December 2, 2008
[Berkman] Chris Dede on Immersive interfaces and education
Chris Dede is giving a Berkman lunchtime talk on using the new immersive environments. [Note: I’m live-blogging, which means IO’m not checking for errors, and that I’m missing stuff, getting things wrong, paraphrasing, etc.]
Why immersion? “Immersion is the subjective impression that one is participating in a comprehensive, realistic experience.” Immersion can help learning by providing multiple perspectives, situated learning, and shifts in identity. Chris is interested in how we can make meaning out of complexity, using immersive interfaces in middle schools.
He sketches three types of immersive interfaces:
1. Augmented reality. You’re in the real world — you’re not an avatar — with a device that lets you overlay the real with the virtual. Entertainment and education can be anywhere. He shows a bit of his middle school math curriculum called “Alien Contact,” which uses mobile phones. Aliens have landed outside the school. The students explore the area (the real physical area), interviewing virtual characters and using mathematical and literacy skills. Students see different pieces of evidence based on their roles (FBI agent, linguist, computer expert, chemist), and have to collaborate to see the entire picture.
2. Alice-in-Wonderland, like SecondLife. Chris’ project has its own MUVE (multi-user virtual environment). This is partial immersion because you’re sitting in front of a monitor. He shows a clip about RiverCity. It’s a 3D simulation of a 1880 town battling infectious diseases. The students have to figure out what’s going on, learning the scientific method.
Situated learning — e.g., a medical internship — i s another example. You learn by doing and by watching people who know what they’re doing. Chris is using a virtuated environment to created a distributed-learning community.
3. Full immersion. Head-mounted displays. E.g., NewtonWorld, where you can see how balls interact, varying mass, velocity, etc. Similarly for MaxwellWorld.
He opens up the discussion.
Q: Would this work with university students? More sophisticated students?
A: A virtual ecosystem can be easy enough for a middle school student, but you can also imagine one complex enough for a university or graduate student.
Q: Complex environments are hard to create.
A: The good news is that the tools are being created by the entertainment industry. We then re-fit them our purposes. E.g., the authoring shell for the game Oblivion is very powerful. Within 5 years we’ll probably be able to build mixtures of emergent behaviors and scripted behaviors that are really compelling.
Q: Why did you make RiverCity historically situated. Doesn’t that make less obviously relevant to the kids.
A: We needed our kids to be experts. Even the least sophisticated kid today knows more about medicine than the most sophisticated person in the 1800s. [I love this idea.] Also, I wanted to show you could teach multiple things at the same time: science, history, English…
Q: [jz] Harvard Libraries have an outpost in SecondLife but not in Wikipedia. There seems to be something about participating in virtual places. Do you think of Wikipedia as an immersive environment? What would it mean to make it so? And would it improve it?
A: Wikipedia doesn’t work for sensory immersion, actional immersion (being able to fly, e.g.), but it might for symbolic immersion (what you get late at night if you’re reading a horror novel), depending on what you’re reading about or co-creating. A better example might be a Harry Potter fan fiction site. You can imagine putting the Wikipedia for HP inside a virtual HP world — your HP avatar could write an entry in the inner Wikipedia. And would it be better? Lectures are generally better in the real world. But it’d take a lot of discussion to answer your question fully….
Q: Some manuscripts can only be experience in a group via a virtual environment.
A: Yes. You could set up a virtual museum exhibit that brings together works, and that might let you explore the artist’s world. Or, for Van Gogh, what the world looked like a schizophrenic.
Q: How can you keep up with the commercial environments so that the educational ones don’t look old fashioned?
A: It depends on what factors matter. In terms of fidelity, many studies show that you need high fidelity in the parts where the experience requires it — e.g., teaching how to read X-rays — but you can have low fidelity for the parts not directly related to what you’re teaching. If it’s engaging, users don’t care about the low fidelity. None of the 15,000 students who have used RiverCity have complained that it’s too cartoon like, even though it’s not even remotely as photo realistic as the games they play.
Q: Metrics?
A: All of these projects measure gains carefully. They’re research projects. Typically the research shows that if it’s well designed, you get gains in learning…which is what research shows for just about educational technique.
Q: [me] First, I love the idea that in RiverCity, students are treated as experts. How much of this would you do in a day? How much of this is the film strip break in the day?
A: It varies developmentally. For young children, I’d do very little. You learn over and under by crawling, not by having your avatar do it. As they get older, maybe 15-20% of the day? It depends on the topic, the age of the students, etc…For my courses, I’d use the virtual environment at the beginning to let them see the scope of the landscape. In the middle, they’d do a formative experience inside the virtual environment: Here’s what I understand so far. At the end, you’d do a summative experience.
Q: [ethanz] Have people done side by side studies of these environments and other creative interventions, including teachers putting in an enormous of creativity into changing a lesson plan. Your examples tell us about engaged teachers more than about virtual environments, perhaps.
A: It’s a question very relevant to policy. One of the considerations: RiverCity’s cost for 30 kids is about the same as for 3,000 kids. But even the most skilled teacher could give students the sense of going back in time. Where the world is not doing much more than lecturing, you’re right to be skeptical. How are we testing this claim? We have control conditions for RiverCity and Alien Contact. The control conditions are paper-based games. We found a strong difference in engagement. In RC, we found a big difference in learning; in AC we’re breaking even in learing, but it’s a first gen project.
Q: I teach law. You are expected to immerse students into being just, fair and convincing. That’s entirely inter-human. To what extent could this virtual, artificial interface enable the inter-human relation, or perhaps hinder it.
A: Immersive interfaces aren’t equally powerful for all subjects. I don’t know the answer to your question, but one of the thigns we can do in RC is have two people can be in the same room and have different experiences. E.g., you could build a pre-Civil War environment. Two avatars walk down the street together. They see the same things, but one is a slave and one is a slave-holder. That leads to an interesting conversation.
Q: [charlie nesson] Can you establish a transfer of skills from games to real world skills?
A: I’m skeptical about claims of far transfer. The evidence there is weak. I’m quite more convinced about near transfer. So, saying that you’re good at World of Warcraft and thus you’d be good as a lawyer isn’t going to get you too far. It might mean that you can make fast decisions, but WoW aggression probably doesn’t correlate with aggression in the courtroom. The first is a near transfer, and the second is a far transfer.
Q: Has there been a lot of research on this?
A: Not that I’ve found. Closest you get to this is the military that has evidence that military skills transfer to civilian life, and many of those skills are gained by simulation.
