Abstract
We investigated specificity of learning as a function of motion direction and speed.
In 2AFC, five participants trained to discriminate which of the two radial inward motion stimuli was faster (1.93 versus 2.17 deg/sec, viewing distance: 114 cm). Five other participants trained with outward motion stimuli. Pre- and post-training, four psychometric functions of speed discrimination were measured per trainee (pedestal speed: 1.93, 2.17 deg/sec) × (direction: inward, outward). All 10 trainees improved their speed discrimination sensitivity after eight days of training. Interestingly, this learning transferred completely between inward and outward radial motions. This transfer was training specific, since little improvement was found from five control participants who skipped the training but the psychometric functions were measured similarly.
Experiment 2, conducted with 12 fresh trainees, was similar except that specificity to pedestal speed, rather than direction, was investigated. Specifically, pre- and post-training, inward radial motion stimuli were used to measure psychometric functions for the inward trainees at the viewing distance of 114 cm, and at 57 cm and 228 cm with the same stimuli. In effect, speed discrimination was measured when all stimulus speeds were doubled or halved. Likewise, outward stimuli were used for the outward trainees. All 12 trainees improved their discrimination sensitivity after eight days of training at viewing distance 114 cm. This learning partially transferred to viewing distances of 57 cm and 228 cm. Namely, discrimination was improved at both distances, but not as much as at the trained viewing distance (the interaction was statistically significant, p = 0.03).
In summary, we found that speed discrimination learning of radial motion transferred completely between inward and outward directions. However, this learning was partially specific to pedestal speed, when the viewing distance was manipulated while the stimulus remained unchanged.