Abstract
Twenty older (mean age was 70.6 years) and younger observers (mean age was 23.1 years) participated in a study designed to investigate stereoscopic shape discrimination. The stereoscopic surfaces (approximately 24.6 minutes peak-to-trough binocular disparity) to be discriminated all had sinusoidal depth modulations, where the resulting peaks and troughs formed concentric circles, were radially-oriented, or were arranged like an “egg-crate” (i.e., binocular disparity = sin(x) * sin(y)). The visibility of the stereoscopic surfaces was degraded to varying degrees by embedding the surfaces in volumetric “noise” (e.g., 50 percent of the stereoscopic points were located on a smooth 3-D surface, whereas the remaining 50 percent of the points occupied random locations in a cubical volume surrounding the depicted surface). While the discriminability of all observers was adversely affected by the presence and magnitude of the volumetric noise, the older observers' performance suffered more (e.g., the younger observers' performance was 30 percent higher than that of the older observers when 40 percent of the stereoscopic points defined a smooth surface and 60 percent of the points comprised noise). The performance of all observers was facilitated when the random-dot stereograms were dynamic (updated at 35 Hz) instead of static. However, the improvement in performance for dynamic stereograms was larger for the younger observers and smaller for the older observers. The results demonstrate that while older adults generally possess good stereopsis, their shape discrimination performance is not as robust or accurate as that of younger observers, especially when stereoscopic surfaces are degraded by volumetric noise.