Spatial predictability modulates oculomotor deficits in low persistence displays

Poster Presentation 43.404: Monday, May 22, 2023, 8:30 am – 12:30 pm, Pavilion
Session: Eye Movements: Individual differences, novel measurement

Clara Kuper1,2,3 (), Xiuyun Wu1, T. Scott Murdison1; 1Realty Labs, Meta Platforms Inc., 2Department of Psychology, Humboldt-Universität zu Berlin, Germany, 3Berlin School of Mind and Brain, Humboldt-Universität zu Berlin, Germany

Head-mounted displays (HMDs) violate real-world spatio-temporal properties by presenting light in discrete frames at persistence levels <10% of the frame illumination period (a.k.a. duty cycle). This challenges the visual system, which extracts spatial information from the flow of light hitting the retina. For example, low-persistence HMDs cause exceedingly hypometric saccades (Goettker et al., JSID 2020). A follow-up study could not replicate this effect during reading (Wu and Murdison, Display Week 2022), suggesting a link between low-persistence stimuli and their spatial predictability. Here, we demonstrate that spatial cues modulated the oculomotor effects of persistence. We tested 25 participants in a novel behavioral paradigm. Participants made sequential saccades to targets presented on an augmented reality experimental display prototype with eye tracking. The task was performed at 4.5, 18, or 90% duty cycle at 90 Hz. Prior to each target, 0, 1, 2, or 3 cues indicating potential target positions were presented. A valid cue was always included if cues were presented. On ⅓ of the trials participants saw a limited-duration discrimination target (landolt c) and had to report its opening direction. We evaluated saccadic gain, saccadic latency, and discrimination accuracy at all persistence levels and cue numbers. We found reduced saccadic gain with lower persistence, replicating previous results. Lower persistence also increased saccadic latency and decreased discrimination accuracy. When the movement target was predictable, the effect of persistence on saccadic gain was reduced. Therefore, by introducing additional spatial references we may be able to re-instantiate normal saccadic targeting performance despite limited display persistence. Our results extend previous findings, suggesting that low persistence HMDs not only caused stronger undershoots, but also increased preparation time for saccades. This suggests that oculomotor effects of persistence are driven by movement planning during fixation, rather than by an intra-saccadic artifact.