Studying Situated Learning and Knowledge Transfer in a Multi-User Virtual Environment
This project utilized multi-user virtual environments (MUVEs) as a vehicle to study (1) classroom-based situated learning and (2) the ways in which virtual environments may aid the transfer of learning from classroom contexts into real world settings. Situated learning, a major theory about cognition and education, centers on apprenticeship in "communities of practice" (moving from newcomer to expert within a sociocultural structure of practices). MUVEs are a promising medium for fostering and assessing classroom-based situated learning because they can support immersive, extended experiences incorporating modeling and mentoring about problems similar to those in real world contexts.
This project extended an educational MUVE developed with prior NSF funding; its curriculum was based on a problem-solving community in which students gained knowledge through co-interpreting data with other participants who have varied levels of skills. This project studied whether such a virtual environment can sufficiently replicate authentic contexts and multi-leveled communities of practice to provide students with classroom experiences in situated learning.
In addition, the limited ability of students to apply school-acquired knowledge to real world settings is a longstanding, crucial problem that a controlled study of instructional design for situated learning could help to clarify. "Transfer" is the application of knowledge learned in one situation to another situation and is demonstrated if instruction on a learning task leads to improved performance on a transfer task, typically a real world setting requiring expert performance. MUVEs provide a unique experimental vehicle for studying how situated affordances (supports for particular activities created by relevant properties of the situation) and constraints (regularities that are invariant under the transformation that changes the learning situation into the transfer situation) affect students' capabilities for learning and transfer. Situated learning involves constellations of architectural, social, organizational, and material vectors that shape particular settings.
This project developed and studied multiple variants of a learning environment in which these vectors were manipulated, allowing investigation into near and far transfer. The project developed MUVE-based curricula centered on alternative models of situated or constructivist learning and assessed their differential outcomes for student motivation and learning, as well as interface usability and classroom implementation.
We worked with classrooms with high proportions of ESL and free-and-reduced-lunch students, matching experimental to control classrooms to determine the relative efficacy of our approach. The intellectual merit of this activity was that studies of situated learning and knowledge transfer address NSF objectives such as fundamental research on learning theory, the development of new methodologies for studying learning and teaching, and the enhancement of students' abilities to apply academic knowledge in real world contexts. The broader impacts of this activity are that MUVEs may be a promising complement to more conventional kinds of computer-based instruction, particularly for low-performing students unmotivated by conventional pedagogy and skeptical about their ability to learn science. Research on educational MUVEs may also provide insights into virtual environments for "edutainment," enabling home-based educational usage of the many entertainment media utilizing virtual contexts and communities.
Curriculum
The River City curriculum unit was based on students collaboratively investigating a virtual "world" consisting of a city with a river running through it, different forms of terrain that influenced water runoff, houses, industries, and institutions such as a hospital and a university. River City contained over fifty digital objects from the Smithsonian's collection, plus "data collection stations" that provided detailed information about water samples at various spots in the world.
River City was typical of the United States in the late nineteenth century; we used museum artifacts to illustrate building exteriors and street scenes from that period in history. Content in the right-hand interface-window shifted based on what the participant encountered or activated in the virtual environment. Dialogue was shown in a text box below these two windows; members of each team could communicate regardless of distance, but in-team dialogue was displayed only to members of that team. To aid their interactions, participants also had access to one-click interface features that enabled the avatar to express (through stylized postures and gestures) emotions such as happiness, sadness, agreement, and disagreement. Multiple teams of students could access the MUVE simultaneously, each individual was able to manipulate an avatar through their computer.
In our pilot implementations, each class was divided into teams of two to four students, which were "sent back in time" to this virtual environment. During their time in the MUVE, students answered questions in a Lab Notebook, which the teachers later used for assessment purposes. The Lab Notebook started with questions that guided exploration of the environment and developed mastery of the interface, building towards later investigations that were content specific and required completing a data table or graph based on the water samples encountered in River City. The Lab Notebook asked the class to help the city solve its environmental and health problems, which were directly related to middle school science content. To accomplish this, the students were required to collaborate and share the data each team collected.
Beyond textual conversation, students could project to each other "snapshots" of their current individual point of view (when someone had discovered an item of general interest) and could also "teleport" to join anyone on their team for joint investigation. Each time a team reentered the world, several months of time had passed in River City, so learners could track the evolution of local problems.
At the end, students were asked to write to the mayor of River City describing the health and environmental problems they had encountered and to suggest ways that might improve the life of the inhabitants. Learners were engaged in a "participatory historical situation" in which they could apply tools and knowledge from both the past and the present to resolve an authentic problem. In this "back to the future" situation, students' mastery of 21st century classroom content and skills empowered them in the 19th century virtual world.
Through data gathering, students observed the patterns that emerged and wrestled with questions such as "Why are many more poor people getting sick than rich people?" Multiple causal factors were involved, including polluted water runoff to low-lying areas, insect vectors in swampy areas, overcrowding, and the cost of access to medical care. Throughout the world, students encountered residents of River City and "overheard" their conversations with one another. These computer-based "agents" disclosed information and indirect clues about what was going on in River City. The phrases "spoken" by these agents evolved over time.