Abstract
Motion direction learning transfers substantially more to a new location when trained with a staircase method (Wang et al., JoV, 2014) than with a single-level method of constant stimuli (MCS) (Ball & Sekuler, 1982). Learning also transfers to an opposite direction with a TPE (training-plus-exposure) procedure when staircases are used (Zhang & Yang, 2014), but the transfer fails with the single-level MCS (Liang et al., JoV, 2015). We suspect that training with a single-level MCS, which uses a pair of fixed motion stimuli with a single-level direction difference, may allow observers to pick up extra local cues for direction judgments. These cues may not be readily available at a new location or direction. We first replicated strong location specificity in peripheral motion direction learning with the single-level MCS using a pair of moving dot patterns (Δdir = 10°). Learning transferred little to contralateral and diagonal retinal quadrants with the transfer index (TI) = 0.26. To disturb the potential local cues, we slightly jittered the directions of the stimulus pair every presentation within a range of ±4° while keeping Δdir = 10°. The direction jitters significantly increased learning transfer (TI = 0.64). We also replicated the null transfer results with TPE training in foveal motion direction learning (Δdir = 3°) with the single-level MCS (Liang et al., 2015). Learning failed to transfer to an orthogonal direction (TI = 0.25). Again if we jittered the stimulus direction within ±2° while keeping Δdir = 3°, significantly more transfer to an orthogonal direction was enabled by TPE training (TI = 0.82). These results are consistent with the local-cue learning hypothesis regarding motion direction learning with the single-level MCS, which is responsible for part of the specificity in motion direction learning literature. They also demonstrate the importance of using appropriate psychophysical methods in training to reduce specificity in perceptual learning.
Meeting abstract presented at VSS 2016