Errors of Measurement: Regression Toward the Mean

In 1886, Francis Galton published an article titled “Regression Towards Mediocrity in Hereditary Stature.” Interested in heredity, Galton had obtained measurements on heights of 205 sets of parents and their 913 adult children. He noticed that if he selected families where the parents were tall, the average height of the children was less than that of their parents, whereas if he selected families where the parents were short, the average height of the children was greater. Galton called this “regression towards mediocrity”; it is now known as“regression towards the mean,” as the term mediocrity has acquired disparaging connotations. The same thing happens with the children: for tall children, the mean height of their parents isless; for short children, the mean height of their parents is greater. This is a statistical, not a genetic, henomenon. This entry discusses how regression toward the mean works, providing several examples.