Traditional measures of alertness and reaction time do not capture the dynamic integration of perception and action, where some form of perceptual judgement (i.e. judgement of an object's spatial parameters in relation to another object) is required to examine this relationship.
PURPOSE:To assess the reliability, within-subject variability, and systematic bias associated with a novel measure of reaction time, the perception-action coupling task (PACT).
PARTICIPANTS: 9 males and 7 females (Age (yrs) = 27.8 ± 3.6) participated in four identical testing sessions. METHODS: The PACT, performed on a touch-screen tablet, requires participants to make judgements on whether a virtual ball affords "posting" in a virtual aperture, both of varying sizes. There are 8 possible ball to aperture size ratios., and aA full cycle of the PACT lasts 5 minutes and consists of 12 randomized presentations of each ratio. For each session, participants completed 9 cycles with a 15-minute break every three cycles. Reaction Time (RT), movement time (MT), and initiation time (IT) were calculated from response parameters.Accuracy of judgements (ACC) was calculated as the percentage of correct responses from each cycle. Systematic bias was determined by repeated-measures ANOVA, reliability was assessed with intra-class correlation coefficients, and within-subject variability was assessed with coefficients of variation. RESULTS: Initiation time was found to have the highest learning effect, requiring the elimination of three cycles to eliminate systematic bias (F = 2.417, p = .056). All other variables required one or less cycles (F = .408 -1.729, p = .167 -.910). All variables showed acceptable reliability (ICC = .775 -943) and within-subject variability () with only one cycle, after elimination of the first 3 cycles. CONCLUSIONS: With a 3 cycle (15-minute) familiarization period, the PACT was found to be stable and reliable in assessing RT, MT, IT and ACC during perceptual judgements.