Elucidating the Effects of Augmented Reality Head-Worn Display Factors on Performance, Trust, and Behavior

Augmented reality (AR) display characteristics have the potential to either enhance or impair users' spatial abilities and performance. While previous work included studies of spatial performance with various display factors, evidence for objective performance differences is limited due to compensatory behaviors employed by users and overall behavioral differences. In general, it is challenging to document the effects of display factors on task performance as they depend on users' task behaviors, which in turn depend on users' reliance and trust in the technology, which are also affected by the display factors. In this paper, we present two within-subjects experiments (each N=20) in which we aim to elucidate some of the interrelations between two AR display factors (field of view and visual contrast) with objective task performance and subjective assessments of reliance and trust, while controlling for different behaviors. Participants performed a ${360}^{\circ }$ search-and-selection task in a unique hybrid setup, in which we simulated a controlled task environment by having participants stand inside an immersive CAVE-like space while at the same time wearing a head-worn display that overlaid AR tags over the simulated environment. Specifically, we evaluated three fields of view (${43}^{\circ }$, ${93}^{\circ }$, and ${143}^{\circ }$) and three visual contrasts (0.05, 0.25, and 0.5). We controlled for four different behaviors: AR-Only (only relying on AR), AR-First (prioritizing AR over real world), Real-First (prioritizing real world over AR), and Real-Only (only relying on real world). By controlling for these behaviors, we were able to show objective and subjective benefits of larger fields of view and visual contrast. We illustrate how the controlled behaviors relate to users' subjective reliance and trust in an AR system, and why it is important for researchers and practitioners to understand these subjective and behavioral aspects.