Applications of Quantum Probability Theory to Human Causal Reasoning

了解人们如何推理因果关系是认知科学和决策的中心主题。过去，许多关于人类因果推理的研究都集中在学习单一因果之间的关系。然而，大多数现实世界的因果结构更为复杂。例如，单个结果可能有多个原因，其中原因本身相互作用。当面对这样复杂的问题时，人们如何做出因果判断呢？有证据表明，人们对复杂因果系统的判断常常偏离经典概率论的规范规则。该项目将探索量子概率在人类因果推理中的应用。量子概率论是经典概率论的非交换模拟。特别是，量子概率可以解释秩序效应，这违反了经典概率论的交换规则。当判断结果的可能性取决于处理原因的顺序时，在因果推理中就会出现顺序效应。先前的工作已经证明，量子概率方法可以解释判断和决策研究中的许多自相矛盾的发现。当前的项目将进一步发展量子概率论，将其应用于因果推理领域。该研究计划的广泛和长期目标是（1）增加人类对复杂因果系统判断的理解；（2）为从量子概率论原理构建人类因果推理模型提供新的基础。人类因果推理是包括发展心理学、决策和学习在内的许多领域的一个重要课题，因此拟议的研究有可能推动一系列学科的知识发展。拟议的研究还具有造福社会的潜力。许多研究人员目前正在研究使用贝叶斯因果网络来帮助决策者（例如情报分析师）处理大量数据。虽然贝叶斯决策工具具有巨大潜力，但仍需要进行研究来理解和改进与这些工具交互的分析师的因果判断。该研究计划通过旨在研究复杂因果推理并通过学习改进因果判断的实验来解决这个主题。拟议的研究还拓展了不断发展的量子认知领域，旨在为社会和行为科学开发量子概率动态系统。

Understanding how people reason about causes and effects is a central topic in cognitive science and decision-making. In the past, much of the research on human causal reasoning has focused upon learning the relationship between a single cause and effect. However, most real world causal structures are more complex. For example, there can be multiple causes for a single effect where the causes themselves interact. When faced with complex problems such as this, how do people make causal judgments? There is evidence that people's judgments about complex causal systems often deviate from the normative rules of classic probability theory. This project will explore the application of quantum probability to human causal reasoning. Quantum probability theory is the noncommutative analog of classical probability theory. In particular, quantum probability can account for order effects, which are violations of the commutative rule of classic probability theory. Order effects occur in causal reasoning when the judged likelihood of an effect depends on the order in which the causes are processed. Previous work has demonstrated that the quantum probability approach can account for many paradoxical findings in judgment and decision research. The current project will further develop the quantum probability account by applying it to the area of causal reasoning.The broad and long term goals of this research program are (1) to increase the understanding of human judgments about complex causal systems and (2) to provide a new foundation for constructing models of human causal reasoning from the principles of quantum probability theory. Human causal reasoning is an important topic in numerous fields including developmental psychology, decision-making, and learning, thus the proposed research has the potential to advance knowledge across a range of disciplines. The proposed research also has the potential to benefit society. Many researchers are currently investigating the use of Bayesian causal networks to help decision-makers, such as intelligence analysts, process large amounts of data. While Bayesian decision tools have great potential, research is needed to understand and improve the causal judgments of analysts who interact with these tools. This research program addresses this topic through experiments designed to investigate complex causal reasoning and to improve causal judgments through learning. The proposed research also adds to the growing field of quantum cognition aimed at developing quantum probabilistic-dynamic systems for social and behavioral sciences.