Recap: Nick Yee Visits Metanomics

Download the video (Quicktime)
Download the audio (MP3)
Read the transcript
Metaversed video archive at SLCN
Subscribe to the Metanomics '08 feed

Last Monday, March 17rd, 2008, Robert Bloomfield and Metanomics hosted Dr. Nick Yee, who applies his Communication Ph.D. from Stanford to conduct experimental research on social interaction and self-representation in immersive virtual environments at the Palo Alto Research Center (PARC).

Broadcasting live on SLCN.tv from CMP's 4-sim Amphitheatre as part of Life 2.0 Conference, Dr. Yee discussed his research in detail.

ROBERT BLOOMFIELD: The way I see it we can organize your work into three different categories. There are the studies of how virtual environments affect the way we perceive ourselves and others; there’s the demographic work on who participates in games and why; and then there’s the work on social interactions in virtual worlds and, particularly, your research on guilds in World of Warcraft.

Interview highlights and analysis are available after the jump.

The interview covered in most detail Dr. Yee's research on the Proteus Effect.

NICK YEE: We labeled the effect the Proteus Effect from the Greek god Proteus, who was able to change his physical form at a whim. So there’s this passage where it talks about he’s changed from a tree into a lion and so forth, and we used the Proteus Effect label to talk about an effect that we had found in a series of experiments. We were looking at whether, and in what ways, the avatars that people use in virtual environments change the ways that they behave. And a lot of people who had been looking at computer-related communication had been looking at features of the environment, such as anonymity or the lack of social cues and so forth. But, for me, looking at online games in virtual worlds, I was real interested in the self representation issue as a fundamental issue that one of the big things about being online is that you can literally change your height, your weight, your gender at a click of a button. You know, what does that mean when we have these flexible Pro-tean bodies? And so that’s really what the effect labels. And so, in a nutshell, what we found was that the avatars that peo-ple choose have a huge impact on how they behave.

Bem’s theory of self perception essen-tially says that, in a lot of cases, we’re not sure how to think about ourselves or how we really feel about things. And what we do is we observe our own actions and our own behaviors from a third party perspective, and infer from the observed actions how we would behave. So there’s this one early well known study where when you put people in black uniforms, they become more aggressive. They pick more aggressive games to play. There’s an-other study where, when they put someone in a nurse’s uniform, they’re less likely to use and this was in a teacher/learner paradigm where they use electric shocks. This was the pre Milgram days. They found that when you put someone in the nurse’s uniform, they’re less likely to use the electric shocks than if you put them in a uniform that resembled a Ku Klux Klan uniform. And so drawing from that research we were saying, “What happens in the Virtual World? Can we leverage this research in the Virtual World where, not only can we manipulate costumes or what people wear, but literally what people look like”? And so that’s where the attractiveness and the height research came out of was leveraging that.

Virtual world research methods are still in their nascent stages, and it is worth noting how Nick's experiments are able to assess complex psychological theories through manipulations of virtual environments that are not possible in 'real' worlds.

NICK YEE: A lot of the studies that we’ve run at the lab focus on the fact that everyone thinks they share the same virtual space, but in a lot of cases they don’t. And so we’ve run a lot of studies where we’ve had these kinds of manipulations. We actually did another study with the height, where we did exactly that, where the participants thought they were tall, but the research assistant saw them as just average height. So again, it’s all in the head of the participant.

Of course, as you said, there’s this real interesting case where what if the participant thinks they’re tall, but everyone else thinks they’re short? So there are all these really interesting ways of using the virtual space to run these really cool studies that can tease apart really complicated theories in psychology.

In one experimental study where we put participants in either attractive or unattractive avatars, the participants in the attractive avatars were willing to walk closer to a confederate stranger, so a research assistant posing as another user in the environment, and they were also--the participants in the attrac-tive avatars were also more willing to disclose personal information about themselves via introductory prompts. So that’s just in a nutshell how we came to that term of a way of thinking about what it means to have these very flexible fluid bodies.

individual choice, which plays an important role in interpreting the significance of research conducted in virtual worlds that relies on the

As Yee's research often relies on the exogeneous assignment of avatar characteristics to determine their effect on individual psychology, audience participants were quick to point out questions about endogeneity in appropriately interpreting individuals' actions. The unanswered element appears to be the role of individual choice which are normally invoked in selecting avatar characteristics.

NICK YEE: That’s something on a “to do” list. So you know, of course, looking at user behavior in avatars, you’ve really also got to look at the choices. So in our studies, we focus on the effect side of the issue. But absolutely, the choice side is just as important. But it also happens to be a lot messier, because choice studies are inherently nonexperimental so they’re a very different class of studies altogether, but they’re equally important. Some of the findings we’ve seen is that it has a lot to do with personality variables, it seems. So whether someone creates avatars that are similar or dissimilar to themselves physically seem to have some personality variables interwoven in there. So it’s definitely very important. It’s a very important part of the issue. In our studies, we focus very much on that effect side, but the choice side is definitely equally important.

As virtual research methods continue to be developed, the discussion about the proper interpretation of virtual research evolves. However, it is worth noting that this emphasizes the abundance of open questions to be explored through additional research in virtual spaces, not only in the applications of virtual worlds to established academic questions but in developing novel research tools to leverage virtual spaces and assessing their validity.