被试操作任务（SPT）范式与动作记忆研究

百余年来,记忆研究所使用的材料一直以语言材料为主。然而,言语记忆是否记忆的全貌?使用浯言材料进行的研究结果是否能够推而广之．应用到其他形式的记忆材料上去?在实验室中如何对言语记忆以外的记忆进行深入研究呢?正是为了回答这些问题．Cohen(1981)所设计的被试操作任务(subject performed task,SPT)实验范式引起了大量相关研究。     

动作记忆：记忆研究的新范畴

动作记忆是伴随动作操作和意识卷入过程的一种记忆活动, 其强调意识卷入状态和操作情景对记忆活动的影响。以往的研究均采用SPT范式探讨动作记忆及其影响因素, 并提出非策略加工、多通道加工、动作编码和情景整合理论以解释SPT效应, 但由于关注角度不同, 所提的理论模型不能全面揭示动作记忆的加工机制。部分研究试图从来源监控方面探究动作记忆的内在机制, 但研究仍然存在分歧。     

操作表征在被试操作效应中的调节作用：ERPs研究

本研究考察操作表征(operational representation)在被试操作效应中的调节作用。选取结构性/功能性操作表征短语各160个,分析再认判断中的脑电差异。结果表明,动作编码比语词编码诱发更显著的P1、N2pb和N400;在500~700ms时间窗,相比于语词编码,对结构性和功能性操作表征短语进行动作编码均可显著诱发了中央区更负的平均波幅;在700~900ms时间窗,仅对于结构性操作表征短语,动作编码比语词编码诱发了前额叶更正的平均波幅。研究为动作编码的语义反馈加工机制提供了数据支撑,也为动作信息再激活与动作表征观点的整合提供解释方向。     

SPT效应的自动突显机制:来自输出监测的证据

研究采用"对回忆的再认"范式,从输出监测的角度考察了操作效应的提取机制。实验1结果显示,被试操作任务条件下的自由回忆的系列位置成绩缺乏首因效应,却拥有扩展的近因效应;被试操作任务条件下的"对回忆的再认"成绩显著差于语词任务下的成绩,差异具体表现在组块2~9、10、11和12上,表明在上述组块,被试操作任务条件下的提取存在自动突显,即操作效应得益于这些组块在提取时的自动突显。实验2结果显示,在类别测验下,被试操作任务和语词任务的自由回忆成绩拥有相似的系列位置曲线,而"对回忆的再认"结果同实验1。研究认为自由回忆的系列位置效应与自动突显机制之间不存在直接的因果关系,而"对回忆的再认"范式则能敏感地测量出操作效应的提取机制。     

记忆编码与特异性效应之间关系的ERP研究

通过事件相关电位（ERPs）探讨了在“学习–再认”模式条件下的记忆编码与特异性效应之间的关系。选择首要特异性特征为深、浅加工,次要特异性特征为高、低词频。研究结果表明,（1）行为数据支持了低频词的再认比高频词更精确;深加工的再认比浅加工更精确,并且深加工存在着显著的字频效应,浅加工没有显著的字频效应。（2）低频词比高频词有更大的ERP晚正成分（LPC）;在低频词条件下,存在着显著的相继记忆效应,随后再认正确的ERPs比随后再认不正确的ERPs更正。因此,加工方式和词频间接地影响着记忆编码的形成。     

分类学习与混合学习下的注意促进效应比较

注意促进效应(ABE)是指目标探测性质的干扰会促进与之同时进行的背景刺激的记忆编码,产生比分心拒绝下更优的记忆成绩。Spataro等人(2017)对此提出项目特异性解释,认为目标探测主要促进的是对项目的特异性信息而非关系性信息加工。本研究采用混合学习和分类学习的方式形成对背景刺激的特异性信息和关系性信息加工。结果表明,与混合学习相比,分类学习下的ABE有所减少(实验2),甚至消失(实验1),表明当编码过程中对背景刺激的加工主要依赖于关系性信息时,目标探测所产生的促进效应会被削弱,从而为ABE的项目特异性解释提供更为直接的证据。     

动作记忆SPT优势效应的理论探讨

操作条件下的记忆效果好于语词条件下记忆效果的现象被称为动作记忆SPT效应。以往研究先后提出非策略加工、多通道加工、动作编码及情景整合理论来解释SPT效应,但这些理论解释仍存在矛盾和分歧,并阻碍了当前动作记忆领域的研究进展。为了解决目前的理论困境,本文详细地阐述了各理论的核心内容、发展历程及其存在矛盾和分歧的原因,并提出应以加工过程与加工对象相结合的视角来建立新的理论模型,以进一步促进实证研究的展开。     

熟悉性对被试操作效应的作用:来自无线索回忆再认的证据

在动作事件的再认记忆中,被试操作后的记忆成绩优于单纯的词语记忆(SPT效应)。人们认为操作促进了回想加工,而熟悉性是否对SPT效应起作用仍然存在分歧。研究采用无线索回忆再认范式,考察了信息提取失败时熟悉性对SPT效应的作用。结果显示:(1)SPT和VT两种编码都引发了稳定的无线索回忆再认效应(RWCR效应)。(2)与VT编码相比,SPT编码引发更强的熟悉性加工,表现出无线索回忆再认的记忆优势。上述结果表明,即使无法正确提取细节信息,在SPT编码中被试凭借熟悉性辨认的能力仍然高于VT编码。该结果从信息提取失败的视角为熟悉性对SPT效应的作用提供了更为直接的证据。     

汉语转折复句的命题表征项目互换效应研究

采用句子一图画验证任务(sentence-picture verification task)探讨了汉语转折复句的命题表征项目互换效应。结果表明,汉语倒装转折复句的项目表达顺序为“(但是)B→虽然A”,命题表征中项目存在着互换的倾向,所形成的命题表征为“虽然A→但是B”。本研究结果初步表明,读者理解汉语转折复句可能是一个按照“虽然A(事实让步)→但是B(转折)”固定方向进行系列认知加工的过程。     

两类常见的类别变量调节效应分析

类别变量在心理学和其他社科研究领域经常遇到,当自变量或调节变量为类别变量时,应当如何分析调节效应呢？详述了多类别变量的被试间设计和两水平被试内设计（因变量重复测量2次）的调节效应分析方法,并给出了分析流程。先进行调节效应的显著性检验,后用选点法或Johnson-Neyman法进行简单斜率检验。多类别变量被试间设计的简单斜率检验是先进行整体检验,后进行配对检验。用两个实际例子演示如何进行类别变量的调节效应分析,最后展望了两类设计的类别变量调节研究的拓展方向,例如更复杂的类别变量的调节模型等。     

The association of motor information and verbal information: a new perspective on the mechanism of the SPT effect

Current perspectives on the mechanism underlying the Subject Performed Task (SPT) effect have not yet reached a consensus. This study investigated whether the association between motor and verbal information is a key factor that explains the SPT effect. Experiment 1 and 2 tested whether motor and verbal retrieval memory performance were different when the association level was varied. Results showed that when subjects used low association learning materials, memory performance in either motor or verbal retrieval condition was significantly better than performance in the combined retrieval condition. Experiment 3 examined whether the level of association between motor and verbal information affects the SPT effect. Results revealed that the SPT effect disappeared when low association learning materials were used. Our findings imply that the high association level between motor and verbal information enables mutual activation of the two information types, leading to improved memory performance.     

Enactment supports unitisation of action components and enhances the contribution of familiarity to associative recognition

Memory for actions is usually better following subject-performed tasks (SPT) than verbal tasks (VT). We hypothesised that enactment unitises the components of actions such that familiarity can support associative recognition following SPT. To examine this hypothesis, participants studied verb–object pairs in a SPT or VT condition. During testing, they discriminated between intact, recombined and new items and made Remember/Know judgments; additionally, their EEGs were recorded. Associative recognition was better following SPT than VT. Early frontal event-related potentials (ERPs) were graded according to the item status following SPT, but no such effects were found after VT. Similarly, the late parietal ERPs were graded following SPT, whereas these effects were smaller and did not differ between intact and recombined items following VT. We conclude that enactment unitised the action and object so that familiarity could contribute to associative recognition and that recollection became sensitive to the amount of the matching associative information.     

Recognising relations: What can be learned from considering complexity

Analogy is an important cognitive process that has been researched extensively. Functional accounts of it typically involve at least four stages of processing (access, mapping, transfer, and evaluation); however, these accounts take the way in which the base analogue is understood, along with its relational structure, for granted. The goal of this paper is to open up a discussion about how this process (which we will call “relational recognition”) may occur. To this end, this paper describes two experiments that vary the level of relational complexity across exemplars. It was found that relational recognition tasks benefit from increased complexity, while mapping tasks suffer from it.     

动作操作与怪异性对联结记忆的影响

被试操作任务（SPT）编码后的记忆成绩通常优于语词任务（VT）编码后的记忆成绩,此类现象被称为SPT效应。本研究考察动作与物体之间已有的联结程度对联结再认的SPT效应的影响。学习阶段,被试使用SPT与VT编码识记日常动作短语（动作与物体配对,如“切西瓜”）和怪异动作短语（如“种锤子”）;测试阶段,要求被试区分“旧”、“重组合”和“新”短语。结果发现,当识记日常动作短语时,SPT条件的联结再认成绩显著高于VT条件的成绩,但是,当识记怪异动作短语时,SPT条件与VT条件的联结再认成绩无显著差异。此外,仅在VT而非SPT条件,怪异动作短语的联结再认成绩显著高于日常动作短语的成绩。研究结果表明,动作操作和刺激材料的怪异性均可促进动作与物体的联结记忆,但是两者不能对联结记忆产生叠加的促进作用。     

Relational processing following stroke

The research examined relational processing following stroke. Stroke patients (14 with frontal, 30 with non-frontal lesions) and 41 matched controls completed four relational processing tasks: sentence comprehension, Latin square matrix completion, modified Dimensional Change Card Sorting, and n-back. Each task included items at two or three levels of relational complexity. Relational processing was impaired in the stroke groups. This was due mainly to items at the intermediate ternary-relational level of complexity. Less complex binary-relational items and more complex quaternary-relational items (the latter are difficult for adults generally) were less sensitive to stroke status. Impairment was greater in frontal than non-frontal stroke patients. Positive inter-correlations among measures supported the domain-general nature of relational processing. Implications for assessment and intervention are discussed.