Habitual behavior is thought to emerge with extended training and reduced sensitivity to outcome devaluation. However, little is known about how habit-like oculomotor responses adapt when devaluation is implicit or embedded within a previously learned context. We examined this in a novel oculomotor learning task involving visual shape-reward associations with both standard and overtrained stimuli. Twenty-six participants completed a shape-color learning task while their eye movements were recorded using an eye-tracker system (1000 Hz). The task involved 11 blocks, including training, intra-block reversal (implicit stimulus-reward changes), and classical devaluation phases (explicitly instructed reward changes). Statistical analyses were performed using linear mixed-effects models on accuracy and response time (RT) measures. As expected, higher accuracy and faster responses for overtrained versus standard-trained stimuli were observed during training, confirming stronger learning. In the classical devaluation phase, overtrained stimuli elicited significantly more errors compared to standard-trained stimuli, relative to the performance in the training phase. This indicates stronger resistance to goal-directed updating. The effect was more pronounced during intra-block reversal of associations, where reward contingencies changed without warning. While RTs were not affected by classical devaluation, intra-block reversal significantly increased RTs for overtrained stimuli, relative to RTs in the training phase. This suggests a higher cognitive cost for overriding well-learned habitual responses when changes are unpredictable. These findings provide new evidence for the behavioral rigidity associated with overtraining of oculomotor behavior and suggest that unexpected outcome changes impose an additional switch cost on habitual oculomotor behavior.
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号