In their paper “Judgment Under Uncertainty: Heuristics and Biases” (1974)2, Daniel Kahneman and Amos Tversky identified three different kinds of heuristics: availability, representativeness, and anchoring and adjustment. Each type of heuristic is used for the purpose of reducing the mental effort needed to make a decision, but they occur in different contexts.
The first type of heuristic is the availability heuristic, which was touched upon in the example of judging the frequency with which tornadoes occur in Kansas relative to Nebraska. Kahneman and Tversky define this heuristic as a mental shortcut for making frequency or probability judgments based on “the ease with which instances or occurrences can be brought to mind” (p.1127).3
The availability heuristic occurs because we can call certain memories to mind more easily than others. The example that Kahneman and Tversky give is that participants asked if more words in the English language start with the letter K or have the third letter K, most participants responded with the former. In actuality, it is the latter that is true, but it is much harder to think of words that have K as the third letter than it is to think of words that start with K.4 In this case, our memories of words that begin with K come to mind more readily than do memories of words with the third letter K.
A second type of heuristic is the representativeness heuristic. We often rely on this heuristic when making probability judgments. We tend to classify events into categories, which, as illustrated by Kahneman and Tversky, can result in our use of this heuristic. When we use the representativeness heuristic, we make probability judgments about the likelihood that an object or event arises from some category based on the extent to which the object or event in question is similar to the prototypical example of that category.5 For example, if someone we meet in one of our university lectures looks and acts like a stereotypical medical student, we may judge the probability that they are studying medicine as highly likely, even without any hard evidence to support that assumption.
The representativeness heuristic is associated with prototype theory.6 This prominent theory in cognitive science provides an explanation for object and identity recognition. It suggests that we categorize different objects and identities in our memory. For example, we may have a category for chairs, a category for fish, a category for books, and so on. Prototype theory posits that we develop prototypical examples for these categories by averaging every example of a given category we encounter. As such, our prototype of a chair should be the most average example of a chair possible, based on our experience with that object. This aids in object recognition because we compare every object we encounter against the prototypes stored in our memory for identification. The more the object resembles the prototype, the more confident we are that it belongs in that category. Prototype theory may give rise to the representativeness heuristic, as this heuristic occurs in situations when a certain object or event is viewed as similar to the prototype stored in our memory, which leads us to classify the object or event into the category represented by that prototype. To go back to the previous example, if your peer closely resembles your prototypical example of a med student, you may place them into that category, based on the prototype theory of object and identity recognition. This, however, causes you to commit the representativeness heuristic.
Anchoring and adjustment heuristic
The third type of heuristic put forth by Kahneman and Tversky in their initial paper on the topic is the anchoring and adjustment heuristic.7 This heuristic describes how, when estimating a certain value, we tend to give an initial value, then adjust it by increasing or decreasing our estimation. However, we often get stuck on that initial value – which is referred to as anchoring – which results in us making insufficient adjustments. Thus, our adjusted value is biased in favor of the initial value, which we have anchored on.
In the example given by Kahneman and Tversky, participants were presented with a question, such as “estimate the number of African countries in the United Nations (UN)”. A wheel labeled with numbers from 0-100 was spun, and participants were asked to say whether or not the number the wheel landed on was higher or lower than their answer to the question. Then, participants were asked to give an estimate of the number of African countries in the UN. Kahneman and Tversky found that participants tended to anchor on the random number obtained by spinning the wheel. So, when the number obtained by spinning the wheel was 10, the median estimate given by participants was 25, while, when the number obtained from the wheel was 65, participants’ median estimate was 45.8
A 2006 study by Epley and Gilovich, “The Anchoring and Adjustment Heuristic: Why the Adjustments are Insufficient”9 investigated the causes of this heuristic. They illustrated that anchoring often occurs because the information we anchor on is more accessible than other information because we have just encountered it. Furthermore, they provided empirical evidence to demonstrate that our adjustments tend to be insufficient because adjustments require significant mental effort, which we are not always motivated to dedicate to the task. They also found that providing incentives for accuracy led participants to make more sufficient adjustments. So, this particular heuristic generally occurs when there is no real incentive to provide an accurate response.
Quick and easy
What these types of heuristics have in common is that they all allow us to respond automatically, without much effortful thought. They provide an immediate response and do not use up much of our mental energy, which allows us to dedicate mental resources to other matters, which may be more pressing. In that way, heuristics are efficient, which is a big reason why we continue to use them. That being said, we should be mindful of how much we rely on them because there is no guarantee as to their accuracy.