【作者】 凤四海；

【导师】 黄希庭；

【作者基本信息】 西南大学 ， 基础心理学， 2006， 博士

【摘要】 时间知觉即“知觉到的现在”，其约3秒钟的时间范围是产生“现在感”的基础，同时也是信息得以进行时间整合并形成统一的知觉对象的基础。时间知觉的时间范围提供了一个人类有意识信息加工活动时间窗口，也反映了人类动作行为的基本时间组织。本研究依据“分段综合模型”对时间知觉的概念界定，在综述经典时间知觉研究和信息加工计时特性有关研究的基础上，提出时间知觉反映了有意识信息加工的时间限度的假设；同时直接以视觉信息的注意加工为对象，设计了五个实验系统考察视觉信息的选择性注意计时特性，揭示其时间限度。实验一～实验三分别采用空间视觉搜索、空间快速序列视觉搜索以及标准快速序列视觉搜索三种范式考察视觉搜索中的注意计时特性，并同时考察空间位置(实验一、二)、搜索任务类型(实验一)、显示集容量(实验一、二、三)、项目间时间间隔(实验二、三)、项目可区分度(实验一、二、三)等因素对于视觉搜索作业的影响。实验四、五采用快速序列视觉呈现双目标任务考察注意瞬脱现象及在改变任务目标和任务类型、项目呈现时距条件下注意瞬脱效应的普遍性，进一步揭示视觉信息加工中注意的计时特性。本研究条件下得出如下主要结论：1空间搜索中，特征搜索和联合搜索都包含并行和串行加工成分；被试作业反应时随显示集容量增大而增大，随项目可区分度减小而增大；2空间序列搜索中，联合搜索正确率随项目呈现时距和可区分度增大而提高，目标未出现时正确率更高；注意资源可能在各加工阶段间进行动态的分配；3标准序列搜索中，被试作业正确率随项目呈现时距增大而提高；该条件下项目间的遮蔽抵消了难度效应；目标未出现时正确率更高，同样也反映了注意的灵活性；4标准序列呈现双任务作业范式中，双目标任务条件下，被试对第二个目标的加工相对单目标任务在第一个目标呈现后200～400ms出现损伤，表现出明显的注意瞬脱效应；5采用相同范式，变化三种项目呈现时距(60、80、100ms)，改变目标项类型和作业类型，仍分离出较明显的注意瞬脱效应；注意瞬脱所反映的不是知觉组织阶段的加工特点，而是知觉后乃至工作记忆加工的计时特点；6视觉搜索和注意瞬脱反映了不同的注意计时特性，注意瞬脱牵涉到对两个序列目标的觉察，此时注意转换的时间限度为约200～400ms。视觉搜索只牵涉到单个目标的检测，以反应时为指标得到约100～150ms每项的注意转换速度估计，但其中可能杂揉了空间并行加工和外显眼动的影响；以正确率为指标，在空间序列搜索中得到注意选择的时间解析度至少可达到80～160ms每项；在遮蔽比较明显的标准序列搜索中，该时间解析度至少也可达到160ms每项。7视觉搜索更多地反映了注意的早期选择瓶颈；注意瞬脱则更多地反映了注意的晚期选择瓶颈；可以尝试用一个多阶段注意选择理论来整合这两个领域的争论。本研究从理论上辨析了经典时间知觉的研究内容，阐明了时间知觉本质上反映了有意识信息加工的时间限度，因而可以据此统合经典时间知觉和当代认知神经科学对信息加工计时特性的研究。据此，本研究从注意的有意识觉察本质出发，首次直接以注意的计时特性为研究对象，突出了注意乃至有意识信息加工时间维度上的限度；并尝试用一个整合的多阶段理论揭示有意识信息加工存在普遍但分层级的时间限度，丰富了时间知觉研究和分段综合模型。本研究还揭示出时间维度上的注意选择普遍存在于信息加工的各阶段，因而视觉搜索和注意瞬脱领域中的一些争论得到了较好的整合。本研究的创新点还表现在将视觉搜索和注意瞬脱等范型引入时间知觉领域；综合采用空间、空间序列、标准序列视觉搜索、注意瞬脱等多种实验范式以及反应时、正确率多种指标系统考察注意的计时特性及其影响因素，体现了分段综合模型所建议的多范式、多方法、多因素的综合研究取向在揭示时间知觉本质中的生命力。最后在指出本研究内容、方法等方面不足的基础上给出了今后进一步改进的方向。更多还原

【Abstract】 Time perception, being regarded as perceived present, has a three-second around temporal extent which is underlying the sense of present. Also, it is substrate for temporal integration of informations from which unitary and coherent percepts are established. The range of perceived present represented a basic temporal organization of humans’ motion behaviors, as well as providing a temporal window for the activities of conscious information processing.This research, beginning with the definition of time perception based on range-synthetic model, reviewed classical researches on time perception and updated researches related with temporal processing of information. Then, it was suggested that time perception represented the time limits of conscious information processing. Based on this hypothesis, five experiments were designed to explore the timing mechanisms, especially the time limits, of selective attention for visual information processing.Three paradigms, named as spatial search, spatial RSVP search, and standard RSVP search, were used respectively in first three experiments to explore the atttentional timing mechanism underlying visual search. Also, the effects of spatial location (Expt. 1 and 2), search task (Expt. 1), setsize (Expt. 1, 2 and 3), SOA between items (Expt. 2 and 3), and discrimination of items (Expt. 1, 2 and 3) on the performance of visual search were examined. Experiment 4 and 5 both used standard RSVP paradigm with double task to explore the phenomenon called attentional blink (AB). Further, the availability of AB effect under conditions with different target identities, task requirements and presenting durations of items was examined to uncover the attentional timing mechanism of visual information processing.The present research concluded that:Firstly, both parallel and serial component was founded in feature and conjunction search in spatial search paradigm, in which the RT of subjects’ performance also increased with the increase of setsize and decrease of discrimination among items.Secondly, the correct ratio of subjects’ performance in conjunction search increased with the increase of both SOA and discrimination among items, with higher correct ratios under the target-absent condition, in spatial RSVP search paradigm which reflected dynamic allocation of attention resource in multiple stages of processing.Thirdly, the performance of subjects increased with the increase of SOA and the effect of discrimination was lost because of the masking between items in standard RSVP search paradigm. Also, higher correct ratios under the target-absent condition were argued as reflecting the flexibility of attention processes.Fourthly, the performance of second target (T2) detection was impaired in condition of dual-target task compared with single-target task in the time range of 200 to 400 ms after the onset of first target, in standard RSVP paradigm with double tasks, which represented remarkable attentional blink effect.Fifthly, with the same paradigm, significant AB effect was found under the conditions of different target identity and task requirement, which suggested that this effect represented the attentional timing characteristic of postperceptual and even working memory processing, rather than in the stage of perceptual organization.Sixthly, attentional blink and visual search represented different attentional timing mechanism. The former involved the awareness of two targets and the time limit of attention switching here was estimated as 200 to 400 ms around. The latter, visual search, involved the detection of single target and the speed of attention switching was estimated as 100 to 150 ms per item, which may include the effect of spatial serial processing and overt eye movement, with the index of reaction time. On the other hand, the temporal resolution of attention selection was estimated as 80 to 160 ms per item in the spatially serial search and 160 ms per item in the standard serial search with remarkable masking effects, respectively, both with the index of correct ratio. Lastly, visual search and attentional blink represented more on the bottleneck of early attention selection and of late selection, respectively. Thus, a theory of multi-stage selection was used to integrate the debates between the two fields.Based on the theoretical analysis of the contents of classical time perception studies, the present research demonstrated that time perception in nature represented the time limits of conscious information processing. It may be used to integrate the fields of classical time perception and of temporal processing of information in contemporary cognitive neuroscience. Based on the nature of conscious awareness, the attentional timing was directly studied, which emphasis the limits of attention and even conscious information processing in temporal dimension. Also, an integrated theory of multi-stage was used to indicated that there exist universal but multiple levels of temporal limits, which expanded the studies on time perception and range-synthetic model. The present research also indicated that the attention selection in the time dimension occurred in multiple stages of information processing, which made it possible to integrate debates in visual search and attentional blink. At last, the present research demonstrated it’s value by introducing visual search and attentional blink paradigms into field of time perception and using varied experimental designs and indices to explore timing of attention and factors affect it. This last point represented a synthetic research approach with multiple paradigms, multiple methods and multiple factors suggested by range-synthetic model, which will show its vitality in uncovering the nature of time perception.As the end, furth research suggestions were provided based on the deficiencies of present research on the research contents, methods, and so on.更多还原