That is a really poor sampling of scores, though. That makes their "Beethoven" example an outlier from someone who wouldn't even be in the top if you included the full spectrum of scores. There was also no link to a study, just a tweet saying there was a study and a URL that doesn't work. This smells a lot more like someone trying to push their own agenda than objective data analysis.

What about the converse? Are the stars of conservatories of music particularly good at math and such? Interestingly, most music here is geared toward millennials and older gens.

Families who insist you study 12 hours a day for higher SAT scores are also likely to believe in 'correct' sort of music and enforce it until that's all you like.

So the widths of the labels come from the length of the name in them, not the spread of SAT scores they correspond to.

And I guess the position of the label is the average for people who say they like that band. Why didn't they plot the spread ?

And where is the data used to make this ?

Looks bogus.