Abstract
The visual span for reading is the number of text letters that can be recognized accurately without eye movements, and is hypothesized to be a sensory bottleneck limiting reading speed. It has been shown that the size of the visual span in peripheral vision can be enlarged through training on a letter-recognition task. This enlargement is accompanied by an increase in reading speed, and is primarily due to a reduction in the crowding effect (He & Legge, VSS, 2012). However, the proposed linkage between reading speed, visual span, and crowding appears to be inconsistent with a dissociation between crowding and reading speed (Chung, Vision research, 2007). Chung found that after training on identifying crowded letters in peripheral vision, the spatial extent of crowding was reduced without significantly increasing maximum reading speed. To address this discrepancy, we repeated but modified Chung’s uncrowding training: whereas Chung trained subjects to recognize flanked letters on the midline at 10 deg in the lower visual field, we moved our training 3 letter spaces leftward, where the impact of crowding on visual span is larger than on the midline. Training consisted of four days of the flanked-letter recognition task, approximately 1500 trials/day. In pre- and post-tests, visual span profiles, RSVP reading speed and the spatial extent of crowding were measured at 10 deg in the lower visual field. Consistent with Chung, we found a significant reduction of the spatial extent of crowding. Unlike Chung, our training significantly enlarged the visual span by 3.1 bits and improved the reading speed by 18%. Our results suggest that when uncrowding training is performed at a location where crowding has a more profound effect on visual span, the training can both enlarge the visual span and increase the reading speed, supporting the linkage between reading speed, visual span, and crowding.
Meeting abstract presented at VSS 2013