社会与工程心理学研究室知识库

与当前任务无关的刺激自动吸引注意的现象被称为注意捕获（Attentional capture）。 注意捕获的早期研究认为该过程是完全由刺激驱动的自下而上的过程，但随着研究的不断深入，有大量研究者认为注意捕获受自上而下的注意控制定势的调控（Attentional control setting）并据此提出了关联注意捕获理论（Contingent attentional capture），其中注意控制定势可以将注意限定到任务目标特征的各个维度。关联注意捕获理论受到大量研究的支持，是注意领域的热点。
目前，在基于特征的关联注意捕获研究中，任务目标全部都是反应信号，即被试需要对具有某种特征的目标进行反应进而建立任务相关的注意控制定势。此类反应信号固然非常重要，但环境中还有另一类与我们的生存休戚相关的视觉信息：停止信号。在生活中，我们需要不断调整自身的行为以适应环境的变化，其中重要的一部分就是根据停止信号抑制当前行为。而一次成功抑制的必要条件就是注意能够指向停止信号。因此，从实际生活的角度，注意对停止信号的选择机制是有关人类生存能力与适应性的重要研究领域；从理论研究的角度，一方面对停止信号的加工是完整的认知过程而过去的研究只关注了反应抑制的部分， 基于停止信号的注意捕获的研究对探索停止信号的完整加工过程具有重要意义； 另一方面基于停止信号特征的关联注意捕获研究是对关联注意捕获领域的重要补充；
综上所述，本论文基于关联注意捕获领域，以停止信号特征为切入点，考察（1）停止信号特征能否建立注意控制定势；（2）若（1）成立，此类注意控制定势具有何种时间特征；（3）若（1）成立，此类注意控制定势具有何种 ERP 效应。
研究一结合了停止信号任务，在空间线索化范式的基础上增加了停止试次。在停止试次中，停止信号的延迟参数为零，即停止信号与执行信号同时出现，被试需要抑制自己的反应。利用该实验范式，研究一的两个实验验证了具有停止信号颜色特征的任务无关线索能够以自上而下的方式捕获注意，这种注意捕获与信号频率和颜色亮度无关，并且不限于特定的颜色组合。研究一的结果表明，被试能够建立基于停止信号颜色特征的注意控制定势。研究一中另一个效应是停止信号色线索对被试整体反应的损害效应包括更长的反应时和更高的错误率。
研究二探讨了基于停止信号特征的注意捕获在不同线索-目标SOA下的捕获/返回抑制效应。结果发现，在108毫秒SOA条件下，停止信号色线索稳定地产生了反应时的线索化效应和错误率的线索化效应。在624毫秒条件下，停止信号色线索依然能够产生反应时和错误率的双线索化效应，同时其对目标反应时的整体损害大幅下降，但这种损害仍然存在。在 1200 毫秒条件下，停止信号色线索只能产生错误率的线索化效应不能产生反应时的线索化及逆线索化效应，同时其对目标反应时的整体损害完全消失。
研究三的行为结果与研究一基本一致。线索诱发的 N2pc成分从 ERP 的角度证实了之前的行为研究中线索注意捕获的结果。在线索呈现后的 180-250毫秒时间窗内，停止信号色线索出现了 N2pc效应，表明注意资源被分配到了其所在的位置，即其都捕获了注意。
以上结果表明：（1）停止信号颜色特征能够建立注意控制定式；（2）其注意捕获能够在108毫秒和 624毫秒SOA条件下发生， 但在1200毫秒的条件下没有产生返回抑制效应；（3）停止信号色线索能够诱发 N2pc效应。

Attentional capture refers to a phenomenon that irrelevant stimulus involuntarily receive attentional priority. Concerning the driven source of attentional capture, early researches proposed that attentional capture is a completely bottom-up, stimulus-driven process. However, with the deepening of related researches, a large number of researchers believed that attentional capture is modulated by the top-down attentional control settings, and accordingly they proposed the contingent attentional capture theory. In this theory, the attentional control settings limit attention to each feature dimension of the target of current task. The contingent attentional capture theory is supported by a large amount of studies and has been a research focus in attention for several years.
So far, among feature-based contingent attentional capture studies, the targets in all experimental designs have been go signals. Participants were required to respond to such go targets and set up attentional control settings pointing to the feature of the go targets. Despite the importance of the go signals, in the living space there is another kind of visual stimuli that shouldn’t be ignored: the stop signals. In life, we need to constantly adjust our behavior to adapt to changes in the environment, and one of the most important adjustment we take is to inhibit current behavior according to stop signals. Such condition is necessary to a successful inhibition as stop signals receive attentional priority. Therefore, practically the mechanism of attentional selection on stop signals is an important research field involving human survival and adaptability; from the perspective of theoretical research, the contingent attentional capture based on the feature of the stop signal is a critical supplement to both attentional capture and response inhibition theory because the former lacked attempts to set up attentional control settings based on the stop signals while the latter lacked focus on the attentional process on the stop signals.
In summary, this study focused on how contingent attentional capture interact with the feature of stop signals, and was designed to examine: (1) whether the attentional control settings based on the feature of stop signals could be set up; (2) the time characteristics of such attentional control settings should (1) be true; (3) the ERP characteristics of such attentional control settings should (1) be true.
In the first study, we combined the spatial cuing paradigm and the stop-signal task, created a new paradigm for the exploration of the attentional capture based on the color feature of stop signals. In addition to the original go trials in the spatial cuing paradigm, the new paradigm included stop trials in which the target (go signal) and the stop signal were presented simultaneously. The two experiments in the first study proved that the task irrelevant cues with the color of the stop signals captured attention in a top-down way, and this attentional capture effect was not due to the frequency and the luminance difference between the stop signal color and neutral color. Further, this attentional capture effect was not limited within certain color combinations. The first study also revealed that cues with the color of the stop signal debased the overall reaction speed and accuracy of targets.
In the second study, we measured the stop-signal-color-based attentional capture effect under different SOAs and tested whether there were inhibition of return in longer SOA. The results showed that there were cueing effect for both reaction time and error rate in the stop-signal-color cue condition in 108ms SOA and 624ms SOA. The debasement, caused by the stop-signal-color cue, for the overall performance of targets were significantly weakened when SOA was 624ms. When SOA was 1200ms, there were neither cuing effect nor inhibition of return in the stop-signal-color cue condition, and the stop-signal-color cues had no effect on the performance of targets.
In the third study we implemented an ERP experiment to examine the N2pc effect elicited by the stop-signal-color cues. The behavioral results largely replicated what we got in the first study. The stop-signal-color cues elicited an N2pc effect in a time course of 180ms to 250ms after its onset, indicating that attentional resources were allocated to the position of the stop-signal-color cues.
These findings suggest that (1) attentional control settings could be built based on the color feature of stop signals; (2) the attentional capture effect modulated by such attentional control settings could occur at 108ms SOA and 624ms SOA; (3) the stop-signal-color cues could elicit N2pc effect.