Year: 2007
Type: article
Abstract: Cognitive Decision Theory: Developing Models of Real-World Decision Behavior Timothy J. Pleskac ([email protected]) Michael R. Dougherty ([email protected]) Indiana University, 1101 E. Tenth St. Bloomington, IN 47405 USA University of Maryland, 1145 Biology-Psychology Bldg. College Park, MD 20742 USA Jerome Busemeyer ([email protected]) Jorg Rieskamp ([email protected]) Indiana University, 1101 E. Tenth St. Bloomington, IN 47405 USA Max Planck Institute for Human Development, Lentzeallee 94, 14195 Berlin, Germany Josh Tenenbaum ([email protected]) Massachusetts Institute of Technology, 77 Massachusetts Ave. Cambridge, MA 02139 USA Keywords: Cognitive modeling; judgment and decision making; uncertainty; Bayesian; risk taking; real world (Libby, 1985), and mechanics’ diagnoses of auto failure on the basis of symptoms (Mehle, 1982). In all of these cases, the generation of diagnostic hypotheses serves as the lynchpin for inductive inference, for evaluating the likelihood of various hypotheses, and for searching for information in the environment to test hypotheses. Although a great deal is known about how people assess the probability of particular hypotheses and how they search for information to test hypotheses, much less is known about the role of hypothesis generation in these processes. In this talk, I will present an overview of a new cognitive theory of inductive inference, HyGene (Thomas, Dougherty, Sprenger, & Harbison, 2007), and illustrate the implications of hypothesis generation both for understanding judgment, and for understanding basic memory processes. For 300 years, the areas of economics, statistics, and probability theory, have held the decision sciences captive with the likes of Bernoulli, Laplace, Von Neumann, Savage and others, developing normative or rational models of decision making. A weakness of this approach, however, is that it disregards the cognitive capabilities of the agent and only considers the environment in which behavior takes place. In contrast, cognitive scientists have often failed to recognize that the processes they're trying to model take place in an uncertain world involving serious consequences. Thus, cognition in the real world needs to confront decision- making topics. In this symposium, we survey several recent examples of the emerging field we call Cognitive Decision Theory – a branch of decision theory in which models of judgment and decision making are built from principles of cognition rather than axioms of rationality. As the talks in the symposium illustrate, the application of cognitive models to understanding judgment and decision making addresses the weaknesses of both cognitive and decision science. On the one hand, models of decision making become more plausible with the realization that the cognitive system places constraints on the types of processes (or heuristics) that are assumed to guide actual decision behavior. On the other hand, models of cognition are challenged to re- conceptualize how their laboratory-derivations can be scaled up to account for cognition in the real world. The Adaptive Decision Maker: Selecting Strategies According to the Environment Jorg Rieskamp How do people make probabilistic inferences, such as inferring which of two patients needs more urgent treatment? Bayesian models have been proposed as normative benchmarks for such tasks. However, it has been demonstrated that for many real-world inference tasks, simple heuristics can perform astonishingly well and sometimes even outperform Bayesian models. If people behave adaptively, it can be predicted that they will select simple heuristics in environments where the heuristics outperform cognitively more demanding strategies. To describe the adaptive process of strategy selection, a computational model will be proposed that specifies how people select strategies. The strategy selection learning theory predicts a strategy selection process on the basis of reinforcement learning. The theory was supported in several experimental studies and makes better predictions than alternative accounts. Diagnostic Hypothesis Generation & Human Judgment Michael Dougherty* & Rick Thomas Inductive inference is a fundamental component of everyday reasoning, both by lay people and by professionals. Indeed, hypothesis generation processes occur in any task that involves taking “data” and formulating possible explanations of that data, including clinicians’ generation of diagnoses on the basis of symptoms (e.g.,Botti & Reeve, 2003), auditors’ diagnosis of sources of accounting errors
Language: English
Subfield: General Decision Sciences
Field: Decision Sciences
Domain: Social Sciences
Open Access status: green
Award: Modeling in Cognition
OpenAlex ID: W2767721654