Confirmation Bias Illustrated
One of the reasons I sometimes have trouble writing lately, and am less inclined to write about controversial subjects, is that it is clearer to me every day how so many people base their opinions on a combination of their personal observations and what others have told them. That is to say, they see what they have been taught, and most of them aren’t inclined to even discuss the possibility that their perceptions may be flawed. The case in point that inspired this post was something that just happened in a discussion group I belong to. Someone posted a story about how a hospital in Australia conducted a study that showed that psychiatric patients acted like werewolves far more often on full moon nights than other nights. And the media dutifully reported on this study and passed the word along. This theory, that crime, madness, etc. are more common on the full moon is called the lunar effect. It wouldn’t surprise me if many of my readers are familiar with the theory in one form or another. It’s been studied extensively by scientists over the past few decades.
So I felt inclined to post my response to the group. I pointed out that this was one small very poorly controlled study by a single hospital, and worse, it appears that the study was designed to get lots of publicity, judging from the amount of werewolf lore the study’s author was touting. And furthermore, that while little “studies” like this often demonstrated the lunar effect, large careful rigorous studies involving vast amounts of patients and data … show no such effect. In fact modern science has basically concluded that the lunar effect is at best folklore, since it can’t be demonstrated in anything resembling a rigorous controlled study.
And how did the other members of the group respond? So far they’ve all posted stories about how in “their nursing home” or “their emergency ward” or whatever … “everyone” noticed that there were more problems on the night of the full moon. Did I debate with these people, and try to point out to them that considerable scientific effort had been made to demonstrate the lunar effect,and that the empirical and scientific evidence strongly suggests that there is no such effect? Of course not, it would have gotten ugly. Few, if any, of them would have been persuaded that they have been fooling themselves into seeing something that wasn’t there.
So what is going on here? Are scientists just stupid, and refusing to see an obvious effect? Well, that’s an argument that is used so often on the fringes of science that I no longer cringe when I hear it. While scientists and science aren’t perfect, they have a remarkable track record: From exploring the furthest reaches of the Solar System and peering billions of light years into space, to understanding biology and medicine to the point where humans are routinely living to over 100 now, to developing the technology that made the computer (or cell phone!) that the gentle reader is using right now … science works! In fact it works a hell of a lot better than religion, prayer, wishful thinking, or “common sense” ever did. I digress, but for me science is about as empirical as it gets.
What I think is going on here was originally called “counting the hits and ignoring the misses” and is now called confirmation bias. People have a powerful tendency to both notice things that fit their preconceived views and not notice things that might contradict them. And, frankly, it’s been my observation that most people are so inclined to reinforce their beliefs this way that it’s a waste of time to debate them. However, to stop efforts at debating people based on this observation would make me guilty of the same bias, so I’m going to plough right along.
I am not however going to debate the validity of the lunar effect. I am more interested in what the confirmation bias means. It’s such a common and all encompassing bias that it seems to me it must say something about the way we perceive the world. And here there is some fertile ground for speculation. For what we all forget most (or all) of the time is that the reality we see is heavily filtered by our brains and sensory organs … as the below illustration shows:
I am pretty sure that the vast majority of people will say that it is “obvious” to them that the squares labelled A and B are two very different shades of grey. And yet when they are connected by same coloured bars on the right, it’s clear that they are the exact same colour. Are our eyes and brains that easy to fool? No, what is going on here is confirmation bias illustrated. IE from the various clues in the picture our eyes and brains construct the image we see, and our eyes and brains are designed to create a meaningful image we can interpret … not present us with an exact photographic reproduction of what we are seeing. Or as the description of the illusion puts it, our eyes and brains are designed to discern actual shapes and surfaces which are much more important than minor changes in shading. And they alter what we see to reinforce this interpretation!
So I’m suspecting that confirmation bias is probably simply a side effect of how our brains construct reality, and for the most part where it leads us astray in any evolutionary sense has been unimportant. IE the caveman that could spot the sabre toothed cat in the grass and wasn’t fooled by the subtle interplay of light and shadow has a much higher chance of survival than one who can’t make out the shape of the cat until it is leaping for his throat. The fly in the ointment though is that when we come to civilization and culture, a case can be made that this tendency could have negative consequences. In fact a case can be made that modern propaganda and advertising are increasingly able to utilize this (and other) biases to get people to believe stuff that just aint so!
That however is a topic for a future post. Next post … I review Gran Torino. Rent and watch it now gentle reader, unless spoilers are of no matter .
(The above image was put into public domain by the copyright holder, and may be reproduced and distributed freely. Copyright © Edward H. Adelson 1995. It’s really a pleasure that someone made and put something like this into the public domain, especially considering how copyright laws have been perverted by Congress the past few decades. My thanks to Mr Adelson, I’ll be reviewing more of his work.)