We Really, Really Want to be Right

Confirmation bias and how we misread evidence

Imagine, you have been gifted two bottles of spring water. It comes from an artesian spring and people have been talking about it. It is packaged in a glass bottle with an actual cork. You sit down for dinner and take a sip. This is the best water you have ever tasted, almost life changing. Then, as you clear the dishes you notice the bottles on the counter, unopened. You have been drinking normal tap water. This is an example of confirmation bias.

Confirmation bias is the tendency for a person to see new evidence as confirming one's pre-existing beliefs.

Confirmation bias is real

Even professional scientists who do experiments for a living are in danger of being affected by it. Regardless of how unbiased a person tries to be, they usually have either a belief in what the outcome will be, or a desire to see a certain outcome. Either situation can sway the results. The way this happens can be subtle, often invisible.

The person may find a reason to exclude results that do not support their beliefs. They may pay extra attention to the results that do support their beliefs. They may inadvertently influence the experiment.

Confirmation bias can go into overdrive if the idea being tested is a strongly held pet theory. Then almost everything seems to reinforce their preexisting belief.

Human nature

Confirmation bias exists because of how humans perceive and analyze the information they obtain. It is important to realize that this is happening subconsciously. People are not intentionally doing it.

This can affect more than an individual person. An entire team, or even the entire country can fall victim to it.

Avoiding confirmation bias

The issue is so important that scientists have developed tools to avoid it, and the result is much better science. For example rigorous rules have been developed to decide when data can be excluded. Rather than going by feeling they use statistics to analyze whether results are true or not.

The gold standard

In many cases confirmation bias simply wastes resources or time. But in some areas the stakes are higher, like in medicine. In these situations, research often uses a technique called the double blind experiment which has become sort of a gold standard for avoiding confirmation bias. In pharmaceutical testing neither the patient, nor the doctor, knows if a real medicine, or a placebo (containing no active ingredient) is being used. This demonstrates the length that people go to avoid confirmation bias.

Confirmation bias in everyday life

What does this mean in our everyday lives? The same tendency toward confirmation bias can interfere with our interpretation of what we see. The result can be bad decisions, or at least unjustified decisions. I mentioned the bottled water as an example. But there are other more serious situations that arise.

For example in health care or retirement planning, where a bad decision can have dire consequences.

Confirmation bias is a completely normal human behavior. It is not feasible to set up double blind experiments before making any decisions. But understanding that confirmation bias could be at work should reduce your chance of falling victim to it.

Look Again Ideas