# How should the sample-size tag be used?

is one of our most popular tags (54th), categorizing 232 threads as of July 14, 2013. However, the topic sample size is very broad and not very distinctive, in my opinion. There are a number of other tags that are related to this topic or overlap some aspects of the topic, including:

At present the tag wiki excerpt for sample-size reads:

Use this tag for questions where it is desired to know an appropriate sample size to perform a test or analysis or when the sample size is relevant to the question.

I'm not sure this provides the ideal guidance on the use of the tag. It seems to me that if someone is trying to figure out what sample size they need for their study, the most appropriate tag might be power-analysis, although I can also see how someone might say that tag should only be used for questions about power analyses (say, the difference between a-priori and post-hoc power analyses) instead.

Moreover, skimming down the list of questions tagged with sample-size, perhaps about half are tagged with one of the above as well, and it's not clear what having the tag sample-size in addition adds. There are, of course, cases where the other tags do not occur and sample-size seems appropriate, but quite often the manner in which each of those questions is related to sample size seems fairly idiosyncratic.

I'm wondering how threads should best be organized with respect to these categories.

For what it's worth, my suggestions would be:

• Copy the excerpt and tag wiki from big-data (which is somewhat more informative) to large-data (which seems to be the preferred tag), and make big-data a synonym.
• My preferred tag wiki excerpt for sample-size might read:

This tag is very ambiguous. Please use only if one of the following is not more appropriate: [small-sample], [large-data], [power-analysis], [power], or [underdetermined].

However, this may not be the best approach, and I'd prefer to open this up to community discussion here than make changes on my own with the only oversight being conducted by two suggested edit reviewers.

• Your suggestions seem reasonable to me. – Gala Jul 15 '13 at 9:23
• It seems that, even without the sample size tag, there's a lot of overlap among the others you listed. In practical situations, power is rarely discussed without considering sample size (except in math. stat. homework problems perhaps). I wouldn't be opposed to deleting one or more of these tags but that's probably more work than it's worth just to prevent a little content overlap. In any case, I don't think that's a good tag wiki excerpt since it doesn't actually tell you anything about the tag. If there's going to be an excerpt that basically says "this tag sucks" then just delete the tag. – Macro Jul 15 '13 at 15:53
• That's a good point, @Macro. My actual position is only 2/3rds of the way to "this tag sucks", though, & deleting the tag is outside of my purview. Perhaps you have a suggestion for a reasonable excerpt? FWIW, there is some precedent for an excerpt like this (see table on SO). – gung - Reinstate Monica Jul 15 '13 at 16:05
• I'm not really sure. I think a mix between the current excerpt and what you've written is on the right track. If I had to write something, I might say: "Use this tag when the substantive part of the question is inherently related to sample size. For example, questions about power as it relates to the sample size or about methods used for small sample sizes would be appropriate. Please use only if one of the following more specific tags is not better suited: ....". – Macro Jul 15 '13 at 17:27
• @gung, I saw due to the approval of this thread you suggested a new tag excerpt and it was accepted. So, maybe you could answer your own question or add and edit to it, saying the issue was handled. It is just a detail, but here stays the suggestion/idea. Thank's. – Andre Silva Jul 29 '13 at 16:53