Predictive text entry in immersive environments

One of the classic problems with immersive environments is data entry: with a head-mounted display (HMD) the user can no longer see the keyboard. Although for many applications data entry is not a requirement, for some it is essential: communicating in collaborative environments, entering filename to which work can be saved, or accessing systems controls. Combining datagloves and a graphically represented keyboard with a predictive spelling paradigm, we describe an effective text entry technique for immersive environments; we explore the key issues when using such a technique, and report the results of preliminary usability testing.