Wikipedia defines regression to the mean as the phenomenon that if a variable is extreme on its first measurement, it will tend to be closer to the average on its second measurement—and if it is extreme on its second measurement, it will tend to have been closer to the average on its first. In simpler terms, things even out over time.
Let’s say you have a really terrible headache. You happened to drink a glass of orange juice and three hours later you feel much better. You would then attribute your recovery to the Vitamin C in orange juice. Now you start drinking juice every day to prevent headaches and voila, you haven’t had a headache in two weeks! Orange juice = No headaches.
But there’s a logical fallacy here—if symptoms are excessively severe this week, then next week they should be less severe simply by random fluctuations. If treatment is only sought when these symptoms are at their worst there will almost always be a coincidental recovery. This appears even if the treatment has no effectiveness whatsoever.
It’s simple enough to understand, but I’m stunned that I never actually knew about this as an actual principle until now. It kind of makes me wonder about all the statistics classes I took in school; what the heck were they teaching us?! I mean this is useful, real-world stuff that biases our thinking in every conclusion we draw and every decision we make. For that reason alone, I’m determined to learn more about behavioural science and statistics this year. I’ve started with Thinking, Fast and Slow, and next on the list is Predictably Irrational: The Hidden Forces that Shape Our Decisions. If anyone has any more suggestions, I’d love to hear them.
Image credit: http://www.cubsrelated.com/