内隐记忆研究的实验方法探索：重学──分离法及其测量效应

为客观计量初学后潜留的内隐记忆效应和探索内隐记忆叠加效应的内在机制,创用了“重学──分离法”。结果发现：“重学──分离法”能有效地测得初学后内隐记忆的保存量;重学后内隐记忆的叠加效应明显,增值为６３．４％;内隐记忆效应与年龄、学历（知识水平）之间无确定的对应关系。     

心境一致性对大学生外显和内隐记忆影响的实验研究

以96名大一本科生为被试,采用120张情绪效价人物图片,运用加工分离范式研究心境一致性对大学生内隐记忆的影响。结果发现：（1）心境一致性效应不仅存在于外显记忆,而且也存在于内隐记忆;（2）心境一致性效应在外显记忆和内隐记忆上没有性别差异;（3）在心境一致性实验中,自然抑郁组被试可能存在外显记忆缺损。     

试论内隐记忆的性质和理论解释

沿着内隐记忆研究的历史发展轨迹阐述了内隐记忆现象的客观存在;深入分析了内隐记忆这一概念,并将其与一些相关概念（诸如外显记忆,无意识记,内隐学习,启动效应,实验性分离现象）进行了比较;概略地介绍了内隐记忆的研究状况及理论解释;最后提出应建立一个意识记忆和无意识记忆相统一、各种合理的理论解释和理论模型相一致的记忆系统的理论模型。     

汉字内隐记忆的实验研究（Ⅱ）：任务分离和反应倾向

在实验Ｉ的基础上,操纵学习时的呈现时间（２５０ｍｓ,５００ｍｓ,７５０ｍｓ）和注意变量（目标和非目标字）,运用任务分离方法及间接证明逻辑,比较了再认测验和整字偏好测验在不同条件下的辨别力（Ａ′）和反应倾向（Ｂ″）。结果：（１）双向功能分离的出现表明再认测验和偏好测验存在本质差异。根据“过程纯净”假设,可以推证内隐记忆存在于偏好测验中。此时,匹配比较法的标准未满足。（２）偏好测验中,５００ｍｓ呈现目标字时,有显著正向启动效应：７５０ｍｓ呈现非目标字时,有显著负向启动效应。（３）内隐记忆和外显记忆发生时,都具有稳定的心理状态。但是前者的报告标准较松宽,而后者的较严格。     

知觉和概念加工范式对内隐与外显记忆发展的影响

控制知觉材料的熟悉性和概念范例的典型性,以考察它们对内隐与外显记忆发展的影响。结果表明：（1）知觉内隐记忆的启动效应不存在年龄和熟悉性效应,而知觉外显记忆成绩则具有明显的年龄和熟悉性效应;（2）在概念内隐记忆任务上,典型范例的启动效应不存在年龄差异,非典型范例的启动效应则存在随年龄而增长的启动效应：无论是典型范例还是非典型范例的外显记忆成绩均随年龄增长而提高。     

内隐记忆的初步实验研究

本研究探讨在无意义字母串学习中内隐记忆的某些特点,并兼用信号检测论进行分析。实验按A、B两组材料系列中所含关键字母“SCT”的比例不同,分成六个组,每组11名,共66名被试。结果表明:1.各组差异明显,合“SCT”比例高的学习材料,启动效应明显,从而证明了存在着一个独立的内隐记忆过程;2.本实验新发现,内隐记忆过程中,心理状态波动甚微,情绪状态稳定,有别于外显记忆过程。     

熟悉与陌生品牌启动效应的实验研究

本研究应用任务分离法对熟悉品牌和陌生品牌启动效应年龄特征进行研究,结果发现大学生被试外显成绩高于内隐成绩,且两者有显著差异。进而再采用二次比较法同样以大学生为被试,对熟悉品牌启动效应进行实验,结果表明:(1)学习前带符号组与不带符号组的成绩之间都有极其显著的差异,学习后带符号组与不带符号组之间也有极其显著差异;带符号与不带符号的两组学习前后的成绩均有极其显著的差异;在二次比较法的应用过程中我们发现:(2)如果学习前后的被试为同一组,则会产生“学习污染”。以陌生品牌为材料,用二次比较法在老、青、少三组被进行实验,并将三个年龄段的被试随机分成两组:学习前组(A)和学习后组(B),结果显示:(3)陌生品牌中存在着启动效应,不同年龄段的消费者外显记忆有差异,但内隐记忆没有差异。     

中学生心境一致性记忆的实验研究

本研究采用任务分离程序（DRM）和实验室诱发情绪的方法,通过2（抑郁组、正常组）×3（积极、中性、消极词汇）×2（回忆、残词补全）的多因素混合实验设计来深入探讨抑郁作为一个个体变量,对外显记忆和内隐记忆有何影响作用。实验1中首先考察了外显记忆中的心境一致性效应,结果显示存在心境一致性记忆效应,实验2中考察了内隐记忆中的心境一致性效应,结果表明,在内隐记忆中也存在心境一致性效应。     

内隐记忆研究的新进展──任务分离范式动向

内隐记忆研究的新进展──任务分离范式动向叶阁蔚,杨治良（华东师大心理系）一、概述内隐记忆是指人们不能够回忆其本身却能在行为中证明其事后效应的经验（ＲＯｅｄｉｇｅｒ,１９９３）．重复启动效应是当前内隐记忆研究的主流。人们为了研究的方便,给出了内隐记忆的...     

“攻击性行为”社会认知的实验研究

采用两种不同的方法对人类“攻击性行为”社会认知的某些特点进行探索。结果表明：①再认过程中,被试对被攻击者的辨别率（Ａ＇）明显高于攻击者,Ｂ″也具有明显特征。证明人类的外显记忆更多倾向记忆“灾难现象”。②偏好测试中,被试偏好攻击者明显多于被攻击者。证实了“攻击性行为”社会认知实验性分离的存在。也就是说学习攻击性图片在人类记忆中,产生了“攻击性行为”社会认知的启动效应。同时也证实了内隐与外显记忆之间存在任务分离的加工演变过程。③偏好测试中出现被试对陌生人（中性人物）的偏好率明显高于攻击与被攻击者的总和。我们认为这是间接启动效应,是外显记忆和内隐记忆在选择上发生冲突继而导致倾向一致的结果。     

提取干扰对不同类型内隐记忆的影响

以往研究关于提取阶段的干扰是否会影响内隐记忆存在着异义, 其中一个重要因素可能源于所采用的内隐记忆测验类型的差异。本研究采用学习-测验范式, 通过4个实验, 分别考察了提取干扰对识别式知觉内隐测验、识别式概念内隐测验、产生式知觉内隐测验和产生式概念内隐测验的影响, 以期对提取干扰和内隐记忆之间的关系有着更全面的了解。结果表明: (1)词汇判断任务(识别式知觉)和语义分类任务(识别式概念)的启动效应在提取干扰下消失了, (2)而产生式词汇判断任务(产生式知觉)和产生式语义分类任务(产生式概念)在提取干扰下仍发现了明显的启动效应, 但与无干扰条件相比, 启动效应量也有着明显的减少。由此可见, 不同类型的内隐记忆都会受到提取干扰的影响, 相比于产生式启动, 识别式启动更容易受到提取干扰的破坏。     

提取干扰对知觉和概念启动的影响

通过学习—测验范式，分别以词汇判断、语义分类任务为测验，探讨提取干扰对不同识别式记忆的影响差异。结果发现，实验1验证了识别式知觉内隐记忆的提取干扰效应，实验2发现识别式概念内隐记忆在提取干扰下也明显降低；两个实验启动量联合分析发现，相比无干扰条件，概念启动比知觉启动受到提取干扰的破坏更大。上述结果说明识别式内隐记忆的提取干扰具有普遍性，提取干扰产生了不同识别式内隐记忆之间的分离，识别式概念内隐记忆更容易受到提取干扰的影响。     

Word priming without awareness: A new approach to circumvent explicit memory contamination

A major methodological limitation arising in the experimental study of implicit memory is that tasks that are characterized as implicit memory tests can be seriously contaminated by the use of covert explicit memory strategies. Given the evidence indicating that brief presentation of words (below the awareness threshold of subjects) can produce semantic priming, we wondered whether rapid visual presentation of primed words might provide an avenue to produce word priming without explicit memory contamination. Normal subjects were tested for word priming on a speeded category membership decision task. Explicit or implicit encoding procedures were used in four different experiments. Results demonstrated that brief presentation of words can indeed offer a means of producing word priming in absence of explicit recognition or recall of the primed words presented during the study phase. They also showed that such priming is equivalent in degree to the priming measured when using either a conventional implicit memory design or an explicit encoding procedure prior to the study of the primes.     

The effect of articulatory suppression on implicit and explicit false memory in the DRM paradigm

Several studies have shown that reliable implicit false memory can be obtained in the DRM paradigm. There has been considerable debate, however, about whether or not conscious activation of critical lures during study is a necessary condition for this. Recent findings have revealed that articulatory suppression prevents subsequent false priming in an anagram task (Lovden & Johansson, 2003). The present experiment sought to replicate and extend these findings to an implicit word stem completion task, and to additionally investigate the effect of articulatory suppression on explicit false memory. Results showed an inhibitory effect of articulatory suppression on veridical memory, as well as on implicit false memory, whereas the level of explicit false memory was heightened. This suggests that articulatory suppression did not merely eliminate conscious lure activation, but had a more general capacity-delimiting effect. The drop in veridical memory can be attributed to diminished encoding of item-specific information. Superficial encoding also limited the spreading of semantic activation during study, which inhibited later false priming. In addition, the lack of item-specific and phenomenological details caused impaired source monitoring at test, resulting in heightened explicit false memory.     

Older adults encode more,not less: evidence for age-related attentional broadening

Previous work has shown that older adults attend to and implicitly remember more distracting information than young adults; however, it is unknown whether they show a corresponding decrease in implicit memory for targets in the presence of distracters. Using implicit memory tests, we asked whether older adults show a tradeoff in memory between targets and distracters. Here, young and older adults performed a selective attention task in which they were instructed to attend to target pictures and ignore superimposed distracter words. We measured priming for distracter words using fragment completion and for target pictures using naming time. Older adults showed greater priming for distracting words compared to young adults, but equivalent priming for target pictures. These results suggest that older adults have a broader attentional scope than young adults, encompassing both relevant and irrelevant information.     

The effects of generation on auditory implicit memory

The generation manipulation produces the classic functional dissociation between explicit and implicit memory. This dissociation has been explained by the transfer-appropriate processing (TAP) framework, which emphasizes the overlap in cognitive processes operative at encoding and at retrieval. However, the vast majority of implicit memory studies have been conducted in the visual modality; in the auditory modality, the effects of generation have never been investigated. In four experiments, we examined the effects of generating from semantic and nonsemantic cues on auditory implicit tests. Generating from antonyms produced a reversed generation effect on priming in auditory perceptual identification and word stem completion, while producing the traditional positive effect on explicit recognition. Generating from definitions, as well as from rhymes, also produced a reversed generation effect on auditory priming. These results are critical for more fully characterizing auditory priming and are consistent with an extended TAP analysis.     

A neural network model of implicit memory for object recognition

People name well-known objects shown in pictures more quickly if they have studied them previously. The most common interpretation of this priming effect is that processing is facilitated by an implicit memory trace in a perceptual representation system. We show that object priming can be explained instead as a bias in information processing, without recourse to an implicit memory system. Assumptions about psychological decision-making processes and bias were added to a neural network model for object identification, and the model accounted for performance both qualitatively and quantitatively in four object identification experiments.