Base-rate respect by intuition: Approximating rational choices in base-rate tasks with multiple cues

Abstract

Although intuitive-automatic processes sometimes lead to systematic biases in judgment and choice, in many situations especially this kind of processes enables people to approximate rational choices. In complex base-rate tasks with repeated outcome feedback we observed choices which were in line with the Bayes' solution in 86% of the cases and which were made within a relatively short time (i.e., 2.2 seconds). The results indicate reliance on extremely wellcalibrated intuition. This view is supported by the findings that choice proportions are almost perfectly predicted by posterior probabilities (r = .93), and that error rates, response times and confidence ratings are highly correlated with inconsistency in the provided information. Our results support the hypothesis that parallel constraint satisfaction models may account for the processes underlying intuition and make the application of simple heuristics and deliberate strategies very unlikely. Taking an interdisciplinary perspective, implications for economic and psychological modeling are outlined. Base-rate Neglect and IntuitionTo be successful in a fundamentally uncertain world, individuals have to make a multitude of judgments and decisions based on probabilistic information. Research on judgment and decision making has repeatedly shown systematic biases in probability assessment and in applying the rules of probability (Gilovich, Griffin, & Kahneman, 2002;Kahneman, Slovic, & Tversky, 1982;Reyna, 2004;Slovic, 1987). One of the most prominent and well established findings is individuals' neglect of base-rate probabilities. 1 This has been, for instance, demonstrated in the following task (e.g., Bar-Hillel, 1980): Two cab companies, Blue and Green, operate in a given city. Eighty-five percent of the cabs in the city are Blue, and the remaining 15% are Green. A cab was involved in a hit-and-run accident at night. A witness later identified the cab as a Green cab. The court tested the witness' ability to distinguish between Blue and Green cabs under nighttime visibility conditions. The witness was able to identify each color correctly about 80% of the time, but confused it with the other color about 20% of the time. What is the probability that the accident was caused by the Green company?To answer this question, both the base rate of green cabs (15%) and the validity of the eyewitness (80%) have to be taken into account. However, most people consider only the latter and give the answer that signals neglect of the base rate, namely 80%. There have been several attempts to explain base-rate neglect and to provide solutions to solve the problem (for an overview see Barbey & Sloman, 2007). In line with earlier approaches (e.g., Tversky & Kahneman, 1983), Barbey and Sloman proposed that base-rate neglect is due to associative / intuitive (i.e., system 1) judgment strategies that fail to adequately represent the structure of the problem. Our research empirically tests and expands this intriguing theory. In line with Hammond, Hamm, Grassia, and Pearson (...