Rational behavior in decision making. A comparison between humans, computers and fast and frugal strategies

Rational behavior in decision making. A comparison between humans, computers, and fast and frugal strategies Chris Snijders and Frits Tazelaar (Eindhoven University of Technology, The Netherlands) Real life decisions often have to be made in "noisy" circumstances: not all crucial information is available, some of the information that is available is redundant, and it is usually far from obvious how to sensibly combine the useful information that is there. Under such circum- stances, the average human is not likely to make correct decisions. There are several theories or ideas as to how one might make good decisions, even under circumstances that are noisy. One idea is that, typically, humans acquire experience over time. Through experience people learn, perhaps even without being aware of it, which bits of information are important and how to combine them. A second stream of thought suggests that in such noisy circum- stances, it is better to use computer models to make a decision (this idea has been popularized by Meehl). Computer models are insensitive to useless information and are able, given enough input from previous similar decisions, to accurately infer what the best decision under particular circum- stances is. A third stream of thought suggests that humans tend to use simple decision- strategies to deal with complex and noisy information, and moreover that these strategies are remarkably accurate (Gigerenzer's "fast and frugal strategies"). We set out to compare these three kinds of decision making: experts, computer models, and fast and frugal strategies, in a set of seven experiments based on real-world data regarding purchasing decisions in organizations.