Abstract
In studies by Patla, Vickers, and colleagues [1997, 2003, 2007] a gaze behavior termed ‘travel gaze’ was reported during walking tasks: the observer's point-of-regard was directed to the floor a fixed distance in front of the observer and traveled at the same rate as the observer as s/he walked forward. This travel gaze (vs. a sequence of saccades and fixations) is surprising because stationary observers are typically unable to make smooth pursuit eye movements without a moving target, and the reported behavior is similar to smooth pursuit in the absence of a target. It was also surprising that the observers reportedly maintained their gaze on the floor; the majority of fixations were reported to fall only 1–2 steps in front of the observer.
We hypothesized that the “travel gaze” and fixations on the floor may have been artifacts of the instrumentation and/or the experimental setup. In the 2003 experiments, observers walked 10-m paths marked with 1) evenly spaced footprints, 2) unevenly spaced footprints, or 3) no markings. While the original experiments reported randomizing the order of presentation, there was no analysis of an order effect.
We replicated the 2003 study using a Positive Science wearable eyetracker. Three observers performed the even-, uneven-, and no-footprint conditions. In addition, they performed a no-footprint condition under low light to eliminate visible texture that might inhibit travel gaze, and a condition in which a richly-textured, high-contrast pattern was present providing fixation targets not associated with any goals.
We did not observe travel gaze episodes; when footprint targets were visible observers made clear fixations with intervening saccades. Without visible footprints, path fixations fell dramatically as gaze shifted away from the path.
Our belief is that the original reports were due to temporal averaging of small saccades and order effects of the visible markers.