Back to Vol. 2, No. 3 Table of Contents

Using the World Wide Web to Build Learning Communities in K-12

Douglas N. Gordin
Louis M. Gomez
Roy D. Pea
Barry J. Fishman

Northwestern University
School of Education and Social Policy

Full Screen

Many are the conditions which must be fulfilled if the Great Society is to become a Great Community .... The highest and most difficult kind of inquiry and a subtle, delicate, vivid and responsive art of communication must take possession of the physical machinery of transmission and circulation and breath life into it. When the machine age has thus perfected its machinery, it will be a means of life and not its despotic master. -- John Dewey (1938)

Table of Contents

Abstract

Social accounts of learning and human knowledge have led to attempts to reorganize schools as learning communities. This paper examines the utility of the World Wide Web for aiding in the construction of school-based and work-based learning communities. An ordered list of interactions is provided to characterize the depth of students' entry into new learning communities. Current offerings on the World Wide Web are then surveyed in terms of these categories. Finally, proposals are advanced for enhancing the architecture of the WWW to facilitate its use for the creation and operation of learning communities.

Introduction

Social accounts of learning and human knowledge (Dewey, 1938; Kuhn, 1962; Toulmin,1972; Vygotsky, 1978; Wittgenstein, 1967) have led to attempts to reorganize schools so as to form learning communities(*1). In general, two different senses of learning communities have been advanced. First, are school-based learning communities formed by students and teachers who engage in long-term projects. Students work on these projects collaboratively, thus allowing them to learn from one another and letting the problems to be solved dictate the knowledge that must be acquired, thus frequently and purposefully crossing disciplinary boundaries (Hill, 1982). Second, are work-based learning communities where students learn the practices of a discipline or profession, with the historical example being apprenticeship. Here the profession defines a "community of practice" that the student aspires to join; this term refers broadly to the practices of a field, its social organization, and its mores. This method of learning through apprenticeship has been characterized as becoming a legitimate peripheral participant (Lave & Wenger, 1991). A primary goal of the common school system is to provide emerging citizens with opportunities to join these diverse work-based communities. However, access is regulated by norms of language, behavior, knowledge and too often, race, sex, and creed. Networks can provide a powerful means to surmount these barriers.

Essential to all types of learning communities is access to resources and communication. In this light, it is argued herein that computational communication technologies (CCT; *2), and, in particular, the World Wide Web (WWW), can play a vital role in the formation and successful operation of learning communities. While the need for such resources is vital in all fields of endeavor, we concentrate here on science education. This specialization is not intended to limit the generality of our argument, but rather to deepen it by presenting one case in some detail. We believe compelling accounts could also be written for other disciplines, such as Art, History, and Physical Education.

Role of Communications in Learning Communities

Learning communities are difficult to build because they have high social and material requirements. School-based learning communities that are attempting to engage in long-term inquiry require access to specialized information, to practitioners, to relevant data and analysis tools, and the means to create arguments and representations so as to negotiate disagreements through them. For example, Dewey proposed that students investigate the history of cotton production and the impact of those changes on the experience of the workers. Such a survey requires students to trace through the history of technological innovations, interview current workers and investigate the lives of former workers, and to express and discuss their results. A CCT could provide essential support for projects of this sort, for example, by providing ready access to historical documents, books and reference material; provide the means and medium to communicate with appropriate workers; and through helping students to structure their discussion via specialized hypertextual media that reify argumentative structure (O'Neill, Gomez, & Edelson, 1994).

The needs of students in work-based learning communities are similarly demanding, including the need to learn about community practices and products, find participants in the community who are looking for an apprenticeship, build and store an accessible portfolio of work, and reflect on and critique their apprenticeship experience. Clearly, a CCT could be useful here as well. It could be argued that apprenticeships have successfully been conducted for thousands of years without the need for CCTs, yet this would miss the point that these new connections to work-based learning communities are not necessarily designed to turn the students into life-long professional practitioners but rather to provide an opportunity for them to understand and investigate a field -- or more to the point, to investigate many fields. The success of apprenticeships historically has demanded years of service where the initiate was immersed in all aspects of a profession. The hope is that shorter more focused and more varied experiences can serve to educate students on a variety of fields, thus providing them with the ability to integrate diverse expertise and to choose a field based on experience rather than inheritance or chance.

In learning communities students must connect to communities of practice (e.g., when a school-based learning community decides to build a sturdier or more agile kite it should profitably engage the knowledge of aerodynamics, kite manufacturing, and the weather). Hence, it is useful to characterize the types of interactions students may have with communities of practice. The following list is designed to describe increasing levels of interaction with a community of practice or fuller levels of participation:

  1. Students access published work, such as papers, figures, presentations, and analyzed data.
  2. Students access tools and raw data, such as equipment for production, laboratory procedures and materials, community developed data sets, and analysis tools.
  3. Students engage in dialogue with community members, either written or oral, such as exchange of letters, email, telephone calls, and interviews.
  4. Performance of joint coordinated activities between students and work-based learning communities, such as performing an experiment together or planning collaborative activities.
  5. Incorporating students' work into published archives of a community's work (e.g., when scientist collect and use students' data).

This list will be used to organize the discussion found below of how the World Wide Web (WWW) can be used to form learning communities. The intended value of these categories of use is that they can be used to guide next generation WWW application design and future technology adoption strategies. Hence, it is necessary to briefly describe the WWW and the rapid process by which it has become a dominant technology.

The World Wide Web

In the early and mid 1980's the Internet and inter-networking were dominated by the technical and research community. It was a special-purpose information preserve for the support of communication and inquiry. Starting in the late 80's and continuing into the early 90's the Internet experienced explosive growth. For example, in 1987 the number of Internet hosts exceeded 10,000, in 1989 the number of hosts exceeded 100,000, and in 1992 there were over a 1,000,000 Internet hosts (Zakon, 1994). Much of this growth did not come from more technologists and researchers joining the network but from non-traditional entrants to the network community. In particular, students and teachers were introduced to email, news groups, listservs, ftp, and gopher, thus causing a significant part of the network's growth. Similarly, the WWW, the most important recent innovation in IP technology, has gained a significant foothold in the technical community and is now expanding rapidly into the world beyond the research community, including the K-12 community of students and teachers.

Paralleling the exponential growth of the Internet overall has been the growth of the WWW: A little over three years ago (June, 1993), there were but 130 WWW servers on-line worldwide, a year and a half later there were over 10,000 WWW servers and today (8/96) there are over 200,000 web servers; further the number is anticipated to double in six months (Gray, 1994). This decentralized development of information resources, for everything from scientific databases to corporate products to traffic and weather reports, is an exciting development for lifelong learning and for education and training.

A national survey of schools' access to the Internet (U.S. Department of Education, 1995) emphasizes the difficulties of providing equitable access. The survey, conducted in the Fall of 1995, showed 50% of schools had some access to the Internet, but only 19% channel that access to five or more classrooms. Those schools that do have the Internet do not necessarily have all of its services, nor provide them for students use: Of the schools with Internet access, 93% have Email and 80% have WWW, but only 41% of the schools with Email access provide it to their students, while 70% of the schools with WWW access allow student access. The gap between the "info-rich" and "info-poor" is demonstrated by the statistic: Schools with large proportions of poor students are only half as likely to have access to the Internet versus students serving wealthier populations (i.e., 31% vs. 62%). Nonetheless, 74% of schools without access are planning to obtain it in the future.

In the discussion that follows we survey several current applications of WWW technology in K-12 learning situations. The survey follows the categories proposed above to describe students increasing participation in communities of practice. Our survey is not meant to be exhaustive. Indeed due to the decentralized nature of the network an exhaustive report may be impossible. We do however attempt to describe several example K-12 uses of the WWW that are illustrative of the breadth of use and which might be indicative of where educators and others learning community builders may be taking the WWW.

Following our brief use-survey we will speculate on how current WWW technology and its potential evolution support the needs of learning communities. Finally, we use both our characterization of WWW use and speculation of its evolution to argue that there are specific needs and identifying characteristics for WWW servers developed for educational purposes. Of course, it is likely that any networked information resource can provide some learning advantages (as can any book in the library) whether it is specifically planned for educational uses or not. The potential value of these resources notwithstanding, it seems to us that a broad set of guidelines for WWW servers designed to promote learning will spur both new application design and widespread technology adoption in schools and other learning communities.

Access to Resources

The search for reference and archival information for reports, projects, and other activities is a large part of school work. Thus, students commonly go in search of information in their school and public libraries. Often, a short passage in an encyclopedia meets a student's need, when her goal is to merely summarize commonly accepted information and standard interpretations. However, when students engage in school-based learning communities they must do more than be passive collectors of previously digested information. In pursuit of long-term projects, or in the defense of controversial hypotheses, students must marshall relevant information. The point of the information is for the student to use it and not to merely report it; i.e., for the student to develop, state, and defend a position and not merely to echo the status quo. It is in these information seeking situations that we believe the WWW will be most valuable to the K-12 community. Its value stems from the way organizations can now act as publishers for themselves, allowing them to quickly place their new work on-line and to clearly express their own point of view. These distinctive points of view are in one sense biases and therefore inimical to the student knowing the "truth." Yet, this can serve as an antidote to overly sanitized textbooks. Before students commit to a belief, they should become aware of the range of existing opinions and to identify. at least, some of the ones they do not accept. Hence, as students combine information from more traditional sources with WWW sources, they originate their own point of view, including critiques of alternatives.

Accompanying these broader freedoms there are dangers. There is the danger that students will only pay attention to information that reinforces their prior beliefs; there is the danger students will be misled by bigoted or short-sighted points of view; there is the danger that students will use information they obtain inappropriately. Solutions to these problems lie primarily with the creation of an environment that emphasizes critical review of students' ideas. In addition, teachers need to guide students' explorations of the WWW holdings; techniques to help teachers do this are discussed below.

Due to the decentralized means of publishing on the WWW, its diversity nearly defies exposition. Examples are used to illustrate this enormous array of information. However, these examples can barely scratch the surface of what is available. Our aim is to show their utility to students, rather than to provide even a representative summary.

The information resources for K-12 education provided by the WWW can be divided into the following five categories:

  1. Libraries
  2. Museums
  3. Government Information
  4. Curriculum and Activities
  5. Indices

These categories are a mix of information sources and information format types. However, this mixture accurately reflects the dominant ways that material is currently organized within the WWW. Entries are jointly organized by institution (e.g. universities or libraries) and form (e.g. plain-text article or hyper-text index).

Libraries

The potential for communication networks to amplify the reach of libraries is being eagerly pursued, as can be seen in the variety of offerings on the WWW. For example, the North Carolina State University library (Morgan, 1994) provides links to books and periodicals. Perhaps the quintessential vision of on-line information access for the "information superhighway" is placing the Library of Congress on the network. The difficulties in converting its current vast holdings of paper books into digital media also illustrates the distance that must be covered. However, the first step in this journey has already been taken as can be seen by the Library of Congress's home page (1994). The general vision is simple and powerful: place the holdings of the Library of Congress on-line coupled with flexible search engines, thus providing unprecedented access. Students could pursue virtually any topic, no matter how obscure or popular, without having to worry that the book is not in the inventory of their local library or had already been checked out. Already, WWW locations are appearing on the index cards of many new books. Further, new types of libraries are being created as on-going work is immediately placed on-line. For example, the National Academy of Sciences (1996) makes available its voluminous reports on a wide variety of research areas.

Similarly, other standard library resources are now available on-line, notably the New York Times (1996), often called the United State's "newspaper of record."

Museums

Museums, like libraries, are great bastions of memory in our society. They too have come on to the WWW providing provocative and useful resources for students. For example, the Exploratorium Science Museum in San Francisco (1994) provides "virtual-exhibits" designed to aid in the understanding of scientific principles and concepts. A Museum of Paleontology has been erected on the WWW by University of California at Berkeley. A diverse collection of dinosaur images and information provides a particularly welcome resource for K-12 students who are often fascinated by our gigantic ancient predecessors. Through these new venues, museums and libraries are continuing their primary task of providing access to cultural artifacts and descriptions of nature. However, the possibilities of access at a distance to so much of their collections introduces order of magnitude differences in the level of access that can be achieved. Of course, these possibilities come at the cost of redefining what is being delivered. Providing a picture of a dinosaur is much different than standing before a fifty-five foot high reconstructed skeleton: The agreed upon "stand-in" for nature has been transformed. As great traditional art museums like the Louvre (1996) are creating WWW versions of their collections, new types of libraries are coming into existence, for example, a growing collection of astronomical images (Digital Image Library, 1996) from University of Illinois's National Center for Supercomputing Activities. This data base of astronomical images is being built up from the contributions of astronomers: We call sites where contents are assembled from the contributions of its users generative, since its consumers are also its producers.

Government Information

Government information provides a treasure trove for those hoping to understand our society and to those seeking to play watchdog. For example, through accessing government information and statistics, students can re-evaluate newspaper editorials and study the human impact on the environment. Towards this later end, the Environmental Protection Agency (1994) provides numerous data sets and statistics. Similarly, the United States Geological Survey (1994) provides information and curriculum units. A superb resource for accessing government data has been made available by the United States Bureau of the Census (1994) which provides standard summary reports or the ability to dynamically construct your own queries (e.g., by zipcode). Through analysis of these governmental data sets students can examine changes in our environment and evaluate what populations will be affected, including both the number of people and their demographic characteristics. For example, one project might ask students to acquire a deeper sense of their own local area by summarizing its characteristics and contrasting them with other areas. Another valuable resource for such activities comes from an enormous collection of declassified military spy photographs collected over the past 25 years (United States Geological Survey, 1995a) in the hopes that students and researchers can use them to detect changes in our local and global environments. The legislative branch of government is publishing the full-text of legislation and the congressional record through the (United States Congress, 1995).

Curriculum and Activities

The examples above were intended to illustrate the utility of these data sets and also to suggest the tenor of potential projects: Students can go beyond merely reporting facts and figures to using them in service of testing or promoting a theory. Often students need guidance to effectively use the information that can be found in remotely accessible libraries, museums, and government data bases. A traditional way to package such guidance is through pre-specified activities or a curriculum. Thus, the WWW provides a new medium to deliver these curricula. An exacting procedural example is provided through a lab experiment to dissect a frog, turned into a WWW interactive session by the Instructional Technology Program at the Curry School of Education, University of Virginia (Kinzie, 1994). Ample use has been made of images that illustrate the steps involved in the procedure. In addition, snippets of video are provided that show students going through these steps, thus providing a model to imitate. This multi-media presentation permits specification of the laboratory procedures well beyond what can be provided in a textbook. Further, the text, images, and video are flexibly intertwined, thus providing a more integrated context than a videodisk. An intriguing contrast is the Visible Human Project (1996) that provides cutaway views of human anatomy without however providing structured activities to accompany the images. More activities are available at the Explorer Project (Unified Network Informatics Technology for Education at the University of Kansas, 1994) which provides extensive listings of educational material in the areas of math and science some of which can be obtained directly over the WWW. One of the most comprehensive listings of activities is found at the ASKEric (1996) server, sponsored by the U.S. Department of Education. This server provides a virtual library with hundreds of activities that were donated by teachers and are organized by content area. The activities range from mini-activities that can take one class period to longer curriculum units. This site too aims to be generative by encouraging its users to send in their own activities to be included in the database.

Indices

The lack of a standardized indexing scheme for the WWW is often bemoaned. There is no Dewey Decimal System or exhaustive card catalog. The decentralized organization and inclusion of multi-media and computational services defies existing categorization schemes. However, there are two primary techniques for conducting searches. First, static indices or "trees" have been created that provide a taxonomic key to resources. The most famous are Yahoo (1996) and the WWW Virtual Library (1996) (the bibliographic links take you to their science education or education entry). Often indices found at related sites provide a more focused and better edited list. For example, the list of science and math links at the Lawrence Hall of Science (1996) is superb. Second, dynamic search engines create an index based on key words. The creation of these indices are based on exhaustive searches of the WWW as conducted by "robots" or "spiders" (e.g., Eichmann, McGregor, and Danley, 1994) that crawl through the WWW and compose traces of its paths and by-ways. In particular, the Lycos search engine provides a powerful facility and was used extensively in the preparation of this paper. However, pattern matching on text is not sufficient. Richer and more flexible indexing strategies are needed. In fact, what have emerged are trailblazers (as proposed by Vannever Bush, 1945) who map out whole networks of associations that are worth following -- this paper itself can be viewed as an attempt to trailblaze. In addition, pragmatic indices can be created that are designed to help accomplish tasks (as opposed to subject indices that group together material on a similar topic). Alternatively, students' exploration of the WWW may be best facilitated if they were able to flexibly and explicitly list their goals when accessing a link, thus providing the opportunity to maintain an external memory and executive assistant. This style of assistance is important since students often wander off-topic when using the WWW.

Access to Analysis Tools and Raw Data

A more sophisticated level of access for students to communities of practice provides them the ability to generate custom representations and obtain access to the underlying data, thus permitting novel analyses and interpretations. Often, the form of the data is dictated by the representations employed. These representations are often very complex in that they are highly encoded and are layered on top of other representational systems (Latour, 1988). Not yet present in any of these systems is an attempt to provide scaffolding for novices through the provision of rich contextual information that would deconstruct the representations and provide examples of how these images, data, and tools have been successfully used in practice. This is unfortunately similar to how scientific laws have been presented in didactic classrooms -- formulas presented absent the situations in which they provide explanatory power. Students, for their part, are then puzzled why these abstractions are so highly valued (Toulmin, 1953). Nonetheless, the appearance of these dynamic tools is an important step in moving education beyond a model of student as consumer to a model of student as active inquirer. The examples presented here are grouped by domain as follows:

  1. Weather maps
  2. Scientific visualizations of climate
  3. Interactive Mapping

Weather Maps

Purdue University's Department of Earth and Atmospheric Sciences (1995) has developed a server that allows students to select weather images and data that is updated hourly, thus allowing students to do weather prediction and investigation.

Scientific Visualization

The tools used by an expert community often serve to help define the basic questions and assumptions used. In particular, scientists in recent years are employing scientific visualizations(*3) (SciVs) as a primary means to communicate discoveries and to perform speculative investigations. The extent to which they have permeated scientific practice is demonstrated by picking up any recent issue of Science or Nature and counting how many SciVs can be found. A common use of these visualizations on the WWW is to provide the ability to browse data sets in order to refine one's choice of which data to request. For example, the Lamont-Doherty Earth Observatory of Columbia University (1995) will dynamically render climate data sets as directed through a vast array of parameters as well as provide the underlying data to the user. Similarly, the Pacific Marine Environmental Laboratory of the National Ocean and Atmospheric Administration (1995) renders several oceanic data sets and will deliver the data in spreadsheet form. Another example comes from the Learning Through Collaborative Visualization (CoVis) Project at Northwestern University (1993). Practicing atmospheric scientists were interviewed and their methods and data analyzed, leading to the design of specialized educational software where students utilize SciVs of climate and the earth's radiation budget (Learning Through Collaborative Visualization Project, 1995a). These visualizations are provided at multiple levels of detail so as to aid students in detecting large and small grain patterns within the global climate system. Individual points can be queried for their value; by choosing different levels of resolution these data points can refer to relatively small areas (e.g., 2.5 degrees square) or large areas (e.g., a single scalar for the whole earth).

Interactive Mapping

Evaluating the data that can be obtained over the WWW often requires flexible tools that allow the iterative selection of data and analytical operations on it. Tools like this are becoming increasingly common on the WWW through the use of forms. An example of a WWW server providing interactive analysis of data (even before the introduction of forms) is provided by Xerox's map server (1994). This server facilitates the interactive exploration of digital maps through the operations of selecting a part of the globe, zooming in on the globe, and specifying what aspects to portray (e.g. rivers or political boundaries). In this manner, not only is information provided, but an environment for querying and investigating that information is specified. A fascinating example has emerged of an application, called the World-Wide Earthquake Locator (Gittings, 1995), that combines the information on recent earthquakes with the Xerox map server to provide customized maps of where recent earthquakes have occurred. Hence, by providing applications on a common global platform new applications can be built by linking others together. This philosophy of linking pieces to create larger wholes is fundamental to the methodology of computer science and has been a primary goal of many operating systems (e.g., Ritchie & Thompson, 1974) -- the WWW now provides such a global operating system for interactive applications. The recent emergence of standards for inter-operability (e.g., Java) should substantially enhance the ease with which such applications can be created.

Forging Learning Communities through WWW Communication Media

This section focuses on how the WWW is being used to forge learning communities between students, teachers, schools, and professionals. Here we investigate how the WWW is used as an interactive(*4) communication medium and on sites whose primary goal is to establish learning communities. Interactive mechanisms are an extremely active area of development on the WWW, and include forums which are similar to newsgroups (messages can be posted and read asynchronously); chat sessions which emphasize synchronous communication through entering short messages; MU*s or multi-user environments where users can interact with one another by talking, emoting (presenting a virtual emotion) or programming the environment; and the incorporation of higher-bandwidth facilities like radio, telephony and video-conferencing. These facilities allow opportunities for permanent hypermedia presentations and high bandwidth synchronous interaction not supported by earlier Internet services, like Usenet newsgroups and Internet Relay Chat (IRC). We expect these advances to be one of the most important developments for the establishment of learning communities. The examples provided here are grouped into the following categories:

  1. Connecting teachers together
  2. Connecting students together
  3. Connecting parents and local communities with schools

Connecting teachers with one another

Developed a learning community among teachers is particularly important since teachers are often shut-off from professional contact with their fellow teachers and researchers due to their being in separate classrooms and their extremely busy schedules. CCTs can provide vital links between teachers, thus allowing them to share professional advice and to adopt new practices and methods. Mechanisms of fostering such change are crucial if new technologies are going to be successfully incorporated into classrooms without having to wait for an new generation of teachers to enter the profession. An exemplary project is Web66 (1995a) which helps teachers and schools to start publishing on the Internet by setting up ftp, gopher, and WWW servers. Web66 supports interactive communication through a listserv that connects educators managing WWW servers. Another excellent service provided by Web66 is a comprehensive list of K-12 WWW sites (Web66, 1995b).

The Global Schoolhouse Project also sought (the project ended in 1994) to build a community between schools. Their WWW server (Craighead, 1994) provided an asynchronous electronic spine as a repository of diverse information including: a description of the project's goals, a list of its participants organized by school with home pages for teachers and students, a "hot" list of software being used by the project (i.e., clicking on the name of a piece of software transfered it via ftp), and access to mailing lists and curriculum.

A similar effort, though state-led, is the Texas Education Network (Tenet) project which is seeking to forge community and expertise amongst the Texas educators. One aspect of this is their WWW server (Texas Education Network, 1994) where a wide range of information can be found, including legal, technical, and practical aspects of telecommunications. In particular, they have assembled a variety of Acceptable Use Policies which describe the privileges and limitations governing a student's access to the Internet resources provided by a school. These rules become all the more complex when schools collaborate, leading students operating under different policies into interaction.

Perhaps even more important than officially sponsored collaborations are grass-roots efforts, like the Teachers Helping Teachers site (1996). This site, and similar ones like it, provide the beginnings of an informal and organic means for teachers to empower one another.

Connecting students with one another

This category is broad in that it refers to connections both within a single classroom and across classrooms. The Gonzaga College High School in Washington DC (1995) is part of the Earth System Science Community and is using the WWW to enable students to file progress reports where they record their activities, list problems or questions, and enumerate the resources they employed that day. Teachers similarly file reports. These reports are available to the entire community, indeed to the entire WWW, though selected portions (e.g., the teachers' reports) are password protected. These communication facilities are integrated into a comprehensive set of services that provide relevant scientific instruction and data, bibliographies, and relevant papers for students to use in accomplishing their project. These facilities provide an excellent example of how the WWW can be used to forge a school-based learning community whose daily work products are remotely accessible to mentors. It is critical for mentors to have access to the on-going work of students and the classroom in order to build the communicative context under which they can provide targeted assistance to students' questions.

Another approach is used by the Knowledge Integrated Framework project which supplies a Networked Evidence Database (1996) where students debate issues. These on-line debates are structured to help students see different ideas, engage in argumentation, and learn how to use evidence to solve a scientific problem. Assistance to help teachers integrate this resource in their classroom is offered, including summer workshops, collections of evidence for students to use, examples of prior student debates, and curricula.

Connecting parents and local communities with schools

Parents can provide a vitally needed source of mentoring for students and a source of support for teachers. For example, parents can provide students with an audience interested in their work. Traditionally the primary audience for a student's work is her teacher, often overwhelming teachers with the responsibility of being an audience for too many students. When a school uses the WWW to make student work available to the broad-based audience of the Internet, students suddenly have a new, and greatly expanded audience for their work.

A number of schools have started to make student work available in this way. A notable examples is the Hillside Elementary School (1994) in Cottage Grove, Minnesota, where students create their own personal WWW pages, in effect giving the students a personalized presence on the Internet. In an analogous attempt to allow parents access to school life many schools, especially in urban areas, have begun using "voice mail" systems. These systems allow parents to phone the school, and through a series of touch-tone commands, learn about their child's attendance record or daily assignments. These systems have been very successful in alerting parents to the work their students should be doing, and aids parents in monitoring homework and academic progress. If a school is able to make student work available on-line, a leap is made from showing parents what students should be doing to what students are doing. Beyond seeing their own child's work, parents would be able to view the work of other students. This allows parents to see how their child compares to others, so that they can make a self-determination of academic progress that is based upon quality of output, not on grades or class rankings. Furthermore, through the addition of a forms-based interface, parents could provide feedback on or critique their child's work, adding another dimension to its evaluation.

Local communities serve as another primary audience and source of resources for schools. The WWW can serve schools as an alternative mechanism to communicate with their local communities. For example, the Hillside WWW server is also used by the entire community showing statistics and giving a calendar of events (Hillside, 1994). Similarly, the Houston Independent School District is publishing the design of its new lab school (Perkins & Castellanos, 1994).

Collaboration between Students and Work-Based Learning Communities

A straight-forward and compelling example of the WWW linking students to practitioners is provided by the Ask an Expert site (Pitsco, 1996). This site provides a list of hundreds of volunteers who will provide their diverse expertise to answering students' questions. Many other sites provide similar, but more focused expertise. For example, the United States Geological Survey (1995b) advertises its "Ask-A-Geologist" service and the AskERIC site promises its staff will respond to questions about obtaining educational resources within 48 hours.

The helping hand can reach the other way as well. The Global Lab Project at the Technical Education Resource Center (TERC) is linking classrooms around the globe to provide scientists with data that they collect. One project had students around the United States collect experimental data on acid rain and then pool the data for analysis by experts in the area. Recent efforts are seeking to have students collect data on levels of ozone and other quantities crucial to observing patterns of global environmental change. The international flavor of the project is conveyed by their WWW server (TERC, 1994) which contains movie clips from Global Lab classes in Nigeria, the Czech Republic, Mexico, and Washington, D.C. A strength of TERC's projects is that they have found ways for students' work to be of genuine use and of interest to scientists, thus providing the basis for a relationship based on mutual benefit. This is consonant with the philosophy of former chief scientist, Robert Tinker (1994), who seeks to transform students and schools from isolated islands of concept learning into powerful social resources.

Pursuing similar goals, Vice-President Gore (1993) is promoting Project Globe which has called for scientists to organize and lead students to collect local environmental data (e.g., air and water pollution levels, tropospheric ozone levels, and amounts of carbon-dioxide emitted).

Often WWW servers are used to publicize efforts that are occurring using other communication facilities. One of the most fascinating of these is the JASON project which seeks to build community by having students remotely participate in scientific field trips or "telepresence." Such trips are accomplished through highly advanced robots that are capable of exploring the depths of oceanic vents or the inner molten reaches of volcanoes. Students participate by suggesting courses of action and asking questions. Jason's WWW server (Feldman, 1994) aids in this process by providing a basis for teacher in-service education, dissemination of curriculum, and multimedia recapitulations of past explorations.

In summary, WWW servers provide valuable aid towards the goal of enabling schools to forge learning communities. WWW servers connect students and scientists to their mutual benefit. Also, these servers provide central repositories of project information, software, expert tools, home page biographies, and mailing lists. In this way, the WWW servers provide an enabling communication infrastructure (*5). WWW sites are evolving as a medium for newly formed educational communities to consolidate their identities, build shared vocabularies and expertise, and pool their resources.

The Need for WWW Educational Server Development Guidelines

To date most resources on the WWW has been developed without the specific goal of fostering K-12 learning communities. Yet, as the above examples show, even without this aim WWW resources have been developed that are of extraordinary value to schools and learning communities. In order to continue to advance in this process of facilitating the construction of learning communities through the WWW an explicit characterization of how to accomplish this goal is needed. The beginning point in this process is a recognition that Educational WWW Servers (EWS) are needed that embodies a distinctive architecture. In particular, an EWS must embody the following features:

  1. EWSs need to provide "just-in-time" curriculum for students. This style of curriculum differs from the standard prearranged sequences of material found in most textbooks and is organized so that students get access to resources they need in order to solve their current on-going problem. This style of problem-solving is no different than that employed by adults in the completion of their tasks. When adults need to build a bookshelf they usually do not engage in a four month sequence on wood working, but look up bookshelves in the index of a "how-to" book. Similarly, students will access curriculum resources in order to achieve their short and long-term goals. Necessary components of such curriculum are:

    • Information resources such as those surveyed in section 4 above.

    • Activity structures must be provided that aid students to learn the conceptual structures of a domain (e.g., There is overarching category that unites visible light with imperceptible radiation of longer and shorter wavelengths, that is, the electro-magnetic spectrum) and to build important artifacts for the investigation of physical phenomena (e.g., a pyrometer that can detect infrared radiation). These activity structures should identify materials required as well as place the activities into a social structure (e.g., performed as a small group or whole class). This is in contrast to the assumed model of use on WWW servers today, namely, a single individual accessing information.

    • Assessment procedures designed to aid students and teachers in understanding a student's level of competency in order to direct them to the resources and activities described above. These assessments can also be keyed to educational standards such as Goals 2000 or district-based mandates. In short, an EWS must have more than content, it must also have that content placed into larger contexts of use. In addition, assessment procedures can help students monitor their own progress, thus promoting their executive and meta-cognitive abilities to lead their own learning.

    Matching resources to teaching and learning tasks is the primary design objective of a EWS and can provide rich opportunities for important new design work in computer science, networking, and learning sciences, particularly around issues of information access, intelligent agents, teacher development, domain cross-referencing, and cognitive development in the subject areas. Some preliminary ideas in this direction are enumerated in the section below on a Geosciences EWS. In addition, work-based learning communities can aim to provide versions of their tools and data sets that are surrounded by a rich contextual background to help ease the difficulty of learning their use. This would involve, for example, making the units of the data explicit and their meaning explained in everyday terms (e.g., a joule is the amount of energy spent in lifting a lemon one meter, Emiliani, 1992).

  2. The EWS should offer facilities for authoring and browsing user commentaries. These commentaries provide a living dimension to the EWS, so that it is continually open to the design work that its users are doing in seeking to utilize its resources in their local situations and for their specific purposes. Commentaries can discuss the usefulness of specific resources and the activity structures designed to provide context for their use, and suggest modifications that "make them work better." This design goal of developing a self-improving or cybernetic EWS can provide a major advantage over any previous static educational innovation in curriculum. We would argue that it provides more of a free market for teaching and learning resources which is evolutionary in its nature (survival through use in selection) rather than top-down from instructional designers -- which is often too abstract and ill-suited to the immediate concerns of local situations. In addition, these authoring facilities provide a means for teachers and students to originate their own curricular ideas and to engage in collaborative knowledge construction and debate. The investigation of how to structure hypermedia documents to facilitate learning communities has been an active area of investigation (O'Neill, Gomez, & Edelson, 1994) that can be utilized here in designing mechanisms to foster both divergent and convergent lines of thought.

  3. The EWS also offers a facility for teachers and students to contact others on the network who share a particular interest with them. These contacts can be as simple as WWW entries containing email addresses or they can grow to be as complex as WWW-based protocols that assist teachers and students in making contact with people that have appropriate expertise at the time it is needed in a learning situation. The WWW is now poorly developed in its interactive communicative functionality. In particular, little has been done as regards either infrastructure or content to set up opportunities for work-based learning communities. Explicit attention must be paid to this issue. A primary critique of current educational practice is that it does not "transfer" to the work-place. As critiques of the notion of transfer have pointed out, connections between subject matter that transfer are not inherent, rather they are constructed via analogies and social agreement (Pea, 1987).

    Educators and working professionals must join together to forge connections between their, too often disparate, learning communities. Opportunities can be found for students to investigate issues and potentialities that the work-based communities do not have the time or energy to pursue. Building on Simon's (1969) insight that economic decisions are limited by bounded rationality we recognize that all decisions are made heuristically within an ecological domain where all relevant resources are limited. Our schools provide a huge resource for performing additional research on these decisions. An obvious example is the researching of alternative energy sources, such as solar power. As technologies rapidly evolve the economies of scale for energy production change -- students can provide an ever ready pair of eyes inspecting such possibilities for fiscal viability. In addition, professionals work in established paradigms and modes of thought. These paradigms guide productive work, yet they can also blind one to productive alternatives.

    Students' perspectives can lead to the fruitful production of new hypotheses that incorporate more divergent influences or randomness than a fuller member of a learning community. This random variation is the prerequisite for the evolutionary selection of ideas to occur. The professional community can easily critique and prune such ideas; the role of our students can be to generate and design new possibilities. Through these designs and the subsequent critiques profound learning can occur. Imagine how much more valuable it would be for each class of students to propose and evaluate an alternative source of energy, rather than for each individual student to be successful in passing a standardized competency exam. In addition to critiquing the work of students, work-based learning communities could be formed that would provide for mentoring of students, provide opportunities for joint work, and mechanisms to acknowledge when the work of students was sufficiently useful that it warranted further investigation or incorporation into a community's corpus.

    Overall, it must be emphasized that EWSs will only develop value for education insofar as educators and other educational agents (e.g., parent, peers, and workers) take action to advance teaching and learning design, usability, and promote their broader adoption. These actions can seed a learning web, and establish new distributed communities for educational reform.

  4. Educational servers fundamentally need powerful search engines for specifying teaching and learning tasks by educators which then guide intelligent agent-based search throughout WWW servers. The kinds of considerations in designing such searches include query by progressive refinement around subject domain, topic within domain (cross-referenced interdisciplinary curriculum resources probably requiring indexing of some kind), and grade level appropriacy (but ultimately learner level irrespective of grade). The key here is turning up resources that are "good enough," so that educators and users find that the time and effort costs of browsing and search are worth it, relative to the benefits derived for teaching and learning.

The Geosciences Server: a Sample Educational WWW server

EWSs, if they evolve, will take many shapes; below, we offer a characterization of one now being designed, that is, the Geosciences Server being built as part of the CoVis project. The CoVis Geosciences server (Learning Through Collaborative Visualization, 1995e) is being developed by multiple institutions that include Northwestern University, University of Illinois, Exploratorium Museum, and select K-12 schools. Part of the design effort is to seek to assure compatibility of server resources, materials, and activities with the leading state frameworks and emerging national science education standards. Further, we are attempting to create a generative site by providing means for users to give commentary on existing resources, discuss and present new ideas, and contribute their own resources.

The ultimate goal of the server is to develop a model for how communications technologies can aid the establishment of learning communities. The resources provided by the server are divided into activities, a "teacher lounge," a "student lounge," community news, general information about the CoVis project, and a "what's hot" bulletin board. However, before these resources can be sensibly presented, it is important to understand that these resources serve as the support for a much wider matrix of social activity. Supplemental to designing this EWS the CoVis Project recruited 150 teachers, from 50 different schools around the nation, to join in establishing a networked learning community. Before the first year's activities commenced (in Fall, 1995) the teachers participated in summer workshops and received a paper resource guide containing an initial draft of what was put on the Geosciences EWS, including, curricular activities, pedagogical philosophies, and contact information. Unlike many University sponsored projects the curricular activities are not our primary focus. Instead, they serve as an initial platform around which we can begin to construct our collective common ground. Initially, four CoVis Inter-School Activities (CIAs) were constructed, each was designed to be conducted simultaneously at many distant schools.

The CIAs were designed to include a variety of components to model what we believe to be good ingredients for project-science activity. In particular, each CIA included structured activities and open-ended project work, the opportunity for students to ask questions of experts via "tele-mentoring," and a dedicated listserv where teachers' experiences and ideas mixed with logistical directions and support from CoVis staff. A large part of CoVis's role in promoting this community was in brokering these events, thereby providing logistical support and leadership. The logistical support included items such as lining up mentors and learning resources. For example, one of the CIAs was a five week unit on global warming (Learning Through Collaborative Visualization, 1995b). Structured activities were designed (e.g., analyzing rising levels of carbon-dioxide) and supporting materials provided. For example, data sets were available to be downloaded as spreadsheets documenting carbon-dioxide levels at Mauna Loa over the past 35 years and temperature changes over the near and long term. The open-ended project for this unit suggested students either focus on global consequences (e.g., rising sea level from melting ice caps) or a specific country's policy decisions. Accordingly, mentors were recruited with expertise in global climate change or the policy of an individual country. Each teacher was then matched with volunteer mentors for their students to contact. A recent innovation is a mentor data base (Learning Through Collaborative Visualization, 1995c) designed to support the endeavor of enrolling mentors and matching them to students.

The teacher lounge area provides additional resources that teachers can use to customize the Geosciences ESW and to record their classroom activities. One resource allows teachers to create indices of WWW resources. Using Teacher Link Lists (Learning Through Collaborative Visualization, 1995d), teachers can construct their own view on the WWW, thus providing them a means to structure their students' explorations. These lists are automatically placed in the Student Lounge area.

Through use, the Geosciences EWS will evolve: As teachers contribute their own projects and activities, as they create indices customized to their interests, and as students' projects appear the EWS will adapt to its daily users. A primary goal of ours, as EWS designers, is to find new means to provide for this style of evolution. However, this evolution should not come at an expense in coherency. Hence, some balance between editorial oversight, the recognition that making changes requires expertise and time, and our desired grass-roots contributions must be sought.

It is important to emphasize that the Geosciences EWS does not and can not support community alone. Rather, it is the material counterpart and support to numerous social initiatives, conversations, relationships, and effort. For example, some of the teachers lacked computers or Internet access when they joined CoVis. This meant part of our work in brokering this community was helping them obtain computers and/or Internet access. In some case, this meant providing expertise and in other cases it meant buying computers and subsidizing Internet service. The important point is that the Geosciences EWS does not stand alone. Rather, it is a material support to our on-going work, a conduit for our communications, and a repository of our past ideas and efforts.

Conclusion

The WWW is still in turbulent flux. As it grows and matures many different communities will customize it and shape it to meet their specific needs. The K-12 teaching and learning community will be no exception. We have attempted to point out here that the needs of this community are diverse and that already the K-12 community has found resources on the WWW that are valuable. The demands of education are especially rigorous, particularly when we consider the development of both school-based and work-based learning communities. We have argued that there is a long way to go before the WWW can serve as a major tool for building educational communities. Specifically, the support for interactive communications is woefully underdeveloped, but is receiving significant attention and growth. We have provided a progressive list detailing increasing depth in the formation of learning communities -- we urge that architectural innovations for WWW interactive communications be evaluated with respect to this list. Hopefully, these suggestions will be used to spur thinking and development around customizing the WWW and its applications to meet needs of schools and education more broadly wherever it occurs.

Bibliography

Footnotes

*1 Of course, it can be said that all communities are "learning communities." Hence, implicit in distinctively labeling some communities as learning commuities is a normative notion of what knowledge these communities should aspire towards. For example, in an ethnography of high school life (Eckert, 1989) found students primarily learning from one another, but almost none of this learning went beyond learning the social customs of high school culture. In contrast, within a learning community students engage in acquiring conceptual structures and explanatory systems useful to understand and transform their physical, social, and psychological world.

*2 This term refers to increasing integration of computational processes within communication media. The primary characteristic of a computational communication medium is the ability to not only transmit and receive information, but also to specify the execution of complex procedures. Such integration has become commonplace not only in the Internet, but also in telephone systems through the introduction of automated interaction systems. In contrast, term "computer-mediated communications" is much broader, encompassing not only CCTs but also any communication modality that utilizes computers whether or it allows the embedding of complex procedures (e.g., e-mail or newsgroups).

*3 As used here, a scientific visualization refers to an image rendered through high-speed computer graphics that is based on a numerical data set that describes some quantity in the world. For example, temperatures around the world can be rendered as a color image, typically mapping cold temperatures to blue hues and warmer temperatures to yellow and red hues.

*4 The WWW can be considered interactive in that pages are retrieved in response to clicking on hypertext pointers. In contrast, we reserve the term "interactive" for dialogic interactions, with the paradigmatic example being two people in conversation. Similarly embraced are epistolary exchanges such as found on newsgroups and email. This point is belabored here so as to avoid the conflation between clicking on hypertextual links and dialogue. Dialogue, in contrast to clicking on hypertext links, is social and as such opens the possibility for negotiation of meaning and disagreement as well as the deepening of human relationships which form the obvious center of any learning community.

*5 This is currently true, of course, only in principle, since access to the Internet is present in only a few schools. Hence, a priority for the education community and for the general body politic must be to ensure equity of access, else the learning communities we seek to build will be closed to segments of the population -- this inequity will injure not only the disenfranchised but will also diminish the quality and range of the learning communities we build by restricting the diversity of thought they possess.

Acknowledgments

We are grateful for research support of the Learning through Collaborative Visualization (CoVis) Project by the National Science Foundation Grant RED-9454729, and the Illinois State Board of Education/Eisenhower Program. We would also like to thank our colleagues from the CoVis Project and community of users for extended discussions of these issues, and continual useful feedback on design, rationale, and pedagogical issues.

About the Authors

Douglas Gordin is currently a graduate student in the Learning Sciences Program at Northwestern University's School of Education. He formerly worked in R&D at AT&T Bell Laboratories and IBM's T.J. Watson Research Center. His research interests include designing educational media, adapting scientific visualization for student inquiry, and designing WWW-based collaborative environmental science curricula.

Louis M. Gomez is currently Associate Professor of Education and Computer Science at Northwestern University. He formerly was Director of Human-Computer Research at Bell Communications Research. His research interests include the applications of computing and networking technology to teaching and learning, human-computer interaction, and computer-supported cooperative work.

Roy D. Pea is currently the Director for the Center for Technology in Learning at SRI International where he also holds the position of Principal Scientist. Formerly he was Dean of Northwestern University's School of Education and Social Policy. His research interests include the design of educational media to support collaboration and scientific inquiry, the scientific understanding of learning and teaching, and assessing new technology applications in real-world learning environments.

Barry J. Fishman is currently the Project Manager of the Learning through Collaborative Visualization Project at Northwestern University where he holds the position of Assistant Professor (Research). His research interests include the effects of adding high performance computing and communication to science education and the design and evaluation of large scale educational interventions.