One of the more common techniques for measuring relationships between variables is the well-known Pearson correlation coefficient. The corresponding handy test statistic is taught in just about every beginning course in elementary statistical analysis and its sheer simplicity often leads researchers to 'discover' some wonderful association between things that nobody had even thought existed. But it isn't all smooth sailing since quite often the implication is that one variable somehow causes the other and there is ample evidence of spectacular failures in this regard. This paper looks at some of the more fascinating conclusions reached by researchers and the reader can decide for themselves whether they feel that there is really something in it or it is all just nonsense. The scrutiny of the type outlined in this paper is essential if to uncover dangerous false conclusions that may be unwittingly quoted or even used by practitioners.