关键词: 奖赏学习, 情绪面孔, 注意捕获, 干扰

Abstract:

Reward facilitates performance and improves cognitive ability in many tasks. In recent years, with the deepening of the research on reward learning, it has been found that the stimulation after reward learning can always get priority in processing. Reward learning refers to establishing an association between characteristics of stimuli and rewards through accentuation training. At the same time, emotional stimuli were consistently prioritized for attention as task-irrelevant stimuli, compared to non-emotional stimuli. As significant stimuli, both emotional and non-emotional stimuli can affect attention processing. At present, the research on reward is mainly focused on the attention processing of physical stimuli with low visual features, such as stimulus color, shape, spatial position, and neutral face. Emotional faces are more salient than the above stimuli, and the top-down perceptual priming is faster. The association of emotional face attention processing advantage and different levels of reward signals may further strengthen the individual's inhibition of distractor, attention modulation, and enhance the processing of target. Therefore, this study uses the associative learning paradigm to establish the learning association between rewards and emotional faces. In the test task, it is further investigated that whether establishing association of different reward value can regulate the allocation of attention resources and influence the processing of non-target emotional stimuli on the target task.

Twenty-nine participants (14 males, 15 females) were recruited according to the sample size calculated by G-power. Participants completed a line segment orientation task. Emotional faces were presented as non-target interfering stimuli in the experiment. First, the participants completed the baseline task, and then performed training and test. During the training phase, participants were given value feedback when they correctly responded to the target to establish reward association between emotional faces (happy, fearful and neutral) and reward (low reward and high reward). The test would begin in 30 minutes after the training. The test was similar to the reward training, but did not present any information related to reward value.

At baseline, the mean reaction times (RTs) of the three emotional faces were significantly different, with that of fear slower than those of happy and neutral. The reward training results indicated that RTs to high reward were faster than low reward and no reward. There was no significant difference between emotions. The results of the test phase were similar to those of the reward training. As the reward level increased, the emotional faces responded faster. The RTs to fearful faces were the fastest compared with happy and neutral faces at the high-reward intervention. There was no significant difference between the three emotions under low reward.

In summary, these findings suggest that reward can effectively regulate cognitive resources and reduce the deleterious of processing of task-irrelevant emotional faces during target identification. Moreover, different rewards have diverse moderating effects on task-irrelevant emotional faces. As the reward value increased, the interference of fearful faces decreased in target recognition.

Key words: reward learning, emotional face, attention capture, distracters