The first roundtable, "Designing Interfaces for Learning", I hosted together with Catherine Caws from the University of Victoria in Canada, and it drew a crowd of Education designers, researchers, engineers and Web developers. Without a doubt, there's a shared feeling that something drastic needs to change with regard to the interface design of the most commonly used educational software. One delegate suggested that interface design will become the key competitive differentiator among Learning Management Systems in the next few years. Another suggested LMSs were going out the door entirely.
Obstacles to research on learning interfaces
Why has learning interface design taken so long to get off the ground? One delegate pointed to the prevalence of commercial systems. The economic strategy in place tends to see a particular commercial system selected, purchased and then integrated into an organization, after which its use is mandated. Innovators attempting to experiment with new designs or interfaces can not use the mainstream commercial systems to do so, and they are heavily discouraged from developing their own systems for this research, as these seem to be competitive or redundant to the invested technology. In other words, a university invests millions into a system like Blackboard, after which they want you to find a way to use it. They don't want you building something else to do the job. This closes the door on research and experimentation on interface design issues.
"Educational Data Mining" or ways in which the interface can collect, interpret and feed back data to the learner, instructor or group, was a running theme throughout both sessions. In the second session, "Touch me, listen to me, move me: learning through multi-modal interfaces" the chairs demonstrated technologies that both collect user data and then interpret and represent that data in visual forms as feedback. Chair Judy Kay described this process as "exploiting digital footprints and turning them into an open learning model". Rafael Calvo of LATTE showed technology for the automatic detection of emotional data from learners using browser-based learning software.
The social interface
One delegate highlighted the often very solitary nature of online learning and the importance of having interfaces that support the social side of learning. Another participant suggested the interface should cue and support social interaction, whereas right now, at best, commercial systems just copy cues from Facebook. Catherine Caws suggested that the importance of the social is the reason Moodle is so popular - because really it is a community.
It was noted that students tend to continue using the tools they already use in their social lives for learning (ie. Facebook and email) over the provided discussion forums. Judy Kay touched on the notion of an entire learner's ecosystem as one that includes all of their personal devices and tools.
Andy Dong, diretor of the Design Lab at the University of Sydney, co-chaired the second session, and restated the premise of the first session, that is: the need for learning interfaces to be considered separately from other interfaces, ponting out that standard HCI principles won't always apply. As an example, he cited common HCI goals including designing for ease of use and task completion and how they can, in fact, conflict with learning goals that sometimes specifically require challenge or productive difficulty.
Visions for the future of learning interfaces
One delegate said she wished people would see learning environments as the classroom, or as extnesions of the classroom, rather than as a separate set of tools. The novel interfaces demonstrated by the design and technology groups included multi-touch tabletops, full-body interactive kinect applications, as well as interfaces that responded to ambient sound input or to user facial expressions.
Perhaps the most interesting comments were on the potential for a more personal approach to the design of learning environments. A personal learning space (closer, perhaps, to an ePortfolio or the personalized magazine app, Zite, than to an LMS) could be the way of the next decade.
I can't help but be inspired by the concept of an evolving customisable intelligent environment directly relevant to, and personalized for, the individual learner. In this futuristic scenario, the interface designer collaborates with the user to design the space. The designer designs for flexibility, adaptability and evolution of the space. It is like a room in which the furniture, content and tools can be moved around and replaced as the learner's needs and expertise change and grow. Surely all the data mining, intelligent systems and touch interfaces could be taking us in precisely this direction. But it's clear we've got a long way to go.