I agree with Beatrice, there is no unethical sample size. Also, the the sample size is defined by the expected statistical accuracy, the economics and the feasibility.

Until you clearly define your goals, method, and limitations, publish your results in the rational details (e.g. sample size and implications), do the proper statistical analysis and interpretation, you are fine. Even if you cannot conclude some things (e.g. you cannot reject or justify your hypothesis) your research can be very useful for others to develop better research method.

Ethical Considerations of Adequate Sample Size

Please let me know if these references/sites are helpful to you:

1. Ethics and Sample Size | American Journal of Epidemiology

https://www.academic.oup.com/.../2/105/256528/Ethics-and-Sample-Size

... if their sample size is not large enough to ensure adequate ... ethics and sample size ... to the ethical considerations in sample size selection ...

Site reference:

https://www.academic.oup.com/aje/article/161/2/105/256528/Ethics-and-Sample-Size

2. Ethical Dilemmas in Sampling

www.jswvearchives.com/spring2010/2dattalo.pdf

Ethical Dilemmas in Sampling ... ethical if participants agree to take identified risks and to bear, ... draw sample size for adequate

Site Reference:

http://www.jswvearchives.com/spring2010/2dattalo.pdf

Dennis

Dennis Mazur

There are no unethical sample size. If you have only a few e.g. patients with a rare disorder, do a qualitative study then.

I agree with Beatrice, there is no unethical sample size. Also, the the sample size is defined by the expected statistical accuracy, the economics and the feasibility.

Until you clearly define your goals, method, and limitations, publish your results in the rational details (e.g. sample size and implications), do the proper statistical analysis and interpretation, you are fine. Even if you cannot conclude some things (e.g. you cannot reject or justify your hypothesis) your research can be very useful for others to develop better research method.

It is more on doing a good science which make sense and a sample size that truly represent the population. I think reading more about statistics in your field would be very helpful. for example there are informative books on determination of sample size in clinical studies, health science and engineering. It is important to know what you are after, what statistical test going to run on the data and decide on sample size. In my opinion the ethical discussion here is not directly relevant. or if your sample size is not a good example of the population then you might call it unethical.

Sample Size Calculations in Clinical Research by Chow et al is an exmaple

Please also see an old book : how to lie with statistic ( It is more on examples of bad statistics and how to avoid)

As Amin's first post indicates it's important to separate your methodological concerns from ethical ones. Sample size (and indeed the 'nature' of how you sample - cf Beatrice and Robert above) - depends upon a combination of statistical (probability) theory and method. Thus, e.g. surgical research and ethnography often deals in small samples, while population 'claims' require larger samples but could vary between 'random, selective and convenience' samples. The ethical issues are: whatever sampling method your 'theory' suggests, might it harm anyone (individual, group or society)? And, if your sampling is poorly done, might that, again do any harm as a result of poor quality research?

This is a statistical question that does have ethical implications. if you are unable to recruit enough subjects for your expected effect, comparison, statistical tests, etc., then your study should be cancelled partially due to ethical reasons. Our assumption should always be that enrollment in a trial is a burden on the subject because they are being exposed to unknown risks. Without a reasonable expectation of results that benefit people generally, it is unethical to place those burdens of unknown risks on our subjects.

Th first ethical question before taking a sample is "what is the purpose of the sample?" Is the research a statistical enumeration of the past, then statistical sample size is available in many reliable textbooks. On the other hand, if the sampling is for the purpose of developing a theory then it is a matter of abduction i.e.testing weak to strong hunches) which will involve several ethical questions. Note, statistics does not predict, only a theory predicts. ON this topic read Charles Sander Pierce and W. Edwards Deming.