Abstract
The limitations of our action capabilities impose a critical constraint on successful performance. When braking to avoid a collision, for example, the deceleration required to stop (the “ideal deceleration”) must be kept below the maximum possible deceleration. Ideal deceleration is optically specified, but maximum deceleration is a property of the observer's body or vehicle for which there is no information. Fajen (VSS 04; submitted) proposed that calibration to brake strength involves learning to detect information about ideal deceleration in intrinsic units of maximum deceleration. The focus of the present study is on the rate of recalibration to changes in brake strength. Participants viewed computer generated displays simulating approaches to a stop sign, and used a joystick as a brake to stop as closely as possible to the sign. When brake strength was manipulated as a randomly presented within-subjects variable in Experiment 1, there was evidence of partial recalibration on a trial-by-trial basis. In Experiment 2, trials were presented in ten blocks of 25. In Blocks #1 through #4, brake strength was either weak (Group 1) or strong (Group 2). Participants in Group 1 initiated braking at lower values of ideal deceleration (in extrinsic units of m/s2), but ideal deceleration at onset expressed in intrinsic units as a percentage of maximum deceleration was the same for both groups, suggesting that participants completely calibrated to the strength of the brake within the first block. In Blocks #5 through #10, brake strength was switched for both groups to a moderate level between weak and strong. Both groups recalibrated, but Group 1 took longer to completely recalibrate. The results suggest that observers are capable of rapidly recalibrating to changes in brake strength, but that the rate of recalibration depends on factors such as the direction of change in brake strength.