语义相关比例对汉字刺激质量与语义启动交互作用的影响

发布时间：2018-05-08 03:42

本文选题：语义相关比例 + 语义启动　；参考：《西南大学》2014年硕士论文

【摘要】：刺激质量和语义启动的相互作用有两种模式,即交互作用和加法效应。在一些条件下,刺激质量和语义启动表现为交互作用,在一些条件下,刺激质量和语义启动表现为加法效应。多阶段激活模型和交互激活模型都可以解释刺激质量和语义启动的相互作用。在语义相关比例为0.5的条件下,多阶段激活模型和交互激活模型都假设刺激质量和语义启动应表现出交互作用,但在语义相关比例为0.25的条件下,两模型的理论假设出现分歧,多阶段激活模型预测刺激质量和语义启动表现出加法效应,交互激活模型预测刺激质量和语义启动仍然表现出交互作用。交互激活模型和多阶段激活模型最大的区别在于交互激活模型包含串联加工,本文的研究目的即是检验串联加工的正确性,解决多阶段激活模型和交互激活模型的争论,考察多阶段激活模型和交互激活模型哪一个能够更好地解释词汇识别过程。本文包含两个实验,均使用汉字作为实验材料,实验一在语义相关比例为0.5的条件下,研究刺激质量和语义启动的相互作用,与以往的研究进行对照,为实验二奠定基础。实验二使用汉字作为实验材料,在语义相关词对所占比例为0.25的条件下,研究汉字识别中刺激质量与语义启动在N400上的相互作用,考察串联加工是否正确,解决多阶段激活模型和交互激活模型的争论。两实验采用语义启动范式,要求被试判断靶字和探测字是否为同一字,在靶字呈现时不需立即作出反应,直到探测字呈现才需作出反应,为延迟反应的同一字判断任务,可以避免决策相关成分P300对N400的混淆。实验一包含两个因素,语义启动(语义相关和语义无关)、刺激质量(模糊5水平和模糊6水平),语义相关比例为0.5,在被试完成同一字判断任务的同时,记录对靶字的脑电反应。行为结果表明,语义启动主效应显著,语义相关靶字比语义无关靶字的反应时短,正确率高：刺激质量主效应显著,模糊6水平的靶字比模糊5水平的靶字反应时短,正确率高；刺激质量和语义启动的交互作用不显著。脑电结果表明,刺激质量和语义启动在N400上出现交互作用,启动条件下的刺激质量效应比无启动条件下的刺激质量效应更大；语义相关靶字比语义无关靶字引发更小的N400,出现语义启动效应；模糊6水平的靶字比模糊5水平的靶字引发更小的N400,出现刺激质量效应。汉字出现刺激质量的N400效应,说明汉字识别中存在串联加工,而英文没有刺激质量N400效应,汉字和英文具有不同的内在加工机制,英文可能存在语音夹层。实验二包含两个因素,语义启动(语义相关和语义无关)、刺激质量(模糊5水平和模糊6水平),在实验一的基础上增加280对语义无关字对作为填充刺激(不进行分析),使语义相关比例达到0.25。行为结果与实验一基本一致,语义相关靶字比语义无关靶字正确率高,出现语义启动效应；模糊6水平的靶字比模糊5水平的靶字反应时短,正确率高,出现刺激质量效应；刺激质量和语义启动的交互作用不显著。脑电结果表明,刺激质量和语义启动在N400上表现出交互作用,启动条件下的刺激质量效应比无启动条件下的刺激质量效应更大。按照多阶段激活模型的预期,在语义相关比例为0.25,并且被试意识到语义相关比例低的情况下,刺激质量和语义启动应该表现出加法效应,与本实验结果不相符,不支持多阶段激活模型。本文的实验结果,在语义相关比例为0.5和0.25的条件下,刺激质量和语义启动均出现N400上的交互作用,刺激质量和语义启动的交互作用不受语义相关比例的影响,表明串联加工是正确的,符合交互激活模型的预期,支持交互激活模型。
[Abstract]:There are two modes of interaction between stimulus quality and semantic priming, namely, interaction and addition effects. In some conditions, stimulation quality and semantic priming are interacted. In some conditions, stimulation quality and semantic priming are additive effects. Both multistage activation model and interactive activation model can explain the quality of stimulation and the interaction activation model. The interaction of semantic priming. Under the condition of semantic correlation ratio of 0.5, both the multi stage activation model and the interactive activation model assume that the stimulus quality and the semantic priming should show interaction, but the theoretical hypothesis of the two model is disambiguation under the condition of the semantic correlation ratio of 0.25, and the multistage activation model predicts the quality of stimuli and the language. The interaction activation model predicts the interaction between the stimulation quality and the semantic priming. The biggest difference between the interaction activation model and the multistage activation model is that the interactive activation model contains the series processing. The purpose of this study is to test the correctness of the series processing, and to solve the multistage activation model. The argument of interactive activation model, which examines the multi stage activation model and the interactive activation model which can better explain the process of vocabulary recognition. This paper contains two experiments, all of which use Chinese characters as experimental materials, and the experiment one studies the interaction between stimulus quality and semantic priming under the condition of the semantic correlation ratio of 0.5, and the previous research. The experiment two lays the foundation for experiment two. Experiment two use Chinese characters as experimental materials. Under the condition that the proportion of semantic Related words is 0.25, the interaction between stimulus quality and semantic priming in Chinese character recognition is studied, and the correctness of series processing is investigated, and the dispute of multi stage activation model and interactive activation model is solved.
Two the experiment adopts the semantic priming paradigm, which requires the subjects to judge whether the target word and the detection word are the same word, and do not respond immediately when the target word is presented, until the detection word is presented before the response is required. It can avoid the confusion of the decision related component P300 to the N400 by the same word judgment task for the delayed response.
Experiment one consists of two factors, semantic priming (semantic correlation and semantic irrelevance), stimulation quality (fuzzy 5 level and fuzzy 6 level), and semantic correlation ratio of 0.5. The EEG response of the target word is recorded while the subject is completed the same word judgment task. The behavioral results show that the semantic priming effect is significant and the semantic related target words are more than the semantic irrelevant targets. The response time of the word is short and the correct rate is high: the main effect of the stimulation quality is remarkable, the target word of the fuzzy 6 level is shorter than the target word with the fuzzy 5 level, the correct rate is high, the interaction effect of the stimulation quality and the semantic priming is not significant. The EEG results show that the stimulation quality and the semantic priming appear on the N400 interaction and the stimulation quality effect under the starting condition. The effect of the stimulation quality is greater than that in the non start condition; the semantic related target words are smaller than the semantic unrelated target words, and the semantic priming effect appears. The fuzzy 6 level target word is smaller than the fuzzy 5 level target word, which leads to the effect of the stimulation quality. The N400 effect of the Chinese character's stimulation quality shows the existence of the Chinese character recognition. Serial processing, while English does not stimulate the quality of the N400 effect, Chinese characters and English have different internal processing mechanisms, and there may be voice dissection in English.
Experiment two contains two factors, semantic priming (semantic correlation and semantic irrelevance), stimulation quality (fuzzy 5 level and fuzzy 6 level), and adding 280 pairs of semantic independent word pairs as filling stimuli (no analysis) on the basis of experiment one, which makes the semantic correlation ratio reach 0.25. behavior result basically consistent with experiment one, semantic related target word ratio Semantic unrelated target characters have high correct rate and semantic priming effect; the target word of fuzzy 6 level is shorter than the target word with fuzzy 5 level, the correct rate is high, the stimulation quality effect is high, the interaction effect of the stimulation quality and the semantic start is not significant. The EEG results show that the stimulation quality and the semantic priming show interaction and starting conditions on the N400. Under the expectation of the multi stage activation model, when the semantic correlation ratio is 0.25, and the subjects realize that the semantic correlation ratio is low, the stimulation quality and the semantic priming should show the addition effect, which does not agree with the results of the experiment, and does not support the multistage excitation. The experimental results of this paper are that the interaction between stimulus quality and semantic priming is N400 under the conditions of semantic correlation ratio of 0.5 and 0.25. The interaction between stimulus quality and semantic startup is not influenced by the semantic correlation ratio, which indicates that the series processing is correct, conforms to the expectation of interactive activation model and supports interaction activation. Model.

【学位授予单位】：西南大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：B842

【参考文献】