Fast transfer of cross-modal interval training needs attentional engagement

Abstract

Here we examined whether improvement of the sub-second timing discrimination could be acquired through fast cross-modal training. To achieve this, we used a visual Ternus display (Ternus, 1926) as probe. In visual Ternus display, dependent on the inter-stimulus interval (ISI) between the two frames, participants could perceive either ‘element motion’ (short ISI) or ‘group motion’ (long ISI). In between the pre-test and post-test of discriminating the two percepts of visual Ternus (with ISI from 50 ms to 230 ms), observers received a task-relevant training, in which they were required to compare the which of the two auditory intervals or tactile intervals was longer (with auditory pitch or tactile intensity randomized). On the other hand, a second group of observers received a task-irrelevant training, in which they were required to discriminate which pair of the beeps was higher in pitch or which pair of the taps were stronger in intensity, although the time intervals in auditory/tactile pairs were tantamount to the intervals in the task-relevant situation. The results showed that after quick (about 15 mins) interval training, observers improved the sensitivity of categorizing the visual Ternus display, with a dominant percept of ‘group motion’, indicating subjective separation of the two frames was improved. However, this benefit was absent in the task-irrelevant condition (pitch or intensity discrimination). It suggests there might be an amodal sub-second interval representation; the transfer of interval timing training could be acquired shortly but needs task-relevant attentional engagement.