I'm pretty sure this is related in Peopleware (if you haven't read it, you should), but I'm not sure. My copy is in the office.
A group of programmers were asked whether they preferred music or not as they were programming. They were then separated into two groups, each group having roughly equal numbers of those who did prefer music while programming, and those who didn't.
Group A was then played acceptable music as they solved a programming problem, and gropup B was left in silence.
As I recall, the only programmers that performed badly were those who preferred silence, but were played music. That's pretty reasonable. If you prefer music but get silence, apparently it doesn't matter too much. If you prefer silence but get given music, apparently it damages your productivity.
But there was a twist. The programming task given was a series of transformations to a set of data, and was designed so that the net effect was to return the data to its original form.
The only programmers who noticed this were in the group without music.
OK, so the conclusions are clear. Based on this story, half-remembered from a book, without references, and without detailed data about the experiment, we can conclude that if you're happy with music then it won't hurt your productivity, but ...
The suggestion is that there is some part of your creativity that is inhibited by having music on, whether you prefer music or not.
This doesn't, of course, address the issue of using music to mask background noise and distractions, and there are many, many points to make, but it's interesting.
it would be interesting to see if there was a difference between music with words and music without-- whether the words themselves occupied a part of the brain that could otherwise be used toward the task at hand. I listen to music when I code, but only fast-paced instrumental-only music.
the music I like is not especially popular (or listenable, some might say), so I don't think I would like a communal music system regardless of the content.
Interesting. In our case, all the engineers wanted to play music, so the point is slightly moot. It'd be interesting to upload some silence mp3s and see if they get voted for.
This seems to me like something that could completely ruin productivity.
Just let everybody use headphones, that way they can put songs they like on their computer and only listen to those. The way this seems to be set up, all that everybody is going to end up doing is voting for songs that they like so that those songs will be played next, and they will be paying more attention to getting votes in than the work they should be doing.
Ok, so, we did the headphone thing for a while, but we like to all talk to each other, and waving vigorously to catch someone's attention gets old fast. Likewise, chatting in our IRC channel is not quite the same as talking directly.
Concentration-wise, we all seem to do OK - if someone needs to zone out for a while, they still can put on headphones, or possibly just use all 5 votes to queue up 5 long classical pieces.
Of course, every team is different, and interacts differently, so your milage may vary.
I hardly ever listen to music while really coding. If i'm doing the same ole boilerplate stuff, then yeah, but if i'm doing data transforms or trying to decipher bug reports, its not a good idea to have music on while trying to comprehend. But since I only code 10% of the time, the rest is perfect for music listening, especially trying to drown out the insanity that is my office. Or signaling for people to walk away.
I recall someone's suggestion to wear ear buds, even if your are not playing anything. The social norm is to avoid interrupting such a person -- keeps at least some of the interruptions at bay.
I've found some CD's of "white noise" and I intend to try playing this. In the past, I've used a fan to take the edge off of background noise. One of the CD's is for "fan noise"; maybe this will help.
I'm cautious of playing anything too loud for too long. Even lower level noise can be damaging when heard for extended periods of time.
A group of programmers were asked whether they preferred music or not as they were programming. They were then separated into two groups, each group having roughly equal numbers of those who did prefer music while programming, and those who didn't.
Group A was then played acceptable music as they solved a programming problem, and gropup B was left in silence.
As I recall, the only programmers that performed badly were those who preferred silence, but were played music. That's pretty reasonable. If you prefer music but get silence, apparently it doesn't matter too much. If you prefer silence but get given music, apparently it damages your productivity.
But there was a twist. The programming task given was a series of transformations to a set of data, and was designed so that the net effect was to return the data to its original form.
The only programmers who noticed this were in the group without music.
OK, so the conclusions are clear. Based on this story, half-remembered from a book, without references, and without detailed data about the experiment, we can conclude that if you're happy with music then it won't hurt your productivity, but ...
The suggestion is that there is some part of your creativity that is inhibited by having music on, whether you prefer music or not.
This doesn't, of course, address the issue of using music to mask background noise and distractions, and there are many, many points to make, but it's interesting.
Some references on the net:
Slides 16 and 17 from http://www.ecst.csuchico.edu/~tyson/classes/630.s07/11.pdf
Search for the reference to Peopleware on:
http://discuss.fogcreek.com/joelonsoftware1/default.asp?cmd=...
http://discuss.joelonsoftware.com/default.asp?biz.5.593799.3...