Q: What would be the best article type for my research with a small sample size?

Detailed Question -

I collected prevalence data from a small sample (n = 66). I would like to publish the work. I was wondering whether a short communication or brief research report are appropriate formats. I think a full research paper may not be appropriate as my sample size is small and analyses is limited to descriptive stats.

2 Answers to this question
Answer:

Hi Florence – Welcome to the forum! Your question is apt, and it’s great to know that you are being careful about publication planning. We have referred your question to our design/stats experts, and we shall get back to you once they provide a response. To help them along, it might help to provide some more information about your data/research if possible.

For now, you may also go through the following resources related to your query:

Until we get back…

Comment

Answer:

Hello again, Florence. Our design/stats expert got back – here’s their response.

In fact, n = 66 may be adequate: it all depends on what you are measuring, what is the size of the population, what (and how many) ‘treatments’ are included in the experiment, and so on.

Another question is the confidence level you have in mind. Leaving aside such rules of thumb as ‘30 is the minimum sample size’ or ‘10% of the population is a good sample,’ how the sample is chosen is more important than the mere size of the sample. In medical journals, for example, many case reports are based on even a single patient. If as an author you are convinced that the results are worthwhile, you should submit your paper to a journal.

The category of publication – whether a short communication, a letter to the editor, or a full-length paper – depends not on sample size but on other considerations, which are usually explained in the journal’s instructions to authors. Let the reviewer decide whether or not the sample size is adequate.

Incidentally, as you may be aware, the tables of Student’s t test (to be used for small samples) go only up to a sample size of 30. Beyond that, the normal approximation (a standard deviation of 1.96 for 5% level of probability, or p <0.05) is acceptable, which means that a sample of 66 is not small.

Hopefully, that should give you the push to go ahead and submit to your target journal. As an additional measure, you may also check this through a presubmission inquiry with the target journals. Learn more here: How to write a presubmission inquiry

All the best for your submission!

Comment