Confirmation Bias
We readily see the evidence that supports what we believe and pass over the evidence that contradicts it without noticing. This lesson explains how confirmation bias works and the small habits that can be built against it.
Beginner · 11 min
In this lesson you will learn
- Define confirmation bias and distinguish its three faces
- Explain how selective attention and selective memory distort our sense of evidence
- Analyse how beliefs like “retrogrades always bring mishaps” feed themselves
- Carry the habit of seeking disconfirming evidence into daily life
One tendency, three faces
Confirmation bias operates at three stages. Selective search: we look for information that will support our belief and fail to ask the question that would test it. Selective interpretation: we read an ambiguous event in line with our belief; the same event looks like two different pieces of evidence to two people with different beliefs. Selective memory: examples that fit the belief stay vivid, while those that do not fade away. Combined, the three create the feeling that “the evidence keeps piling up” even when no strong evidence exists.
The three stages usually work simultaneously and unnoticed; that is why, to the person looking from the inside, the process seems perfectly fair.
How a belief feeds itself
Someone who believes “Mercury retrograde always brings mishaps” records every delay, every broken device, every misunderstanding during the retrograde; the very same mishaps on ordinary days get attributed to no one's retrograde. Trouble-free retrograde periods are not counted either, because “nothing happening” is not an event the mind holds on to. And so only one column of the ledger fills up. Yet there are four columns: mishaps during retrograde, smooth days during retrograde, mishaps outside retrograde, smooth days outside retrograde. Testing the claim requires counting all four; the mind, left to itself, counts only the first.
This four-column table is the everyday form of a simple but powerful tool known in statistics as a contingency table; even filling it in roughly is usually enough.
The bias lives in everyone, on every side
Confirmation bias is not the problem of one viewpoint or one level of education; believer and sceptic alike see the examples that favour their own view more easily. Indeed, knowledgeable people can defend their biases better, because they produce more skilful objections to opposing evidence. That is why the remedy is not declaring “I am unbiased”, but building arrangements that push the bias out of the way data is collected. The blind assessment you met in the previous module is exactly such an arrangement. Impartiality is not a personality trait; it is a working method.
Counter-habits
At the daily scale, three small habits help. Writing in advance: note your expectation before the event; this closes off the memory's later editing. The disconfirming-evidence question: every time you defend a belief, ask “which examples do I know that don't fit it?” Counting all four columns: whenever you say “Y happens when X happens”, check whether you have also counted the times Y happened without X and X happened without Y. These habits do not eliminate the bias; but they stop it from making decisions silently.
What the three have in common is that they lean on mechanism, not willpower: however good the intention, the most reliable safeguard against memory and the filter of attention is moving the record outside the mind. A small notebook beats the firmest vow to “stay neutral”.
Common mistake: seeing the bias only in the other side
Most people who learn about confirmation bias spot it first in other people's views; their own views seem exempt because they feel “already proven”. But that is exactly how the bias works: from the inside, the evidence always looks balanced. The honest question is: when did I last find a piece of evidence against my own view convincing? If no answer comes to mind, it is worth going looking.
Today's small practice
Start a one-week “prediction journal”. Whenever you form an expectation — “this meeting will go badly”, “this is my lucky week” — write it down in advance, in a single sentence. Once the outcome is known, note what actually happened next to it. At the end of the week, look at your hit rate. Most people find their written predictions were less accurate than they remembered; this little journal is the antidote to selective memory.
Summary
- Confirmation bias works in three forms: selective search, selective interpretation, selective memory
- Beliefs feed themselves because days when “nothing happens” are never counted
- Testing a claim means counting all four cases; the mind counts one column
- The bias lives in everyone; knowledge can even sharpen its defence
- Writing in advance and asking for disconfirming evidence are practical counter-habits
Check your understanding
Questions and answer order are shuffled on every attempt. Results stay in your browser.