"An increase of 10 dB reduces productivity by approximately 5%."
I'm surprised it isn't more. As someone who is particularly distracted by noise and interruptions, I usually have some music without lyrics playing in noise cancelling headphones whenever I'm in an open office environment.
The real question is when are companies going to pick up on research like this and make changes to maximize productivity. What do offices at FAANG look like?
My productivity increases with regular brown noise -- it "covers-up" the irregular traffic and train noises outside my apartment so my brain doesn't pay attention to it. Brown noise is also pleasant to the ears.
Music: it depends. Any kind of forgettable music like muzak or background instrumental/classical helps. But anything with lyrics distracts me because it engages the language processing part of my brain -- I need that part to do my work. But I know many people who can work while listening to rock and roll.
Also the article makes a difference between cognitive function and effort-task performance.
If I was a mechanic or someone doing a mechanical task, I wouldn't mind blasting heavy metal in the background -- it gives me a "metronome" to sync to. But I wouldn't do the same if I was a tax accountant where I need the cognitive headspace to do careful work.
If we create a quiet office environment, everybody gets to pick their preferred noise via headphones.
I worked as a cook for years while I was younger. There was always a radio on (this is like 1980s-90s). Typical blue collar environment with tons of noise from machinery, raucous people, and the radio.
While doing prep work for the next meal time the music didn't bother me at all (so go through various fridges and the menu and make lists of tasks to prepare then execute on the list interrupted by deliveries, dishwashers quitting in the middle of their shift, and so on - peel carrots, chop onions, clean squid...).
While working the actual rushes on the line (preparing individuals meals to order as part of a team, timing courses to go out and maximizing throughput) the music was extremely annoying to me. My coworkers literally would go crazy without it so it stayed on but it was a major problem for me.
I eventually left cooking for comp sci (long story involving a health condition and a privileged life allowing me to do this) and discovered "psytrance". Finally, I understood how music could make you more productive. Since then I've gotten better at tuning out music with lyrics but it still bugs me when coding or writing.
I wish I could turn to the same genre of music to increase my performance. Like you lyrics are out with exceptions (for some reason I can sort of tune out vocal trance), but some programming I find lend itself well to something like bossa nova, while others I need something highly repetitive like house, and I can't determine ahead of time which will be better. And what really throws me is that usually something like funk will slow me down, but there are times it's the only thing that works. This would all be fine, if I could just roll through the dial, but it takes at least 15 minutes before I know, and it's distracting messing with music until I find something suitable. usually I opt for no music if I'm somewhere quiet, even if there's a good chance music might make me more productive.
Classical music doesn't work for me. It's far too interesting and I'll catch myself pausing my work cause I got distracted by it! Same with jazz and the like.
It has to be some repetitive music/noise so it doesn't catch my attention.
Here is one way to formulate this. You are an information processing machine with certain throughput. Article says 10 dB worth of energy going into your ear causes an average of 5% decrease in your throughput, in the form of processing cost to filter irrelevant bits out. What we don’t know is what that 10dB stimulus is representative of. 10dB of baby crying vs 10dB of equally distributed, unpatterned brown-noise vs 10dB of low frequency predator sounds are by definition are not going to get assigned the same attentional saliency and will have different processing costs. Furthermore, your individual processing machine will accrue certain biases throughout, so same stimulus will not have the same processing cost for everyone. A typical example would be combat PTSD biasing processing towards gunfire, so it would cost more to filter out fireworks later on. In sum, there is a lot of aggregation and averaging going on in this calculation and this is why %5 feels too little a cost for 10 dB.
The study examines a factory workplace which may already have a high-level of background noise. Perhaps the effect is non-linear? Anecdotally, two or three coworkers talking near me doesn't feel more distracting than one coworker on the phone.
Well, the dB scale is already deeply nonlinear. Adding a 50dB conversation to 40dB ambient noise results in… 50.4 dB total noise. Add two more people each talking at 50dB (at the same time!) and the total noise is about 55dB.
I'm surprised it isn't more. As someone who is particularly distracted by noise and interruptions, I usually have some music without lyrics playing in noise cancelling headphones whenever I'm in an open office environment.
The real question is when are companies going to pick up on research like this and make changes to maximize productivity. What do offices at FAANG look like?