Abstract
Repetitive exposure to a visual feature results in two opposite behavioral consequences: performance deterioration (PD) due to visual adaptation and increased performance on the feature, namely, visual perceptual learning (VPL). Whether visual adaptation relates to VPL formation process has been a long-standing controversial issue. To resolve the controversy, we measured both PD and VPL on an orientation detection task during training. The experiment consisted of 1-day pre-test, 5-day training, and 1-day post-test stages. One group of subjects (N=7) was presented with stimuli only in the non-dominant eye while the other group (N=6) in the dominant eye. For the non-dominant eye group, during daily training, the mean performance decreased, indicating that PD occurred on each day. The degree of PD decreased with increasing training days. In contrast, performance at the post-test stage was significantly higher than during the pre-test stage, indicating that VPL occurred after the 5-day training. Importantly, the degree of reduction in PD over the 5 days of training was significantly and highly correlated with the degree of VPL across subjects. On the other hand, for the dominant-eye group, neither significant reduction in PD during daily training nor VPL occurred. Since early performances on each day were almost constant in both groups, it is unlikely that a daily performance gain due to VPL occurred. Thus, the significant correlation between the PD reduction and VPL in the non-dominant group cannot be explained by the hypothesis VPL causes larger reduction in PD. These results are rather consistent with the hypothesis that PD reduction is necessary for VPL to occur. To our knowledge, these are the first results that show the causal relationship between PD and VPL.
Meeting abstract presented at VSS 2017