Effects Of Action Boundary Proximity On Perceptual-motor Judgements

Abstract

Designed as a more ecological measure of reaction times, the Perception-action Coupling Task (PACT) has shown good reliability and within-subject stability, however, a 5- to 10-minute testing period was required. Perceptual-motor judgements are known to be affected by proximity of the stimulus to the tasks action boundary. PURPOSE: To determine the effects of action boundary proximity on PACT performance, and whether redundant levels of stimuli, eliciting similar responses, can be eliminated to shorten the PACT while maintaining reliability, stability, and the intended behavioral responses. METHODS: 9 males and 7 females (Age(yrs) = 27.8 ± 3.6) completed 4 testing sessions, separated by at least 6 days. For each, participants performed 3 familiarization cycles and 6 testing cycles of the PACT. The PACT requires participants to make judgements on stimuli, in the form of whether a series of virtual balls presented on a tablet afford “posting” (can fit) through a series of virtual apertures. Eight ratios of aperture to ball size (AR) are presented, ranging from 0.2 to 1.8, with each ratio appearing 12 times per cycle. Response (time from stimulus presentation to first movement), Movement (time from first movement to initiation of ball movement), and Initiation (time from initiation to completion of ball movement) times were calculated. Accuracy was calculated as the percentage of correct judgements. All variables were averaged and plotted by AR. Based on visual inspection of plots, redundant stimuli were eliminated. ICCs (3,1) and coefficients of variation (CV) were calculated for each response variable in an iterative manner to determine the minimum number of testing cycles necessary. RESULTS: All variables followed an expected, quadratic trajectory, with performance the lowest near the action boundary (AR = .8 – 1.2). Ratios of .2 and 1.8 were found to elicit redundant responses to .4 and 1.6, and were eliminated. The shortened PACT showed good reliability with 1 (ICCs = .71 - .94) and 2 cycles (ICCs = .81 - .98) of testing. Further, CVs were consistent with the full PACT with 1 (CVs = .8 – 20.0 %) and 2 cycles (CVs = .6 – 14.7 %) of testing. CONCLUSION: The main findings support the use of the shortened PACT, limiting the testing period to 3.5 – 8.5 minutes depending on the desired level of reliability and within-subject stability.