← Research library
ResearchWeak evidence

Fifteen students per group, a block-tapping test, and one frequency that moved it

Ahmadi and colleagues (2021) · Avicenna Journal of Neuro Psycho Physiology · Summary by Sentiosonics research agent
Year
2021
Stimulus
Beta
Design
Controlled pretest-posttest study with an unstimulated control group
Sample
n = 60

Summary

This is a small study that reports a clear result, and the gap between those two facts is the whole point of grading. Sixty Iranian undergraduates were divided into four groups of fifteen. Three groups heard twelve minutes of binaural beats in the alpha, beta or theta range; the fourth sat without any audio. Everyone took the Corsi block-tapping test before and after, which measures how long a sequence of highlighted blocks you can reproduce.

The group means tell the story plainly. The beta group, hearing 240 Hz in one ear and 255 Hz in the other for a 15 Hz beat, went from 5.63 to 6.64. The theta group went from 5.43 to 6.03, the alpha group from 5.50 to 5.83, and the unstimulated control from 5.60 to 5.66. Analysis of covariance on the post-test scores, with the pre-test removed, found a significant difference between the stimulated and control groups, with eta squared of 0.14. On the post-hoc comparisons only the beta group separated from the control.

So the claim the paper makes is narrow: twelve minutes of a 15 Hz beat raised Corsi span, and alpha and theta beats did not. We grade it weak, which is what a score of 3.3 means on our rubric, and the main drivers are fifteen people per arm, a control group that received nothing at all rather than a matched sound, and no blinding. With no sham audio, the beta group’s gain and the control group’s flat line are equally consistent with entrainment and with the difference between doing something and doing nothing for twelve minutes before a retest.

The paper’s statistics are reported inconsistently, and a reader should know. The post-hoc paragraph states that the beta and control groups differed significantly while printing P=0.30 next to that claim, then states that the other pairs did not differ significantly while printing P<0.05 next to those. Those two p-values are transposed relative to the sentences they sit in. The abstract has a similar problem in its methods line, listing four stimulation frequencies for three groups, where the body describes three protocols at 9.5, 15 and 6 Hz. The direction of the intended result is not in doubt, since the group means and the covariance analysis agree with the abstract. What is not checkable from the published text is the exact post-hoc p-value for the one comparison the conclusion rests on. We have graded what the paper reports and flagged what it reports twice in two different ways.

Limitations. Fifteen per group, one session, one twelve-minute exposure, and a retest minutes later, so nothing here speaks to repeated or prolonged use. Convenience sample of undergraduates at one university. No sham, no blinding, no follow-up, no EEG, so despite the title there is no measurement in this study of whether any brainwave synchronisation actually occurred. Three stimulation arms compared against one control multiplies the comparisons without a stated correction. Read beside the within-subject trial that tested two kinds of memory at once, which found a visuospatial benefit and a verbal decrement from a 10 Hz beat, and the 1,000-person home-use experiment, where beats lowered reasoning scores, this sits at the optimistic end of a literature that does not agree with itself.

Our analysis

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.

Study designControlled pretest-posttest design with four parallel groups. The authors describe it as quasi-experimental and recruited by convenience sampling, while also saying participants were randomly divided, so we do not score it as a randomized trial.
Sample size60 students in total, which falls in the rubric's 50-99 band. Scored one point below that band because the four groups are 15 each, so every comparison the paper draws rests on 15 against 15.
Control conditionThe control group was, in the paper's words, not stimulated. No sham audio, no matched listening time, so attention and the experience of wearing headphones are not controlled.
BlindingNo blinding reported for participants or testers. The outcome is a performance task rather than a questionnaire, which is why this is not scored lower.
Outcome measureThe Corsi block-tapping test, a standardised and widely used measure of visuospatial working memory span.
ReplicationSingle study. No replication of this protocol found.
Conflicts of interestFunded by an internal research grant from Azarbaijan Shahid Madani University (contract 97/85-2). No conflicts declared and no commercial interest in the result.
Peer review & venuePeer-reviewed open-access journal published by Hamadan University of Medical Sciences, with no PubMed record and five citations to date. A local lower-tier venue rather than an indexed one.
Overall grade · 3.3 / 5Weak evidence

Graded by our research agent on Nov 2, 2026.

Get updates from the field

A monthly letter: new research, graded, and the claims worth checking. No hype.

Monthly, one click to unsubscribe. See our privacy policy.