Sunday, November 30, 2014

Confirmation Bias: I see it everywhere

Once I learned about Confirmation Bias I started seeing it everywhere.


Confirmation bias is like that. It wasn't until I bought an old Volvo station wagon that I began seeing them all over. Suddenly everyone was driving a Volvo. Or was it just me? Well, there are actually two things going on here. One is what psychologists call Inattentional Blindness. In inattentional blindness we simply fail to see certain aspects of our environment because we are paying attention to other things. There are lots of experiments showing this feature of our visual system, but the most famous is the selective attention test involving six people passing two basketballs around. By now you've probably seen it, but if you have not, then go ahead and watch it now:


Inattentional blindness is not confirmation bias, but the two seem closely related to me. If you have not been paying attention to something in your environment you will probably discount it from how you figure the world really is. But once you notice it, then you might even start looking for it. At this point you've moved into the realm of confirmation bias.

Confirmation bias is accused of causing all kinds of problems. It apparently affects us constantly as we go about our lives. But there are ways around it. Scientists are very worried about how it can interfere with empirical [based on observation] research. The empirical method is one of the basic building blocks of modern science so a human bias that interferes with that can do a lot of damage to the scientific enterprise.

My favorite way of thinking about science is to conceive of it as an ongoing attempt at overcoming human biases. Scientists have been developing more and more sophisticated ways of using statistical methods to make sure their data collection is a true random sample (or as close to random as possible). In medicine they use double-blind experimental methods to make sure the person conducting the research does not know who is receiving which medicine, and therefore she cannot knowingly, or unknowingly influence the experiment. There are other methods used for overcoming confirmation bias, but perhaps the most powerful method that the scientific method espouses is independent confirmation. If one team of researchers does a study and finds a particular result they then publish their research so that other teams can try to replicate the study. If no one can replicate it then the scientific community as a whole will not accept the conclusions of the original research. This is why scientific consensus is the greatest proof of a theory.

There's something you can do to help you start to mitigate the effects of confirmation bias as well as other biases bouncing around in your head: start a surprise journal. Get yourself a journal and start taking note of things that surprise you; things you don't expect. Look at these things. They can be clues to ways in which your own thinking might need to be changed.

My wife and I were driving down the 75 to Atlanta one spring and, being from Chicago, she said that there were a lot more license plates from Illinois than from other northern states. Suspecting confirmation bias at work I told her to get out a piece of paper and start keeping track of every northern license plate we came across. This is really the best to be sure that your observations are accurate. By the time got Atlanta she had one more license plate from Illinois than from Indiana. So, although she actually did get more Illinois license plates, she was surprised at how many Indiana vehicles she wasn't noticing.

So if you're really curious, you have to do the hard work of keeping track of everything. This is easy while driving down a freeway for a couple hours. It gets a lot harder when keeping track of every encounter you have with other people for months on end in order to figure out if the full moon really does cause people to act strangely. That's why so many people continue to believe that the full moon actually has a real effect. Even though it has been studied numerous times and there's no statistically significant differences between when the moon is full and when it is not, many people continue to notice more strange behavior.


No comments:

Post a Comment