Regression to the mean: treatment effect without the intervention

Abstract

Regression to the mean (RTM) is a widespread statistical phenomenon. It is a group phenomenon that occurs whenever an extreme group is selected from a population based on the measurement of a particular variable. If a second measurement is then taken for the same group, the second mean will be closer to the population mean than the first measurement. This decrease (or increase) can be mistakenly attributed to a treatment effect; the conclusion can be drawn that an effect results from treatment when it in fact results from chance. Any intervention that is aimed at a group or characteristic that is very different from the average will appear to be successful because of RTM. It is therefore important that any genuine reductions because of the treatment are separated out from the effect of RTM. If the problem is ignored then this will lead to errors in the interpretation of results and, potentially, decisions made on the evidence of those results. This paper highlights the importance of the issue and its effects on many common clinical, public health and managerial decisions.