The findings genuinely conflict
There is no shortage of studies on background noise and cognitive performance, but they do not converge. Some find improvements on memory and attention tasks; others find no difference, or interference.
That inconsistency is not a sign of sloppy work. Too many variables move at once: task difficulty, noise level, the participant's baseline attention, and the ambient noise the room already had. So both 'white noise improves concentration' and 'white noise does nothing' overstate what is known.
Two mechanisms people propose
When researchers try to explain why some people benefit, two ideas come up repeatedly.
Stochastic resonance
In some systems a moderate amount of noise makes a weak signal easier to detect. The hypothesis that this applies to neural systems underpins reports that moderate noise helped people with attention difficulties in particular.
The word doing the work is 'moderate'. Even in this account, louder is not better: past a point the effect disappears and only the interference remains.
Arousal and the inverted U
Performance tends to be poor when arousal is very low, poor again when it is very high, and best somewhere in between. In a silent room that makes you drowsy, background sound can lift arousal into a useful range. If you are already keyed up, the same sound pushes you past it.
This is why the same person gets different results at nine in the morning and the night before an exam.
Masking is the less controversial part
Set the question of enhancement aside and masking is on firmer ground. Speech is unusually hard to ignore because it carries meaning, which is why a nearby phone call disrupts you more than music does. Background noise reduces the intelligibility of that speech, so your brain stops trying to decode it. Sound-masking systems in open-plan offices exist for exactly this reason.
A second, less discussed benefit is contrast. In a silent study room, one scraped chair is an event. With a low noise floor, sudden sounds are proportionally smaller and less startling.
Testing it on yourself
Since individual variation dominates, your own data beats anyone's review. A week is enough.
- Compare within the same subject and the same time of day. Mixed conditions tell you nothing.
- Run odd-numbered sessions in silence and even-numbered ones with pink noise.
- Record two things per session: measurable progress (pages, problems) and how often your mind wandered.
- Compare after a week. If there is no difference, you have learned something useful - skip the noise.
- Treat music with lyrics as a separate condition entirely. Incoming language changes the picture.
Two cautions
Claims that specific frequencies or binaural beats put your brain into a chosen state are not well supported. Sound is an environmental adjustment, not a mechanism that manufactures concentration.
Watch volume and duration. Hours of loud audio is a hearing risk with no upside; quiet enough to recede into the background is better on both counts.