The Regression Fallacy
If you don't understand the idea of regression to the mean you are much more likely to make thinking errors that will cause you great difficulties. Here are some links to definitions and some examples that will make it all clear.
regression and the regressive fallacy
The regressive fallacy is the failure to take into account natural and inevitable fluctuations of things when ascribing causes to them (Gilovich 1993, 26). Things like stock market prices, golf scores, and chronic back pain inevitably fluctuate. Periods of low prices, low scores, and little or no pain are eventually followed by periods of higher prices, scores, pain, etc. To ignore these natural fluctuations and tendencies leads to self-deception regarding their causes and to post hoc reasoning. For example, a professional golfer with chronic back pain or arthritis might try a copper bracelet on his wrist or magnetic insoles in his shoes. He is likely to try such gizmos when he is not playing or feeling well. He notices that his scores are improving and his pain is diminishing or gone. He concludes that the copper bracelet or the magnetic insole is the cause. It never dawns on him that the scores and the pain are probably improving due to natural and expected fluctuations. Nor does it occur to him that he could check a record of all his golf scores before he used the gizmo and see if the same kind of pattern has occurred frequently in the past. If he takes his average score as a base, most likely he would find that after a very low score he tended to shoot not a lower score but a higher score in the direction of his average. Likewise, he would find that after a very high score, he did not tend to shoot a higher score but rather would shoot a lower score in the direction of his average.
and from this article one of my favorite examples of the regressive fallacy
Do You Sabotage Yourself?
Kahneman captured his first great insight by observing his own students. In the late 1960s, he was teaching a class on the psychology of training to flight instructors in the Israeli air force. Concerned at how the instructors screamed obscenities and pummeled trainees' helmets until they cried, Kahneman told his class that research on pigeons showed reward to be a better motivator than punishment. One flight instructor burst out, "With all due respect, sir, what you're saying is for the birds." He heatedly told Kahneman that trainees almost always did worse on their next flight if they'd been praised--and tended to fly better just after getting yelled at. Kahneman was dumbstruck. He realized he was staring into the face of a profound misperception: The flight instructor believed that his own praise or criticism caused the trainee's performance to reverse. In reality, Kahneman knew, chance alone dictates that an unusually good or bad event is typically followed by a much more ordinary one--what statisticians call "regression to the mean."Still want more? From Fallacy Files Weblog
Comments
Hmmmmm... the example at http://gncurtis.home.texas.net/regressf.html was poorly chosen :-) It ignores the very real possibility (though by no means certainty) that athletes from a predominately poor society who suddenl;y come into great wealth may easily become more enamored with their new found wealth and status than with the athletic ability that got them there and by doing so fail to maintain their athletic peak - thus their performance would suffer and it would be a direct result of obtaining that wealth/status. In other words it is a bad example because it ignores human behaviour.
That was a great read... I know a lot of people guilty of those misconceptions.
Vinny, Glad you enjoyed it. Be on guard everyone is susceptible to errors like this.
Doug, I don't disagree that wealth might have an impact on an athlete though it doesn't seem to slow down Americans that come from relatively poor backgrounds and become multi-millionaires. But regardless of any impact that may have the regression to the mean will still take place. I suppose that a new mean may evolve but once again the regression will take place. I'm not sure if it is good or bad to have messy real world examples. It does make it less clear, but maybe that is good training for identifying the cases where it still applys.