Maarten L. Wijnants, Radboud University Nijmegen, the Netherlands
Anna M. T. Bosman, Radboud University Nijmegen, the Netherlands
Fred Hasselman, Radboud University Nijmegen, the Netherlands
Ralf F. A. Cox, Radboud University Nijmegen, the Netherlands
Guy C. Van Orden, University of Cincinnati, OH

Abstract: When people perform repeated goal-directed movements, consecutive movement durations inevitably vary over trials, in poor as well as in skilled performances. The well-established paradigm of precision-aiming is taken as a methodological framework here. Evidence is provided that movement variability in closed tasks is not a random phenomenon, but rather shows a coherent temporal structure, referred to as 1/f scaling. The scaling relation appears more clearly as participants become trained in a highly constrained motor task. Also Recurrence Quantification Analysis (RQA) and Sample Entropy (SampEn) as analytic tools show that variation of movement times becomes less random and more patterned with motor learning. This suggests that motor learning can be regarded as an emergent, dynamical fusing of collaborating subsystems into a lower-dimensional organization. These results support the idea that 1/f scaling is ubiquitous throughout the cognitive system, and suggest that it plays a fundamental role in the coordination of cognitive as well as motor function.

Keywords: fractal scaling relations, nonlinear dynamics, motor coordination, degrees of freedom, task complexity