Robert D. Carlitz

Department of Physics & Astronomy

University of Pittsburgh

Pittsburgh, PA 15260

Mario Zinga

Pittsburgh Public Schools

Pittsburgh, PA

(January 31, 1994)

ABSTRACT

A project currently underway in the Pittsburgh Public Schools is seeking to develop new environments for teaching and learning using the technology of wide area computer networks. The history of this project offers lessons for other school districts who might wish to develop similar resources for their own use. From this history we extract a set of guidelines that can be followed by these other school districts.

I. Background

A. National Networks

B. The School Component

C. Major Themes

A. National Networks

The teaching environments discussed in this paper use the technology of wide area computer networks. The largest such network, known as the Internet, dates back to 1969. At that time the Advanced Research Projects Agency of the US Department of Defense (DoD) undertook an experiment to see if computer scientists working for the DoD could use their computers to form a communications network (called the ARPAnet). This experiment succeeded, not just in the technical sense of forming an operational network, with hundreds of computers performing in a coordinated manner, but in a larger and perhaps unforeseen social sense, whereby the people linked by this new network grew together in a cohesive online virtual community. This community had a mail system, and from the technical roots of this mail system there developed aspects of a real culture: recreational assets, intellectual debates, shared interests and curiosities. The technology was extended to support these aspects of the community, and the community expanded and flourished, not ignoring its original basis in computer science, but no longer a one-dimensional undertaking used by its members solely in support of a single professional activity.

The natural evolution of this community was to broaden and expand beyond the boundaries of computer science and beyond the confines of DoD-supported research. Thus, as the DoD withdrew support from their research network, another - even larger - network was in place to supplant it. The new network was known as the NSFnet, since it was founded by the National Science Foundation. The initial focus of the NSFnet was on access by researchers to a set of supercomputer centers which the agency had recently established. But, with the lessons of the ARPAnet in mind, the NSFnet was structured to reach a much wider audience and to serve a much wider range of needs.

Like its predecessor, the NSFnet was based upon a set of published public protocols, which could be implemented by any vendor for any piece of computing hardware. Access to the NSFnet was open to anyone involved in education and research, through the intermediary of a number of regional networks that were established for this purpose. The NSFnet quickly grew to include the participation of all major universities and research centers in the United States. The same protocols were adopted by research centers in other countries, and a comparable growth of what was by now called the Internet occurred on a worldwide basis. Within these universities and research centers, network usage was not restricted to the sciences, and educators and researchers in other fields found the online community equally attractive and compelling. At the federal level programs were being developed to formalize and extend the network infrastructure, and discussions began on the structure for a National Research and Education Network. Legislation currently before the Congress will help to implement this program, which is now known as the National Information Infrastructure.

B. The School Component

This was the stage where many school networking projects, which extended network connectivity to grades K-12, first began. People on the Internet had asked whether the "E" in NREN referred to all education or only to education at the college and postgraduate level. The nature of the Internet culture meant that there could be only one answer to this question. If nothing else people on the Internet wanted their own children to be able to use this resource; and the egalitarian forces in the larger society also mediated for universal access to the network. It is a premise of the present paper that universal access is a desirable social goal, a valuable asset for our educational system, and a practical technical aspiration.

Beyond these broad aspirations there are specific reasons for considering wide area networks as a fundamental tool for educational reform. The present paper describes the evolution of a project currently under way in the Pittsburgh Public Schools. This project, which is known as Common Knowledge: Pittsburgh, aims to develop computer networks in the service of education throughout the city school district. We are finding in the course of this project that technology can support the basic educational goals espoused by the school district. In this manner we hope to play a significant role in the district's restructuring process.

The name Common Knowledge: Pittsburgh (CK:P) is based upon a term first used in a talk that Carlitz gave at the National Net-91 conference in April of 1991. The phrase is meant to invoke the idea of "common knowledge" in two senses - one having to do with a body of information shared by all members of the community, and the other with the idea that the means of accessing network resources should become common knowledge to the community at large. Thus the project has at its basis the idea of universal access and the idea of using network access as a fundamental educational tool. The following sections describe the process through which these ideas have evolved into a dynamic educational activity in the city.

C. Major Themes

A major part of this paper will be devoted to presenting a chronology of the development of Common Knowledge: Pittsburgh (CK:P), with particular attention to successes that the project has had and to difficulties encountered along the way. Our purpose is to derive some lessons from our experience and offer them as guidelines for other school districts interested in developing their own school networks. Table I contains a chronology of significant dates in the evolution of CK:P. The table indicates the sections in this paper where the events noted are discussed.

A few themes will occur repeatedly in our discussions. One has to do with the idea of teachers and students as researchers. This concept fits well with efforts at educational reform which encourage problem-solving by both teachers and students, and it points at the basic skill to do research, which will become increasingly important in training the workforce for an information-oriented society. Wide area computer networks provide access to vast quantities of information, and they create the opportunity for individuals to disseminate their own information resources. There is an obvious need for the ability to sort through this information, assess its quality and integrate relevant information with one's own body of acquired knowledge.

A second broad theme has to do with the process of institutionalization of the use of network technology in the support of education. This is the process by which new educational initiatives are accepted and assimilated. We feel that it is important to consider the process of institutionalization if any innovation is to be of lasting significance. Wide area networks are likely to be of long-term value in our nation's schools and merit a careful introduction into the educational environment.

We believe that we have a model which can assist the process of institutionalization. Our model starts with pioneering teachers who are given access to new tools and encouraged to experiment. It proceeds to involve the larger population of teachers in a series of steps, with appropriate opportunities and incentives along the way. The first step encourages a grass-roots approach, with the participation of the most adventurous and pioneering teachers. The activites of this group - and all those who follow in the project - are coordinated in teams which provide the teachers with people of complementary ability. This approach provides a fcous and avoids the problem of having highly-motivated pioneers become isolated from their colleagues as they explore individual projects. As activities expand to involve the larger population of teachers, administrators at all levels are brought into the process. Along with the participation of these administrators comes the support of a whole range of structures within the school district, which assist the teachers in their new undertakings.

II. Preliminaries

A. Researchers Make Contact (1989)

B. Pioneers in the Classroom (1990)

C. Teaching the Teachers (1990)

D. Moving Forward (1991)

E. Continuing Programs (1992-)

A. Researchers Make Contact (1989)

In Pittsburgh, as in other cities around the world, the question of Internet access for children and their teachers was beginning to be raised by researchers already on the network. One of the authors of the present paper (Carlitz) is a theoretical physicist and had been familiar with the network for several years. He posted questions about school networking to several Internet news groups early in 1989 and quickly found a handful of people with similar interests, even though none of the news groups in existence at that time was structured to deal specifically with this sort of topic. Patt Haring, an educational consultant in New York, quickly moved to fill this gap. She established an informal mailing list in May of 1989 using a computer at the City University of New York. The mailing list was entitled KIDSNET and had an initial readership of a half dozen people. As word of this new resource spread across the Internet during the summer of 1989, interest grew and many people asked to join the KIDSNET discussions. By the end of the summer a more formal support structure was needed, and the mailing list was moved to computers at the University of Pittsburgh. Carlitz has supervised its operation since that time, and the mailing list continues to be a valuable online resource, now known as KIDSPHERE.

The two authors of this paper met when Carlitz began to make inquiries in the Pittsburgh Public Schools about possible interest among district teachers in Internet resources. A phone call to the district's office of Computers in Education produced a list of teachers likely to be interested in this activity and quickly led to a meeting with those teachers. Zinga and others in this initial group were provided with accounts on computers at the University of Pittsburgh as part of a research project which Carlitz had initiated at the university.

B. Pioneers in the Classroom (1990)

The Pittsburgh teachers joined the KIDSNET mailing list and used it to contact other teachers online and to develop ideas for applying this new resource in their classrooms. Although the initial technology was often awkward to use, the teachers were able to develop a variety of interesting applications. Zinga initiated pen-pal exchanges for his fifth grade students with several other classes around the world. His students moved from pen-pal introductions to specific classroom projects. One set of students was corresponding with a group in Ireland. They began to exchange their favorite jokes, and both sides were puzzled by the fact that the others' jokes weren't funny to them. The Pittsburgh students had an American frame of reference, while the Irish students were more European, and the language they used was subtly different. The ensuing discussions gave both sets of students not just an insight into the culture of the other group, but also insights into the nature of language and of humor - topics with which many college-level students would wrestle uncomfortably.

With another group of students Zinga developed a collaboration with a school in Bologna, translating incoming messages for his students from Italian to English. The following year the Italian students were studying American geography and asked the Pittsburgh students to act as mentors. Zinga's students worked as a team, debating the best order of presentation and cross-checking that information which each team member supplied. In Zinga's research class, students used the network as part of a project on "Struggles for Freedom." In response to a network query they located an eyewitness to the original Woolworth's lunch counter sit-in in North Carolina. The students conducted online interviews and used the material from this source as the basis of a play they wrote and produced.

At Peabody High School Len Donaldson initiated joint projects in social studies with a high school in Japan. These projects led to a close relationship between the students at his inner-city site with what became a sister school in Nishinomiya. Donaldson's students watched anxiously during the typhoon season and kept in close touch with their counterparts in Japan through regular exchanges of electronic mail. The following year one of his students traveled to Japan, visited the sister school and sent electronic mail back to Pittsburgh from the Japanese classroom. At another Pittsburgh high school Regine Fougeres initiated exchanges in German with schools in Canada and Germany. She discovered Freenet resources and foreign language newsgroups and made these resources available for classroom use as well.

As activity in the school district grew, it began to attract the attention of supervisors, who heard of exciting new resources and received requests for computing equipment to support this new activity. Network access accounts were created for all those supervisors who were interested in exploring the Internet. The supervisory personnel began to use the network for e-mail contacts internally and with colleagues outside the school district. Staff in the office of Computers in Education were particularly active on the network and began providing regular online support to teachers across the district.

C. Teaching the Teachers (1990)

By the end of the 1989-1990 school year, there was enough networking activity in the Pittsburgh Public Schools to justify some attempt at formalizing its structure. Zinga and Carlitz went with Regis Schilken, a special education supervisor, to meet with Associate Superintendent Stanley Herman in May of 1990. Herman approved the idea of attempting to develop wide area networks as a tool for students and teachers throughout the school district. An effort would be made to involve more teachers in the process, and a search began to find the funding necessary to carry out this activity.

In the fall of 1990 Carlitz and Zinga invited teachers around the region to join the Pittsburgh Public Schools teachers exploring the network. Interested teachers were given guest accounts as part of Carlitz's research project, and a series of monthly lectures was initiated. The number of guest accounts eventually grew to over 200, with representation from 20 school districts around the region and from the Diocese of Pittsburgh. Since that time several of these districts have developed their own programs for Internet connectivity. Recently the Allegheny Intermediate Unit, which serves all school districts in Allegheny County, has also set up an Internet connection. The Intermediate Unit is a likely mechanism for the majority of school districts in the region to develop their Internet access.

The lecture series taught us several things. First of all, there was a great interest on the part of teachers to join the network community and gain access to Internet resources for both themselves and their students. There was at the time no formal support mechanism within their school districts to facilitate this process, and this placed what proved to be an excessively large burden upon us as the only line of support for these teachers. The problem was compounded by the fact that we were using the same system used by all students and faculty at the University of Pittsburgh. This system offered few concessions to novices and tended to exacerbate support problems. As a result the drop-out rate was moderately high, and fewer than half of the initial participants remained online six months later.

D. Moving Forward (1991)

A donation by Sun Microsystems to Carlitz's department at the University of Pittsburgh enabled us to solve many of these problems in one fell swoop. Although the donated computers ran the Unix operating system in their native mode, and this operating system can be difficult for beginners to master, we were able to insulate our users from the raw operating system through a menued interface which highlighted commonly-used resources and made them simple for everyone to use. These menus had been written by Tracy LeQuey Parker for the Texas Education Network. With some minor refinements we found them to make an excellent environment for our group of networked teachers. With the easy-to-use menu system, the need for telephone support was dramatically cut, the drop-out rate declined, and the large majority of new account recipients maintained their network access.

While these activities were under way in Pittsburgh, there was a parallel flurry of activity relating to school networking on the national scene. The interest in local projects such as ours and state projects such as the one in Texas was beginning to influence national planning. The National Science Foundation held a meeting to discuss how computer networks can support the NSF's efforts at the reform in the teaching of mathematics and science. EDUCOM, a consortium of university computer centers, hired John Clement to explore possibilities in the area of school networking. In the spring of 1991 the Technology Education Research Center (TERC) in Cambridge, MA, sponsored a meeting to consider the possibility of forming a consortium of state departments of education which might facilitate school networking and educational reform. This was to lead to the establishment of the Consortium for School Networking. Later that year, in an event that set the stage for major expansion of existing wide area networks, Congress passed legislation sponsored by Senator Al Gore which defined a program to develop a National Research and Education Network (NREN).

E. Continuing Programs (1992-)

These initiatives have proved to be the basis of much of what is currently happening in the area of school networking at the national level. A program for testbeds in school networking has been developed as part of the program for the Applications of Advanced Technology in the National Science Foundation. This activity has led to a new program called Networking Infrastructure for Education, which is sponsored jointly by the NSF's divisions of Education and Human Resources and Computing, Information Science and Engineering. That program will likely be a model for new efforts in the US Department of Education and the Department of Commerce. Support for related activities will probably come from the Department of Energy and the National Aeronautics and Space Administration. With the elevation of Clinton and Gore to the White House the NREN program has gained added stature - and a new name, the National Information Infrastructure.

The outcome of EDUCOM's K-12 program and the TERC meeting was the development of a new organization known as the Consortium for School Networking (CoSN). CoSN's first Board meeting was held in Pittsburgh in January of 1992. The founding Chairperson of CoSN was Connie Stout, the head of the Texas Education Network; Clement served as the founding Executive Director until moving to the NSF in July of 1993. The new Chairperson, chosen at the Consortium's general meeting in Dallas in November, 1993, is Gwen Solomon from the New York City Board of Education. CoSN has taken a leading role in organizing conferences, workshops and demonstrations of school networking. A recent national workshop, jointly sponsored by CoSN and the Federation of American Research Networks, brought together teachers and networkers to develop models for the support of school networking. In general CoSN serves as an important catalyst for school districts, vendors and regional networks as these groups work to develop networks which can adequately serve the school population..

III. Implementation

A. Constructing a Collaboration (1991)

B. Change in the School District (1992)

C. A Formal Beginning (1993)

D. Initial Sites (1993)

E. Expansion and Institutionalization (1993-94)

A. Constructing a Collaboration (1991)

The task of putting together a formal collaboration was orchestrated by Carlitz and Zinga. Their talents were complementary, in that Carlitz was familiar with the network and its use in the the university research environment, while Zinga was familiar with education at the classroom level and with the administrative structures of the Pittsburgh Public Schools. Through their joint efforts Common Knowledge: Pittsburgh grew to be a collaboration involving the Pittsburgh Public Schools, the Pittsburgh Supercomputing Center (PSC) and the University of Pittsburgh. The project's educational component has its roots in programs developed by teachers in the school district; technical support is provided through personnel at the PSC; and assessment and project management are provided through the university. This collaboration developed and matured as a proposal to the National Science Foundation was prepared in 1991.

The preparation of a formal grant proposal for the NSF required a fundamental shift from the essentially volunteer grass-roots activity that had marked our networking activities prior to the start of the 1991 winter semester. At that time Herman convened meetings of the academic support staff under his direction, and this group was charged with developing projects which could be included as part of an eventual formal proposal. Herman's group included representatives of the various curriculum content areas, a third of whom were already familiar with the Internet activities of teachers in their area. The focus of all the project ideas was to be upon specific curricular activities, so that network use would eventually become a standard part of work in the classroom.

Involvement of the curriculum support staff was a first step toward a larger goal of CK:P. This was the idea of institutionalizing the use of networking technology by familiarizing all levels of school personnel with the educational potential of network resources. Herman announced that communications among his staff would shift from FAX, which had been the standard for urgent messages, to electronic mail. The transition took place over a period of weeks, with successive memos sent out by both FAX and e-mail and numbered in a declining sequence "20, 19, 18, ...," and with the promise that after message number 1 the FAX distribution would be discontinued.

Technical support for a formal proposal required access to solid networking expertise. Given Carlitz's long association with the fellow physicists who directed the Pittsburgh Supercomputing Center, it was natural to develop the collaboration in this direction, drawing upon the considerable networking resources of the PSC. The PSC's principal focus is the provision of high-performance computing for scientific research. Networks are a key means to this end, and the PSC had played an important role in setting up PREPnet, Pennsylvania's regional network. Through its High School Inititiative the PSC had some experience in the pre-college environment, and individual staff members had a strong personal interest in community and school networking. In conversations with Eugene Hastings, a member of the PSC Communications group, the general framework of a school networking project for Pittsburgh began to emerge.

The assessment component of the collaboration was developed by Janet Schofield of the Learning Research and Development Center at the University of Pittsburgh. Schofield was familiar with technological and sociological issues relating to public education and had worked extensively in the environment of the Pittsburgh Public Schools. Her involvement insured that the project would be able to maintain a rapid pace of development while maintaining necessary educational standards at each stage of implementation.

Development of the proposal for Common Knowledge: Pittsburgh continued throughout 1991, and a proposal was submitted to the NSF in January of 1992, in time for consideration as part of the NSF's new program for testbeds in school networking. Following the approval by the Pittsburgh Board of Education of the concept of a five-year program to develop networking technology to serve students and teachers throughout the city, the school district hosted the first meeting of the Board of Directors of the Consortium for School Networking in January, 1992. This event helped solidify the collaboration by bringing all participants together for a demonstration of the sort of connectivity that CK:P proposed to establish. The demonstration was set up by PSC staff and was viewed by school district personnel and by invited vendors. The link to CoSN helped indicate the national scope of school networking efforts, and the initial Pittsburgh demonstration has been repeated at several national education conferences.

B. Change in the School District (1992)

1992 was a year of transition for the Pittsburgh Public Schools. The Superintendent, Richard Wallace, who had helped bring the school district to the distinguished position it holds among major urban school districts, retired after ten years of service. The new Superintendent, Louise Brennen, had served within the district for many years as an Associate Superintendent; she came to office with a dramatic new charge. Under her leadership the school district has made a major commitment to restructuring and site-based management. There has been a wholesale reorganization of the district's administrative staff in preparation for this organizational shift. With changes in the responsibilities of the support staff, there have been many changes in staffing. Herman, who had played such an important role in getting CK:P started in the Pittsburgh Public Schools, accepted a superintendency in the neighboring Woodland Hills School District, and some of the content specialists retired from the school district.

Herman's place on CK:P was filled by Lloyd Briscoe, the principal at Zinga's school. Briscoe's interest in networking had several roots. Previously he had been the principal at Greenway Middle School, which had served as a teacher training center for the school district. Before leaving Greenway, he had learned of the Internet through the activities of one of his special education teachers, and had been planning to implement these activities on a wider scale at Greenway. His interest in educational equity fit naturally with the broad scope of the proposed CK:P project, and his meeting with Zinga led to a natural working relationship.

The NSF reviewed the proposal that had been submitted in January, 1991, and suggested that the project might proceed with a modified strategy of deployment. It was suggested that the original five-year district-wide plan might be presented as a two-year pilot, to be followed by a three-year full-scale implementation. The project's emphases upon universal access and long-term institutionalization could be retained, as was the focus on having curricular needs drive the use of technology. A revised proposal was submitted in June, 1992, and was approved for funding at the start of the 1993 calendar year. The project was set to go forward under a restructured school district - in many senses of the word. It was a transition which proved to be beneficial for the project in a number of ways.

First of all, the project's focus upon curricular projects developed by teachers in the school district fit well with the district's new emphasis on site-based management and local initiative. This enabled the project to retain widespread support in the district at a time when major structural changes were occurring. Secondly, the new Superintendent made it clear that technology would be a high priority for her administration. This emphasis on technology is a reasonable priority for a city like Pittsburgh, which itself is undergoing a massive transformation from its former role as a center for heavy industry to its present role in the developing technologies of the information age. But within the school district the emphasis has a different character, recognizing the manner in which technology can support such fundamental educational goals as outcomes-based education, educational equity, and the empowerment of students and teachers.

Third, it became clear that the technology of wide area networks could be a valuable tool in the school districts efforts at restructuring and site-based management. During the period between the original grant proposal (in January, 1992) and the revised proposal (in June, 1992), the teams charged with designing the first two restructured schools in the city were able to complete their site plans. Both schools requested local area networks following the model developed for CK:P as part of their schools' technology implementation. Funding for these facilities has since been provided by the school district.

C. A Formal Beginning (1993)

When CK:P began its formal operation in January, 1993, it was able to start putting together the infrastructure necessary to accomplish its broad goals. Carlitz was able to concentrate upon project coordiation. Gwendolyn Huntoon, the head of the PSC's Communications group, became the technical project manager for CK:P; and Eugene Hastings, who had helped develop the project's overall architecture, was named project engineer. Technical support staff were hired, and work began on developing servers for the project's initial sites. Assessment personnel were hired, and an initial assessment framework was developed.

On the educational side, postings were prepared for a project manager and support staff. Given the changes going on in the school district, this was necessarily a gradual process, but over a period of time it was possible to recruit an education staff with long experience in the local school district. The people selected had developed the necessary networking expertise in the process of working to develop the grant proposal itself and were familiar with organizational aspects of the district from their previous on the job experience. Richard Wertheimer, a mathematics supervisor who had organized a science and mathematics project for the proposal, became the educational project manager. Zinga was selected to coordinate curriculum support. Priscilla Franklin, who had worked in the district's office of Computers in Education, was hired as a project teacher.

These hirings were spread over a period of months. Both Zinga was formally hired for the project Herman was able to arrange for release time. This enable Zinga to work with Huntoon and Carlitz and keep the educational and technical components of the project coordinated. Zinga put together an office for the educational project staff as they were being hired, with the help of Briscoe, who was able to find space in Woolslair School.

Initial efforts to coordinate school district and PSC activities on the project involved the establishment of two large committees. One focused upon educational issues, and the other on technical topics. These committees involved representatives of all members of the collaboration and helped everyone develop an understanding of the project's overall focus and direction.

D. Initial Sites (1993)

A set of initial implementation sites was chosen from among the fifteen project ideas that had been included in the original grant proposal. The selection was made on the basis of likelihood for success and included those sites at which the most teachers had been developing a familiarity with the network and its use in the classroom. There were two high school sites and two elementary school sites.

Westinghouse High School proposed to extend a science and mathematics program to involve student consultation with online mentors. The program already included mentors from the Westinghouse Electric Corporation; online interactions would obviously facilitate this process. Wertheimer had worked with personnel at Westinghouse for some time and had been instrumental in putting together the Westinghouse proposal. He was now in a position to help facilitate its implementation.

Schenley High School proposed to use network resources for its foreign language programs. Fougeres and other teachers at the school had already initiated exchanges of student correspondence over the network and had located a wide range of useful foreign language resources online.

McCleary Elementary School is a restructured school for which the school district has provided a local area network, a school server and a router. This equipment served as an in-kind contribution by the school district to the NSF grant. Teachers at McCleary were developing an interdisciplinary project on architecture with a specific emphasis on monuments. Students were to use the network to locate relevant materials and to disseminate their findings.

The Woolslair Elementary Gifted Center was chosen to serve as the project's educational office, as a beta test site, where new hardware and software can be initially installed and tested, and as a resource center for other teachers in the school district.

The process of site preparation was complicated by two factors. The NSF funding had not provided for all of the user devices required at each site. These devices were to be covered by parallel proposals to local foundations, but these proposals could not be completed until the NSF funding was assured. This delay meant that it was initially necessary to rely upon loaned equipment and donations. We decided to stretch our resources by making use of previously-existing equipment at those sites which had suitable devices in sufficient numbers. Older equipment at Schenley and Westinghouse was thus pressed into service, necessitating compromises in the user interface for those sites and significantly increasing demands on the technical support staff.

Network teams were established at each site, with a focus on preparation for the 1993-94 academic year. CK:P staff facilitated this process in several ways. During the summer of 1993 representatives of each site team attended a workshop which provided instruction on the availability of educational resources on the Internet and the use of tools to access these resources. The workshop had two parts - two and a half days at the beginning of the summer and two days at the end of the summer. During the intervening weeks teachers worked online to familiarize themselves with resources introduced at the workshop and to discover new resources which could be shared at the second session. Teachers who did not have their own equipment at home were loaned equipment from the project or the school district and were supplied with modems which had been donated to the project by Science Resources for Education (a non-profit organization based in Berkeley, CA). Equipment for the workshop and for teachers' use over the summer was provided through a donation from Apple Computer. This equipment has seen extensive use since that time - at the beta test site, at teacher's homes, at other school sites and at remote demonstrations.

Following the workshop school network teams were invited to reconsider their projects in light of what they had learned. With the help of CK:P staff they draw up action plans for the fall term. The various sites evolved to have their own unique character. Schenley's activities were focused on a single lab of older computers tied to the network; Westinghouse's activities were scattered in classrooms around the school; and McCleary involved devices on a local area network that ran through the entire school.

Support for the network teams has continued through the school year in the form of regular site visits by CK:P's educational and technical staff and by demonstrations at the CK:P beta test site, where equipment is initially configured and evaluated. During this period the network teams have gained familiarity with the available technology, while CK:P staff have learned how this technology actually works in the classroom setting and what steps are necessary to develop a production capability in that setting.

Demonstrations of the networking architecture used by CK:P were offered both locally and at educational conferences nationwide. These national demonstrations were held under the auspices of the Consortium for School Networking but were carried out with the assistance of CK:P staff members and occasionally with hardware support through the CK:P project. In this manner information about the project's architecture was disseminated widely, and CK:P staff had the opportunity to look for and correct weak points in the proposed school configurations.

E. Expansion and Institutionalization (1993-94)

In preparing for the project's second year, we have worked to strengthen the element of site-based management and initiative. The mechanism that has been employed to develop second year sites is one borrowed from the school district's plan for school restructuring. (There is also an element of imitative flattery following the district's successful pursuit of NSF funding for the project.) An internal competition has been set up for which a request for proposals (RFP) was circulated in October. All schools in the district were invited to submit proposals for network projects. Proposal guidelines suggested that submissions should follow the outline used by the most successful components of our original NSF grant proposal. These involved the statement of an educational problem, the identification of elements of the problem which could be resolved through networking technology, the specification of a plan for implementing this technology in the classroom context, and, finally, identification of some of the plan's technical details. The emphasis is clearly upon using the technology in support of the curriculum and not on acquiring technology to see how it might get used. This focus guarantees effective use of the requested equipment and gives participating teachers a clear buy-in to the proposed activity.

The RFP process has been a popular one in the school district. Several hundred teachers have visited CK:P's beta test site to get a first-hand look at the Internet and to see how it could be introduced into their classrooms. Over 40% of the city's 80 schools have submitted proposals, the result of thousands of hours of efforts. The challenge that the project now faces is not one of finding valid educational applications and the teachers to implement these applications, but the task of funding the hardware needed for these applications, installing this hardware on the scale demanded by participating teachers, and supporting the teachers' classroom projects.

Since the NSF could not be expected to provide capital equipment for the school district on any significant scale, support for hardware purchases has been sought from local foundations. An award by the Buhl Foundation funded equipment purchases at our first-year sites, and a proposal to the Heinz Foundation is currently in preparation to cover the second-year sites.

Continued evaluation of our administrative efficiency has led to a shift in the administrative structure of CK:P. The original large educational and technical committees have been replaced by smaller working groups. These groups function similarly to the network teams at the individual sites. They involve representatives from the project's educational, technical and assessment teams and focus upon specific issues and activities. These groups are currently working to develop image handling capabilities for an online art collection, to create an online library catalog (with a facility for including student reviews), to investigate new educational resources on the network, and to codify policies for student use and ethics. A core group with representatives of all elements of the collaboration meets weekly to oversee the project's progress. All meetings involve the participation of the project's assessment staff, who have also used mailing lists, online surveys and individual surveys to track this progress.

Student use of the network has developed in two ways. First there has been the steady development of targeted projects at our initial school sites. There have also been teachers from other school sites who have heard about the possibility of using the network and have submitted requests for student use to CK:P, backing up their requests with specific proposed curricular applications. This approach mirrors that of the RFP process and underscores the role of this process in the goal of institutionalization.

As CK:P has developed structures for implementing its goals the school district has initiated an important and complementary process. A task force for instructional technology has been formed to develop a long-range plan for technology implementation. The technology plan will provide the Board of Education with guidance with technology purchases over the next few years. CK:P's role in this effort is to provide a working model of networking technology, and a model for staff development and the staged implementation of new technology.

IV. Lessons

Change takes time

Provide resources for teachers

Integration with curriculum

implementation teams

Internal competition for resources

Support of administrators

Technology plan

Supporting infrastructure

Coordination of academic and administrative computing and networks

The process that has been followed in the development of Common Knowledge: Pittsburgh contains a number of useful lessons. Some of them reflect accidental discoveries that have been made in the course of the project, and others confirm the basic soundness of the original project design. All of these lessons are useful not just for technology planning inside the Pittsburgh Public Schools but for other districts around the country. Indeed, a striking feature of conversations that we have had with our counterparts in other school districts has been the universality of the problems we face in implementing appropriate educational technology in these districts. Optimistically one might hope that this common set of problems will admit a common set of solutions. In any case we hope to offer suggestions that might be tried fruitfully in other environments.

The first and most obvious point is that change takes time. A study of the timeline for CK:P (see table) shows that each significant stage of development took about a year to complete. When teachers encounter new technology, it takes time to assimilate the value of this technology for their teaching activities. And when teachers attempt to incorporate the technology into a modified curriculum, it takes time for them to become comfortable with a new approach to teaching and to assess its value in their own scheme of education. When systems experiment with new approaches to learning or with new organizational mechanisms, it takes time for them to implement changes and to assess the value of these changes.

Against this conservative background we can offer suggestions for how to mediate change in an efficient manner. It clearly makes sense to provide resources for pioneers. While it might seem like a frill for some districts to encourage teachers within their system to experiment with new approaches to education, the cost of this exercise is minimal compared to the expenses that can be incurred by false starts and dead-end approaches to change. The pioneering teachers in any school district are a valuable resource - one to be cultivated and used as a means of exploring new paradigms and models.

This is not to say that experimentation should be open-ended and undefined in terms of its goals. Rather there should be a close integration with the curriculum of any experimental efforts. This integration is the key to whether innovations are used only occasionally or employed regularly in the classroom.

The means of achieving this integration is the development of implementation teams. The teams should involve teachers and administrators who know how to guide activities through the school bureaucracy, and people with the technical expertise necessary to make needed decisions in this area and guide detailed implementation. A continued educational focus assures that project activities are comprehensible to teachers in the school district, and that technical experts are viewed as facilitators in the educational process.

The mechanism in CK:P for encouraging an integration with the curriculum has been the RFP process described previously. This internal competition for resources has additional benefits. As technology is deployed, it is placed in sites where teachers are committed to using it. This maximizes the efficiency of technology expenditures while rewarding individual initiative and encouraging effective site-based management.

An internal competition cannot work, and no effective planning can be carried out, unless the implementation teams have the support of administrators in the school district. While grass-roots efforts must be at the heart of any effective educational reform, such reforms can never pass out of the classrooms of the first adopters and across the entire school district unless they are clearly endorsed at all levels of the school administration. In the case of CK:P this support has been provided for two reasons. First, if not foremost, has been the recognition that the technology of wide-area networks fits in with the district's strategic goals, in terms of individualized instruction, site-based management and school restructuring. Secondly, the administration has seen the early adoption of new technology as a means of securing additional funds in a time of stringent local budgets. This second point applies to any district adopting advanced technology, since such programs are typically attractive to local industry and can encourage new collaborations with local funding sources. Networking technology works particularly well in this context, since it allows the collaborators to be linked directly through the very technology that the collaboration is helping to develop.

For an administration to endorse elements of technology which are to be applied district-wide, it is a necessity that these elements be embedded in some larger technology plan. CK:P helps the planning process by providing a model for implementation and a research base upon which to build. These resources are available to school districts nationwide through the online resources of the project. A CK:P gopher can be reached at gopher.ckp.edu by those with current Internet access. The focus of CK:P is on the instructional use of technology. District-wide technology planning must consider larger issues, notably the coordination of academic and administrative computing and networks. Obviously there needs to be a level of interoperability between these two systems so that teachers can provide administrative data and can access this data as needed. On the other hand there is a need for system security in administrative systems which may go beyond that required fro academic systems.

Any technology plan must include not only the elements of necessary hardware and software but also a supporting infrastructure adequate to maintain equipment that has been purchased and to assist teachers in its effective use in the classroom. In CK:P much of this support structure has been supplied through the project's staff, but there will be a continuing need for this type of support beyond the period in which external funding for it is available.

V. The Common Knowledge Model

Educational equity

Common protocols

Educational resource

Primary means of information distribution

Common Knowledge: Your City's Name Here

The enabling legislation for the National Research and Education Network was passed by Congress and signed into law by President Bush in December, 1991. Further legislation was delayed by the election campaign of 1992, since Senator Gore, now the Democratic vice-presidential nominee, had been the principal architect of this legislation in the Senate. With the election of the Clinton/Gore ticket in 1992 the prospects for intensive development of a network that reaches all sectors of society became very bright. There exist several bills before the present Congress which would implement various portions of this program, which is now known by the name of the National Information Infrastructure.

While the full extent of the educational component of the NII has not yet been defined, there are a few outlines that one can presently discern. Insofar as educational networks are a tool for educational equity, they should probably be deployed so as to reach all classrooms in all schools across the country. This would be an ambitious national undertaking, but one that is realizable in technical terms and affordable in terms of its likely cost/benefit ratio.

To achieve the benefits of universal access, the networks which serve our nation's classrooms should follow the same common protocols observed across the Internet. This will allow for interoperability of all systems and services on the national network and will encourage the development of many new networked educational resources.

The present architecture of Internet information servers is rich enough to encompass such multi-media services as interactive electronic textbooks. Full-fledged instructional resources are likely to be developed in some profusion, along with more specialized tools with which teachers can construct elements of their classroom presentations. All in all the Internet is likely to become the prime educational resource in a majority of subject areas.

As this happens there will be another factor encouraging extensive applications of school networking. Since the Internet is such an economical and efficient mechanism for information dissemination, it is likely that the federal government and many state governments will adopt it as their primary means of information distribution. At this point it will become essential for school districts to attain good Internet connectivity if only to maintain their essential contacts with the governmental authorities under whose jurisdiction they lie.

We see that educational and administrative applications will tend to drive school districts in the direction of Internet connectivity. This picture presupposes the existence of a lively and widespread National Information Infrastructure. If the NII does indeed develop in this fashion, school districts will find it possible to develop programs in collaboration with other local groups seeking to gain their own Internet connectivity. These groups will include local government and industry as well as other educational entities.

How, then, should these school districts develop their own programs for Internet access and for the use of the resources of wide area networks in their curricula? We would like to offer the experiences of Common Knowledge: Pittsburgh to these school districts and suggest that Common Knowledge: Your City's Name Here be adopted as a model for their network development. The resources of CK:P and other networked school districts will be online for the use of CK:YCNH. We invite collaborations with these other school districts and look forward to an online community which is fully inclusive.

Chronology

May		1989	start of KIDSNET mailing list		II-A
Sep		1989	KIDSNET at Univ. of Pittsburgh		II-A
Mar		1989	meeting with teachers			II-A
Sep		1989	first classroom use in PPS		II-B
Jun		1990	presentation to Assoc. Supt. Herman	II-C
Sep-Dec		1990	lectures for Pittsburgh teachers	II-C
Sep		1990	donation from Sun Microsystems		II-D
Nov		1990	meeting at NSF				II-D
Feb		1991	EDUCOM K-12 project			II-D
Jan		1991	presentations to PPS support staff	III-A
May		1991	TERC meeting, formation of CoSN		II-D
Apr		1991	National Net �91			 II-E
May		1991	PPS support staff required to use net	III-A
Nov		1991	first discussions with PSC		III-A
Dec		1991	passage of NREN legislation		II-D
Jan		1992	CK:P proposal submitted to NSF		III-A
Jan		1992	first meeting of CoSN Board		II-E
Jun		1992	revised CK:P proposal sent to NSF	III-B
Sep		1992	Louise Brennen becomes PPS Supt.	III-B
Jan		1993	CK:P funding approved by NSF		III-B
Feb-Jun		1993	reorganization of PPS support staff	III-B
Jan-May		1993	initial CK:P staffing			III-C
Feb		1993	donation from Apple Computer		III-D
Jun		1993	summer workshop				III-D
Oct		1993	Instructional Technology Task Force	III-E
Oct		1993	CK:P funding from Buhl Foundation	III-E
Oct		1993	RFP process begins			III-E
Oct		1993	CoSN/FARnet meeting			II-E
Nov		1993	first general meeting of CoSN		II-E
Jan		1994	evaluation of CK:P internal proposals	III-E

Table I. Chronology of major events in the development of Common Knowledge: Pittsburgh.