跳转到主要内容

regression toward the mean

n. 1. The tendency of an observation to be closer to the mean of a population than a prediction of the observation using a sample mean as a basis for prediction. 2. The higher likelihood that an unlikely result will be followed by a more likely result. Thus if an athlete has an outstanding performance, his or her next performance is more likely to be close to the average than to be similarly outstanding. This is the basis for the gambler's fallacy in which chance events are believed to be self-correcting.