Inter-dependence between mathematically independent variability components in human multi-finger force control.

In human movement control, inherent error or uncertainty in the controller, motor system, and sensory system causes variability in motor outcomes. Previous studies have suggested different methods to quantify and examine independent components of the motor variability of motor outputs in a redundant motor system. While these motor variability components are mathematically independent, it is unknown if these components are behaviorally independent among subjects. The aim of this study was to investigate inter-relations between mathematically independent motor variability components in multi-finger force control among subjects. Nineteen healthy subjects performed two tasks, producing a constant force and a sinusoidal force for 12 s over 12 trials. We used the hierarchical variability decomposition (HVD) model to quantify mathematically independent variability components, i.e., online task-relevant variance (onTRV), online task-irrelevant variance (onTIV), offline task-relevant variance (offTRV), and offline task-irrelevant variance (offTIV) for each subject. The correlation analysis performed among all subjects showed that online and offline motor variability components were positively correlated, while task-relevant and -irrelevant variability components were not. The results indicate that these mathematically independent motor variability components existing in multi-finger force space are dependently scaled among subjects, more noticeably between online and offline controls, suggesting a non-independent behavioral relationship between these mathematically independent components among subjects.