Regression to the mean is a statistical phenomenon that occurs when a variable that is extreme on its first measurement tends to be closer to the average on its second measurement, and vice versa. This concept was first identified by Sir Francis Galton in the late 19th century while studying the relationship between the heights of parents and their children.
The principle of regression to the mean states that if a variable is extreme on its first measurement, it will tend to be less extreme on its second measurement. This doesn’t imply a causal relationship but rather is a result of the statistical tendency for extreme values to be followed by more typical values.
For example, imagine a student who scores exceptionally high on a test. If the test is retaken, the student’s score will likely regress towards the mean (or average) score. This doesn’t necessarily mean the student’s ability has changed, but rather that the first score was partly due to variables that did not repeat exactly in the same way, such as luck, exceptional circumstances, or measurement error.
In practical terms, regression to the mean warns against drawing conclusions about causation from correlation, especially in situations where chance or variability plays a significant role. It’s a crucial concept in fields such as psychology, medicine, finance, sports, and any area involving data analysis and interpretation, reminding us to be cautious in attributing cause to what might be natural statistical variations.
See also: Hot hand fallacy