ED VS. COGNITIVE BIASES, PART 1: HINDSIGHT BIAS
Posted in Ed vs. Cognitive Biases on February 19th, 2008 by EdLo and behold, I was planning to start a new series this week and current events (if we can so euphemistically label mass murder) provide a picture-perfect example. But first let's talk about what hopefully will be an interesting series.
Cognitive biases are patterns of deviation from rational judgment that occur under a given set of conditions. It's an event or scenario that warps your judgment, which would be sound (or less unsound) otherwise. One reason I find them so interesting is that, unlike logical fallacies, they affect everyone and do so subconsciously. I see flawed logic largely as a matter of brainpower; as intelligence declines, the propensity to make arguments that make no sense increases. Cognitive biases sneak up on you. They're almost like our brain's way of rebelling against us, and they can seriously F your S up.
Let's start with Hindsight Bias, a.k.a. the "I Knew It All Along" phenomenon. It is the tendency to view what has already happened as more likely without realizing that retrospective knowledge of the outcome is affecting one's judgment (Plous 1993).
In short, humans have a very difficult time objectively evaluating how information about an outcome affects their memories. Take a simple example like the Super Bowl. If someone asked before the game, 'What are the odds of a Giants victory?' you'd say x. If I asked you today, 'What did you think were the odds of a Giants victory?' you'd say x+n. Knowing the outcome affects how you recall your estimate the odds in hindsight. Ample research (let me know if you care about a list of academic citations, which I'm happy to provide) shows this phenomenon in buying decisions, elections, medical diagnoses, foreign affairs, and more.
Cue the news.
Nearly every news item about the NIU gunman has quoted the DeKalb police chief, who noted that friends thought The Gunman's behavior became "erratic" in the weeks leading up to the shooting. Pure hindsight bias, of course. Let's say that instead of going on a killing spree, he simply went about his normal day. If you asked all of his friends today "Have there been any changes in This Guy's behavior?" they would be highly unlikely to note anything or offer any serious concerns. However, knowing that he seemingly came unhinged and shot 20-some people, small (or perhaps even imagined) deviations in his behavior are impregnated with meaning. Every email, every phone call that didn't get returned, and every day he showed up to work 3 minutes late suddenly becomes a "sign," obvious harbingers of what was to come. You simply cannot ask people "Did John Doe act strangely or differently last week?" right after he cracks and kills half a dozen people. And himself. The way to get useful information would be to wait (get some distance from the events) and make a conscious effort to question whether the "erratic" behaviors really happened and are being blown out of proportion.
In the wake of such a jarring series of events, that's a lot harder than it seems.