单一启动单一探测范式中负启动效应的发生机制

【作者】 王军妮；

【导师】 王勇慧；

【作者基本信息】 陕西师范大学 ， 基础心理学， 2014， 博士

【摘要】 日常生活中,受信息加工系统的限制,我们需要从数以万计的海量信息中选择需要的信息而忽略无关的信息。这些被忽视的信息再次被注意时其加工会受损,即个体对先前受到忽略的刺激词的反应比对一个新异刺激的反应要慢,错误率要高,这一现象称为负启动(negative priming)。经典的负启动范式通常包含启动和探测两个顺次呈现的显示,两种显示均包含目标和干扰项,且探测显示中的目标项与启动显示中的干扰项相同或相关,被试的任务是对目标项进行反应。一般认为,正是由于在启动显示中选择目标的同时忽略了干扰项才导致了负启动效应的发生。然而,近年来的一些研究表明,启动显示中仅有一个词(没有选择和忽略),通过操纵其它条件(重复试次的比例、忽视指导语、启动词的意识以及启动-目标SOA),也能引发负启动效应,这对传统的选择抑制说提出了挑战。上述条件提示,自上而下的抑制机制(预期或忽视指导语)似乎对引发单一启动单一探测范式下的负启动起关键作用。但以往研究在探讨这一机制时存在一些明显不足：预期策略是否能够在操纵相关试次比例的单一启动单一探测中发挥作用不得而知；足够长的启动-目标SOA是抑制形成条件的说法并不能解释所有的相关研究；当掩蔽紧随启动词之后呈现时,自上而下的抑制(忽视指导语)到底起什么作用并没有得到澄清。本研究采用启动和目标词具有语义相关关系的汉字材料,由被试做分类判断,探讨了单一启动单一探测范式下操纵相关试次比例时,由被试的预期所导致的自上而下抑制加工对负启动的影响；在等长的启动-目标SOA条件下,启动和掩蔽以及掩蔽和目标之间的空屏(ISI)在抑制形成中的重要作用；以及当ISI位于掩蔽和目标之间时,自上而下(由忽视指导语引发)和自下而上(由掩蔽刺激引发)抑制的相对作用。具体来说,本研究分为三大系列,共由十个实验组成。研究一采用单一启动单一探测范式,操纵以往较少受关注的相关项的比例以及增加刺激集的数量,考察在这些条件下由预期所导致的自上而下抑制机制的特点。由两个实验组成,实验一将相关试次的比例设置为20%和80%,结果在20%时发现了语义负启动,80%时发现了语义正启动。实验二在相关项低比例基础上进一步操纵了刺激集的大小,发现仅有小的刺激集可以诱发语义负启动。研究二主要探讨在等长的启动-目标SOA条件下,启动和掩蔽或掩蔽和目标之间的间隔(ISI)在抑制形成过程中的作用。由五个实验组成,在这五个实验中,首先在屏幕中央呈现一个简短的(16ms)受到忽视的启动汉字,之后是掩蔽刺激(##),最后出现目标汉字并要求完成分类判断。ISI插入在启动和掩蔽之间(实验三和实验七)或者插入在掩蔽和目标之间(实验四——实验六)。实验三探讨是否在启动和掩蔽之间插入一个空屏间隔(470-ms启动-掩蔽ISI)比未插入间隔(0-ms启动-掩蔽ISI条件)更有利于抑制的形成,从而引发负启动效应。结果发现,插入ISI发生了显著的负启动效应,而未插入ISI时是显著的正启动效应。实验四探讨是否将与启动和掩蔽之间相同空屏间隔插入掩蔽和目标之间(470-ms掩蔽-目标ISI)相比未插入间隔(0-ms的掩蔽-目标ISI)同样会诱发负启动效应。结果在470-ms ISI时仅出现了激活的衰退,而0-ms ISI和实验三类似,发现的是正启动效应。实验五考察是否延长掩蔽-目标ISI(700ms)会相较短掩蔽-目标ISI条件(470-ms)更利于引发负启动,结果发现了长ISI下显著的负启动效应,而短ISI条件下的模式与实验四的相同条件一致。实验六继续使用实验五中的长掩蔽-目标ISI条件(为了与本实验另一条件作对比,称之为114-ms mask/700-ms ISI),与等长启动-目标SOA条件下的短掩蔽-目标ISI (700-ms mask/114-ms ISI)比较,以验证长掩蔽-目标ISI条件下的负启动效应并非由相应变长的启动-目标SOA引发,而确是因增长的ISI所致。结果发现,前者如实验五出现了显著的负启动效应,而后者则没有。因为实验三至实验六的结果证明,不论ISI插入启动和掩蔽之间还是掩蔽和目标之间,都能引发负启动效应,只是插入掩蔽和目标之间的ISI需要的时间更长。实验七则操纵不同时长的启动-掩蔽ISI(70ms,170ms,470ms),探讨是否可以在更短时长的启动-掩蔽ISI条件下诱发负启动效应,结果只在470ms时发现了类似于实验三的显著负启动效应。研究三探讨当ISI位于掩蔽和目标之间时,抑制效应究竟来源于自上而下的忽视指导语还是自下而上的掩蔽。即这两种成分在负启动效应产生中的相对作用。由三个实验组成。实验八没有忽视指导语,使用研究二部分的“##”作为掩蔽刺激,考察启动刺激处于弱激活条件下的效应,结果发现长掩蔽-目标ISI条件下确有显著的负启动效应发生。但该结果不能排除被试自身采用的忽视策略对抑制形成的作用。实验九则运用密度更大的掩蔽刺激“(?)”,也没有忽视指导语,考察当启动刺激处于无意识状态无需采用忽视策略条件下的效应模式,结果发现,两种ISIs下负启动效应均不显著。实验十仍然使用“(?)”作为掩蔽刺激,对比了有忽视指导语和无忽视指导语时的效应,结果只在有忽视指导语时发现了显著的负启动。综合研究一、二和三的结果,我们得出以下结论：1.在单一启动单一探测范式中操纵相关试次比例时出现的抑制效应确实由采用自上而下的预期策略所致,而非以往认为的默认效应。但这种自上而下预期策略的实施受刺激集数量的调控,较小的刺激集可以产生较大的抑制过程,较大的刺激集则诱发较低的,甚至不会诱发抑制过程,这与刺激集大小而导致的启动词特征的突显性有关。2.抑制的形成取决于启动和掩蔽或掩蔽和目标之间ISI(空屏)的时长。但ISI位于启动和掩蔽之间(强激活)所需的时长要短于位于掩蔽和目标之间(弱激活)。支持了本研究提出的观点：足够长的空屏是抑制形成的先决条件,其它干扰信息(启动或掩蔽)的存在将会占用认知资源并影响抑制的实施。同时也支持了以往的理论：对强激活刺激的抑制要快于对微弱激活刺激的抑制。3.ISI位于掩蔽和目标之间时的抑制效应来自于自上而下的忽视指导语。单独使用无意识掩蔽策略无法引发抑制效应,还需附加自上而下的忽视策略才能引发明显的抑制效应。更多还原

【Abstract】 In daily life, we need to select the desired information from hundreds of thousands of mass information and ignore irrelevant information becasuse of a limited processing system. Subsequently, the ignored material is subject to impaired processing. That is, one may have a slower reaction time or less accurate responding to previously ignored objects than to a new object. This phenomenon is known as negative priming (NP). A typical NP experiment often involves two consecutive displays labeled "prime" and "probe". In each display, targets and distractors are included, and the targets in probe display are identical or related to the distractors in prime display. Participants are asked to response to the targets in each display.It is generally believed that the presence of selective conditions in prime display leads to the occurrence of negative priming. However, some recent studies have shown that when the prime display consists of a single word (no choice and ignore), NP can also be triggered by manipulating some other conditions (e.g. the proportion of repeated trials, ignore instructions, consciousness of the primes, prime-target SOA), which challenges the traditional view on selective inhibition.The above-mentioned conditions suggest that it seems the top-down inhibitory mechanisms (expectation or ignored language) play a key role in triggering the negative priming. However, there are some obvious deficiencies in previous studies to explore this mechanism:it is unknown whether the expected strategies can play a role in manipulating the proportion of related trials in single-prime and only target paradigm; these ideas on the long enough prime-target SOA helpful to information of inhibition can not explain all of the relevant researches; when a brief prime is followed immediately by a pattern mask, the role on the top-down strategy (ignore instruction) in triggering single-prime NP has not been clarified. In current study, the materials of Chinese charaters are used, and participants are required to make a binary animate/inanimate judgment on a single target that was preceded by a semantically related or unrelated prime. We explore the influence of the top-down inhibitory process resulting from the expectation on the negative priming underlying the proportion of related trials in single-prime and only target paradigm; and verify the crucial role of the blank interval (ISI) between prime and mask or between mask and target in inhibitory formation with the constant prime-target SOA; and then examine the relative role of the bottom-up (triggered by ignore instruction) and top-down (resulted from mask) inhibition when ISI is located between mask and target.This study contains a series of ten experiments for three parts.The first part, using the single-prime and only target paradigm, manipulates the proportion of related trials which is less concerned previously and increases the quantity of the stimuli set, and then explores the features of the top-down inhibitory mechanism from expectations. It contains two experiments. In Experiment1, the proportion of related trials is set to20%and80%. Semantic negative priming effect is observed in20%condition, whereas the reliable positive priming effect is observed in80%. In Experimnet2, we manipulates further the quantity of the stimuli set based on the low proportion of related trials, and finds that only the small stimuli set can induce a reliable negative priming.The second part focuses on the influence of interstimulus intervals (ISIs) between prime and mask or between mask and target on inhibitory formation in the constant prime-target SOA conditions. It consists of five experiments, In all five experiments, a brief (16ms) and ignored single prime appeared at center, and followed by a mask (##), and then a single target appeared which entailed semantic categorization. An ISI could be inserted either between the prime and mask (Experiment3and7) or between the mask and target (Experiments4-6). The objective of Experiment3is to examine whether it is more conducive to the formation of inhibition when a blank interval (470ms prime-mask ISI) is inserted between prime and mask than no interval (0ms prime-mask ISI), thus triggering the negative priming effect. The results show that a significant negative priming effect is observed by inserting ISI, whereas a reliable positive priming effect is found without ISI. Experiment4is to examine whether the same interval as that between prime and mask in Experiment3inserted between mask and target (470ms mask-target ISI) is similarly more benificial to triggering negative priming than without ISI (0ms mask-target ISI). As the results, attenuation of activation is found in470-ms ISI condition, whereas the reliable PP is found in0-ms ISI as that in Experiment3. Experiment5focuses on whether prolonged mask-target ISI (700ms) is more conducive to trigger negative priming than short mask-target ISI (470ms). The results show that a significant negative priming effect is found in the long ISI, while the same pattern of result as that of Experiment4is observed. To verify the negative priming in long mask-target ISI condition is indeed caused by the lengthened ISI, rather than by the corresponding lengthened pime-target SOA, Experiment6continues to use the long mask-target ISI in Experiment5(In order to compare with the other condition in current experiment, it was called as the114-ms mask/700-ms blank) and compared it with the short mask-target ISI (700-ms mask/114-ms blank). It is found that a significant negative priming effect in the former condition as that of Experiment5, while does not in the latter. Because the results in Experiment3-6demonstrate that either ISI inserted between prime and mask or between mask and target, can lead to negative priming effect, just that shorter duration is required for the latter, Experiment7manipulates the variable length of prime-mask ISI (70ms,170ms,470ms) and explores whether negative priming can be observed in the shorter length of prime-mask ISI. The result is that NP is found only in470-ms prime-mask ISI condition as that of Experiment3.The third part explores whether the inhibitory effect is caused exactly by the top-down ignore instruction or the bottom-up mask when ISI is located between mask and target. In other words, that is to examine the relative role of these two inhibitions in triggering negative priming effect. This part is composed of three experiments. Experiment8continues to use the masking "##" as that of the second part without ignore instruction, and explores whether negative priming can only occur in weak activation of prime. A significant NP is observed in long ISI condition. But the result can not rule out that the inhibitory formation is affected by the ignore stratigies used by participants themselves. Experiment9uses a masking stimuli of greater density ("2") without ignore instruction too, and explores what is the pattern of results when prime is unconscious without ignore instruction. Non-significant negative priming is observed in two ISIs. Experiment10still uses "(?)" as a masking stimulus and compares the inhibitory effect both with and without ignore instruction. Only a significant NP is observed when there is ignore instruction.According to the study results above, we make conclusion as below:1. When the proportion of related trials is manipulated in single prime and only target paradigm, the inhibitory effects indeed result from the top-down expected strategies, instead of the default effect previously thought. Nevertheless, the implementation of inhibition relates to the size of stimuli set. Smaller stimulus set can trigger a greater inhibition, while the larger stimulus set can induce lower, or even can not induce it. This relates to the salience of prime’ feature caused by different sizes of stimulus set.2. Formation of inhibition depends on the length of ISI between prime and mask or between mask and target. But the required length of ISI is shorter when it is located prime and mask (strong activation) than between mask and target (weak activation). This supports our proposed viewpoints that long enough ISI is a prerequisite for the formation of inhibition, and that the remained existence of other interference information (prime or mask) will occupy cognitive resources and affect the implementation of inhibition. Simultaneously, this result also supports the viewpoints that the speed of inhibitory formation is faster when prime word is activated stronger than weaker.3. When ISI is located between mask and target, the inhibitory effect comes from the top-down ignore instruction. Unconscious masking strategy alone can not lead to NP, instead that NP is triggered by adding a top-down ignore instruction in unconscious condition.更多还原