可替代原因形态对因果推理的影响

发布时间：2019-03-21 08:57

【摘要】：因果推理是一种高级认知加工过程，个体需要在充分理解因果结构、因果强度和相关背景知识的基础上做出最优选择。可替代原因（alternative cause）指的是在多因一果的简单因果关系中影响结果发生的唯一核心原因（focal cause）以外的其他因素。可替代原因在因果推理中扮演着重要的角色，是因果推理中需要认真推敲的内容。已有研究表明，因果信息的呈现方式，因果联结强度，因果结构等都会对因果推理产生影响，那么是否可替代原因的强度，结构类型和状态也会对因果推理产生影响呢？本研究基于可替代原因的视角，借鉴FernbachRehder（2013）的研究，通过调节可替代原因的强度，可替代原因的表征结构，可替代原因的状态，依次探讨了不同的可替代原因条件对因果推理的影响。具体分为三个实验研究：实验1以大学生为被试，采用2（核心原因强度：40%，80%）×3（可替代原因强度：25%，50%，75%）的被试间设计探讨在核心原因强度存在差异的条件下可替代原因强度对因果推理的影响。为了避免回忆误差的影响，实验以纸笔的形式呈现给被试相应的因果信息及相关的因果结构，，让被试对P(E/C)，P(C/E)，P(E/-C)，P(C/-E)做出概率判断。结果显示，被试对核心原因强度敏感；反向预测推理（由结果没有出现推测原因）对可替代原因的强度敏感；在诊断推理中存在核心原因强度和可替代原因强度的交互效应。上述结果说明被试可以利用可替代原因的信息对不同类型的因果推理做出判断，在强核心原因条件的诊断推理中弱核心原因的存在反而使得推理结果降低（发现弱证据效应）。在实验1的基础上，实验2采用2（可替代原因表征：明确，不明确）×3（可替代原因强度：25%，50%，75%）的被试间设计进一步探讨了可替代原因的表征和强度对因果推理的影响。实验以纸笔的形式呈现给被试相应的因果信息及相关的因果结构，让被试对P(E/C)，P(C/E)，P(E/-C)，P(C/-E)做出概率判断。结果显示，可替代原因的表征并不会对因果推理产生影响；反向预测推理中出现对可替代原因强度的敏感；但是在预测推理和反向诊断推理中我们发现可替代原因表征和强度的交互效应。上述结果说明推理者不能对可替代原因结构的状态进行整合，可替代原因的强度没有对推理产生影响可能是因为被试认为不明确条件中的可替代原因是不存在的，不明确表征可能会使得强可替代原因条件下的推理出现较大偏差。在实验1和实验2的研究基础之上，实验3以大学生为被试，深入探讨了在结构明确的条件下，可替代原因出现，不出现两种推理条件中可替代原因强度（25%，50%，75%）对因果推理产生的影响。根据我们提供的因果关系和结构信息让被试对P(E/CA)，P(C/EA)，P(E/C-A)，P(C/E-A)，P(E/A)，P(A/E)做出概率判断。结果显示，在预测推理中，被试对可替代原因的强度敏感；在可替代原因没有出现的条件下，预测推理中也出现了对可替代原因的敏感。结果表明人类在因果推理中，并不是没有考虑可替代原因的影响，对替代原因的忽视是由于在最后的概率判断中出现的表征错误和计算步骤的缺失。本研究从不同的可替代原因条件出发，全面考察了可替代原因对于因果推理的影响。通过这三个实验，本研究得出三个重要结论：首先，对可替代原因的忽视并不仅仅是因为对已有知识的检索困难。第二，对可替代原因的忽视并不是由于推理中的懒惰，也就是说只有当可能的判断误差很小时人们才会放松推理的标准。第三，明确可替代原因的表征并不一定能够提高推理准确度。研究结果显示当相关知识以简单、具体的形式呈现时，人们更容易进行因果推理。
[Abstract]:Cause and effect reasoning is a kind of advanced cognitive process, and the individual needs to make the best choice on the basis of full understanding of the cause and effect structure, the cause and effect strength and the related background knowledge. Alternative cause refers to other factors other than the single core cause that affects the occurrence of a simple causal relationship for a single fruit. The alternative causes play an important role in the cause and effect reasoning, which is the content that needs to be carefully knocked out in the cause and effect reasoning. The existing research shows that the form of the causal information, the causal link strength, the causal structure and so on will have an effect on the cause and effect reasoning, and whether the strength, structure type and state of the cause can affect the causal inference. Based on the perspective of the alternative causes, this study is based on the study of FernbachRehder (2013), and by adjusting the strength of the alternative reason, the characteristic structure of the alternative reason, the status of the alternative reason, the effect of different alternative causes on the cause and effect reasoning is discussed in this order. In particular, it is divided into three experimental studies Study: Experiment 1 was tested with 2 (core reason:40%,80%) and 3 (an alternative reason:25%, The experimental design of 50%,75%) discusses the cause-and-effect reasoning under the condition of the difference in the core reason intensity In order to avoid the influence of the memory error, the paper presents the cause and effect information and the related causal structure in the form of a paper pen, and makes the test to the P (E/ C), P (C/ E), P (E/-C), P (C/-E). Rate judgment. The results show that the participants are sensitive to the core cause intensity, and the reverse predictive reasoning (the result does not show the reason for speculation) is sensitive to the intensity of the alternative causes; the core reason intensity and the alternative cause strength are present in the diagnosis and reasoning. The results show that the participants can use the information of the alternative reason to judge the cause and effect of different types, and the existence of the weak core reason in the diagnosis and reasoning of the strong core reason can cause the inference result to be reduced (the weak evidence is found). Effect). On the basis of experiment 1, the experiment 2 uses 2 (alternative cause characterization: clear and ambiguous).3 (alternative cause strength:25%,50%,75%) is further explored for the characterization and strength of the alternative causes for cause and effect The result of the experiment is presented to the cause and effect information and the related causal structure in the form of a paper pen. The result of the experiment is P (E/ C), P (C/ E), P (E/-C), P (C/-E). The results show that the representation of the alternative causes does not have an impact on the cause and effect reasoning; the inverse predictive reasoning is sensitive to the strength of the alternative causes; however, in the prediction of the reasoning and the reverse diagnostic reasoning we find alternative causes for characterization and strength The results show that the reasoner can't complete the state of the alternative reason structure, but the strength of the alternative reason is not affected by the reasoning, which may be due to the fact that the participants are not aware of the alternative causes in the condition Is not present, it is not clear that the characterization may result in a reasoning under the condition of a strong alternative. On the basis of the research of experiment 1 and experiment 2, the experiment 3 is based on the research of the experiment 1 and the experiment 2, and the experiment 3, under the condition that the structure is clear, can be substituted for the reason, and the cause and effect of the alternative reason (25%,50%,75%) in the two reasoning conditions are not present The effect of reasoning. Based on the causality and structure information we provide, it is made to P (E/ CA), P (C/ EA), P (E/ C-A), P (C/ E-A), P (E/ A), P (A/ E), P (C/ E-A), P (E/ A), P (A/ E), The results show that in the predictive reasoning, the participants are sensitive to the strength of the alternative causes; in the absence of the alternative causes, the predictive reasoning also appears to be The result shows that human being in the cause and effect reasoning does not consider the effect of alternative causes, and the neglect of the alternative causes is due to the characterization errors occurring in the final probability judgment. The result of this study is to investigate the alternative causes from different alternative causes. Through these three experiments, three important conclusions are drawn: first, the neglect of the alternative causes is not only due to the fact that Second, the neglect of alternative causes is not because of the laziness in the reasoning, that is, only when the possible judgment error is very small The criteria for reasoning are relaxed. Third, the characterization of an alternative cause does not necessarily The results show that when the relevant knowledge is presented in a simple, concrete form, people
【学位授予单位】：西北师范大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：B842

【参考文献】