Confirmation bias is the tendency of all humans to seek out (and interpret) information, data, evidence, and sources that supports or confirms what we already believe and ignore (or discredit, downplay, etc.) information, data, evidence, and sources that disconfirm what we already believe.
Suppose I tell you, “I want you to guess my rule. I will give you a string of 3 numbers, and you need to guess the rule I am following in my list of three numbers. You may concoct your own strings of 3 numbers, and I will tell you whether or not they fit my rule.”
The string of numbers is 2, 4, 6
The usual responses are as follows:
People become entirely convinced after amassing their body of evidence that the rule is counting up by twos. However, there is something wrong about this approach, because I will end up saying… counting up by 2 is not my rule!
The typical methodology employed involves amassing evidence that supports your hypothesis, namely that the rule is counting up by two. However, this fails to take into account that there are loads of hypotheses that are compatible with the string of numbers “2, 4, 6”.
After all, going from 2 to 4 is an increase by 100%. Going from 4 to 6 is an increase by 50%. Following the pattern, one could increase 6 by 25%, then increase 7.5 by 12.5%, and so on. Or, the pattern could be “multiply by two, then add two…” repeated indefinitely. Or it could also be something entirely different. Perhaps it is merely “listing positive integers”. In that case, listing -6, -4, -2, for instance, wouldn’t fit the rule.
The key to the experiment is not to concoct examples that repeatedly confirm your suspected rule; it’s rather to concoct examples that disconfirm the supposed rule, and to do this as much as possible. The following approach tries to disconfirm the previous rule each step of the way, ultimately arriving at the correct answer:
Notice that we arrived at the rule “counting up” from observing that we finally found a counter-example to the rule — a string of three numbers that disconfirm the hypotheses we had been concocting.
This is a great experiment that exposes implicit confirmation bias, and its findings are applicable across a wide range of phenomena, from homeopathy to astrology to psychics to anti-vaxers. It’s only through an understanding of our own confirmation biases that we can at least try to transcend our cognitive limitations and arrive at truth.
It would be entirely remiss of me to do an entire series on critical thinking without paying due attention to informal fallacies, which are essentially mistakes in reasoning.
Instead of going through a long list of fallacies (most of which my readers will already know about since they will be philosophically inclined), I will place below a list of the best sites to learn about logical fallacies in depth. I hope you enjoy!