The indispensability of precollision kinematics in the visual perception of relative mass

Abstract

The visible kinematic pattern that occurs when objects collide contains information about the relative mass of the objects. Recently, Gilden and Proffitt (1989) have claimed that perceivers are limited to the use of simultaneous elemental cues and that therefore mass discrimination judgments are independent of whether the precollision epoch is visible or not. The present paper argues that their experimental results are irrelevant because crucial parameters were held constant, and that false conclusions were drawn because their data were not treated for what they are: threshold measurements obtained with the method of constant stimuli. Three experiments are reported, showing that a high level of performance is attainable only when the whole event is visible, and thus that the effective information is extended over time. In addition, modified PROBIT analysis was used to determine what information observers use when the precollision epoch is occluded. Analytically complex invariants, based on vector components in the abstract collision-axis system, proved to have the best fit to the data. Thus the elementaristic premises in Gilden and Proffitt's cue-heuristic model for relative mass perception are defeated.

Heuristic judgment of mass ratio in two-body collisions

“…In this situation, the basis of judgment will be predominately determined by an analysis ofspeeds. Runeson and Vedeler (1993) in their Experiment 1. Length of arrow represents speed, and all collisions are drawn to the same scale.…”

“…The displacement of the point of subject equality in mass ratio toward large values of rnB/rnA is apparently due to the displacement of speed equality to rnB/rnA> 1.8. IfB is moving faster than A, subjects are Generalization of the Initial Conditions: Experiment 2 Experiment 2 was intended by Runeson and Vedeler (1993) to provide definitive support for their conjecture that observing the precollision kinematics is required for optimal judgment of mass ratio. As their method of occlusion in this experiment was to totally erase the precollision history of both objects, it is not surprising that they found that people were sometimes extremely inaccurate in their judgments in the occlusion conditions.…”

“…Gilden and Proffitt (1989) pursued this work and developed separate angle and speed heuristics that could entirely account for the judgments of mass ratio observed in their studies as well as in those reported by Todd and Warren (1982). Runeson and Vedeler (1993) have reopened the issue with data from three experiments which they argued were inconsistent with a heuristic framework. The issue, as set forward by Runeson and Vedeler, is that Gilden and Proffitt (1989) reached a number of erroneous conclusions because of errors in their treatment of data and because they held certain crucial parameters constant in their experiments.…”

“…

The logic ofjudging relative mass from a two-body collision is developedfrom data presented by Runeson and Vedeler (1993

…”

Perspective distortion of trajectory forms and perceptual constancy in visual event identification

“…determine specific kinematic (i.e., motion) patterns and that observers can determine the nature of an event from the resulting patterns of motion. The specific nature of the patterns that allowed recognition of the human motions was not described, although a potential source of information about relative mass in collisions was described and investigated (Runeson, /1983Runeson & Vedeler, 1993), with some ensuing controversy (Gilden & Proffitt, 1989, 1994Runeson, 1995). 1 Following an investigation of trajectories as information about lifted weight (Bingham, 1987b), proposed that trajectory forms are the information used to identify events.…”