Reconsidering the avatar : From user mirror to interaction locus

• Main Author: Jää-Aro, Kai-Mikael
• Format: Doctoral Thesis
• Language: English
• Published: KTH, Numerisk analys och datalogi, NADA 2004
• Subjects: virtual environments; computer-supported collaborative work; user representation; human-computer interaction
• Online Access: http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-3707; http://nbn-resolving.de/urn:isbn:91-7283-688-1

This thesis is concerned with shared virtual environmentsfor collaborative work. An important aspect of shared virtualenvironments is the avatar, the representation of the user inthe virtual world. The proper design of the avatar has been thesubject of considerable research, aimed at allowing the avatarsto express as much as possible of human non-verbalcommunication and, as it were, tie the user closer to thevirtual world. I will go through the historical development of sharedvirtual environments and how the design principles for avatarshave followed the available technology over time. I describeearlier research on extending avatars and environments in orderto better support collaboration in virtual spaces. I will thendescribe a user study where pairs of subjects cooperated on aconstruction task, and the implications for design ofcollaborative applications in VEs that can be drawn from thisstudy. In particular I show how the subjects used the availableresources in the environment to negotiate a sharedunderstanding of the environment and the task. Some of thesubjects had no visible avatars, but still solved the task byusing the environment itself to orient themselves and drawattention to important features of the environment. Following this, I and co-workers have designed virtualenvironments which have had no explicit avatars, nor have usedtraditional methods for navigation in 3D space, but ratherrelied on task-oriented features of the space, such asagglomerations of other users or interesting objects in orderto present a relevant view of the environment. A view positionmay be shared by several users, or be“unoccupied”,merely representing a potential site for interaction. Based on these experiences, I make the claim that atraditional anthropomorphic avatar is neither necessary norsuffcient for successful collaboration in virtual spaces, butthe design of navigation and user representation is contingenton the specific application, some reasonable applications notutilising a user representation at all. === <p>QC 20161027</p>