Monday, November 12, 2007

10 The Proteus Effect and Second Life

Online fantasy games such as EverQuest and World of Warcraft have long attracted a subset of the population. Unlike Second Life, these games allow users to indulge orc-slaying fantasies and explore Lord of the Rings-esque landscapes. Though the cyber universe might seem best suited to providing outlets for virtual adventures impossible in the real world, Second Life has reached a large demographic that prefers buying cyber groceries to attending virtual elf weddings.

When users first sign up for Second Life, they construct an avatar. Every aspect of this avatar’s appearance is left up to the user, and the number of potential avatars is virtually limitless due to the large number of changeable variables offered. Once users are satisfied with their creation, they are encouraged to explore the Second Life world. Users can easily find social events and other attractions to attend. For a fee, they can even acquire property, from a small plot to an entire island. Users are also free to create or build anything they like using tools provided in Second Life. They then retain rights to this virtual creation and can sell it to others in the cyber world.

Yee and Bailenson demonstrated evidence for the Proteus Effect, in which a person’s behavior is governed, in part, by the features of the avatar assigned to them. Their experiments showed that more attractive avatars offered more intimate self-disclosures and taller avatars were more confident negotiators. When I signed up for Second Life, I decided to create a tall avatar. The avatar resembled me in almost every way except height. I explored Second Life, attending a nightclub as well as a shopping mall where I talked to various avatars. While Yee and Bailenson predict that I would behave differently while inhabiting a tall avatar, I found no evidence of behaving more dominantly or self-assuredly in Second Life than I normally would. This could have stemmed in part from my “newbie” status, but I think the nature of this self-experiment makes an accurate behavioral assessment difficult. In addition, Yee and Bailenson’s experimental situation was fundamentally different from the one employed in this exercise. While they studied situations in which avatars were assigned to users, I was allowed to create an avatar with features of my own choosing. This difference in freedom could lead to a difference in behavioral results.

No comments: