Coherence-Based Reasoning: Modelling Attention and Information Search

点击翻译按钮获取中文摘要

Humans strive for coherent representations in perception, problem solving, reasoning, and judgment. Thereby perceptions and knowledge are combined in a way to form harmonious interpretations in order to prepare appropriate responses. Parallel constraint satisfaction models (PCS) implement coherence principles in a mathematically precise way. In judgment and decision making research, PCS models have been applied successfully to describe automatic information integration processes and to predict new phenomena such as coherence effects, that is, shifts in the evaluation of information during decision making. PCS-DM has been developed as a fully specified model for adaptive decision making (Glöckner, Hilbig & Jekel, 2014, Cognition), and it has been shown to account for choices, response times and confidence better than competing models. In the current work we aim to extend PCS-DM by adding a specification for coherence-based mechanisms of attention and active information search, hitherto unspecified aspects of the model. Experimental tests will strive at investigating the extended model's ability to predict attention in decisions with directly accessible information as well as active search for new information.