Jeremy Bailenson, the founding director of Stanford University’s Virtual Human Interaction Lab and co-founder of STRIVR, studies the impact of virtual reality experiences on participants, with a focus on how virtual experiences change participants’ perceptions of self and others. His team has nurtured researchers whose pioneering work has studied the potential for VR to drive lasting changes in behavior and assist people in deepening their empathy for others and their own future selves. VHIL researchers have published numerous papers describing those studies and their potential effects in the areas of social interaction, education, health, and behavior around environmental conservation.

Julie Dirksen, an instructional designer, consultant, and author of Design for How People Learn, interviewed Bailenson for Learning Solutions. The interview has been edited significantly for length.

Julie Dirksen: Using virtual reality for training is new to a lot of people. One of the big questions people have is: When is it worth it? What kinds of learning are possible in VR that maybe other media can't provide?

Jeremy Bailenson: Training has always been the natural application of VR; it started with the flight simulator in 1929. Your question, though, is when is it worth putting on those goggles and blocking out the physical world? In Chapter 10 of my book (Experience on Demand), I present guidelines for when it's worth going into VR.

I've created a nice acronym: DICE. The idea is you should do something in VR if you were [thinking] to do it in the physical world [and] it would be:

  • Dangerous: Like the flight simulator mistakes. People learn by doing, making mistakes, getting feedback, and iterating. That process is dangerous in a plane.
  • Impossible: Things like: I become a woman; I become a person of color; and I experience prejudice while actually being in someone else's body. You can't do that in the real world.
  • Counterproductive: This falls into the bucket of experiences where you learn a very important lesson but it comes at a cost. For example, a father in the 1970s catches his son smoking a cigarette, sticks him in the closet, and forces him to smoke a carton to learn a lesson. You may learn that lesson—but at a serious cost to your lungs. VR gives you the best of both worlds: You can have an experience that changes the way you think about something without having physical harm.
  • Expensive: And rare. A lot of the things we do in corporate training and in the STEM field trip training fall into this [category]—it's really expensive to fly to the top of Mount Kilimanjaro but in VR, it's not [expensive].

In the training space, the way a lot of what we do now becomes expensive is that some trainers are better than others. It's hard to get the best people all of the time. Even the best people have bad days. VR gives you this level of consistency.

We know from learning science that the best way to learn is repeated experiences. You get some variations in the surface details, but the overall lesson is repeated. Organically, those variations actually improve the generalization of transfer because you've now got many instances and you've learned a general category, as opposed to a specific instance. With VR you can start with the baseline perfectly consistent on what you want to have—which is the perfect instruction, the right level of interactivity, the right level of telling vs. doing, and constructivism. But then you can add the variation that you want.

The other thing VR gets you from a learning science standpoint—we now know based on dozens of studies—it's so much better to do shorter learning trials and spread them out. What VR allows you to do is exactly that. You don't have to hire the expert to fly in. You do these lessons on demand, on the learning schedule that's optimized from what we know about how neuroscience works.

J.D.: We're really focusing on behavior change issues. The really hard problems come where somebody knows the right thing to do, but they're still not doing it. What promise do you see there?

J.B.: Yes. There is a reason why the book is titled Experience on Demand.

Before I started publishing a lot of my more exciting research about using VR to do behavior change, I had to establish myself academically in the field. A lot of those studies were about replicating normal face-to-face processes in VR to establish that the brain treats these virtual experiences similar to real ones. For example, you put someone in a room with another person and see if they choose to respect that person's personal space; put a physiological sensor on someone and see if the arousal levels when someone else ostracizes them is similar in VR compared to the physical world.

And they are! Dozens of studies really have established that the brain treats a virtual experience—not identically of course—but in ways that are very consistent with what happens in the real world.

There's a reason why there's a phrase “experience is the best teacher.” If somebody had decided not to put their helmet on because they didn't want to mess up their hair, and they then physically watched their best friend, who made the same decision, have an accident and get injured because of that, I bet that person would wear the helmet.

It's as simple as that: VR gets you out of the abstract, makes it a visceral experience. The smart way for our readers to think about VR is: VR does not provide a “media” experience. It is closer to an actual experience.

When we design our content, that's what I always tell people to think about: Imagine that you actually had all the resources to build a store and create an accident. And they [learners] get to see their own arm get hurt by a box falling onto it because they didn't follow procedures. You get to do that without the injury—and the brain treats it as if it were real.

J.D.: We know that future intent and actual behavior do not match that often or necessarily.

J.B.: With all of our studies, we look at behavior because my interest as a scholar is to study these important behavioral change areas; you know—things like conservation, exercise and health, learning to be better toward others in terms of prejudice and intolerance. All of these are things we say we want to do better.

But it's hard to change your daily behavior. It's hard to pick through every piece of garbage and separate the recyclables in the perfect way. It's hard to carpool in Uber as opposed to taking an Uber X. I mean all of us could do better. Our research in general focuses on using VR as a lever to change what I call these “entrenched behaviors,” ones that are really dug in and hard to change.

In the tree-cutting study, everybody got the same statistics. There were three groups of subjects across a few studies. One read a beautifully written narrative about what it's like to cut down a tree. The second saw a video shot from the shoulder of a lumberjack. And the third actually used a haptic device and wore the goggles to cut down the tree.

The researcher did two things. Before and after, she asked people about their intention to recycle and to conserve paper. Everybody showed a pre/post difference in intention to conserve—everyone says they want to conserve more after learning that toilet paper comes from real trees compared to beforehand.

But then she found a way to measure their behavior afterward. And this is in a number of studies both in Georgia and at Stanford. She found a way to look at their actual paper use. The only group that actually changed their behavior was the VR group. They reduced paper use by 20 percent. So that's best example of an experience which would be “counterproductive” in the real world. Cutting down real trees to teach the dangers of deforestation would be ridiculous.

VR is not a magic tool. It doesn’t solve all the world’s problems. I often get asked, “Does virtual reality cause empathy? Does virtual reality cause behavior change?” My answer is: You would never say that about the medium of the radio or the television. It depends on what you do with it. It's just a medium. And you know it's up to us to figure out how to leverage what makes that medium special.

J.D.: One question that comes up is: The immediate effect is great, but the question with behavior change is how long-lasting it is.

J.B.: We just published the big canonical paper on this. We built a simulation called “Becoming Homeless.” This is about a seven- or eight-minute experience that's designed to show that there is a psychological phenomenon called the fundamental attribution error. Basically this says that when something bad happens to us, we blame the situation; when something bad happens to someone else, we blame their character. Most people do this; it's a human thing.

So, with homeless people, can you reverse that so that people can see that situational factors can actually cause someone to lose their home?

So we built this experience. Fernanda [Herrera, the lead researcher] ran two very large studies. Our behavioral measure was: After you go through we hand you a petition; you actually had to sign a document agreeing to have your personal taxes increased in order to support affordable housing measures.

Immersive VR caused more signatures compared to all the control conditions, including the gold standard of role-playing, the way we do perspective-taking in corporate retreats.

In study two she looked right after treatment; and two weeks out, four weeks out, and eight weeks out. VR continues to outperform the best control conditions from the first study—which was role-playing—even eight weeks out.

But in general, we need more data on the longitudinal side. There's just not enough; there's only been five or six studies that look at anything after a day—and there needs to be hundreds.

Focus on behavior change using AR and VR

Julie Dirksen, along with coauthors Dustin DiTommaso and Cindy Plunkett, examined the potential for training using both VR and augmented reality, or AR, to drive lasting behavior change. Download their report today: Augmented and Virtual Reality for Behavior Change.