(a) Absolute alignment error as a function of initial dial orientation in the tactile (black) and visual (gray) pointer-alignment task, which involved the infinite edge and raised dot, respectively. Lines indicate means across participants’ medians (N = 10). Error bars indicate the standard error of the mean. The greater average alignment error in the tactile than in the visual condition (F1,9 = 12.9, p=0.006) mainly stemmed from smaller errors in the visual trials with initial dial orientations closest to the 0˚ target. That is, there was a significant interaction effect of the initial orientation and sensory condition on the alignment error (F5,45 = 6.5; p=0.0001) besides a main effect of initial orientation (F5,45 = 10.6; p<10−6). A post-hoc examination revealed that the initial orientation did not significantly influence the performance in the tactile condition but that it did significantly influence performance in the visual condition. It also revealed that alignment error was significantly different only for the −20, −10 and 10° initial orientations between the tactile and visual conditions (p<0.01 for all comparisons). (b–e) Time from initial touch to rotation onset, time from initial touch until contact force reached its plateau-like stage, contact force at rotation onset, and plateau contact force during the tactile and visual pointer-alignment tasks. Height of black and white bars indicates mean values across participants’ medians in the tactile and visual condition, respectively, and gray lines indicate median values for each participant and condition. (f) Rotation duration as a function of initial dial orientation for both visual and tactile conditions. Lines indicate means across participants’ medians and error bars indicate the standard error of the mean. (g–h) Frequency distribution of number of movement components and mean number of movement components as a function of initial orientation. Lines indicate means across participants’ means and error bars indicate the standard error of the group mean.