were high-poverty schools supported by federal Title I funds.

The primary focus of Success for All is reading, writing, and language arts. (A complementary program that adds support for mathematics instruction is called Roots and Wings.) The program provides schools with curriculum materials, instructional strategies, and extensive teacher professional development and follow-up training. It also includes one-on-one tutoring for young children who are struggling to read, as well as parent involvement programs. The various components are research-based and designed to maximize the probability that they could be reliably replicated. Slavin pointed out that from the mid-1980s onward, reading instruction has focused on the systematic development of phonics in the context of meaningful text. Slavin noted that this approach is consistent with the consensus opinion of researchers that was to emerge years later (National Research Council, 1998b).

To maximize the probability that Success For All could be replicated, Slavin stated that he chose to emphasize an engineering approach, favoring strategies that could be reliably implemented rather than more cutting-edge approaches. Noting that he considers Success For All to be an exercise in large-scale social change, he added:

A major objective is to help ensure that children get off to a good start in school. Slavin noted that deficits that emerge in the early grades tend to persist as students progress through the educational continuum:

Slavin noted that although implementing Success for All increases per-pupil costs in the elementary grades, he argued that the program is

cost-effective if it is well implemented and succeeds in accomplishing its goals—especially the goal of keeping students out of special education. (According to Slavin, the additional cost of implementing Success for All is approximately $500 per pupil in the first year and about $200 per child thereafter.) He noted that a large percentage of students enter special education on the basis of poor reading skills.

The collection and monitoring of achievement data are integral to the Success For All model. Slavin, in collaboration with various researchers, has used these data as the basis for a variety of publications documenting the performance of the program. Highlighting data for more than 60,000 students enrolled in 111 schools in Texas that had used the program between 1994 and 1997, he reported that students in the program outperformed other students on the state-mandated Texas Assessment of Academic Skills (TAAS) (Hurley et al., 2000). Commenting on this and other evaluations of Success For All, Slavin noted that “there is a great deal of variation in outcomes, depending on the degree to which the program is actually implemented. I think every program on earth has that phenomenon. But even if you average across good implementers and poor implementers, you will find in study after study after study that students in Success for All schools learn more than do students in matched control schools.” Other studies describing the design and performance of Success for All can be found at <www.successforall.net>

While emphasizing the importance of teacher professional development in Success for All, Slavin offered that the program outcomes, as with other reform models, vary depending upon teachers’ skills and commitment:

Bertha Rubio, principal of Crockett Elementary School in San Antonio, Texas, discussed how Success for All was being implemented at her school. She described Crockett’s 894 students as 96 percent Hispanic, 46 percent limited-English-proficient, and 96 percent economically disadvantaged. Prior to adopting Success for All in the 1997-1998 school year, 65 percent of Crockett’s 3rd and 5th graders were passing the Texas Assessment of Academic Skills (TAAS) reading exam. She noted that the

TAAS, as a state-mandated criterion-referenced exam, is an important factor driving policies and practices for Crockett and all other schools throughout the state.

Its 65 percent pass rate earned for Crockett an “acceptable” rating from the state, and thus Principal Rubio was not worried about being placed on the list of low-performing schools. She was motivated to adopt Success for All by her desire to further improve Crockett’s TAAS scores. She set a goal of increasing the school’s pass rate from 65 to 80 percent after one year of the program, and then to 90 percent after three years. A pass rate of 80 percent would move Crockett from the “acceptable” category to being a “recognized” school; a 90 percent pass rate would earn Crockett an “exemplary” rating. However, of more fundamental importance to Rubio than ratings was her strong desire to improve the skills of the 35 percent of Crocket students whose reading was so poor that they could not pass the exam.

Under the leadership of then-superintendent Diana Lam, San Antonio elementary schools were required to choose and implement one of several nationally disseminated whole-school reform programs. In consultation with her staff of 100, Rubio chose Success for All. Among the most important reasons she cited was that it is research-based and had been shown to be effective. Also, the program includes assessments every eight weeks to track students’ progress, a strong family support component, tutoring for those students who need it, and—most importantly for Rubio—program materials in both Spanish and English.

Rubio reported that after the first year of implementation, the TAAS pass rate increased from 65 to 71 percent and remained at 71 percent after year 2. At the end of the third year, the pass rate for 3rd and 5th graders on the TAAS reading exam jumped from 71 percent to 83 percent. The most dramatic gains were registered by students who participated in the program for three consecutive years and by students with limited English proficiency. Because of a high rate of geographic mobility among neighborhood residents, many children transfer in and out of Crockett Elementary School each year and therefore did not receive continuous instruction under the program. The pass rate for the latter group increased from 47 to 74 percent in three years.

Rubio attributed Crockett’s success to the quality of training provided by Success for All program staff, her faculty’s strong commitment to implement the model, and to parental support and involvement:

Conference co-moderator Catherine Snow used a confectionary metaphor to contrast the underlying principles of different approaches to school reform:

Success for All and the general philosophy of reform described by Samuel Stringfield could be thought of as being of the baked apple variety. Snow continued the metaphor, describing the chocolate soufflé:

Snow suggested that the Pittsburgh mathematics reform initiative, described at the conference by Diane Briars and Brian Lord, might be considered a chocolate soufflé-type reform, as is the reform strategy developed by Lauren Resnick, director of the University of Pittsburgh’s Institute for Learning.

Diana Lam, superintendent of the Providence, Rhode Island, Public Schools, agreed that the approach to reform that she and Lauren Resnick were currently pursuing was of the chocolate soufflé variety. Rather than choosing to implement different whole-school reform models in different schools, Superintendent Lam set out to work systematically with all of the schools in Providence. Her goal was to transform commonly heard rhetoric, like “all children can learn,” “closing the gaps,” and “achieving high standards for all,” into reality—a distinctly uncommon reality, particularly in school districts like Providence.

Lam noted that when she arrived in Providence in fall 1999, she knew that she had her work cut out for her. Attendance was very poor, and only

60 percent of students were graduating from high school; 75 percent of Providence students are from low-income families, more than one-third are immigrants who arrived in the United States within the past three years, and more than 7,000 of the district’s 26,000 students change addresses within the city each year. A total of 50 percent of Providence students are Hispanic, 22 percent are black, 10 percent are Asian, and 18 percent are non-Hispanic white.

For Lam, the key to improving learning outcomes is to change the quality of teacher-student interaction. With support from Resnick’s Institute for Learning, Lam implemented a change strategy in which she, as superintendent, taught Resnick’s basic philosophy and principles of learning to the principals, who in turn were expected to teach their teachers, who would put the information into practice as they worked with their students. She referred to this strategy as creating “nested learning communities.” Elaborating on this idea, Resnick asked the conference audience to “think of those dolls from Eastern Europe that nest inside one another. I don’t know which is more important, the little one, the center one, or the one on the outside. The important thing is that they do fit together. They are going in the same direction, they have the same general shape, and are kind of accountable to each other because they have to fit inside each other.” Those being taught are held accountable for learning, but those responsible for teaching also are held accountable. Accountability is key, whether it is teachers working with students, principals working with teachers, or the superintendent working with the principals—learners are held accountable only for what they are taught.

Rather than adopt a relatively scripted reform package like Success for All, Lam chose to pursue reform in Providence based on Lauren Resnick’s more general philosophy and principles of learning. Referring to her previous work as superintendent of the San Antonio Independent School District she recalled that a major focus of her work in Texas was to oversee the widespread implementation of Success for All and other New American Schools reform models (Berends et al., 2001). Comparing that with her rather different focus in Providence, she felt compelled to add “I do like baked apples—just for the record—but I also like chocolate soufflé!”

Central to Resnick’s principles of learning is the premise that the kinds of higher-order thinking and problem-solving skills associated with higher intelligence can be routinely elicited by the environments in which people live and, more to the point, the environments in which children

learn. Thus for Resnick, aptitude (or intelligence) is not something that is primarily determined by one’s genes or whose parameters are set in early childhood:

As a basis for her position, Resnick cited 30 years worth of converging evidence from brain research, social psychology, and cognitive science (National Research Council, 1987; Greeno et al., 1996; Resnick, 1998):

The first principle of learning that Resnick and Lam are using to guide them as they work to construct learning environments that will socialize intelligence is to require:

The second principle of learning is accountability talk. Having a thinking curriculum requires lots of talk—talk among the students and between the students and teacher. But as Resnick put it, “It is not just any old kind of talk. It has to be talk that is accountable—accountable to the community in the sense of building on what others have said and using

each other’s ideas. There also must be accountability to knowledge.” That means that “you are supposed to say what is true, and you expect your talking partners to say what is true. It’s not ok to say any old thing. You challenge each other, and you push back.”

Accountability talk also requires accountability to good reasoning. As Resnick noted, cognitive scientists have found that people have a basic capacity for good reasoning but also that it can be lost if not routinely practiced. Resnick added, however, that good reasoning skills can be rebuilt in social environments that routinely require good reasoning. The other principles of learning are:

• Clear expectations,

• Fair and credible evaluations,

• Learning as apprenticeship,

• Organizing for effort,

• Recognition of accomplishment,

• Socializing intelligence, and

• Self-management of learning.

For a more complete discussion of Resnick’s principles of learning, see Resnick (1998) and also <www.instituteforlearning.org>.

After one year of working to reform the Providence public schools in accordance with Resnick’s principles of learning, Superintendent Lam reported great enthusiasm for these ideas from the school system’s administrators and teachers. Although she did not report any changes in quantitative measures of student learning at this early juncture, frequent classroom visits have convinced her that the principles of learning were quickly finding their way into the classroom: