Sabroso Shot #16 - Spurious Correlations
I love data projects. Particularly when they're unusual or unexpected.
I think one of the most astonishing things about data thought is that people seem to buy it as fact simply because it is presented as data - and there are few things more dangerous than falling for the 'correlation = causation' fallacy.
Tyler Vigen produced an interactive page which digs into some very amusing demonstrations of the correlation = causation fallacy.
These correlations appear to correlate with 90%+ of each other, which is a hugely strong correlation, but the choice of variables acts to show the weakness of 'correlation = causation' in a Python-esque level of ridiculousness.
I'm by no means a data scientist, but I do find data interesting - and I find it doubly interesting when people expose bad science. Ben Goldacre has made a career from doing just that. I recommend his Ted talk as a starting point, but his books Bad Science and Bad Pharma tell you just how messed up the world is because of people using data incorrectly.
As a final fun use of data, I give the example of a man who tracked his heart rate while proposing to his wife. This proposal happened at the Colosseum in Rome, which incidentally is where I proposed to my own wife. (I share the thoughts in the annotations of 'no turning back now' once the Roman Forum comes into sight).
Seeing the creeping tension and nerves turned into a spike on a graph seems silly, but there's something poetic about it I think.
Sign up to the Sabroso Shot today
The Sabroso Shot is full of great ideas for your next campaign, our own musings and commentary on the latest content trends and an occasional showcase of what we're working on.