关键词: 眼跳, 注意控制, 加工水平, 无关信息

Abstract: Posner & Petersen (1990) and Petersen & Posner (2012) described three networks of attention, i. e. alerting, orienting and executive corresponding to different functions of areas in brain. In order to eliminate the task-irrelevant information, these networks work in different patterns which can be presented by eye movements (Lyle and Edlin, 2015). According to The levels of processing model (Craik & Lockhart, 1972), the way in which information is encoded affects how well it is remembered. We hypothesized that task-irrelevant information has different impact on each of the networks of attention, and perceptual and conceptual information affect attentional control in different manner due to their levels of processing. Two experiments were conducted to investigate the effects of task-irrelevant information on attentional control. In experiment 1, a white point-light person walked (biological motion) at the center of the screen facing toward left or right. After 450ms, the color of the point-light walker turned to red or green as a cue to indicate the direction of saccade. Participants were instructed to respond to the color of red or green of the point-light as quickly and accurately as possible by making a saccade toward a target on the left or right side on the screen. The walking direction could be consistent or inconsistent (task-irrelevant) with saccade direction. The experiment 2 was similar to experiment 1 except that the dot were replaced by Chinese characters (“红”，red; “绿”，green). Subjects’ eye positions in those experiments were recorded by SMI Hi-Speed eye tracking system with sampling rate of 350Hz. The saccadic error rate, latency, amplitude, peak velocity, duration and amplitude of each condition were analyzed. In experiment 1, compared to the consistent condition, the inconsistent has significantly longer saccade latency [t(21)=-2.50, p<.05, Cohen’s d=1.091], greater saccadic direction error rate [t(21)=-2.14, p<.05, Cohen’s d=0.934] and greater saccade amplitude [t(21)=-2.54, p<.05, Cohen’s d=1.109]. The results show that the irrelevant information of motion direction increases saccadic error rate, saccade latency and amplitude, and have no effect on saccade peak velocity and duration. The irrelevant information impact on the saccadic plan mainly by capturing attention and indicating the direction of attention. The experiment 2 shows that both motion direction [F(1, 23)=21.68, p<0.01, η2=0.49] and semantic information [F(1, 23)=43.22, p<0.01, η2=0.65] significantly impact saccadic error rates, and their interaction is significant [F(1, 23)=9.49, p<0.01, η2=0.29]. The results indicate that when both irrelevant information of motion direction and semantics presented, saccades were more severely impacted. The saccade latency, peak velocity and amplitude are significantly affected by semantic processing, while the irrelevant motion direction has less impact on them except saccadic error rate. The irrelevant motion direction has stronger effect only when both motion direction and semantics are inconsistent. The results indicate that both alerting and orienting networks of attention are affected by perceptual and conceptual task-irrelevant information. However, the attention system takes longer to exclude the conceptual information than the perceptual ones. The executive of attention is impacted depending on the type of information, in which the similar type of information takes more effect than the dissimilar one.

Key words: saccade, attentional control, level of processing, task-irrelevant information