中文摘要：

为检验返回抑制是否受线索生物学意义的调节,分别在实验一和实验二中以阈上和阈下不同情绪效价（高兴、生气和中性）的面孔为外源性线索,变化同时提示位置的多寡,要求被试对靶子尽可能快而准地作探测反应。实验一发现,三种线索情况下均出现返回抑制且效应量无显著差异;实验二发现,线索为中性面孔时出现返回抑制,线索为高兴和生气面孔时未出现返回抑制。说明返回抑制受线索生物学意义的调节。阈下线索的生物学意义（情绪效价）能得到自动加工,从而影响空间注意的转移和返回抑制机制的功用;阈上线索的情绪效价被清晰感知时,自上而下的注意控制机制使线索的生物学意义被忽略,从而阻碍情绪效价功能的发挥。

英文摘要：

In the classical exogenous cue-target paradigm, when a target is presented in the same location as a cue within a stimulus onset asynchrony （SOA） of 250 ms or less, target detection is facilitated. However, when the SOA is extended to more than 250 ms, target detection is slowed at the cued location. This phenomenon is called inhibition of return （IOR）. IOR is generally assumed to be an adaptive mechanism which enhances the chance of success or survival by maximizing the chance of detecting important information in visual search during human evolution. If so, the time course and/or magnitude of IOR may be affected by biological information of the cue. If a cue provides information relevant to individual’s survival, then attention may not be inhibited from reorienting to the cue. However, previous evidence concerning this issue is controversial. This study is to further examine whether IOR is modulated by emotional valences of face, which convey important biological and social information. We hypothesized that emotional information conveyed through supraliminally presented face cues may not differentially affect the IOR effect, because this information could be suppressed by the top-down control settings demanding no interference from emotion, whereas emotional information conveyed through subliminally presented face cues may affect pattern of the IOR effect, because this information is processed automatically and it escapes from top-down suppression. A total of 32 undergraduate students participated in Experiments 1 and 2, which manipulated the number of simultaneously cued locations. Faces with different emotional valences （happy, angry and neutral） were as uninformative peripheral cues, supraliminally in Experiment 1 and subliminally in Experiment 2, while the participants were instructed to make detection responses to the target, which was presented at one of the cued locations or at an uncued location, as quickly and accurately as possible. Experiment 1 obtained an overall IOR effect but the mag