Confirmation Bias

The great American documentarian Errol Morris published a book last year, A Wilderness of Error: The Trails of Jeffrey MacDonald (Amazon). For those uninformed about the case, Jeffrey MacDonald, a Green Beret doctor called the police for help on February 17, 1970 from his home in Fort Bragg, North Carolina. The police arrived to a horrific scene, with his pregnant wife and two young daughters slaughtered brutally. MacDonald claimed that they had been killed by a band of drug-crazed hippies. Over the next several years, MacDonald went from presumed victim, supported by his ex-in-laws when first implicated in the murders, to the lead suspect. Nine years after the killings he was finally brought to justice, and remains in prison today. Joe McGinniss had followed the case as it unfolded and later wrote a book Fatal Vision made into a blockbuster TV miniseries in the 80s.

I decided to show my students the Morris film The Thin Blue Line earlier this week and happened upon Fatal Vision free on YouTube. As I was doing some other work, I half watched the film and then was drawn to do some research online to find out what ultimately happened to MacDonald. While doing that research, I came upon the following clip from Morris: You Tube. Here and in the book, Morris claims that most of what we know about the MacDonald case is based on pure fiction and that it is quite possible that an innocent man has sat in prison for over 30 years. This, of course, follows the case he compiled in The Thin Blue Line, which actually freed a man stuck on Death’s Row for 10 years for a crime he didn’t commit.

What both cases show us is the considerable danger that emerges from Confirmation Bias – the tendency of people to favor information (or evidence) that supports their beliefs or hypothesees. Morris puts it succinctly in the form of a question: “Does a theory in some ways determine the kinds of evidence that you look for and the kind of evidence that you reject?” The answer to this question is a resounding yes and well known to anyone that studies human behavior. People do not look through innocent or objective eyes, they look through eyes attached to their brains and thus heavily influenced by their tastes, experiences, background, beliefs, values, knowledge and the like. And as people try to make sense of the world and establish their ways of being in it, the ways of seeing they have developed within their cultural milieu heavily influence that ontology.

Turning to the world of politics, we see a fundamental problem with the world we live in today. I have already discussed the issue of selection bias on this blog – where the framing of the available options heavily influences the choices people make. Here I consider the related concept of confirmation bias. Essentially, in a world where it is easier to find news and media that support your general worldview and to avoid or ignore media that doesn’t, can it be argued that confirmation bias has only grown in significance? It certainly appears that we are entering more and more irrational times, where people simply ignore evidence or arguments that challenge their belief systems. Sure this is true of conservatives, with studies showing that they only strengthen their resolve when confronted with counterfactual evidence, but it appears to be true across the political spectrum – with people holding so steadfast to their opinions, there is no room for reason or transformation to actually occur. How can we negotiate in a world where people are trapped in their own ideologies, unable to see outside them? Does democracy really have a deeper meaning if it merely becomes majority rule? What roles do technology and education play in reinforcing, or potentially challenging, these trends? I think these are all questions we should explore in great detail, determining ways to expand the presumed the array of choices available (selection bias) and to give people the open-mindedness and critical thinking skills to overcome our collective penchant toward constantly confirming whatever we believe; as wrong as we may be.