PSYCHOLOGICAL INVESTIGATION OF EYEWITNESS TESTIMONY-COGNITION AND MEMORY FOR FACES IN DAILY CONTEXT.

本研究首先对证人证言中最活跃的研究领域之一，即所谓的“事后信息效应”进行了文献综述。在这篇综述中，我们研究了产生这种效应的潜在机制。为了解释事件后信息效应，人们提出了四种假说：替代假说(Loftus&Loftus，1980)、偏差假说(McCloskey&Zaragoza，1985)、共存假说(Bekerian&Bowers，1983)和源头监控假说。这篇综述发表在日本认知科学学会期刊上。在一系列实验中，使用了一种名为重复启动法的新方法来研究已知面孔的影响和过程。在这一系列实验中，使用了两种质数(完全面孔和不完整面孔)。主要任务是熟悉度判断，被试判断出现的面孔是否熟悉。结果表明，即使是不完整的面孔也有助于熟悉的判断，不完整面孔的判断比完整面孔的判断更长。这可能是由于最初对不完整面孔的知觉加工造成的。这项研究在意大利国际记忆会议上发表。最后，山田和石岛研究了人类面部表情的分类机制。在审查了现有的模型后，山田和石岛山田和石岛进行了一项实验，对这些模型进行了检验。结果表明：1)信息首先映射到情感的语义空间，然后计算给定信息与原型的距离，3)原型所属的最后一个最接近的范畴被选为合适的范畴，从而实现对面部表情的范畴判断。

In this research to investigate one of the most active research fields in eyewitness testimony, so called "Post-event information effect"literature were reviewed first. In this review, we examined the underlying mechanisms which are responsible for producing the effect. We pointed ont that to explain the post-event information effect, four hypotheses have been proposed, they are substitution hypothsis (Loftus & Loftus, 1980), bias hypothesis (McCloskey & Zaragoza, 1985), co-existence hypothsis (Bekerian & Bowers, 1983), and source monitoring hypothesis. This review was published in the Japanese Cognitive Science Society Journal.A new method so called repetition priming method to investigate the effect and process of known faces was used in a series of experiment. In these series of experiment, two types of prime (complete faces and incomplete faces) were used. Main task was familiarity judgment and subjects judged whether presented face was familiar or not. The results showed that even incomplete faces facilitate the familiar judgment and judgments by incomplete faces (followed by complete faces) were longer than those by complete faces. This might result from the initial perceptual processing of incomplete faces. This research was presented at ICOM (International Conference on Memory, Italy). Lastly Yamada and Itsukushima investigated the categorization mechanisms of human facial expression. After reviewing the existing models, Yamada and Itsukushima Yamada and Itsukushima administered an experiment which examined the models. The result indicated that to attain the categorical judgment of facial expression, 1) information first mapped into semantic space of emotion, 2) then calculate the distance of the given information from the prototype was done, and 3) last nearest category to which the prototype belongs was chosen as the proper category..