Data is interesting but it has to yield insight to be valuable. One of the tools we use to generate insight is correlation analyzes where we look at two sets of data and see how closely one tracks the other. We then try and figure out which is the driver of the other. Unfortunately, sometimes this is a fool’s errand. Tyler Vigen has created a series of comical graphs that show “spurious correlations.”
Here is an example:
The correlation factor calculates the degree of tracking, with 1.00 being perfect. Here the factor is 0.992, which is as close to perfect as anything I have seen. Yet, when we look at the two categories we intuitively know that the two data sets are unrelated.
This example proves the title!
If your correlation analyses are not producing guidance on how to improve your business then please contact Sam.