Figure 1.

Example of data processing flow at 80 bpm in the Flexion-on-the-beat and Face-to-face condition. To record knee angle displacement, three reflective markers were attached to the right centre of hip, knee, and ankle joints (and are therefore visible on Participant A but not Participant B). Time series data of knee angle displacement were low-pass filtered (cut-off frequency = 7 Hz). The continuous phase of each participant’s movement was computed using the Hilbert Transform. Both the real and imaginary parts of displacement obtained by the Hilbert Transform were normalized and plotted on a phase plane (with the real part of displacement on the X-axis and the imaginary part of displacement on the Y-axis) to compute the continuous phase of each participant’s movement. The mean and standard deviation (SD) of interpersonal phase angles were then calculated within each trial using circular statistics. Audio signals of the metronome recorded in synchrony with the motion capture data were analysed to determine the beat onset times. Individual beat onset times were then superposed on each phase plane trajectory to compute the mean phase angle of beat time and its SD using circular statistics. We defined maximum flexion as 0° and maximum extension as 180° because the most stable coordination pattern is conventionally defined as 0° or in-phase, and previous studies have reported that flexion-on-the-beat pattern is the most stable pattern. Colour of interpersonal phase angle and beat markers changes from red to yellow with time-course.