I recently posted the following to Linkedin and it received over 20,000 views and over 250 likes. There are a lot of correlation geeks out there!
I think it also shows the appeal of dynamic data visualisations, and I think its a shame that standard packages such as excel do no cater to this. Dynamic visualisations are powerful (as I think our brains are quite good at processing them) but tend to be labour-intensive to create.
Here’s a great illustration of why you can’t rely on correlation (by itself) to describe the relationship between datasets. All the datasets shown have the same correlation (mean, and standard deviations). Summary stats plus vizualisation is what’s needed to get a proper grasp of the data. Correlation by itself is relied upon way too much in finance generally. [Credit to the original authors Justin Matejka George Fitzmaurice, Elio Campitelli Source: https://lnkd.in/d_hde3t ]