5X5 Episode 42

Skepticism 101 - Confirmation Bias
S: This is the SGU 5x5 and our latest series of Skepticism 101. And tonight we're talking about confirmation bias. Confirmation bias is an extremely common psychological phenomenon by which people will confirm beliefs that they already have. They will make observations in such a way that it is in line with their existing beliefs. And they will tend to ignore or reinterpret observations that would disconfirm beliefs that they already have.

J: So, Steve, to clarify that a little bit, I would also add to it that people will bend reality so it would help them continue to believe or reaffirm something that they already believe or want to believe.

S: Yeah, in fact they'll interpret the observations they make in such a way that it confirms their beliefs. Bigotry is a really good example of this because it's a kind of a fixed belief that doesn't necessarily accord with reality. So, for example, people may believe about a certain nationality or racial group, let's just say, for example, that they are immoral. And then they will, every observation that they make that seems to confirm the immorality of that group they will remember, they will note it, they will interpret it as a representation of their low morality. Whereas, if they made the same observation about another group or their own group, they will interpret it differently, not as evidence for immorality. So, the interpretation of the observation derives completely from their preexisting belief in such a way as to confirm that belief.

R: Another common example is when the phone rings while you're thinking of your mother and you pick up the phone and it is your mother. It's an easy way for you to think: "Oh, yeah, that's because of the psychic connection between my mother and I". You conveniently forget about all the times your mother called and you weren't thinking of her or all the times you were thinking of her and she didn't call.

E: And you mentioned the psychic connection, Rebecca. You know, there's also a attribute to this, that those self-proclaimed psychics rely on confirmation bias in the sense that they'll do, say, a cold reading for their subject or their target, which is basically a sophisticated guessing game. But they only have to be successful a couple of times during the whole course of the reading, because they're counting on the person walking away from that meeting remembering, perhaps, the one or two things that did hit. And the target totally forgets the one hundred things that were mentioned that totally missed. So, it's an important aspect of cold reading and psychics.

B: Also, since confirmation bias is obviously part of human psychology, it's obviously, then, been with us for quite a long time. For example, Francis Bacon clearly had confirmation bias in mind when he said: "…it is the peculiar and perpetual error of human understanding to be more moved and excited by affirmatives than by negatives…" In fact, many studies have shown that people generally give an inordinate amount of value to information that's positive or supportive of prior beliefs. This tendency affects memories, such that when we are searching our memories for information about a position we are more likely to recall data that confirms that position.

J: You know, we're all, including skeptics, of course, we're all very susceptible to confirmation bias. It's, this type of selective thinking goes on almost in the background and you have to force yourself to not fall for it. Because, I think, we're constantly trying to deceive ourselves and, you know, exercise confirmation bias. It's one of the hardest ones, I think, to overcome.

S: It's also extremely powerful. As you say, Jay and Bob, it's ubiquitous, everyone does it. It could be extremely subtle. It really locks us into our belief systems. It is what we call the "default mode" of human thinking: this is how our brains will tend to operate unless you really understand confirmation bias at work and make specific efforts to work against it. You guys gave a lot of great examples. There are really mundane examples, as well. My favorite one, which comes from Gilovich, is: why do women think that men always leave the toilet seat up when men think that they always put it down?

R: Men they're stupid. Oh.

S: Men remember it when they put it down, and women only notice it and remember it when they fail to leave it down. So, you can have two people, living in the same house, have completely 100% contradictory memories and beliefs about a very mundane fact: does the man put the seat down or not after using the bathroom. And that's simply confirmation bias at work.

R: And so, the way to overcome confirmation bias is, for instance, to set a camera up in your bathroom and study the results of how often the man does actually leave the toilet seat up. Because, since we're all susceptible to it, that's the nice thing about actual scientific experimentation. It's that it can overcome these very human foibles.

S: That's right. In short, science is all about controlling for these inherent and powerful biases in our observations. So, instead of just counting hits and forgetting misses, science would count everything systematically, so that you can't bias the information. This is just one of those biases that&mdash; science, essentially, is a system designed to counteract and account for all the information systematically, so that you can not be influenced by such biases.