1998 Conference Attendees

Portals is a client-server application for use on a school's web server and an associated set of support materials for teachers, students, and mentors. Portals is designed to facilitate substantive communication between students, their mentors, and their teachers about research projects involving computational problems.

Students use Portals throughout their project work to create "packets" of documents that describe and explain what their projects are about, their hypotheses and predictions, their progress and findings, and their questions. The documents students share through Portals can take several forms: commentaries, concept maps, flow charts, and visualizations. Students choose the forms that best suit the information they want to convey, the stage of their project work, and the nature of their questions.

Teachers use Portals as a way to have access to what students are thinking and doing so they can provide appropriate help and guidance over time. Since Portals is also a tool for archiving student work over time, documenting learning milestones and missteps, and revealing how students go about building their work plan and their understanding, teachers may also find it helpful for assessing students' progress.

Mentors--who may include practicing professionals from a variety of scientific fields as well as teachers within the school or at other schools--use Portals to receive the documents, comment on the work they represent, and return the annotated documents to students. These exchanges occur throughout the school year, with students refining and revising their documents over time based on comments from mentors, discussions with their teachers, and their own growing understanding of the problem.

During 1996-97 we conducted formative research with preliminary versions of the tool and built the tool itself. During the current school year we are implementing Portals in our testbed schools and conducting research on its use in instructional and assessment activities and its impact on the content and patterns of student-mentor-teacher commmunication. We are also investigating the impact of Portals (specifically, its emphasis on multiple representations of knowledge and revisitation of past work through archival features) on how students learn.

Portals is funded by the National Science Foundation's Division of Collaborative Research in Learning Technologies, Grant #CDA-9616990. It has been developed in cooperation with the Department of Energy's Adventures in Supercomputing Program (AiS). The Portals software has been developed in collaboration with Arcus, Inc.

Portals is in large part a response to two years of student learning research we conducted for the Adventures in Supercomputing (AiS) program--schools participating in AiS are now serving as the testbed for Portals). For copies of these research studies see:

In the AiS program (sponsored by the U.S. Department of Energy), students work in small groups to identify and develop computational solutions to problems that they find interesting. They generally work on computational science projects for an entire semester or year; often the nature of the problems they choose to study requires that they collaborate with scientists and other project mentors.

Findings from our research indicated that:

• Teachers and students rely upon mentors for content expertise.
• Teachers want detailed feedback from mentors and external experts about the accuracy and soundness of students' project ideas, methods, and conclusions.
• There are often discrepancies between what teachers and students expect from mentors and what they feel they receive.

Furthermore, data suggested that:

• Particularly when dealing with unfamiliar mathematical models and complexvisualizations that are often fundamental to computational science projectwork, students did not know how to characterize their current states ofknowledge and need or how to express nascent thoughts on the assumptions,limitations and implications of their models.
• Mentors and external reviewers often did not know quite how to interprettheir roles in student project work or just what the students were askingof them.

Mentors described the challenges of communicating with students who werenot generally at a level needed to fully understand the phenomena they wereinvestigating. They also discussed the struggles of wanting to find waysto best explain concepts that students need in order to complete theirprojects while not revealing complete solutions to these projects.

We also found that students and their teachers were facing three centralchallenges when working with computational techniques and technologies tosolve meaningful problems. These challenges were:

1. Shaping students' research interests into tractable problems for investigation;
2. Developing mathematical models to represent these research problems; and
3. Drawing meaningful project conclusions from numerical data and scientific visualizations.

Portals was designed to draw mentors and teachers more closely into students' thinking and collaborative process so that they can more effectively intervene and participate in students' work around these key challenges.

The emphasis on collaboration in Portals (as opposed to emphasizing production) reflects our interest in and enthusiasm about the potential of the WWW for supporting flexible, prolonged and in-depth conversation among students, teachers and mentors around focussed and long-term project work. Our interest in this educational application of networking technologies builds on our constructivist approach to education, which emphasizes inquiry, conversation, and scaffolding--all activities that we stressed in our approach to building Portals. Portals makes central the communicative aspects of students' models of phenomena under study, and their representations of their knowledge. Our understanding of the potential of the WWW to support this kind of work was also built on lessons learned by many other exemplary programs, including CoVis, and Elliot Soloway's Hi-C group.

Our institutional capabilities limit our ability to quickly mock-up even very simple preliminary versions of new software ideas. Working with an institution with a greater capacity for in-house development would give us a better opportunity to try out preliminary concepts at low cost and low risk -- this would also increase our own experience with and understanding of the technical hurdles and opportunities we are working with.

Working with institutions with strong ties to large teacher organizations or teacher education institutions would give us valuable opportunities to learn more about how to most effectively scaffold, contextualize and explain new technological tools in ways that will make them seem sensical, useful, and do-able to larger groups of teachers.