This tweet crossed my path, and it made me think about data, how it is interpreted and often used to support arguments it has no business supporting.
Causation vs Correlation
— Mark Chubb (@MarkChubb3) July 17, 2016
The whole idea is correlation does not imply causation. The connections between the two axes in any of the graphs provided at “Spurious correlations” seem unrelated. But then I thought, what if there really was some connection (regardless how small) between any of the two paired variables.
So here is the assignment I thought might work in a wide variety of disciplines that discuss how data is used to support arguments (in most cases causation is implied).
Students pick a graph/relationship from the site/book they find interesting . They then actually research and find as much data as they can about the two variables and attempt to weave a convincing, data-driven narrative for why the relationship between the two is due to causation and not just a random correlation. I could see this going in directions all the way from creative storytelling (semi-fictional) to some potential leads for more research (factual). How fun is that?!?
I would love to hear ideas for how to take this half-baked idea and make it great!
Thanks to @markchubb3!