Representativeness Heuristic

Heuristics and Biases

The representativeness heuristic is a theoretical construct that forms part of the influential heuristics and biases framework for explaining intuitive judgment in humans pioneered by Daniel Kahneman and Amos Tversky in the 1970s. From this perspective, because of their limited time, knowledge, and computational ability, in general people cannot make judgments according to the often complex normative rules of logic, probability theory, and statistics. Instead, they have to resort to simpler judgment heuristics that exploit natural assessments. Natural assessments benefit from naturally existing propensities of memory and perception that are conveniently available and easily assessed. These judgmental heuristics are often useful, but sometimes they lead to systematic and serious cognitive biases.

In regard to confidence (probability) judgment, the key notion has been variable substitution, according to which the complex judgment of probability is substituted with a simpler natural assessment. Assessment of probability in the sense implied by probability theory and statistics involves consideration of all possible outcomes and their frequencies of occurrence. In the face of this complexity it is proposed that people substitute probability with a subjective variable that is conveniently available and easier to assess. The representativeness heuristic suggests that the degree to which an instance or event is representative of a category is used as a proxy for the probability that the instance or event belongs to the category, as when you assess the probability that a person is a lawyer by assessing how similar he or she is to your stereotype for a typical lawyer.

Although representativeness is a useful guide to probability in many real-life circumstances, because it does not obey the rules of probability theory, use of the heuristic is claimed to produce a number of biases, or cognitive illusions, in probability judgment. Two classical demonstrations of such biases are base-rate neglect and the conjunction fallacy.

Base-Rate Neglect

If you are in Florida and encounter a male with a short haircut driving a Chevrolet Corvette, you may consider the probability that he is an astronaut (assuming that a clean-cut male driving a sports car close to Cape Canaveral fits your stereotype for an astronaut). In this case, probability theory implies that you should take into account the base rate of astronauts, which presumably is low also in Florida, and the modest reliability of hairstyle and car make as predictors of profession (per the celebrated Bayes's theorem of probability theory). Because people use the representativeness heuristic when they assess the probability that the person is an astronaut, it is proposed that they only take into account the similarity between the person and their stereotype for an astronaut, while neglecting the base-rate of astronauts. Because of this base-rate neglect, they get too [Page 658]captivated by the representativeness of the evidence. Likewise, even physicians tend to be too captivated by a positive result of a medical diagnosis test (e.g., for HIV), neglecting the often low prevalence (base-rate) of the disease in the population, thereby overestimating the probability that the patient actually has the disease.

...