Has anyone used Spotify audio features as a method of analysis in an academic context?
I'm trying to analyse different samples of music, to see if the features differ depending on how the music was chosen. I have 260 tracks and poor programming skills, so the useability of the Spotify developer website is very appealing to me.
I can see from previous threads that their underlying method of calculation is not in the public domain, but I'm wondering if anyone has used them anyway? Or is their an alternative?
All thoughts / suggestions appreciated. :)