Abstract
The largest spatial displacement of dots in a two-frame random-dot kinematogram yielding good direction discrimination performance (i.e., Dmax) is reduced in older observers (Roudaia et al., J. Vis., 2010). Dmax depends on dot density, which suggests that Dmax is constrained by the number of false targets in the display (Eagle & Rogers, Vis. Res., 1997). As such, the reduction in Dmax with aging may be caused by an age-related decline in the efficiency of solving the correspondence problem. The current study investigates that hypothesis. Stimuli consisted of white dots (0.04 × 0.04 deg) randomly distributed in a small, medium, or large square patch (height: 6.4, 12.7, or 25.4 deg) on a black background. The number of dots was varied to yield average density levels ranging from 0.025% to 5%. On each trial, two random-dot patterns were presented for 100 ms each, separated by a blank ISI lasting 40 ms. The second pattern was identical to the first pattern, but was shifted within a display window to the right or to the left by a displacement ranging from 0.03 to 5.6 deg. Direction discrimination accuracy of older (mean age: 74 years) and younger (mean age: 24 years) subjects was measured in 105 conditions blocked by patch size. Performance in both groups was best at medium displacements, and declined at smaller and larger displacements. Decreasing density did not affect performance at short displacements, but improved performance at large displacements. Importantly, this improvement was significantly greater in older than younger subjects, such that the effect of age at large displacements greatly reduced with decreasing density. These results support the hypothesis that age-related declines in performance at large displacements are caused by a reduced ability to solve the correspondence problem in motion.
CIHR and Canada Research Chair program grants to A.B.S. and P.J.B.