The 70 dB Creativity Study: What Mehta, Zhu, and Cheema (2012) Actually Found

You have probably seen the claim: a moderate level of ambient noise, around 70 decibels, boosts creative thinking. It comes from Ravi Mehta, Rui (Juliet) Zhu, and Amar Cheema's 2012 paper in the Journal of Consumer Research, “Is Noise Always Bad? Exploring the Effects of Ambient Noise on Creative Cognition”. The finding helped make coffee-shop soundscapes a familiar productivity tool. It also became advice that 70 dB is a magic number.

The actual result is more bounded. The researchers found benefits of moderate ambient noise on particular creativity measures, using a particular mix of environmental sounds. They did not establish a universal listening level. Understanding that distinction makes the study more useful for choosing sound for everyday work.

The Study at a Glance

What the Five Experiments Actually Tested

The paper did not give everyone one battery of creativity tasks. Its five experiments addressed different questions, so their results should be read separately.

In the first experiment, participants solved Remote Associates Test (RAT) problems: find a word that connects three apparently unrelated words. The 70 dB group solved more problems correctly than the 50 dB group, the 85 dB group, or the no-soundtrack control. The control was not acoustic silence: the laboratory still had background sound averaging about 42 dB.

The second experiment asked participants to generate ideas for a new mattress. Ideas received higher creativity ratings at 70 dB than at 50 or 85 dB. Later experiments used alternative uses for a brick and a shoe-polish problem, comparing 50 with 70 dB. Their results favored 70 dB on rated creativity or on originality and appropriateness. Better-rated ideas did not consistently mean a larger number of ideas.

The fifth experiment moved to a student lounge and examined willingness to buy innovative rather than conventional products as ambient noise varied. That was a consumer-choice result, distinct from solving a creativity problem under a controlled 70 dB condition.

These distinctions matter. A correct answer on the RAT is a remote-association result with one target solution. Generating alternative uses is an open-ended, divergent-thinking task. Interest in an innovative product measures something else again. The paper offers several related findings, rather than a single measure of creativity in all its forms.

Why 70 dB Is Not a Universal Optimum

The controlled comparisons sampled particular levels. Finding that 70 dB outperformed 50 and 85 dB does not tell us whether 65 or 75 dB would be better, or whether the same pattern holds for another sound. Three sample points cannot identify an exact optimum across every possible listening condition.

The lower comparison also needs care. In the first experiment, 50 dB was a low-noise condition and the separate control averaged about 42 dB. Describing the whole paper as a comparison of “silence, 70 dB, and loud noise” loses part of its design.

The soundtrack matters as much as the number. Cafeteria chatter, traffic, and construction contain changing events and different spectral features from continuous white, pink, or brown noise. A result with that environmental blend cannot establish that broadband noise at the same level produces the same cognitive effect. The study also provides no direct evidence that a particular noise app improves creativity.

For a related acoustic distinction, see equal-loudness contours: sounds with different spectra can feel differently loud at the same measured sound pressure level. A volume-slider position is not a calibrated measurement at the listener's ears.

The Proposed Mechanism: Processing Disfluency

Mehta and colleagues proposed that moderate noise introduces processing disfluency: a mild difficulty in processing information. In their account, that difficulty encourages a more abstract way of thinking, which can help people reach less obvious associations. At higher noise levels, the processing burden becomes too large and performance suffers.

This is an explanation the paper investigated, not a reason to assume that every mild distraction will help. The broader review by Alter and Oppenheimer (2009) discusses processing fluency as a cue people use in judgment. It provides theoretical background; it does not prove that a difficult font, an uncomfortable chair, or a different noise spectrum will reproduce the ambient-noise result.

The familiar inverted-U shape is useful as a way to think about a balance between stimulation and interference. Our Yerkes-Dodson article explores that broader idea. A similar curve, however, is not enough to establish that two studies share the same mechanism or the same ideal sound level.

What Later Research Adds

The five experiments in the original paper provide evidence within one research program. Independent studies with other participants, soundscapes, and tasks are needed to establish how broadly the result travels.

A 2019 classroom-noise study by Massonnié and colleagues illustrates the limits. It compared silence with classroom noise in children. Older children showed no creative benefit, while some younger children with weaker selective attention performed worse. That is a different population and design from the original adult experiments.

A 2022 study by Mones and Massonnié compared ambient sound with silence during divergent thinking in undergraduates. Ambient sound increased the number of ideas on average, but did not improve originality; the fluency benefit depended on cognitive flexibility. More ideas and more original ideas are different outcomes.

Neither study is a direct test of the original 50/70/85 dB comparison. Together, they give a reason to keep track of the task, listener, and outcome rather than treat ambient noise as a general creativity enhancer. They do not justify saying that independent replications have established an inverted-U curve with a reliable peak at 70 dB.

Creativity and Focus Are Different Questions

A creative task can involve several kinds of work. Brainstorming possibilities, selecting an idea, reading carefully, and checking a calculation place different demands on attention and memory. A sound environment that feels helpful for one stage may be distracting for the next.

For example, Salamé and Baddeley's 1982 experiments showed that unattended speech disrupted immediate recall of visually presented digit sequences. That is a specific verbal-memory finding, not evidence that every form of background noise harms every task. It does show why one positive creativity result cannot settle the question of sound for careful work.

The RAT result also makes a simple “noise helps divergent thinking and harms convergent thinking” rule too neat. The original paper found a benefit on a task with a single correct answer. The useful distinction is the actual work being done and the outcome being measured, rather than a label applied to an entire working day. Our music-versus-noise article discusses other ways background sound can interact with task demands.

How to Use the Finding in Everyday Work

Treat the study as a reason to experiment with your listening environment, without making 70 dB a target you must reach. Compare a comfortable ambient backdrop with a quieter session on a similar task. For brainstorming, notice the usefulness and originality of your ideas as well as their number. For execution, notice errors, interruptions, and how easily you can follow the details.

Keep the comparison practical: use similar session lengths and change one sound setting at a time. A single productive afternoon cannot tell you whether sound caused the improvement, but repeated comparisons can help you choose a setting that suits your routine.

Start at a comfortable low level and adjust only as needed. The paper does not supply a 60–75 dB recommendation for everyday listening, and setting a phone or app volume to “70” does not recreate a measured 70 dB experimental condition. If the sound pulls your attention away from the task, reducing it or returning to quiet is a useful result of the comparison.

Choosing Sound With dpli

The practical case for an adjustable tool is control over your listening environment. dpli Focus generates sound in real time and lets you adjust its spectrum and level. You can choose a comfortable backdrop and change it when your task changes.

That product use is separate from the 2012 finding. The ambient-noise experiments did not test dpli or establish a creativity benefit for generated white, pink, or brown noise. The research supports asking a better question: which listening conditions help you do this particular task comfortably and consistently?

Seventy decibels is one tested condition in a useful study. Its value comes from showing that background sound can interact with creative performance, and from giving us clear boundaries for what remains to be tested.

References