There has been some fierce discussion about this subject in blogosphere and on Programmers.SE. It all started with Joel Spolsky's claim:
[...] putting on headphones to drown out the ambient noise has been shown to reduce the quality of work that programmers produce
I've found following retort (not well referenced though):
- RAND data says exactly the opposite: that programmers who choose to listen to music see a 30% productivity boost. Of course, that’s probably because RAND is taking numbers from tens of thousands of industrial programmers who already have a reasonable place to work, whereas you’re looking at a dozen or so goons in a basement burning a million dollars of VC on ping pong tables and being confused why they can’t focus.
- Maybe you should learn that correlation does not imply causation, and that the productivity drops you see might be a result of the disasterously low quality work environment that you attempt to explain away as acceptable.
I feel both opinons are quite single-sided, but what does science say on the subject?