empirical decision theory the scientific study of human judgment and decision making. A growing body of empirical research has described the actual limitations on inductive reasoning. By contrast, traditional decision theory is normative; the theory proposes ideal procedures for solving some class of problems.
The descriptive study of decision making was pioneered by figures including Amos Tversky, Daniel Kahneman, Richard Nisbett, and Lee Ross, and their empirical research has documented the limitations and biases of various heuristics, or simple rules of thumb, routinely used in reasoning. The representativeness heuristic is a rule of thumb used to judge probabilities based on the degree to which one class represents (or resembles) another class. For example, we assume that basketball players have a ‘hot hand’ during a particular game – producing an uninterrupted string of successful shots – because we underestimate the relative frequency with which such successful runs occur in the entire population of that player’s record. The availability heuristic is a rule of thumb that uses the ease with which an instance comes to mind as an index of the probability of an event. Such a rule is unreliable when salience in memory misleads; for example, most people (incorrectly) rate death by shark attack as more probable than death by falling airplane parts. (For an overview, see D. Kahneman, P. Slovic, and A. Tversky, eds., Judgment Under Uncertainty: Heuristics and Biases, 1982.) These biases, found in laypeople and statistical experts alike, have a natural explanation on accounts such as Herbert Simon’s (1957) concept of ‘bounded rationality.’ According to this view, the limitations on our decision making are fixed in part by specific features of our psychological architecture. This architecture places constraints on such factors as processing speed and information capacity, and this in turn produces predictable, systematic errors in performance. Thus, rather than proposing highly idealized rules appropriate to an omniscient Laplacean genius – more characteristic of traditional normative approaches to decision theory – empirical decision theory attempts to formulate a descriptively accurate, and thus psychologically realistic, account of rationality. Even if certain simple rules can, in particular settings, outperform other strategies, it is still important to understand the causes of the systematic errors we make on tasks perfectly representative of routine decision making. Once the context is specified, empirical decision-making research allows us to study both descriptive decision rules that we follow spontaneously and normative rules that we ought to follow upon reflection. See also BAYESIAN RATIONALITY, DECISION THEORY, HEURISTIC. J.D.T.