Factorization of Force and Timing in Sensorimotor Performance: Long-Range Correlation Properties of Two Different Task Goals.

Abstract

Long-range correlations are often manifested in the form of 1/fβ noise in a series of repeated measurements of the same neural or behavioral variable. Recent work has demonstrated that the magnitude and nature of these long-range correlations reliably capture individual differences and variation in task performance. In sensorimotor timing experiments, task characteristics such as tapping or circle drawing affect these long-range correlations during the production of isochronous time intervals. Such correlations are highly reproducible across multiple trials for the same task but do not correlate between tasks. In the present experiment, we investigate whether two behavioral variables that are simultaneously controlled by the same participant in a given experimental condition can show such differentially organized fluctuations. In order to answer this question, 13 participants were asked to produce repetitive movements with their right index finger at a specified time interval (500ms) and a specified force (8N) in the absence of an auditory metronome and visual feedback of force levels following a synchronization-continuation paradigm. Although participants showed high levels of consistency in the long-range correlations for each task component separately over multiple trials/observations, the long-range fluctuations for force and timing were found to show no correlations with each other for each participant. Cross recurrence quantification analyses (CRQA) revealed that there was limited shared structure between the timing and force time series data. Taken together, these results suggest that complex systems can organize multiple processes in a relatively independent manner while maintaining a high degree of reliability within one task parameter.