Fig 1.
Experimental setup for the immobile condition.
Fig 2.
Visualization of the global coordinative structure (in-phase) (N=18).
This shows the global coordinative structure (N=18). Mean first peak values of the cross-correlation and phase difference between the time series of anterior-posterior motion of the pendulum and each body are shown as lines projected onto the topologically arranged body parts (gray circles). The degree of cross-correlation is expressed by the thickness of the line and the phase difference by color gradients. RP represents the R-pendulum, LP L-pendulum, RW R-wrist, LW L-wrist, RE R-elbow, LE L-elbow, RS R-shoulder, and LS L-shoulder.
Fig 3.
Visualization of the global coordinative structure (anti-phase) (N=18).
Fig 3 represent the same information as Fig 2, except that the data are for anti-phase.
Fig 4.
Motion similarity between the pendulum and body parts (in-phase) (N=18).
This shows the mean and standard deviation (error bars) of the first peak value of cross-correlation (N=18) between the time series of the anterior-posterior motion of the pendulum and each body part by mobility (immobile vs. mobile) in gray and black dots and by frequency in the horizontal axis. Shown on the upper right of each panel is the result of the two-way analysis of variance: M is the main effect of the mobility factor, F is the main effect of the frequency factor, and M × F is the interaction of the mobility and frequency factors (p-value set at 0.05).
Table 1.
Results of two-way ANOVA for first peak value of cross-correlation (in-phase condition).
Fig 5.
Motion similarity between the pendulum and body parts (anti-phase) (N=18).
Fig 5 represent the same information as Fig 4, except that the data are for anti-phase.
Table 2.
Results of two-way ANOVA for first peak value of cross-correlation (anti-phase condition).
Fig 6.
Phase difference between the pendulum and body parts (in-phase) (N=18).
This shows the relative phase between the anterior-posterior motion of the pendulum and each body part (N=18) by mobility (immobile vs. mobile) in color temperature (warm vs. cold) and frequency as color gradient. All data of the 18 participants are shown in the outer circle. Two things are shown on the inner circle: the mean vector (straight line) of the data of all participants, of which the length corresponds to the angle concentration κ, and the probability density function (von Mises distribution) calculated based on the mean vector and concentration κ. Results of the two-way Harrison-Kanji test are shown on the upper right of each panel.
Table 3.
Results of the two-way Harrison-Kanji test for phase differences (in-phase condition).
Fig 7.
Phase difference between the pendulum and body parts (anti-phase) (N=18).
Fig 7 represent the same information as Fig 6, except that the data are for anti-phase.
Table 4.
Results of the two-way Harrison-Kanji test for phase differences (anti-phase condition).
Fig 8.
Results of the phase difference between the left-right pendulum.
The phase shift (panels (a) and (b)), absolute value of the phase shift (panels (c) and (d)), and SDϕ (Fig 8e and 8f) are shown for each phase mode (in-phase and anti-phase). Line plots with markers indicate the mean values for 18 participants, and error bars indicate 95% confidence intervals. Asterisks represent pairs that were significant in the Watson-Williams test (performed as a post-hoc test when the two-factor Harrison-Kanji test was significant).
Fig 9.
Schematic diagram of possible forces at work and compensatory movements during the present experiment by phase mode.
Solid lines indicate the direction of forces caused by swinging the pendulums, which may affect the body, and dashed lines indicate the direction of counter forces caused by possible compensatory movements. This study’s results appear to correspond to the compensatory movements shown in this diagram.