Chapter 12: Seeing What We Expect to See

The Neighbour Who Did Not Wave

Gabriel is forty-eight. A few months ago, a new neighbour moved into the flat across the corridor: a woman of about his age who lives alone and leaves early every morning. On the second day after she arrived, Gabriel passed her on the stairs and said a cheerful good morning. She looked at him briefly, nodded and kept walking without a word.

Gabriel was hurt. "Not very friendly," he said to his wife that evening. After that, he began to notice things. When the neighbour came home, she closed her door quickly. When there was a meeting of the residents, she did not come. When Gabriel saw her in the shop, she seemed to look away. Each small event confirmed what he already thought. "She really does not want anything to do with us," he said.

Then, one evening, Gabriel's wife came home with news. She had spoken to the neighbour properly for the first time, at the bus stop. The neighbour worked night shifts at a care home and often came home exhausted. On the morning Gabriel had greeted her, she had just been told that her mother was seriously ill. She had not come to the residents' meeting because she was at work. And she had been wondering, for months, why the man across the corridor never smiled at her.

Gabriel felt ashamed and also puzzled. He had not invented anything. Every event he noticed really happened. But he had noticed only the events that fitted his first impression, interpreted unclear events in the same direction and forgotten anything that did not fit. He had seen exactly what he expected to see. This chapter is about three closely related patterns that shape what we notice and believe: confirmation bias, the availability heuristic and the halo effect.

Confirmation Bias

Confirmation bias is the tendency to notice, look for, interpret and remember information in ways that fit what we already believe. It is one of the best known and most widely observed patterns in human thinking, and it affects people of every background, level of education and type of work.

Confirmation bias works in at least three ways. The first is in what we look for. When we want to check a belief, we tend to look for evidence that would support it, rather than evidence that might show it is wrong. The second is in how we interpret what we find. When evidence is unclear, we tend to read it in the direction of our existing view. A neighbour who closes a door quickly can be read as unfriendly or as tired. The third is in what we remember. Events that fit our beliefs are easier to recall later, while events that do not fit tend to fade.

In a well-known kind of experiment from the 1960s, people were shown a short series of three numbers, such as two, four and six, and told that the series followed a hidden rule. Their task was to discover the rule by suggesting other series of three numbers. After each suggestion, they were told whether it followed the rule or not. Many people quickly formed a guess, such as "numbers that go up by two", and then tested it only with series that fitted their guess, such as eight, ten and twelve. Each time they were told yes, and they became more confident. In fact, the hidden rule was much broader: any three numbers in rising order. People who tested series that might break their own guess, such as one, two and three, or three, two and one, discovered the real rule much faster.

This simple task shows something important. Looking only for confirming examples can make us feel more and more certain while learning very little. The most useful test of a belief is often the one that could prove it wrong.

Confirmation bias also shapes the questions we ask other people. A parent who suspects that a child has not done their homework may ask, "You have not done your homework, have you?" A manager who suspects a problem with a project may ask only about the things that could be going wrong. The way a question is asked often invites the answer we expect. Asking an open question, such as "How is your homework going?" or "What is going well and what is difficult?", gives the other person room to tell us something we did not expect.

Why We Protect Our Beliefs

Why should our minds work like this? One reason is efficiency. Checking every belief against every possible piece of evidence would be impossible. Once we have formed a reasonable view, it makes sense to rely on it and to need strong evidence before changing it. A mind that changed its beliefs with every new scrap of information would be confused and unstable.

A second reason is that our beliefs are often connected to how we see ourselves and to the groups we belong to. Changing a belief about politics, religion, our family or our work can feel like a threat to who we are or to our relationships with others. So we are understandably cautious about evidence that challenges such beliefs.

In one well-known kind of study, researchers chose a social question on which people held strong opposing views. They gave people on both sides the same set of mixed evidence: some findings that seemed to support one side and some that seemed to support the other. Afterwards, many people on both sides reported feeling more sure of their original view. Each side had found the supporting evidence convincing and the opposing evidence weak. The same information pushed them further apart.

It is worth adding an honest note here. For some years, a stronger claim was popular: that correcting people's false beliefs often makes them believe the false thing even more strongly, a so-called backfire effect. Later, larger studies found that this kind of backfire is rare. In most studies, clear corrections moved people's beliefs at least somewhat toward the facts, even when the topic was emotional. So people are not hopelessly closed to evidence. They are cautious, and they need it presented clearly and respectfully, but they can and often do change their minds.

The Availability Heuristic

The second pattern in this chapter concerns how we judge how common or likely something is. Often, we do this by asking ourselves, without noticing, how easily examples come to mind. If examples come easily, we judge the thing to be common. If they do not, we judge it to be rare. This mental shortcut is called the availability heuristic, because it relies on how available examples are in our memory.

Much of the time, this shortcut works well. Things that happen often usually are easier to remember, because we meet them more. But some things are easy to remember for other reasons: because they are dramatic, frightening, recent, personal or heavily reported in the news. Those things can seem much more common than they really are.

A classic illustration involves a list of names. In one study, people listened to a recorded list of names of real people. In some versions, the people of one gender on the list were very famous, while the people of the other gender were much less well known, although the list contained roughly as many names of each, or even slightly fewer of the famous group. Afterwards, many listeners judged that the list had contained more names of the gender with the famous people. The famous names were simply easier to recall, and what was easy to recall seemed more common. Other studies suggest that even the feeling of ease matters: people asked to recall only a few examples of their own assertive behaviour later rated themselves as more assertive than people asked to recall many examples, a task that had felt hard.

In daily life, the availability heuristic affects how we judge risks. A dramatic accident that is reported widely and shown again and again can make a whole type of travel seem far more dangerous than it is, while common everyday risks that rarely make the news, such as falls at home, can seem small. A friend's story about a terrible experience with a product can outweigh a large amount of calmer information about many people who were satisfied. Vivid stories stick, and what sticks seems common.

The availability heuristic also affects how we judge ourselves and others. In couples and teams, each person tends to remember their own contributions more easily than other people's, simply because they were there for all of their own efforts and only some of other people's. As a result, people often feel that they do more than their share. When each member of a household is asked privately what share of the housework they do, the shares can add up to much more than the whole. Each person is not lying. Each is simply remembering what is most available: their own work.

The Halo Effect

The third pattern is the halo effect. In old paintings, holy figures are often shown with a circle of light, a halo, around their heads. The halo effect is the tendency for one good quality in a person or thing to cast a glow over everything else, so that we see other good qualities that may not be there. It also works in reverse: one bad quality can cast a shadow that makes everything else look worse.

Studies have found that people who are seen as physically attractive are often also judged, on little or no evidence, to be more intelligent, kind or competent. A speaker who is confident and fluent may be believed to know more than a quieter speaker who knows the subject better. A product with beautiful packaging may be assumed to work better. A student who did well in one subject may be expected to do well in all of them, and a student who once caused trouble may be watched more closely for years.

First impressions are especially powerful because of the halo effect and confirmation bias working together. A first impression creates a halo or a shadow, and then confirmation bias leads us to notice and remember what fits it. This is exactly what happened to Gabriel. One unfriendly moment on the stairs created a shadow, and every later event was seen through it.

The halo effect is a serious concern in situations where fairness matters, such as hiring, grading and judging competitions. Research on job interviews has found that unstructured conversations, where the interviewer simply chats and forms an overall impression, are less reliable at predicting how well people will do in the job than structured interviews, where every candidate is asked the same questions and each answer is scored separately against clear standards. Structure does not remove the halo, but it gives each quality its own chance to be judged on its own.

Three Lives and What They Noticed

Precious is thirty-four and has a trip coming up that involves a long flight. A few weeks before she travels, there is a serious air accident that is reported for days, with many pictures and personal stories. Precious finds herself unable to stop thinking about it, and she considers cancelling her trip. She talks to her brother, who suggests that she look for calm, reliable public information about the safety of different ways of travelling. She learns that, by the usual measures, regular passenger flights are among the safest ways to travel, and that over the same distance, the car journeys she makes without a second thought carry more risk. Her fear does not vanish, because fear rarely obeys facts completely. But she understands that the vividness of the news, not the real level of danger, has been driving her feelings. She takes the trip.

Viktor is fifty-three and teaches mathematics. At the start of the year, a new student, Amina, answers his first three questions brilliantly. For the rest of the term, Viktor finds himself expecting excellent work from her, and when she makes mistakes, he assumes they are careless slips. Another student, Oscar, made a rude joke on the first day. Viktor finds himself reading Oscar's later work with suspicion. When Viktor marks the mid-term tests, he tries a new method that a colleague recommended: he covers the names before marking. To his surprise, Oscar's test is one of the best and Amina's has some real misunderstandings that he had not noticed. Viktor realizes that he had been marking his impressions of the students, not just their work. He keeps marking without names and makes a point of speaking with Oscar about his good results.

Leticia is forty-one and has to hire a new person for her small office. In the past, she has relied on her instinct in interviews, and some of her choices worked out well while others did not. This time, she writes down the four skills the job really needs. For each skill, she prepares one question and a short description of what a weak, middle and strong answer would look like. She asks every candidate the same questions in the same order and writes a score for each answer straight away, before moving to the next. One candidate who is charming and confident turns out to score in the middle on most skills. A quieter candidate who seemed less impressive at first scores highly on all four. Leticia hires the quieter candidate, who proves to be excellent.

None of these three people was careless. Precious was frightened for understandable reasons, Viktor wanted to be fair and Leticia wanted to choose well. What helped each of them was a method that separated their first impression or their most vivid memory from the full set of facts: calm information, hidden names and a structured set of questions.

Changing Your Mind Well

If confirmation bias makes it hard to change our minds, then one of the most valuable skills in decision making is the ability to change our minds well: not too easily, at every new rumour, and not too stubbornly, against strong evidence.

One helpful habit is to ask yourself, in advance, what would change your mind. If you believe that a certain plan will work, ask: what result, if I saw it, would make me think the plan is failing? If you believe that a colleague is unreliable, ask: what would I need to see to think otherwise? Writing down the answer before you look at the evidence makes it harder for confirmation bias to change the test after the results are in.

Another helpful habit is to state the opposing view in its strongest form. Instead of picturing the weakest version of an argument you disagree with, try to describe the best version, the one that a thoughtful, well-informed person on the other side would give. If you cannot do this, you may not yet understand the question well enough to be sure of your own view.

A third habit is to notice the feeling of being challenged. When we meet evidence against our views, we often feel a small sting of discomfort, followed by an urge to find what is wrong with the evidence. That urge is natural. But it can be useful to pause and ask: would I examine this evidence so critically if it supported my view? If not, I may be holding the two sides to different standards.

Finally, it helps to treat beliefs a little like tools rather than like possessions. A tool is something we use because it works, and we replace it when we find a better one. A possession is something we defend because it is ours. Beliefs held like tools are easier to improve.

Practical Steps

First, look for evidence that could prove you wrong. When you want to check a belief, ask what you would expect to see if the belief were false, and look for that. Like the people who tested one, two and three in the number task, you will learn more from tests that could fail.

Second, give first impressions time. When you meet a new person, a new place or a new idea, remind yourself that your first impression is based on very little. Try to collect several more observations before forming a firm view, and notice the ones that do not fit your first impression as carefully as the ones that do.

Third, check vivid stories against wider information. When a dramatic story makes something seem common or dangerous, look for calm, reliable information about how common or dangerous it really is. Ask whether the story is memorable because it is typical or because it is unusual.

Fourth, use records instead of memory. Memory favours what fits our beliefs and what was vivid. Simple written records, such as a note of who did which household tasks this week, how often a worry actually came true or how each candidate answered each question, give a fairer picture.

Fifth, judge qualities separately. When you assess a person, a product or a piece of work, list the qualities that matter and judge each one on its own, ideally before forming an overall impression. Hide names or other details that are not relevant when you can.

Sixth, spend time with people who see things differently. Talking with thoughtful people who disagree with you, and listening to understand rather than to win, is one of the best natural protections against confirmation bias. You do not have to agree with them. You only have to let their view reach you.

A Short Exercise: The Other Side Notebook

This exercise helps you practise looking for evidence against your own impressions. It takes about ten minutes a day for one week.

Choose one impression you currently hold that affects your behaviour. It might be about a person, such as "My new colleague does not like me"; about yourself, such as "I am bad at public speaking"; or about a situation, such as "This neighbourhood is unsafe at night." Write the impression at the top of a page in a small notebook.

Draw a line down the page. On the left, write "Fits", and on the right, write "Does not fit". Each day for a week, write down at least one thing you noticed in each column. The "fits" column will probably be easy. The "does not fit" column may take real effort, because confirmation bias makes such events easy to miss. You may need to look for them on purpose: the moment your colleague asked how your weekend was, the time you spoke well in a small meeting, the evening when the street was full of families.

At the end of the week, read both columns. Ask yourself three questions. Is my impression as strong as it felt at the start? Which events in the "fits" column could be explained in another way, as the closed door could be explained by tiredness? Is there a simple step I could take to learn more, such as a short, friendly conversation?

You may find that your impression was correct. That is useful too, because now it rests on fairer evidence. But many people find that their impression softens, and that the softer view leads to better choices and kinder relationships. If you tried this exercise with an impression about yourself, you may also find that you are more capable than your harshest view suggested.

Gabriel did a version of this exercise after his conversation with his wife. Within a week, his "does not fit" column included the neighbour holding the door for an elderly resident, a small plant left outside her door that she clearly tended every day and a quick, tired smile on the stairs that he had never noticed before. The next weekend, he knocked on her door with a cake his wife had baked and an apology.

Questions People Often Ask

"If everyone has confirmation bias, how can we ever know what is true?" Not by any single person being perfectly open minded, which nobody is, but by methods that check our beliefs from outside. Science, at its best, works this way: people test ideas, other people try to repeat the tests, and claims that survive many attempts to disprove them become trusted. In everyday life, we can use smaller versions of the same methods: written records, honest friends, structured judgments and a habit of looking for evidence that could prove us wrong.

"Is it bad to trust my first impressions?" Not always. First impressions sometimes pick up real information quickly, especially for people with long experience in a certain kind of situation. The difficulty is that first impressions also pick up irrelevant things, such as appearance, accent or mood, and that once formed, they are hard to change. A sensible approach is to treat a first impression as a question rather than an answer: "I have a feeling about this person. Let me see whether it holds up."

"Why do vivid stories affect me even when I know the facts?" Because the availability heuristic works through fast thinking, which responds to emotion and images more than to numbers. Knowing the facts engages slow thinking, but it does not switch off the feelings that the story created. This is normal. It can help to remind yourself, gently and repeatedly, of the calm facts, and to limit how often you expose yourself to the same distressing images, especially before sleep. If a fear becomes so strong that it disrupts your life, talking with a qualified professional can help.

Chapter Summary

Confirmation bias is the tendency to look for, interpret and remember information in ways that fit what we already believe. It makes us feel more certain without making us more correct, and the most useful tests of a belief are often those that could prove it wrong. People can and do change their minds when evidence is clear and respectful; the idea that corrections usually backfire has not held up in larger studies.

The availability heuristic leads us to judge how common something is by how easily examples come to mind, so vivid, recent and dramatic events seem more common than they are. The halo effect lets one good or bad quality colour our view of everything else, which is why structured, separate judgments are fairer.

Looking for evidence against your views, giving first impressions time, checking vivid stories, using records, judging qualities separately and listening to people who disagree all help. In the next chapter, we turn to money, and to the curious way our minds sort it into separate boxes.