Effects of Modality, Surface-Type and Surface-Smoothness on the Discrimination of Texture

Abstract

This study assessed the performance of participants in haptic and visual discrimination tasks involving the surfaces: abrasive paper and Japanese sharpening water stones. A recent study using abrasive paper surfaces showed that participants' visual discrimination of smooth stimuli was more accurate than the haptic discrimination of those same smooth stimuli (Bozoglu-Sinclair, 2001). These results differed from those of an earlier experiment, which examined visual and haptic performance in the discrimination of Japanese sharpening waterstones and found that participants in the haptic condition performed better than those in the visual condition with smooth stimuli (Heller, 1989, Experiment 2). In both previous experiments, the participants in the haptic and visual conditions performed equivalently with rough stimuli.

The current study employed a 2 (modality) x 2 (smoothness) x 2 (surface) design using Heller's procedure. Data for both accuracy and inspection time were analyzed. Of interest was whether Heller's results for accuracy of performance would be obtained with a different type of surface (i.e., silicon carbide). As in the experiment by Bozoglu-Sinclair, data for inspection time were also analyzed to determine whether, as in that study, visual inspection time would be shorter than haptic inspection time for examination of both rough and smooth stimuli. In addition, optical profilometry was employed to obtain roughness values in microns for each stimulus.

Results for accuracy differed from previous findings in that the haptic and visual conditions were equivalent for both rough and smooth stimuli. Inspection time results, however, showed that participants in the visual and rough conditions made judgments significantly faster than those in the haptic and smooth conditions respectively. Results from optical profilometry indicated that the manufacturer's scale of micron values led to different ranking of the stimuli for the silicon carbide condition and different roughness values for both silicon carbide and Japanese waterstones. Performance measures seemed to be more congruent with the optical profilometry values than with the original scale. These results highlight the desirability of using optical profilometry in evaluating stimulus materials.