We present empirical results from a study conducted with eleven users with upper-body motor impairments who imagined input devices and corresponding gestures to operate them for performing common tasks in interactive systems. We report a strong preference for embodying devices (80%), primarily through the hands, rather than holding them, and identify ten device archetypes, among which smartphones (36.4%) and remote controls (27.3%) were most prevalent. We also observed a diversity of gestures to operate imagined devices involving both unimanual and bimanual input with little consensus (.069) across participants, which we analyzed in relation to their self-reported motor impairments. Based on these findings, we propose design recommendations for accessible interactions involving imagined input devices, structured through the lenses of ability-based and ability-mediating design. We also outline future work opportunities for imagination-powered accessible computing, in which users' imagination plays a central role.
ACM CHI Conference on Human Factors in Computing Systems