You are not going to settle the science from your desk. You can settle something narrower and considerably more useful, which is whether a particular track earns twenty minutes of your week.

The brainwaves guide walks the evidence and lands on a mixed record. A systematic review of 14 EEG studies found five that fit the entrainment idea, eight that conflicted with it, and one mixed.[4] A meta-analysis of 22 studies reported a medium pooled effect across cognition, anxiety, and pain, varying by beat frequency, timing, and exposure.[5] Neither of those tells you what happens when you put headphones on at 9 a.m. on a Tuesday.

This page is the protocol for finding that out, and most of it is about the one problem that makes self-testing audio harder than self-testing almost anything else.

What a personal test can and cannot settle

Testing one intervention on one person is a real design with a name. N-of-1 trials borrow the methodological elements of group trials to evaluate a treatment in a single patient, and there is a reporting standard for them: the CENT 2015 extension to CONSORT, written around individual trials and series of prospectively planned, multiple crossover N-of-1 trials.[3]

Read the phrase "multiple, crossover" carefully. The formal version alternates between conditions repeatedly and plans the whole thing in advance. Trying a track once, liking it, and buying a subscription shares nothing with that.

What your version can conclude: whether this audio fits the way you actually work. That is a question about you and your week, and it is the question you were asking.

What it cannot conclude: whether brainwave audio works, whether entrainment occurred, or whether your result would replicate in anybody else. A personal log produces a personal decision.

The problem you cannot fully remove

Here is what makes audio harder to test on yourself than caffeine, a schedule change, or a new chair. You will always know whether the sound is playing, and you will usually want it to work.

That combination has been measured, and the result should worry anyone planning to judge a track by feel.

Researchers gave 37 participants a cognitive task alongside tones they had been told would help or hinder performance. Reaction times and success rates showed no significant differences across the placebo, nocebo, and control conditions. Participants nevertheless rated the supposedly helpful tone as significantly more beneficial, and believed it had improved their performance.[1]

Nothing changed in the measurements. The belief formed anyway, and it formed specifically around a tone.

The effect is not hypothetical for beat audio either. A 2024 study played binaural beats to 141 participants while varying what they had been told about the treatment's effectiveness, from 0% to 100% in 25% increments. Learning rates shifted with the instructions rather than tracking a straight line, peaking under maximum uncertainty.[2]

That was a laboratory learning task with deliberately deceptive instructions, and the authors flag that they never checked whether participants believed them. Take the narrow lesson: what you have been told about a track before pressing play is itself an active ingredient.

So a self-test that asks "did that feel better?" is close to guaranteed to return yes. The design below exists to route around that as far as an honest amateur can.

Choose an outcome you cannot talk yourself into

Pick one number, decide it before you start, and make it something you could show another person.

The strongest available option for focus work is the one the Deep Work Runway already uses: minutes from your planned start to your first useful action. It has a clean definition, it is hard to fudge after the fact, and opening the document does not stop the clock.

Other defensible choices, depending on what you are testing:

  • Words drafted, or cases cleared, in a fixed 45-minute block.
  • Number of times you left the task to check something, counted with a tally.
  • Whether you completed the planned output, recorded as yes or no.

Notice what is missing. How focused you felt, how calm you felt, and how productive the session seemed are exactly the measures the placebo research just contaminated. Record them if you like, and keep them out of the decision.

The design

Ten blocks, five with audio and five without, run across two weeks.

Hold the obvious things constant. Same type of task, same time of day, same block length, same location. A test comparing a Tuesday morning strategy block against a Thursday afternoon inbox session measures the difference between Tuesday and Thursday.

Randomize the order rather than alternating. Flip a coin the night before, subject to using five of each. Strict alternation lets you anticipate tomorrow's condition, and anticipation is the variable you are trying to contain.

Change one thing. One track, one volume, one set of headphones for the whole test. Swapping formats midway leaves you with ten blocks and no comparison.

Log the number and nothing else. Date, condition, the one measure, plus a word if something unusual happened. Slept badly. Client emergency at 9:15.

Getting closer to blind

You cannot fully blind yourself to whether sound is playing. You can get closer than most people bother to, and the gap between the two is where a personal test stops being theatre.

Compare two active conditions. Instead of audio against silence, test your candidate track against a second one you have no expectations about, such as plain instrumental music or ambient noise. Both conditions have sound, so the difference in expectation between them is far smaller.

Have someone else queue it. A partner or colleague sets the playlist each morning from an identically named pair. You will often guess, and guessing correctly some of the time is still better than knowing every time.

Write your prediction before each block. One word: better, worse, or same. If your predictions turn out to track your results closely, expectancy is doing visible work in your data, and that is worth knowing before you act on the numbers.

None of this makes the test rigorous. It makes it honest about which parts are not.

Reading the result

Compare the five audio blocks against the five without. Look at the middle of each set rather than the best block in either, since one exceptional morning will otherwise decide the whole thing.

Then apply a threshold you set in advance. A difference that would not change your behaviour is not worth acting on. If your time to focus runs 14 minutes without audio and 12 with it, that gap is well inside the noise of ten blocks and does not justify a subscription.

Be similarly careful in the other direction. Ten blocks cannot prove the audio does nothing. It can show that no difference large enough to matter to you turned up in your own work, which is a sufficient basis for a decision and not a scientific finding.

Four things the log can tell you, each with an obvious next move.

A difference you would notice without the log. Keep the track, and re-test in a few months if the effect seems to fade.

No meaningful difference. Stop paying for it. You ran a fairer test than most of the people writing reviews.

Worse with audio. This happens and it rarely gets mentioned, because nobody sells it. Sound competes for attention, and anything with lyrics or a shifting texture competes harder during language-heavy work. A negative result is a result. Drop it.

The measure itself was wrong. Your blocks kept failing for reasons audio was never going to touch, such as an undefined task or an unresolved decision underneath it. Fix that first and the audio question becomes answerable later.

Where Awakened Mind sits in this

None of the studies above were run on this app, and general research on auditory stimulation does not independently show that Awakened Mind entrains brainwaves or produces any intended state. The app's own terms describe it as a general-wellness tool rather than a medical device.[10]

If you want to run this protocol using one of its programs as the candidate track, that is a reasonable use of it, and the result you get belongs to you rather than to the product. A test worth running is one the product can fail.

Start with five blocks

  1. Write down your one measure and your threshold today. What you will count, and how big a gap would change your behaviour.
  2. Flip for the order of your next ten blocks. Five each way, recorded before the first one runs.
  3. Log the number and your one-word prediction after every block. Nothing else goes in the log.

Beat stimulation research has been described as promising and inconsistent for a decade, with reviews repeatedly unable to name a best frequency, timing, or schedule.[6] That is unlikely to resolve in time to help you decide about Thursday. Ten blocks and a threshold will.

If the honest answer turns out to be that the audio changed nothing, you have lost two weeks and gained a subscription you no longer pay for. The brainwaves guide explains why that result would be entirely consistent with the evidence.

References

  1. Cognition and the placebo effect: dissociating subjective perception and actual performance. PLOS ONE, 2015. Link Limit: 37 participants on a Flanker interference task in a laboratory. Objective performance showed no significant differences while subjective ratings did.
  2. Uncertainty of treatment efficacy moderates placebo effects on reinforcement learning. Scientific Reports, 2024. Link Limit: 141 participants, a laboratory learning task, and deliberately deceptive efficacy instructions. The authors did not measure whether participants believed them.
  3. CONSORT extension for reporting N-of-1 trials (CENT) 2015 Statement. Journal of Clinical Epidemiology, 2015. Link Limit: a reporting standard for formal N-of-1 trials in clinical settings. An informal self-test does not meet it.
  4. Binaural beats to entrain the brain? A systematic review of the effects of binaural beat stimulation on brain oscillatory activity. PLOS ONE, 2023. Link Limit: of 14 EEG studies, five fit the entrainment idea, eight conflicted with it, and one was mixed. No clear trend either way.
  5. A meta-analysis of the effects of binaural beats on cognition, anxiety, and pain. Psychological Research, 2018. Link Limit: 22 studies combining different problems, people, and methods. Does not show that an entrainment mechanism caused any of it.
  6. Auditory Beat Stimulation and Its Effects on Cognition and Mood States. Frontiers in Psychiatry, 2015. Link Limit: studies vary widely in frequency, timing, session length, and masking, with positive, null, and conflicting outcomes.
  7. Binaural beats for stress in non-clinical populations: a systematic review. Advances in Mental Health, 2024. Link Limit: mixed results, no best frequency or schedule identified, and adverse effects often unmeasured and unreported.
  8. Safe listening devices and systems: a WHO-ITU standard. World Health Organization, 2019. Link Limit: covers sound exposure generally. Does not establish safe practice for any specific listening habit.
  9. Musicogenic Seizures. Annals of the New York Academy of Sciences, 2003. Link Limit: rare and highly individual. Available reports do not implicate consumer beat frequencies in the general population.
  10. Awakened Mind Privacy Policy and Terms of Service. Awakened Mind, 2026. Link Limit: the product's own terms, describing a general-wellness tool. Not a research finding.