In a paper titled "Evaluating Computer-Supported Cooperative Work: Models and Frameworks," John M. Carroll, Mary Beth Rosson and Dennis Neale propose an assessment framework for computer-supported collaborative systems based on "activity awareness." The three also have developed a variety of tools-a timeline for document histories, deadlines and project status as well as a concept-map interface-to enhance activity awareness.
In the researchers' model, activity awareness includes such variables as context, communication, coordination and whether the structure of the work is tightly coupled-dependent on frequent communication-or loosely coupled-less need for frequent communication.
"The more shared activity awareness among users of computer technologies, the more effectively a group will function," said Carroll, the Edward M. Frymoyer Professor of Information Sciences and Technology, and lead researcher. "The less shared awareness, the more likely a breakdown will occur."
The findings were presented Nov.8 at the ACM Computer Supported Cooperative Work Conference in Chicago. Rosson is professor in Penn State's School of Information Sciences and Technology, and Neale, the paper's first author, is a doctoral student at Virginia Tech where Carroll previously headed the Center for Human-Computer Interaction.
While researchers have well-established models and procedures to evaluate single-user systems, there is no well-articulated framework for assessing multi-user, computer-supported cooperative systems, Carroll said. To fill that vacuum, he and the other researchers are proposing "activity awareness."
"Awareness is both a process and a product," the researchers wrote. "The more aware people are, the less there is a need to coordinate activities."
The researchers studied sixth and eighth-grade students in two Virginia middle schools during 2002-03 and 2003-04. Both years, the students collaborated on science projects with peers from the other school. The project used a Java-based system called Classroom BRIDGE with features including a real-time interaction editor and an integrated chat tool. In addition to studying how activity awareness influenced collaboration, the researchers also introduced new system features in 2003-04 to support improved awareness.
Observation, interviews and questionnaires revealed that in the first year, students had difficulty collaborating with their remote partners. That breakdown occurred largely because important information known by one group of students wasn't shared with the remote partners. Planned activities were disrupted; schedules weren't followed; and ability of the teams to work together was compromised.
"In the second year, we had better and more stable software that included the timeline version and concept-map interface, and those gave the students a better awareness of what their collaborators were doing," Carroll said.