August Collsiöö: Revisiting a Brunswikian View on Intuitive and Analytic Thought: The Cognitive Processes in Multiple-Cue Judgment

Abstract

Intuitive and analytic thought has been discussed since the time of the ancient Greek philosophers. The current thesis reintroduces an operational definition from Egon Brunswik (1956), conceptualizing intuitive processes as approximate, ubiquitously affected by error, and analytic processes as predominantly deterministic with occasional error. The thesis is focused on multiple-cue judgments, a type of judgment where multiple pieces of information (cues) are integrated to make a judgment regarding some distal factor (criterion). These types of judgments occur constantly in everyday life. You judge the potential helpfulness of a new colleague, the ripeness of an avocado you intend to buy, or the reasonable price of an apartment based on multiple information-baring cues. Research on multiple-cue judgment has often investigated whether people approach such tasks with a rule-based or memory-based cognitive process. The current thesis investigates the intuitive or analytic instantiation of rule-based and memory-based processes in multiple-cue judgment tasks to nuance and deepen the understanding of processing in such tasks. In Study I, a computational model is developed to formalize and test the Brunswikian conceptualization of intuitive and analytic thought. The model correctly identifies the parameters of a generating model and differentiates between more or less analytic processes across both perceptual and conceptual tasks. Study II investigates the effects of verbal and numeric information on learning and cognitive processing in a multiple-cue learning task. Across formats, participants primarily approached the task with an analytic process, and verbal information was beneficial for learning in non-additive tasks by inviting memory-based processing. Study III presents a conceptual framework for understanding more intuitive or analytic instantiations of rule-based and memory-based cognitive algorithms. Rule-based algorithms were implemented by either a fully analytic or intuitive process, and the mode of implementation was strongly driven by the nature of the feedback (deterministic vs. probabilistic). In contrast, memory-based algorithms were implemented more along a continuum. Study IV compares the mainstream fast-and-slow dual-process perspective, which associates intuitive processes with high speed and analytic ones with low speed, with the presented Brunswikian approach. Across four multiple-cue learning experiments, the Brunswikian perspective was supported, whereas evidence pointed against the predictions from fast-and-slow dual-process theory. In conclusion, the thesis presents a novel operationalization of intuitive and analytic thought and shows how investigating the noise in cognitive processes can provide information regarding the cognitive processing of an individual. By analyzing the noise in rule-based and memory-based processes, a more nuanced description of these cognitive processes is presented.