In the planning stage of scientific research, sample size determination is undoubtedly one of the most frequently encountered yet weakest sections in dossiers submitted to ethics committees. Statements such as “30 subjects were deemed sufficient based on similar studies in the literature,” “30 subjects were selected due to time and budget constraints,” or “Our advisor deemed this appropriate” are no longer acceptable justifications by today’s scientific standards. So, why has the number “30” turned into a scientific myth, and why is this approach now considered a methodological error?
Looking at the historical process, the number 30 is a threshold value associated with the Central Limit Theorem in statistical literature. Statistically, it is assumed that when the sample size reaches 30 or above, the sampling distribution approaches normality. However, this does not mean that “30 subjects are sufficient for every study.” The fundamental factor determining the power of a study is not the assumption of the Central Limit Theorem, but the effect size the study aims to detect. If the true effect size in your study is very small, it is impossible to make this effect statistically significant with 30 subjects. In this case, a study conducted with 30 subjects is destined to fail before it even begins.
Ethics committees do not separate the scientific validity of a research project from its ethical responsibility. If a study is underpowered—meaning it cannot reach a significant result because the sample size is insufficient—this study must also be questioned ethically. Risking living subjects or patient data for a study that can be predicted in advance to yield no meaningful results is incompatible with scientific ethical principles. Ethics committees now expect researchers to prove not just “how many people” they will work with, but also based on which “effect size,” “alpha error rate,” and “statistical power” parameters this number was calculated.
The answer to the question “Why 30 subjects?” must be a mathematical necessity, not an opinion. Sample size calculation must be reconstructed each time according to the type of study, the type of statistical test to be used, and the expected effect size. For example, the number of subjects required for a t-test versus an ANOVA design or a logistic regression model is completely different. While 15 people might be sufficient for one, 200 people might be needed to capture the same effect in another. Ignoring this and approaching every research with the number 30 causes statistical blindness.
The greatest risk for researchers is the rejection of the project or a request for revision if the ethics committee notices this unjustified approach. A researcher aiming to publish in a reputable academic journal should provide concrete data such as “A total of 128 participants were calculated to detect an effect size of 0.50 at a power of 0.80 and a significance level of 0.05 using G*Power/PWR software,” instead of an expression like “approximately 30 people” in the methodology section. This is a necessity not only to obtain approval from the ethics committee but also to prove the methodological quality and seriousness of the research.
In conclusion, the cliché of 30 subjects is a habit that has fallen behind modern scientific methodology. If you want the result of your research to be based on a truly observable and reproducible effect rather than just a quest for “significance” (p-value), you should determine your numbers based on statistical power, not tradition. In a scientific study, sample size should be shaped not by the researcher’s budget or habits, but by the nature of the data and the magnitude of the phenomenon you wish to measure. Remember, the ethics committee monitors not only the protection of subjects but also whether the study is scientifically meaningful and value-added. A strong methodological infrastructure accelerates ethics committee approval and increases the chances of your publication being accepted.
