Abstract
Visual crowding, the difficulty in recognizing objects in the periphery amid distractors, can be reduced by perceptual learning. However, it is uncertain how much this improvement transfers to other visual tasks. We examined whether reducing the crowding effect in the peripheral visual field through training transfers to the performance in a word recognition (WR) task (lexical decision task), a rapid serial visual presentation (RSVP) task (reading words that form sentences) and on the FrACT test (four visual acuity and three contrast sensitivity tasks), in three normally sighted samples (N1=36, N2=36, N3=42, respectively). Every sample was randomly assigned to either a training or a control group. The training groups performed a Landolt-C gap detection training daily over 4 days at 10° eccentricity in all four quadrants. A 2-down, 1-up adaptive procedure adjusted the critical spatial distance of two ring-shaped flanker distractors positioned radially and tangentially to the Landolt-C with respect to central fixation. Pre- and post-training, both trained and control (untrained) groups performed the WR (sample 1), RSVP (sample 2) or FrACT tasks (sample 3) at the same positions at 10° eccentricity. Each training group showed a significant (p < .05) decrease in peripheral visual crowding following training. Preliminary results suggest significant (p < .05) improvements in WR and RSVP task performance for the training groups, while no such improvements were observed in the control groups, indicating successful transfer of crowding training. In the FrACT test, on average visual acuity and contrast sensitivity both showed improvements, but no significant differences emerged between training and control groups. The results suggest that training on a Landolt-C crowding task can enhance performance in tasks directly impacted by crowding effects, such as WR and RSVP tasks, but may not extend to more basic single visual features like visual acuity or contrast sensitivity.