Two mistakes appear at opposite ends of research coverage. One treats a small study as proof that an intervention works for everyone. The other dismisses it because it is small. Neither starts with the right question: what was this study designed to establish?
A demonstration can answer a narrow question
Suppose a hypothetical research system produces a defined output in one person under documented conditions. That observation can answer whether the system achieved that task in that case. It cannot, on its own, tell us how often the same outcome would occur across many different people.
The difference is logical, not a judgment that the demonstration was unimportant. A narrow question can be a necessary and valuable one.
More observations are not automatically more participants
A study may record thousands of attempts. Those attempts can reveal within-person consistency, learning or variation. They do not turn one participant into a thousand independently enrolled people.
Our reporting therefore separates the number of people, the number of attempts and the duration of follow-up. If a headline uses a large observation count while hiding a small participant count, ask what unit is actually being counted.
The study purpose changes what to look for
FDA describes early drug studies as examining matters such as safety and dosage, with later studies addressing additional questions about effectiveness and benefit-risk. These are broad descriptions of drug development, not a rigid template for every device or experimental procedure. 1
An article should first describe the purpose of the study it covers. It should not congratulate a safety study for proving efficacy when that conclusion was never supported by its design.
A fictional arithmetic example
If two of three hypothetical participants improve on a defined measure, the observed fraction is two-thirds. Reporting “67% success” without the denominator makes a very small observation look more general than it is.
Adding “in this three-person example” restores the scale. It still does not establish what would happen in a different population or under a different protocol. The fraction is descriptive, not a promise.
Questions that improve the article
Who was eligible and who was actually included? What happened to every participant? Were outcomes defined before the results were examined? Is the comparison appropriate to the claimed conclusion? What follow-up occurred? What information is missing?
These questions are useful at any sample size. A large number does not automatically correct a mismatch between the design and the claim.
The right conclusion may be deliberately narrow
A strong report can say that a result justifies further investigation without calling it either a cure or a disappointment. It can explain what a larger or differently designed study would need to resolve.
Read the brain-to-voice study report as an example of keeping a research demonstration and a population claim separate. The size of the study belongs next to its result, not hidden after the reader has already formed a conclusion.