One way in which Science advances is through the discovery in data samples of linear relationships between variables: these are "scaling relationships". The breaking of such a relationship is often indicative of a change of the underlying physical processes. Understanding the physical processes that underlie such scaling behaviour is key in such diverse areas as astronomy, biology and city planning.
However, correlations between variables are not evidence of causality, and, indeed, may simply be the consequence of producing misleading plots. We illustrate this with examples of significant game-changing plots that have served to change our scientific ideas and plots that are the consequence of misguided data massaging.
The talk ends with an examination question for the audience.
IFIC and Observatori Universitat de València