Figure 2. Alignment accuracy during tactile pointer-alignment trials.

(a) Distribution of the alignment error for the various edge lengths for all trials by all ten participants (108 trials/participant and edge length). Gray segments of the distributions refer to trials with rotation in the wrong direction. The accumulation of data at the ±38° represents trials in which the pointer reached the end of its movement range (see Materials and methods). (b) Absolute alignment error (deviation from the 0° target position) as a function of edge length based on median values for individual subjects (gray lines) and the corresponding data averaged across participants (black line; means ±1 sem). (c) Contours superimposed on a fingerprint – photographed through a flat glass plate – show the 8, 4 and 2 mm edges twice with an orientation difference that corresponds to the average alignment errors with these edges. For reference, superimpose of the fingertip is an array of black dots, laid out in a hexagonal array with a center-to center spacing of 1 mm, which approximately correspond to the spacing of receptive field centers of relevant tactile neurons if uniformly spaced across the fingertip. (d) Proportion of trials with rotations in the correct direction as function of edge length for each participant for all initial dial orientations pooled (gray lines) and the corresponding data averaged across participants (black line). Under the criterion that 75% correct responses define the threshold level, the vertical dashed lines indicates an estimation of the range across participants of threshold of edge length for correct rotation direction. (e) Cumulative frequency distribution of the pointer displacement referenced to movement in the direction of the target for trials performed by all participants with each edge length and initial dial orientation. The vertical dashed lines indicate the displacement required to reach the target position. The dashed segments of the distributions refer to trials with rotation in incorrect direction (i.e., negative displacement values) and are curtailed by the pointer reaching the end of its movement range. (f) Pointer displacement in the correct direction as a function of initial dial orientation and edge length shown as mean values across subjects (±1 sem; N = 10) based on participants’ medians. The dashed horizontal lines indicate the displacement required reaching the target for the 10, 20 and ±30° initial dial orientations. Data are pooled across the 10, 20 and ±30° orientations since there was no significant effect of sign of the orientation on the pointer displacements these initial orientations.