Abstract
The field of human-computer interaction (HCI) is in a period of rapid development. Tangible and embodied interfaces, in which the fingers and/or body are used to manipulate virtual objects, are gaining in popularity in many computing contexts. Because these new interfaces offer a more direct sensori-motor mapping than conventional mouse- or keyboard-based interfaces, it is becoming increasingly important for HCI designers to consider fundamental aspects of the interaction between perceptual, cognitive, and motor processes. With this in mind, we have employed empirical and theoretical advances in ideomotor theory to shape the development of a new embodied interface. The interface was designed to more effectively map the whole body movements of a user onto the movements of a virtual representation of the user (avatar). It was thought that translating the user's movements to the avatar would facilitate performance and allow the user to more easily relate to the avatar and the virtual world. The present paper reports a testing of the new interface in which the user made an avatar complete a series of aiming movements (using both limbs) to virtual targets that appeared in random locations using the embodied, joystick, or keyboard interface. The embodied interface provided clear accuracy and response time advantages over keyboard and joystick interfaces. These data suggest that an embodied mapping of actions to virtual worlds is superior to mappings based on keyboards or joysticks.Acknowledgments: This research was supported by grants from NSERC and NSF.