The Illusory-Correlation Bias

The tendency to inaccurately link an action and an effect.

It's so easy to overestimate a link between two variables—a cause with an effect—even when the link is slight to non-existent. For example, we might link being slim with happiness, assuming heavy women are less happy. Haven't you had the thought, "If I lose my pot belly, I'll be happier." But is it true? Isn't happiness dependent on myriad variables, including our brainstate or moods. And what about astrology, linking distant planets to earthly personalities? Really far-fetched and so many fall for it.

Redelmeier and Tversky (1996) assessed 18 arthritis patients over 15 months, while also taking comprehensive meteorological data. Virtually all of the patients were certain that their condition was correlated with the weather. In fact, the actual correlation was close to zero.

The opposite of an "illusory correlation" is an "invisible correlation," where an actual correlation is not seen. For instance, the link between smoking and cancer was not noticed for years.

After a delightful birthday lunch at Poggio’s in Sausalito, my friend Dawn (a fitness trainer who has hiked the 2,000 mile Pacific Coast Trail twice) and I went for a casual stroll in my hilly neighborhood.
   Back at the house, she gasped, “Your fingers are blue! You might not be getting enough oxygen to your extremities. Take off your shoes, let’s look at your toes.”
    They weren’t blue, which baffled us even more. We talked about calling a doctor, but first I needed to go to the bathroom. While sitting on the john, I glanced down at my hands, which were resting on my new unwashed blue jeans.
     “The jeans did it!” I screamed.
     “No way.”
    “Let’s see if it comes off,” I said, lathering my hands with soap. Then I ran them across my white kitchen counter and, voila, it turned blue. We howled, realized we had been snookered by an illusory correlation.
   According to David Ludden, associate professor of psychology at Lindsey Wilson College in Kentucky, there is a strong urge in humans to seek causes for events in their lives, and we fall for the ones we manufacture even if they’re wrong.
   “No other species makes causal inferences like we do,” says Ludden. “This ability clearly gave humans a strong evolutionary advantage, allowing them to understand and—and hence control—their environments to a greater degree than any other animal. Still, our ability to ascertain causality is not that reliable. For one thing, we are prone to causal illusions—we infer causal relationships where none exist. Furthermore, when we are unable to explain why events occur, we feel distressed, so we tend to make up explanations with little or no evidence to support them. Nevertheless, the advantages of this ability to construct beliefs about causal relationships must have outweighed the negative side effects for early humans.”
   So now I know why Dawn and I quickly fabricated a link between oxygen-deprivation and my fingers. What if we’d gone to the hospital? Would the doctors have figured out the true cause? Or would they have made up an illusory correlation, too?
   My friend Lynn calls this cognitive bias “The Jumping-to-Conclusions Bias,” because conclusions appear quickly and with a feeling of certainty. To learn more about certainty, peruse neurologist Robert A. Burton’s keen tome, On Being Certain: Believing You're Right When You’re Not.