At first I felt like they must have only surveyed professors, since professors love Python. But after looking at their methodology it's pretty obvious what happened, all of their searches are for hits for "X programming" where X is the name of the language. Yeah, I bet "javascript programming" returns less results in most searches than "python programming" for any of several reasons. We can start with the idea that "programming" wouldn't be part of the keyword for most professional developers, but would be for most data scientists and others for whom programming isn't their main job. There's also no animal called a javascript. The formulation "javascript programming" is just generally painfully redundant and awkward.
I don't much care for javascript personally but a result that runs this far against the grain is highly suspect. I would expect the results to be more in line with every other developer survey if they just compared hits on websites dedicated to programming (like stack) for 'python' vs 'js'.
Their github methodology is also weird, they count how many new repos are made that use a language. So every Python "hello world" counts for more than, say, the entirety of Linux. This is probably a better proxy for what language people are learning than what language is most used.
Biting the hand that feeds IT
