Discrimination of the direction of motion on the human hand: a psychophysical study of stimulation parameters.

1. In these experiments we assess the relative importance of the spatial and temporal properties of a moving tactile stimulus in determining the ability of humans to discriminate its direction of motion. Movement along the finger was simulated by applying a series of pulses to adjacent locations on the skin using the tactile array of an OPTACON stimulator. Simulated motion permitted us to vary independently the overall distance moved as well as the spacing, timing, and number of sequential stimuli. Different combinations of spatiotemporal parameters allowed us to further examine the relationship of apparent velocity of motion and sweep duration to behavioral performance. Discrimination accuracy was measured using signal detection techniques to calculate the discrimination parameter d' and PCmax, a bias-free measure of the percent correct identification of the direction of motion. 2. In experiments where the path length was constant, discriminability of the direction of motion increased as the spacing between successive pulses narrowed. Similarly, for a given interpulse spacing, the accuracy of discrimination increased linearly with distance, saturating at perfect performance. These apparent spatial effects on performance actually reflect the total number of stimuli presented to the skin rather than their proximity. Sweeps containing the same number of pulses are equally discriminable regardless of either their spacing or the total distance crossed on the skin. d' values obtained at 1.2-, 2.4-, and 4.8-mm spacings appear indistinguishable when plotted as a function of the total number of pulses in a sweep. 3. Experiments in which both the distance moved and the spacing between pulses was varied randomly confirmed that discrimination accuracy depends on the total number of pulses in a sweep rather than the spatial dimensions of the path traversed. Stimulation of only two points that mark the start and stop locations on the skin appears insufficient to enable subjects to discriminate correctly the direction of motion. Two-point stimulation elicits random performance whether the points lie 1.2 or 4.8 mm apart. Discriminability rises linearly to near-perfect performance when eight or more pulses are delivered sequentially. Extrapolation of the d' and PCmax curves suggests a mean threshold of approximately three points for 75% correct discrimination of the direction of motion across the skin. 4. The relationship of stimulus spacing to discriminability over a fixed path suggests that direction discrimination does not simply involve computation of the location of the start and stop points on the skin or their spatial disparity.(ABSTRACT TRUNCATED AT 400 WORDS)