Distinguishing the noise and attractor strength of coordinated limb movements using recurrence analysis

Michael J. Richardson, Richard C. Schmidt, Bruce A. Kay

Research output: Contribution to journalArticlepeer-review

63 Citations (Scopus)


The variability of coupled rhythmic limb movements is assumed to be a consequence of the strength of a movement’s attractor dynamic and a constant stochastic noise process that continuously perturbs the movement system away from this dynamic. Recently, it has been suggested that the nonlinear technique of recurrence analysis can be used to index the effects of noise and attractor strength on movement variability. To test this, three experiments were conducted in which the attractor strength of bimanual wrist-pendulum movements (using coordination mode, movement frequency and detuning), as well as the magnitude of stochastic perturbations affecting the variability of these movements (using a temporally fluctuating visual metronome) was manipulated. The results of these experiments demonstrate that recurrence analysis can index parametric changes in the attractor strength of coupled rhythmic limb movements and the magnitude of metronome induced stochastic perturbations independently. The results of Experiments 1 and 2 also support the claim that differences between the variability of inphase and antiphase coordination, and between slow and fast movement frequencies are due to differences in attractor strength. In contrast to the standard assumption that the noise that characterizes interlimb coordination remains constant for different magnitudes of detuning (Δ ω) the results of Experiment 3 suggest that the magnitude of noise increases with increases in |Δ ω|.
Original languageEnglish
Pages (from-to)59-78
Number of pages20
JournalBiological Cybernetics
Issue number1
Publication statusPublished - 2007
Externally publishedYes


Dive into the research topics of 'Distinguishing the noise and attractor strength of coordinated limb movements using recurrence analysis'. Together they form a unique fingerprint.

Cite this