Selective attention modulates motion speed learning by suppressing visible irrelevant stimuli

Abstract

Selective attention modulates perceptual learning in motion speed, but previous studies were unable to answer whether selective attention modulates learning by enhancing the attended signal, suppressing the ignored signal, or a combination of the two. We investigated this question by separating the attended/ignored signals in two disparity-defined depths. The learning stimuli contain speed change in one specific direction (target direction) at one specific depth plane (target depth), with another depth plane (non-target depth) containing either nothing (control group), or distractor dots moving in random direction random-direction group) or orthogonal to target direction (orthogonal-direction group). Observers' 80%-correct motion coherence detection thresholds at two depth planes were measured before and after six hourly sessions of speed discrimination. After learning, observers' motion sensitivity increased by an average 23% at target depth plane at all tested directions in all three conditions, as well as at the non-target depth plane when no distractors were present (control condition) (21%). At conditions where clearly visible moving distractors in either non-specific (random) or specific (orthogonal) direction, there is a clear disadvantage at the distractor direction(s) (6%, 0%). Our results strongly suggest that selective attention modulates motion speed learning largely by suppressing visible irrelevant stimuli.