In a recent article in this journal, Maraun, Gabriel, and Martin (2011) critically discuss the received views of the phenomenon of regression towards the mean, proposing changes in the way it should be viewed. Using in part Galton’s original data for the relation between parental height and child height, the present paper suggests that some undiscussed aspects of regression towards the mean are relevant. The usual methodological warnings given in the received view are not, in fact, entirely inappropriate.
