Why and how to act through multiple avatars in virtual reality

In virtual reality (VR), users typically control one virtual body — their avatar. Previous works enable users to act through multiple avatars simultaneously. These works, though, lack a systematic account of why users want to interact with multiple avatars and do not explain how users can manipulate and generate avatars. To address this, we run six workshops with 12 VR experts and develop a design space that captures four fundamental dimensions for acting through multiple avatars in VR: Appearance , Context , User-Avatar Mapping , and Motion Control . Researchers can use the design space to generate novel interaction opportunities involving multiple avatars or analyze existing work. We then run a usability study with 17 participants to understand the practicalities of an interface that integrates parts of the design space, which reveals conceptual and technical challenges that we address through design recommendations. • Our design space explains how users can interact through multiple avatars in VR. • Users want multiple avatars to stand in for themselves, other persons, or things. • Users use multiple avatars to perceive and act in multiple contexts simultaneously. • Challenges with multiple avatars include clutter, syncing, and switching between them.