Blake Borgeson, in blog form

I’ve been reading the blog Overcoming Bias sporadically over the last several months, and I’ve learned a lot from it.  The topic, naturally, is examining ways that cognitive biases affect how we think, and if you haven’t already delved into the ideas, you’ll be amazed when you come to understand just how much these biases lower our practical intelligence.

I just came across one of the earliest experiments to illustrate the strength of a very prevalent example, confirmation bias, and I think it’s really fascinating, not to mention helpful to be aware of.  The experiment was carried out by Peter Wason, and it’s called the 2-4-6 Problem–I’ll go ahead and quote for you:

Among the first to investigate this phenomenon was Peter Cathcart Wason (1960), whose 2-4-6 problem presented subjects with three numbers (a triple):

2

4

6

Subjects were told that the triple conforms to a particular rule. They were then asked to discover the rule by generating their own triples and using the feedback they received from the experimenter. Every time the subject generated a triple, the experimenter would indicate whether the triple conformed to the rule. The subjects were told that once they were sure of the correctness of their hypothesized rule, they should announce the rule.

While the actual rule was simply “any ascending sequence”, the subjects seemed to have a great deal of difficulty in inducing it, often announcing rules that were far more complex than the correct rule. The subjects seemed to test only “positive” examples—triples the subjects believed would conform to their rule and confirm their hypothesis. What they did not do was attempt to challenge or falsify their hypotheses by testing triples that they believed would not conform to their rule. Wason referred to this phenomenon as confirmation bias, whereby subjects systematically seek only evidence that confirms their hypotheses, an explanation he made appeal to also for performance on his selection task (Wason 1968), though he did briefly consider that participants might be using a three-valued rather than two-valued logic. Confirmation bias has been used to explain why people believe pseudoscientific ideas.

It’s interesting for me to see that if I was faced with this problem, my first inclination would be to think, “Maybe it’s just adding two each time…” then choose “1,3,5” as my first test.  Even if I was limited by the number of triples I could guess, I’d feel a strong urge to confirm my hypothesis, when the rational choice, the one with the best chance of the quickest success, would be to try to disprove it.

As we go through life building models of how the world works and why things are like they are, confirmation bias leads people unknowingly to rely on untested beliefs, setting us up for failure whether we’re trying to solve problems, resolve our differences, or just make good choices in life.  Beware! =)