What Are Cognitive Biases?

Cognitive biases are systematic errors we make, rooted in the very principles of how our cognition works. When we talk about cognitive biases, we assume that people are: a) irrational by nature (or at least only partially rational); and b) bound to make systematic mistakes in predictable ways, since each bias is linked to a specific mechanism of our thinking.

The Concept of Cognitive Biases

The first studies of cognitive biases are associated with Daniel Kahneman and Amos Tversky. Initially, they believed cognitive biases mainly affected decision-making and certain other mental processes, such as reasoning and judgment. In a classic example, Kahneman asked two groups to estimate the percentage of African countries in the UN. Before answering, the first group was asked, “Is the percentage of African countries in the UN more or less than 65%?” while the second group was asked the same question but with 10% instead of 65%. As a result, the first group’s estimates were much higher than the second’s. This demonstrates the anchoring heuristic, which can lead to a whole class of similar cognitive biases.

Over time, the term “cognitive bias” has been used to describe a wide range of phenomena, including visual illusions and memory errors, even though these are not truly cognitive biases since they are not related to rationality. For example, many visual illusions shouldn’t be considered cognitive biases because the systematic errors occur at the stage of forming a perceptual image, not at the stage of making a judgment. In contrast, the original cognitive biases described by Kahneman are linked to the mechanism of making judgments, regardless of the content being evaluated.

There aren’t many studies on the ontogenetic aspect—how cognitive biases relate to the development of cognitive processes. However, there is some evidence that cognitive biases may be innate or appear at a very young age. Some support for this comes from the work of Jean Piaget, who described the “A-not-B error.” In this experiment, a toy is repeatedly hidden in container A in front of a child under one year old, and the child reaches for it. After several repetitions, the toy is hidden in container B, but the child often still reaches for container A. This error is similar to a cognitive bias, as it directly results from how human thinking works, though it disappears with age. There is currently no convincing evidence that cognitive biases are formed only through life experience.

Classification of Cognitive Biases

There have been attempts to group cognitive biases or create lists to systematize them. However, most such attempts suffer from incompleteness and redundancy due to the lack of a clear definition of what a cognitive bias is. Generally, it’s useful to distinguish cognitive biases in three types of processes: decision-making, reasoning, and judgment. Within these categories, individual biases can be identified based on which components of these processes they affect.

For example, in decision-making, you first need to imagine alternatives in a given situation and then assess their subjective expected utility. Thus, cognitive biases can be divided into those that distort the representation of the choice situation (what alternatives we see, their number and composition) and those that distort the evaluation of the subjective value of alternatives (which ones seem most beneficial). There’s also a third class related to estimating the probability of outcomes. For instance, the gambler’s fallacy is when someone, after losing 10 times in a row at slot machines, believes their chance of winning the 11th time is higher, even though the outcomes are independent.

Regarding the second class, which involves biases in evaluating attractiveness, there’s the “IKEA effect.” Studies have shown that the more effort we put into creating, modifying, or assembling an object, the more we like it. For example, the value of an IKEA cabinet depends not only on the money spent but also on the effort required to assemble it.

It’s also important to understand that decision-making processes differ across subject areas. Classic examples often relate to economics and resource allocation, while others involve medicine, such as choosing a treatment plan. Cognitive biases in economics and medicine don’t always work the same way. Initially, it was believed that cognitive biases were universal and independent of the area or content of our thinking. However, it’s clear that decisions about a patient’s life differ from economic or everyday decisions.

Approaches to Studying Cognitive Biases

To simplify, classical economic theory assumed people are rational agents. In contrast, Kahneman and his followers believed people are mostly irrational and prone to cognitive biases. Modern perspectives recognize that not everyone exhibits cognitive biases to the same degree; people differ in their susceptibility to biases and, therefore, in their rationality.

The traditional approach is to identify more of these systematic errors and make people aware of them. Since cognitive biases are mostly involuntary, simply being aware of them can reduce the likelihood of negative consequences.

Kahneman divided our thinking into System 1 and System 2. System 1 is impulsive and responsible for most cognitive biases, while System 2 is slow and allows us to consider more alternatives. This is a simplification, but it illustrates that part of our cognition works automatically and can lead to systematic errors, while another part helps us manage these errors. However, people systematically differ in their susceptibility to cognitive biases. Studying individual differences in cognitive biases is a current trend in this field. Some people are more prone to certain biases than others, and not to all biases equally. The main idea is that people are predictably different, and there are patterns that determine who is prone to which biases, though many questions remain.

An alternative approach distinguishes not between two thinking systems (automatic and controlled), but between “hot” and “cold” cognition. Hot cognition refers to phenomena involving emotions and cognitive processes, while cold cognition involves processes largely unrelated to emotions. Accordingly, cognitive biases can be divided into those influenced by emotional modulation and those that are not.

For example, the asymmetry of emotional valence effect shows that when making a logical conclusion, premises outside the immediate reasoning can influence the outcome. If you’re told, “If I get my dream job, I’ll be happy,” and “I got my dream job,” you’d logically answer “yes” to “Will I be happy?” But if you add, “But I lost a loved one,” most people would answer “no,” even though this premise shouldn’t logically affect the conclusion. Interestingly, this only works one way: if the premises are negative, adding a positive one doesn’t change the answer. This bias shows that the emotional tone of information we use in reasoning can directly affect our conclusions.

Classical economic theory assumed decision-makers are rational and that emotions are a main source of systematic errors. For psychologists, this is odd, since many approaches in the psychology of emotions note that emotions evolved for survival and can sometimes enhance rational decision-making, though very intense emotions can disrupt our thinking.

Cognitive Biases in Advertising

Cognitive biases also appear in advertising. However, what’s often called a cognitive bias in advertising may actually be an example of how our attention works. For example, where a character looks on a poster can influence where viewers focus their attention. Attention and memory aren’t directly related to rationality, but there are techniques in merchandising, like product placement in stores, that influence customer choices. Classic studies by Richard Nisbett in the 1960s showed that the order of clothing items on a rack affects what customers choose, with items on the right being chosen more often if arranged left to right. Yet, customers are unaware of how they made their choice.

Many such techniques aren’t true cognitive biases. Still, applied psychology in advertising predates discussions about human rationality and cognitive biases. Some patterns do apply to advertising, especially in decision-making processes.

Research on cognitive dissonance by Leon Festinger and his students shows, for example, that car owners pay more attention to ads for the brand they bought after the purchase, not before. This is because they need to reassure themselves about their decision. So, they notice ads for their own car more often, even though the ads are meant for potential buyers. These studies on cognitive dissonance aren’t directly about cognitive biases. Festinger studied how we try to avoid unpleasant situations and convince ourselves to minimize the discomfort (dissonance) between our beliefs and observations. However, some researchers link these phenomena to specific cognitive biases, such as confirmation bias—the tendency to seek information that confirms our beliefs rather than contradicts them.

Cognitive Biases in the Legal System

Another important area where cognitive biases appear is the legal system, since jury or judge verdicts are examples of decision-making. First, there’s the question of whether the verdict will be guilty or not, and second, what punishment will be imposed. Studies often use vignettes—descriptions of different legal cases—to see how various details affect decisions. Courts are supposed to be fair, meaning similar cases should have similar outcomes regardless of wording or irrelevant details like age or gender. However, we systematically see that such factors influence verdicts.

Recently, more tests have been developed to measure susceptibility to different cognitive biases, aiming to put them on par with intelligence tests. For example, framing tests show how different wordings and descriptions affect choices. Another example is the sunk cost fallacy, where people continue investing in a losing project simply because they’ve already invested a lot. However, it’s important to understand that these results are statistical correlations between certain groups and cognitive biases, and the nature of these relationships is still being studied.

Cognitive biases are not a disease or a deviation from the norm, but since people differ in their susceptibility to various biases, there are no universal “recipes” or training methods. Instead, we need to develop and use different ways to organize our environment—both physical and social—so that it doesn’t encourage cognitive biases and increases the chances of successful outcomes in both everyday life and professional activities.