The controversy stems from this paper by a U Calgary economist about isolating differences in efficiency gains arising from listening to music from each of the band's main singers.
I was actually pointed to this paper by Marginal Revolution, who got it from the same source as Levitt, but it's clearly a joke. The 'paper' is about three-and-a-half pages text. It wasn't published, and nowhere does it suggest it will be published. However, Levitt, famed for being unconventional and humourous about economics, promptly starts bashing the author for being unconventional and humourous about economics.
Oh my. He did subsequently apologize (kinda). The answer to Levitt's query about why he did the experiment? He didn't. It was done by a grad student who subsequently quit the program.
Hat tip: Worthwhile Canadian Initiative
Disclosure: I have never listened to an AC/DC song of my own volition.