What are simple thinking strategies that allow us to solve problems and make judgments efficiently?

III.A. Heuristics for the Judgment of Probability and Frequencies: Availability, Representativeness, and Anchoring and Adjustment

The heuristics most widely studied within psychology are those that people use to make judgments or estimates of probabilities and frequencies in situations of uncertainty (i.e., in situations in which people lack exact knowledge). Most prominent among these are the availability, representativeness, and anchoring and adjustment heuristics.

The availability heuristic leads one to assess the frequency of a class or the probability of an event by the number of instances or occurrences that can be brought to mind or by how easy it seems to call up those instances. For instance, which class of words is more common: seven-letter English words of the form “_ _ _ _ _ n _” or the form “_ _ _ _ i n g”? According to the availability heuristic, to estimate the frequency of occurrences people draw a sample of the events in question from memory. Specifically, for this case they retrieve words ending in –ing (e.g., “jumping”) and retrieve words with “n” in the sixth position (e.g., “raisins”) and then count the number of words retrieved in some period or assess the ease with which such words could be retrieved. They then answer that the more numerous or easier class of words is more common. Because people find it easier to think of words ending with –ing than to think of words with the letter “n” in the next-to-last position, they usually estimate the class “_ _ _ _ i n g” to be more common. This judgment, however, is wrong because all words ending with –ing also have “n” in the sixth position; in addition, there are seven-letter words with “n” the sixth position that do not end in –ing.

The availability heuristic has been suggested to underlie diverse judgment errors, ranging from the tendency to overestimate how many people die from some specific causes of death (e.g., tornado) and underestimate the death toll of others causes (e.g., diabetes) to why people's answers to life satisfaction questions (“How happy are you?”) may be overly influenced by events that are especially memorable.

The representativeness heuristic has been proposed as a means to assess the probability that an object A belongs to a class B (e.g., that a person described as meek is a pilot) or that an event A is generated by a process B (e.g., that the sequence HTHTHT was generated by randomly throwing a fair coin). This heuristic produces probability judgments according to the extent that object A is representative of or similar to the class or process B (e.g., meekness is not representative of pilots, so a meek person is judged as having a low probability of being a pilot). This heuristic can lead to errors because similarity or representativeness judgments are not always influenced by factors that should affect judgments of probability, such as base rates. The representativeness heuristic has also been evoked to explain numerous judgment phenomena, including “hot hand” observations in basketball (the belief that a player is more likely to score again after he or she already scored successfully than after missing a shot) and the gambler's fallacy (the belief that a successful outcome is due after a run of bad luck).

Another heuristic, anchoring and adjustment, produces estimates of quantities by starting with a particular value (the anchor) and adjusting upward or downward from it. For instance, people asked to quickly estimate the product of either 8×7×6×5×4×3×2×1 or 1×2×3×4×5×6×7×8 give a higher value in the former case. According to the anchoring and adjustment heuristic, this happens because the first few numbers presented are multiplied together to create a higher or lower anchor, which is then adjusted upwards in both cases, yielding a higher final estimate for the first product.

Although it has been pointed out that availability, representativeness, and anchoring and adjustment are quite useful heuristics (because they often lead to good judgments without much time or mental effort), most of the large body of evidence amassed that is consistent with the use of these heuristics comes from studies showing where they break down and lead to cognitive illusions or biases (i.e., deviations from some normative standards). This heuristics-and-biases research program has caught the attention of numerous social scientists, including economists and legal scholars. There are good reasons for this attention, since systematic biases question the empirical validity of classic rational choice models (i.e., models of unbounded rationality) and may have important economic, legal, and other implications.

However, the exclusive focus on cognitive illusions has evoked the criticism that research in the heuristics-and-biases tradition equates the notion of bounded rationality with human irrationality and portrays the human mind in an overly negative light, with some researchers even arguing that cognitive illusions are the rule rather than the exception. It has also been criticized that, to date, the cognitive heuristics posited have not been precisely formalized such that one could either simulate or mathematically analyze their behavior, leaving them free to account for all kinds of experimental performance in a post hoc fashion. For instance, it is still an open question of how people assess similarity to make probability judgments with the representativeness heuristic or how many items (e.g., words ending with –ing) the availability heuristic retrieves before it affords a frequency estimate of a class of object (albeit theoretical progress has been made, for instance, by testing whether availability works in terms of ease of recall or number of items recalled). Moreover, the heuristics-and-biases program focuses on human computational capabilities (the first blade of Simon's scissors), largely ignoring the role of the environment by not specifying how such heuristics capitalize on information structure to make inferences. Finally, this program appears to consider heuristics as dispensable mechanisms (that would not be needed if people had the right tools of probability and logic to call on), in contrast to Simon's view of indispensable heuristics as the only available tools for solving many real-world problems.

Kahneman and Tversky have countered some of this critique by drawing a parallel between their heuristic principles and the qualitative principles of Gestalt psychology—the latter being still valuable despite not being precisely specified. Irrespective of the various criticisms, the heuristic and biases program has undoubtedly led to a tremendous amount of research into the idea that people rely on cognitive heuristics made up of simple psychological processes rather than on complex procedures to make inferences about an uncertain world. As a result, this insight has been firmly established as a central topic of psychology.

Introduction

Heuristics

Heuristics are experience-based techniques for generating inferences, such as that do not rely on formalized reasoning. It could be compared to a type of human inference skill that is a ‘shortcut,’ instead of rigorous analytical reasoning, that makes complex judgments easier. One example of a heuristic is the familiarity heuristic: If asked which city, Munich or Dusseldorf, has the greater population, those who do not know the actual population values will answer ‘Munich’ because they have heard of Munich more often; they would be correct (Munich has about 1.3 million people, while Dusseldorf has about 600 000). The impact of heuristics on the evaluation of evidence has been studied by Goldman. Goldman points out that jurors typically use heuristics to evaluate the elements of evidence presented to them. Goldman considers heuristics an alternative to Bayesian inference.

Heuristics are approximate strategies or ‘rules of thumb’ for decision making and problem solving that do not guarantee a correct solution but that typically yield a reasonable solution or bring one closer to hand. As such, they stand in contrast to algorithms that will produce a correct solution given complete and correct inputs. More specifically, heuristics are usually thought of as shortcuts that allow decisions or solutions to be reached more rapidly and in conditions of incomplete or uncertain information—often because they do not process all the available information. Decision heuristics have been studied in different research traditions, primarily one that has focused on when and where verbally described heuristics can break down and yield biases, that is, deviations from classical norms of rationality, and another that has investigated how specific computationally modeled heuristicscan exploit structured information to yield fast and accurate decisions. Heuristics proposed for probability judgments include representativeness, availability, and anchoring-and-adjustment; for choices between alternatives, heuristics include recognition, one-reason decision making, and cue tallying; and for sequential search across alternatives, satisficing (searching with an aspiration level) is a common heuristic approach.

Heuristics

Heuristics are usually contrasted with ‘algorithms’ in problem-solving. Solving a problem by an algorithm or fail-safe rule is supposed to yield an exact, reliable solution that works for every case. ‘Solving’ it by heuristics – by an unintegrated and incomplete set of suggestive ‘rules of thumb’ that work in some cases, but not in all, and not for fully understood or unified reasons – is just as uncreative as solving it by algorithm. However, many people have noticed that heuristic procedures (such as sampling many special cases by trial and error) occasionally lead to insights, sometimes through inductive generalization and analogy with cases in which heuristics succeed, and sometimes because of the stimulus provided by cases in which heuristics (or even algorithms) fail (see the discussion of anomalies, below).

2.2.1 Heuristics

A heuristic is an experiential guide to problem solving that may otherwise be referred to as a mental shortcut. According to Aickelen and Clark (2011), heuristics operate whereby on the basis of experience or judgment they may be more reliable in producing a good solution, although there is no guarantee that the solution will be optimum. These heuristics can take many forms, and their actual utility will be dictated by the situation at hand, the questions being asked of that situation (Is the crime staged? Are these cases linked?), and the type of heuristic used. In this last instance, the quality of the output through use of the heuristic may be heavily influenced by a variety of factors, such as experience with the type of case currently being analyzed. Problems encountered may not be the fault of the heuristic per se, but rather the divergence of the correct statistical approach where serious errors of inference are made (Nisbett, Krantz, Jepson, & Kunda, 1983).