Correlation - Perfectly Spurious?

Business students looking at sales forecasting are likely to come across the important of correlation - and the strength of the relationship between a dependent and independent variable. However, the alert student should also be on the look out for spurious correlation. Suprious correlation is the false presumption that two variables are correlated, when in reality they are not!

Here is a good example of almost perfect correlation (as indicated by the regression line) between the fatality rate on US highways and the quantity of lemons imported into the USA from Mexico.

The policy implications are clear - more lemons should be imported if the US wishes to reduce further its road fatality rate….unless, of course, there is a problem with that correlation