Foundations of Skepticism: Confirmation Bias…..
In a prior post, I described one of the most vital supporting pillars of skepticism: the requirement for increasingly powerful evidence in support of claims that push the limits of plausibility and call into question fundamental aspects of our scientific knowledge. In this post, I’m going to reveal an inherent flaw in our ability to accurately interpret causal relationships in the natural world. This flaw, known as confirmation bias, serves as the driving force behind almost all unfounded belief systems and the primary reason why anecdotes and testimonials are unreliable.
Examples of confirmation bias are all around us. Nobody is completely immune to its sirens’ call, and partial immunity can only occur with awareness of the concept and a constant struggle to avoid its influence. The feeling one experiences when realizing that they are wrong, especially regarding something that they are emotionally invested in, is unpleasant to say the least. Our brains, and thus our minds, have evolved to avoid this feeling even if we must sacrifice an accurate assessment of reality. We do this via three different mechanisms, all of which revolve around our tendency towards biased selection of information that supports what we already believe to be true.
1. The selective collection only of information that confirms (biased collection)
2. The selective interpretation of information in such a way that confirms (biased interpretation)
3. The selective recollection of only information that confirms (biased recall)
The net effect of these mental mechanisms is the overemphasizing of only some of the available data or information regarding a claim. The best example that comes to mind, and one which melds this topic with the practice of medicine, is that of the commonly held belief in so-called lunar effects. This is the hypothesis that the full moon has effects on the human body. These include effects such as increased rates of mental illness and suicide, and spikes in birthrates. High percentages of medical and law enforcement professionals accept lunar effects as a true phenomenon despite the fact that they are highly implausible and have been thoroughly disproven by the collective data of over 100 studies. So why does this pervasive belief persist?
The answer is confirmation bias.
When a strange event occurs, or there is an odd cluster of events, during a full moon, such as the unlikely need for the delivery of a baby in a pediatric emergency department, many will assign a causal relationship (biased interpretation) because they already believe in lunar effects. Others may make the connection for the first time because of prior exposure to legendary effects of the moon as with werewolves or its association with witchcraft, or because, after the fact, they hear of the many unusual things that have happened during a full moon from coworkers (biased collection). They may link the full moon to the strange event simply because humans are uncomfortable with the random nature of the natural world and a full moon is sexy. As storytelling primates, we often force explanations where none exist, and we love a good tale. This is what humans have evolved to do, and we do it better than any other species. At times, we do it too well. When a strange event or cluster of events occurs when the moon isn’t full, no connection is made. We do not interpret this as evidence against the previously assigned causal relationship (more biased interpretation).
Over the years, as the number of shifts in the ED pile up and random weird events continue to occur both during and not during the presence of a full moon, we tend to only remember the former (selective recall). Also, because human memory is malleable and changes with time and multiple retellings as in a childhood game of Telephone, strange events that did not occur during a full moon can drift to fit our beliefs. The occurrence of an ED delivery two weeks before and five days after a full moon two years ago may morph into that night we had two babies born during a full moon. The memory, although as false as a blatant lie, is no less real to the individual recalling it.
All of this is reinforced by the popularity of the belief, and to a lesser degree by misconceptions regarding the actual effects of the moon on the earth. Think tidal forces. Yes, it is true that the human body is largely made of water but in reality I am exerting more of a tidal force, or gravitational pull, on my keyboard as I type this than the moon does from 239,000 miles away. In fact, there is more gravitational pull on the earth during a new moon than a full one because a new moon lines up with the sun. But because of the dramatic visual stimuli that a full moon provides, this post isn’t about new moon madness.
So what are we do to about confirmation bias? Is it really an unavoidable pitfall? Well, perhaps not entirely unavoidable but it is an obstacle that has felled even the most powerful of minds. The best defensive strategy against this hardwired mental defect is an understanding of the fallability of human perception. The best offensive measure is active exposure to information that does not fit within the cozy confines of our personal worldview. And giving scientific discovery more weight than personal anecdotes wouldn’t hurt either.