M.A. Thesis: Psychological Science (Visual Science), CSU Northridge, 2024
Description
When augmented reality overlays a cue to help you find something in the real world, what kind of cue actually helps—and does it come at a cost to your attention or comfort? My thesis tackled that question by bringing a visual-science lens to AR design, an area where new technology often gets deployed before we understand the attentional mechanisms it's leaning on. Using a Microsoft HoloLens 2 and a real, physical search environment (107 custom 3D-printed letters arranged across shelves), I ran a proof-of-concept study comparing three conditions: a 3D arrow, a motion spline (a moving ball that traces a path to the target), and a no-cue control. Participants completed conjunction visual-search trials across three difficulty levels while I measured search time, perceived workload (NASA-TLX), cybersickness, and ADHD symptoms as factors.
When augmented reality overlays a cue to help you find something in the real world, what kind of cue actually helps—and does it come at a cost to your attention or comfort? My thesis tackled that question by bringing a visual-science lens to AR design, an area where new technology often gets deployed before we understand the attentional mechanisms it's leaning on. Using a Microsoft HoloLens 2 and a real, physical search environment (107 custom 3D-printed letters arranged across shelves), I ran a proof-of-concept study comparing three conditions: a 3D arrow, a motion spline (a moving ball that traces a path to the target), and a no-cue control. Participants completed conjunction visual-search trials across three difficulty levels while I measured search time, perceived workload (NASA-TLX), cybersickness, and ADHD symptoms as factors.
What I found:
- Among the AR cues, the motion spline significantly outperformed the arrow on search time (p = .043) consistent with the idea that motion functions as a low-level feature that automatically guides attention.
- Neither AR cue significantly outperformed the no-cue control. While the motion spline was numerically the fastest condition, the difference from searching with no guidance at all was not statistically significant
- Difficulty mattered, but not how I expected. Lower-difficulty trials were actually slower than medium and high ones, pointing to a likely priming effect from the shared spatial environment, a finding that raises interesting questions about how trials influence one another in naturalistic AR setups.
- Static vs. moving cues didn't differ in cybersickness, suggesting the motion spline guided attention without an obvious comfort penalty.
Findings
More than the individual results, the study established the feasibility of testing AR guiding cues in an ecologically valid, real-world setting grounded in attention research—laying methodological groundwork for systematically identifying which cues best support human search behavior. The work bridges basic research on top-down and bottom-up attention with applied AR interface design, with implications for fields from medical diagnostics to industrial quality control.
More than the individual results, the study established the feasibility of testing AR guiding cues in an ecologically valid, real-world setting grounded in attention research—laying methodological groundwork for systematically identifying which cues best support human search behavior. The work bridges basic research on top-down and bottom-up attention with applied AR interface design, with implications for fields from medical diagnostics to industrial quality control.
Supported in part by NIH GM137863.