2.2 Overcoming Cognitive Biases and Engaging in Critical Reflection

Confirmation Bias

One of the most common cognitive biases is confirmation bias, which is the tendency to search for, interpret, favor, and recall information that confirms or supports your prior beliefs. Like all cognitive biases, confirmation bias serves an important function. For instance, one of the most reliable forms of confirmation bias is the belief in our shared reality. Suppose it is raining. When you first hear the patter of raindrops on your roof or window, you may think it is raining. You then look for additional signs to confirm your conclusion, and when you look out the window, you see rain falling and puddles of water accumulating. Most likely, you will not be looking for irrelevant or contradictory information. You will be looking for information that confirms your belief that it is raining. Thus, you can see how confirmation bias—based on the idea that the world does not change dramatically over time—is an important tool for navigating in our environment.

Unfortunately, as with most heuristics, we tend to apply this sort of thinking inappropriately. One example that has recently received a lot of attention is the way in which confirmation bias has increased political polarization. When searching for information on the internet about an event or topic, most people look for information that confirms their prior beliefs rather than what undercuts them. The pervasive presence of social media in our lives is exacerbating the effects of confirmation bias since the computer algorithms used by social media platforms steer people toward content that reinforces their current beliefs and predispositions. These multimedia tools are especially problematic when our beliefs are incorrect (for example, they contradict scientific knowledge) or antisocial (for example, they support violent or illegal behavior). Thus, social media and the internet have created a situation in which confirmation bias can be “turbocharged” in ways that are destructive for society.

Confirmation bias is a result of the brain’s limited ability to process information. Peter Wason (1960) conducted early experiments identifying this kind of bias. He asked subjects to identify the rule that applies to a sequence of numbers—for instance, 2, 4, 8. Subjects were told to generate examples to test their hypothesis. What he found is that once a subject settled on a particular hypothesis, they were much more likely to select examples that confirmed their hypothesis rather than negated it. As a result, they were unable to identify the real rule (any ascending sequence of numbers) and failed to “falsify” their initial assumptions. Falsification is an important tool in the scientist’s toolkit when they are testing hypotheses and is an effective way to avoid confirmation bias.

In philosophy, you will be presented with different arguments on issues, such as the nature of the mind or the best way to act in a given situation. You should take your time to reason through these issues carefully and consider alternative views. What you believe to be the case may be right, but you may also fall into the trap of confirmation bias, seeing confirming evidence as better and more convincing than evidence that calls your beliefs into question.

Anchoring Bias

Confirmation bias is closely related to another bias known as anchoring. Anchoring bias refers to our tendency to rely on initial values, prices, or quantities when estimating the actual value, price, or quantity of something. If you are presented with a quantity, even if that number is clearly arbitrary, you will have a hard discounting it in your subsequent calculations; the initial value “anchors” subsequent estimates. For instance, Tversky and Kahneman (1974) reported an experiment in which subjects were asked to estimate the number of African nations in the United Nations. First, the experimenters spun a wheel of fortune in front of the subjects that produced a random number between 0 and 100. Let’s say the wheel landed on 79. Subjects were asked whether the number of nations was higher or lower than the random number. Subjects were then asked to estimate the real number of nations. Even though the initial anchoring value was random, people in the study found it difficult to deviate far from that number. For subjects receiving an initial value of 10, the median estimate of nations was 25, while for subjects receiving an initial value of 65, the median estimate was 45.

In the same paper, Tversky and Kahneman described the way that anchoring bias interferes with statistical reasoning. In a number of scenarios, subjects made irrational judgments about statistics because of the way the question was phrased (i.e., they were tricked when an anchor was inserted into the question). Instead of expending the cognitive energy needed to solve the statistical problem, subjects were much more likely to “go with their gut,” or think intuitively. That type of reasoning generates anchoring bias. When you do philosophy, you will be confronted with some formal and abstract problems that will challenge you to engage in thinking that feels difficult and unnatural. Resist the urge to latch on to the first thought that jumps into your head, and try to think the problem through with all the cognitive resources at your disposal.

Availability Heuristic

The availability heuristic refers to the tendency to evaluate new information based on the most recent or most easily recalled examples. The availability heuristic occurs when people take easily remembered instances as being more representative than they objectively are (i.e., based on statistical probabilities). In very simple situations, the availability of instances is a good guide to judgments. Suppose you are wondering whether you should plan for rain. It may make sense to anticipate rain if it has been raining a lot in the last few days since weather patterns tend to linger in most climates. More generally, scenarios that are well-known to us, dramatic, recent, or easy to imagine are more available for retrieval from memory. Therefore, if we easily remember an instance or scenario, we may incorrectly think that the chances are high that the scenario will be repeated. For instance, people in the United States estimate the probability of dying by violent crime or terrorism much more highly than they ought to. In fact, these are extremely rare occurrences compared to death by heart disease, cancer, or car accidents. But stories of violent crime and terrorism are prominent in the news media and fiction. Because these vivid stories are dramatic and easily recalled, we have a skewed view of how frequently violent crime occurs.

Tribalism

Another more loosely defined category of cognitive bias is the tendency for human beings to align themselves with groups with whom they share values and practices. The tendency toward tribalism is an evolutionary advantage for social creatures like human beings. By forming groups to share knowledge and distribute work, we are much more likely to survive. Not surprisingly, human beings with pro-social behaviors persist in the population at higher rates than human beings with antisocial tendencies. Pro-social behaviors, however, go beyond wanting to communicate and align ourselves with other human beings; we also tend to see outsiders as a threat. As a result, tribalistic tendencies both reinforce allegiances among in-group members and increase animosity toward out-group members.

Tribal thinking makes it hard for us to objectively evaluate information that either aligns with or contradicts the beliefs held by our group or tribe. This effect can be demonstrated even when in-group membership is not real or is based on some superficial feature of the person—for instance, the way they look or an article of clothing they are wearing. A related bias is called the bandwagon fallacy. The bandwagon fallacy can lead you to conclude that you ought to do something or believe something because many other people do or believe the same thing. While other people can provide guidance, they are not always reliable. Furthermore, just because many people believe something doesn’t make it true. Watch the video below to improve your “tribal literacy” and understand the dangers of this type of thinking.

Sunk Cost Fallacy

Sunk costs refer to the time, energy, money, or other costs that have been paid in the past. These costs are “sunk” because they cannot be recovered. The sunk cost fallacy is thinking that attaches a value to things in which you have already invested resources that is greater than the value those things have today. Human beings have a natural tendency to hang on to whatever they invest in and are loath to give something up even after it has been proven to be a liability. For example, a person may have sunk a lot of money into a business over time, and the business may clearly be failing. Nonetheless, the businessperson will be reluctant to close shop or sell the business because of the time, money, and emotional energy they have spent on the venture. This is the behavior of “throwing good money after bad” by continuing to irrationally invest in something that has lost its worth because of emotional attachment to the failed enterprise. People will engage in this kind of behavior in all kinds of situations and may continue a friendship, a job, or a marriage for the same reason—they don’t want to lose their investment even when they are clearly headed for failure and ought to cut their losses.

A similar type of faulty reasoning leads to the gambler’s fallacy, in which a person reasons that future chance events will be more likely if they have not happened recently. For instance, if I flip a coin many times in a row, I may get a string of heads. But even if I flip several heads in a row, that does not make it more likely I will flip tails on the next coin flip. Each coin flip is statistically independent, and there is an equal chance of turning up heads or tails. The gambler, like the reasoner from sunk costs, is tied to the past when they should be reasoning about the present and future.

There are important social and evolutionary purposes for past-looking thinking. Sunk-cost thinking keeps parents engaged in the growth and development of their children after they are born. Sunk-cost thinking builds loyalty and affection among friends and family. More generally, a commitment to sunk costs encourages us to engage in long-term projects, and this type of thinking has the evolutionary purpose of fostering culture and community. Nevertheless, it is important to periodically reevaluate our investments in both people and things.

In recent ethical scholarship, there is some debate about how to assess the sunk costs of moral decisions. Consider the case of war. Just-war theory dictates that wars may be justified in cases where the harm imposed on the adversary is proportional to the good gained by the act of defense or deterrence. It may be that, at the start of the war, those costs seemed proportional. But after the war has dragged on for some time, it may seem that the objective cannot be obtained without a greater quantity of harm than had been initially imagined. Should the evaluation of whether a war is justified estimate the total amount of harm done or prospective harm that will be done going forward (Lazar 2018)? Such questions do not have easy answers.

Table 2.1 summarizes these common cognitive biases.