Abstract:

Axis-aligned motion (AAM) bias is a tendency for observers to assume symmetric moving objects maintain axis-trajectory alignment and bias their judgments of trajectory toward the axis when this alignment is broken. This doctoral dissertation explored observers' final destination judgments of symmetric moving objects across three contexts, on a computer screen (Exps. 1A-C), in an immersive mediated environment (Exp. 2), and in the real world (Exp. 3). Consistent with prior research, observers in the first and second experiments exhibited AAM bias; stimulus elongation/aspect ratio (Exp. 1) and judgment method (Exp. 2) were shown to impact the magnitude of AAM bias. In contrast, judgments of objects' destinations in Experiment 3 were unaffected by the shapes' axial deviation from trajectory. A Bayesian account of the observed context effects is proposed, and the impact of action (participant mobility) and mediation (domain of stimulus presentation) on the AAM bias are discussed. Axis-aligned motion bias and other perceptual regularities are then reviewed in the broader context of interface design for real-world application areas.

In conclusion, it is asserted that AAM bias is a prototypical example of a perceptual regularity that is a mixed blessing. When present, the bias is perceptually advantageous in perceptual judgment tasks of axis-trajectory aligned moving entities, like most living creatures, and disadvantageous in tasks demanding judgments of axis-trajectory misaligned moving entities, which are typically artifacts.