In design and implementation, the WQS learning environment is not limited to the digital domain but anticipates a rich classroom context that embraces varied, flexible interactions among students and teachers. Consistent with this holistic perspective, we are forming iterative design partnerships with K-12 teachers in which to carry out prototyping, curriculum integration, and evaluation (Garaway, 1995). What we learn will be applied not only to refinement of the WQS prototype, but also other "WebSims" focusing on different pieces of the science curriculum.

Since the project's inception in the spring of 1998, research, design and implementation activities have included:

• Computer models of system relationships between land use and the chemical indicators of water quality within an "archetypal" (i.e. not geographically specific) river basin, and implementation of a workable model within a client/server framework.
• Web interface for interacting with the simulator via the Internet, enabling learners to select locations (subwatersheds corresponding to distinct land uses), choose input variables, including timebase, and view model output in the form of graphs.
• Web-based, interactive "Tour" that introduces learners to the key operations as well as important scientific ideas behind the simulator.
• Digital notebook enabling students to record their observations and articulate explanations and hypotheses as they investigate water quality relationships aswell as providing teachers with a means to structure and assess student investigations.

In addition, to assess the WQS with respect to our stated goals and research agenda, it was necessary to devise a sound pedagogic framework, which we based on the "Jigsaw" approach to peer-group learning.

Having created an initial prototype and "draft" pedagogic framework, we next developed and implemented a number of workshops and labs in which to begin field-testing the WQS. The following summarizes evaluation activities conducted through the summer and fall of 2000.

• AP Environmental Science, Montgomery Blair H.S. Two classes of 11^th graders.
• Honors Biology, Montgomery Blair H.S. Two classes of 9^th graders, totaling 54 students, each consisting of two 1 1/4 hour block sessions. Data collected: notebook entries, prediction statements, video and audio data of student pairs, pre- and post-tests.
• Bowie High School Summer Bridge Program for average level students. The WQS was field tested in classes that utilized GLOBE (Global Learning and Observation to Benefit the Environment) protocols in comprehensive studies of water quality. Data collected: notebook entries, video and audio data of student pairs, pre- and post-tests, and student presentations.

• Two day workshop with 7 teachers at Montgomery Blair H.S., July 2000. Grade levels ranged from 8-12.
• Subject areas included Earth Science, Environmental Science (Regular, AP), Ecology, Biology, Computer Science.
• Teachers worked with the simulator in pairs, following the peer-group learning strategy based on the Jigsaw technique. In doing so, they modeled and tested the same pedagogy that we advocate they employ with their students.
• Focus groups and in-depth interviews with the participants examined their experience working with the simulator, notebook, and the Jigsaw technique.
• Data collected: notebook entries, concept maps, video observations of simulator operations and associated peer group dialog, video and audio with transcripts of focus group sessions and in-depth interviews

Most of the teachers we have worked with during spring and summer, 2000 have agreed to employ the simulator in their classrooms during the 2000-2001 school year.

Based on the preliminary data we would like to conduct further prototyping of the WQS along the following lines:

• Further elaboration of the computational model, with attention to hydrological and biological variables, as well as the possible addition of economic values to help student weigh benefits relative to costs and risks of alternative mitigation strategies.
• Placement of the computational model on a spatial foundation so that changes in indicators may be tracked at multiple locations within and between sub-watersheds.
• Generation of history information providing students, teacher, and researchers with a trace of student activity.
• Development of software support for student exhibits and other culminating activities. This would include specialized tools allowing students to incorporate selected graphs and notebook entries into a web-based report.
• Scaling of server-side capabilities to support collaborative teams of students at many locations, and also collaboration among teachers in iterative development and testing of accompanying classroom materials, including questions for the digital notebook, and sharing of experiences in using the WQS and other "WebSims."

To this end, we are investigating a number of funding opportunities, and are exploring strategic partnerships with other education researchers and scientists.