Three examples of submovement composition of head movement during eye-head combined gaze shifts.

<p>For stationary gaze target (<b>A–B</b>), task-associated head movement, indicated as arrowheads ▴ and ▾, was included up to the end of gaze target display (A) or the onset of correction saccade when correction gaze shifts occurred (B). For flashed gaze target (<b>C</b>), task-associated head movement was included up to 200 ms following gaze end. Note the goodness of fit in <b>A</b> improved by 18.9% from a single submovement (#SM = 1; 19.1% error) to 3 overlapping submovements (#SM = 3; 0.2% error). The goodness of fit in <b>B</b> improved by 6.1% from a single submovement (#SM = 1; 6.3% error) to 2 overlapping submovements (#SM = 2; 0.2% error). Same format as in <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0047565#pone-0047565-g002" target="_blank">Fig. 2</a>.