Abstract
How do people circumvent a barrier en route to a goal? Baxter & Warren (2017) found that participants preferred a waypoint at the end of the barrier that was a closer distance (d) to their starting point and/or a smaller deviation angle (β) from their initial heading. Here we dissociate the deviation angle (β), which would reduce turning, from the visual angle (α) between the waypoint and the goal, which would yield a shorter path. A participant (N=10) walked in a virtual environment while wearing an Oculus HMD. On each trial, they started walking towards an orientation pole, setting their initial heading. After progressing 2m the disappearance of the orientation pole and simultanious appearance of a 3m barrier in front of a goal pole was triggered. In half the trials the orientation and goal pole positions were the same, and in the other half they were positioned at the opposite ends of the barrier, putting β and α in conflict). The barrier's orientation (±45°, ±75° from sagittal) and lateral offset (± 0.4m from start-orientation line) were varied. The left/right response was analyzed using a mixed-effects logistic regression model. All fixed-effects variables (d, β, and α) significantly predict whether the average participant avoids to the right (p< .001). When the right barrier end is 1m closer, the odds of passing to the right are 3.41 higher than the left; when β and/or α is 1° smaller to the right, the odds are also higher (1.15, 1.38; respectively). Scaled to standardized units, however, α has a greater odds ratio (28.44) than β or d (6.01 and 4.38). The results indicate that the distance, heading deviation, and goal angle all influence waypoint selection in barrier avoidance. The goal angle, which often coincides with the heading deviation, is the strongest predictor, possibly because it indicates a shorter path to the goal.
Meeting abstract presented at VSS 2018