You may have noticed lately how all the Republican pundits say Romney will win and all the Democratic pundits say Obama will win. Don’t they see the same data? They can’t both be right. Can they both be wrong?
In a word, yes! How? Confirmation bias.
According to Science Daily, confirmation bias is “a type of cognitive bias and represents an error of inductive inference toward confirmation of the hypothesis under study.” What the heck does that mean? Three things:
- Interpretation – When giving evidence both for and against your belief, the bias tends to focus only on the evidence for the belief and ignore evidence against the belief. If you look at 5 polls but only write and think about the one poll that supports your belief, you may be guilty.
- Search for information – You seek evidence from someone you know agrees with you and do not seek evidence from those that disagree with you. Maybe there is a poll that regularly inflates the results to support your candidate. Is that your first poll of choice? You may be guilty.
- Memory – you tend to remember evidence that supports your argument and forget evidence that contradicts your position. Maybe you say all 5 polls but for some reason only the one poll that agrees with your position sticks out in your mind. You may be guilty.
Let’s dig deeper. Confirmation bias is a misappropriation of evidence, leading to incorrect inductions. Most often, confirmation bias is an automatic response. These automatic responses are based on heuristics. People use heuristics (rules of thumb) to quickly make evaluations. Without heuristics, we would never have time to make all the decision necessary in our lives. We would spend all of our time re-hashing the same evidence, the same logical steps, and the same conclusions. Where heuristics go wrong is when they are applied to emotionally charged ideas. The pain at looking at contrary evidence leads many individuals to ignore that evidence in favor of positive evidence that confirms our original ideas. Individuals do not like dissonance in their thoughts. They want certainty. Unfortunately with confirmation bias, that certainty comes at the expense of objectivity, leading to ill-founded beliefs.
There are two times confirmation bias can be observed, when you first develop an idea and when you are presented with new evidence that contradicts one of your existing ideas. If confirmation bias can be avoided in the former, the easier it will be managed in the latter. However, many people, myself included, have adopted ideas without fully evaluating all the evidence for and against when initially presented to them. Young kids are particularly susceptible. This may require re-evaluation of ideas that were once closely held. I have done this before when I reevaluated my belief in God, which led me to reject my earlier beliefs based on new evidence.
Fortunately, there is a way to avoid confirmation bias – scientific inductive reasoning – as described in The Logical Leap: Induction in Physics. Two principle techniques to scientific reasoning are establishing reliability and validity. Validity can be further sub-divided into both internal and external validity. So how can these be employed to avoid confirmation bias?
1. For evidence to be reliable, new evidence should confirm older evidence. If it contradicts it, chances are something fishy is going on or evidence is not being placed within its appropriate context. Seek evidence from multiple, disparate sources, particularly in emotionally driven complex issues. In highly controversial issues, it’s especially important to gather evidence from sources, both pro and con, and place that evidence within their proper context.
2. Internal validity looks at the internal logic of an argument. Ask yourself if the evidence presented shows only part of the picture. Ask yourself if you have fully considered ALL of the evidence. Ask yourself if the conclusions are as solid as claimed. It may help to break up the argument into all of its component parts and verify that evidence supports each part. In short, play “devil’s advocate” to establish internal validity.
3. External validity verifies an idea is consistent with the wider context of one’s knowledge base. Ask yourself if this conclusion is true, what does it mean for other ideas. Ask yourself honestly what the evidence means for my life. The conclusion from the evidence should integrate with knowledge you already have without contradictions. If it doesn’t, either something is wrong with your new conclusion or something is wrong with your existing ideas.
By employing these methods consistently, you can avoid confirmation bias. Take a step back from the emotional charged idea and be objective. If you do this, you have a good chance at avoiding confirmation bias and achieving objectivity in your thoughts.