A dense electrode cap, thirty-one people, and two minutes of beats that changed nothing
- Year
- 2012
- Stimulus
- Binaural beats
- Design
- Within-subject counterbalanced crossover with white-noise control periods
- Sample
- n = 31
Summary
If a 7 Hz beat entrains cortical rhythms, a high-density electrode cap on thirty-one heads is a reasonable place to catch it happening. This study looked, with white noise as the comparison and a vigilance task running throughout, and reported nothing: no change in cortical power at either frequency tested, no change in vigilance, and no personality trait that predicted a response.
The procedure ran 13 minutes per participant and alternated conditions tightly. Two minutes of white noise, two minutes of beats, a 30-second white-noise rest, two more minutes of beats with the carrier tone moved to the other ear, another two minutes of white noise, then the whole sequence again at the second beat frequency. The beats were 7 Hz, built from a 400 Hz standard tone against a 407 Hz carrier, and 16 Hz, from 400 Hz against 416 Hz, played at 70 dB SPL. Which frequency came first was counterbalanced across participants. Throughout, participants did a continuous zero-back task, with capital letters appearing on screen every five seconds, which both measured vigilance and kept eye movements under control.
Neither frequency produced a significant difference in cortical frequency power compared with the white-noise control periods. Vigilance did not change either, on responses or reaction times. The authors also tested a specific secondary hypothesis, because earlier work with flickering light had found that personality predicted who entrained: they measured the Big Five with the NEO Five Factor Model and found no relationship between any trait and either the EEG or the vigilance outcomes. The conclusion they draw is narrow and well matched to what they did, that a short presentation of steady-state binaural beats is not sufficient to alter vigilance or entrain cortical frequencies at these two bands.
We grade this unsupported on the result override, with a score of 3.8, the highest score of any null in our library. The design is sound, the instrumentation is better than most, the funding is clean and the reporting is careful about its own scope.
Limitations. The exposure is the thing to hold onto: two minutes per condition, which the authors flag in their own abstract and which is far shorter than the fifteen to thirty minutes used by studies that report effects, let alone the hours people play these tracks for. A null at two minutes does not establish a null at thirty. Thirty-one participants with no power calculation leaves room for a small effect to go undetected. Only two frequencies were tested, both on a 400 Hz standard tone. Participants were doing a cognitive task throughout, so this is a test of beats as background to work rather than beats listened to attentively, and white noise is an imperfect comparison for a tone pair. Fourteen of the forty-five people recruited were dropped, mostly for corrupted files and EEG artifact, which is a lot of attrition for reasons unrelated to the question; two of them were dropped because they could not hear the beats at all.
How we graded this study
Each criterion is scored 1–5. Study design and replication count double in the overall score (how we grade). Graded from: full text read.
Graded by our research agent on Nov 16, 2026.