Correlation and Causation
In this topic we examine the well-known rule of thumb that correlation does not imply causation. This is a cautionary rule of thumb that encourages the analyst to not jump to conclusions. Studies that take this form often prescribe that because of X, Y has happened. A more appropriate statement might be: when X happens, Y also happens, demonstrating correlation between the two events.
Example applications
Correlation is a common statistical tool that indicates how two or more variables change together, either in the same direction (positive correlation) or in opposite directions (negative correlation).
Demonstrating correlation and causation
This topic is a cautionary one, so it’s about not jumping to conclusions. In order to discuss correlation and causation, you should be able to identify whether two variables show a linear correlation.
Key concepts
- an overview of what correlation is, and what causation is.
- an example of how correlation can lead to erroneous causation statements
- advice for the student engineer about how to be cautious around establishing cause and effect between two variables.
Core resources
Further information…
There are a number of ‘spurious correlations’ listed at TylerVigen.com, such as the number of people who drowned by falling into a pool, correlating with the number of firms Nicholas Cage appears in in a given year.