2
$\begingroup$

I think something needs to be done about the Statistics tag, given the existence of a stats SE.

If the "Statistics" tag literally means all statistics, then I don't feel it has a place on this site. Not because it's not maths, it just is stats.

The tag description is:

Mathematical statistics is the study of statistics from a mathematical standpoint, using probability theory as well as other branches of mathematics such as linear algebra and analysis.

$1$. What exactly does studying statistics from a "mathematical standpoint" mean? What other standpoint is there? A "statistical" standpoint?! I do not see how this tag distinguishes between "standard" stats questions that belong on stats.SE and "other" posts that would be appropriate here.

But more importantly,

$2$. It's clear that many users are misinterpreting this tag, which seems to imply that the posts ought to be advanced, as simply encompassing any statistics question. And they can't really be blamed, as the tag itself would lead one to believe that.


Further Notes - Don't Read if Too Long

I noticed this because two users asked very basic statistical questions which quite obviously belonged on the statistics forum: this one which I answered and this one which I did not answer, because I do not know enough statistics to even attempt it.

I suggested to the user that they move the post to stats.SE and they explained that they have "seen lots of questions about distributions and statistics posted on here", which makes sense.

There are several other statistics-related tags which don't seem to belong here. Statistical Inference is defined as

The area of statistics that focuses on taking information from samples of a population, in order to derive information on the entire population.

This sounds perfectly fit for stats.SE. The same goes for Descriptive Statistics,

The area of statistics that provides descriptions of data, may it be samples or the population. This includes graphical representations and numerical indicators. No information is inferred from samples about the population, as in inferential statistics.

Fortunately only the "Statistics" tag is used much, but it's quite prolific at that.

This question asked about the policy for migrating between math and stats SE. The consensus seemed to be that no such policy exists. The two arguments cited for keeping stats posts here were that 1) some users want them here and 2) they may get better answers here.

I suppose the former is true, though I'm not sure this justifies their posting on a slightly off-topic forum. The latter though, as far as I can tell, was entirely unsubstantiated. I see no reason to think that statisticians would give lower-quality answers than mathematicians, and indeed I can think of a few reasons for why the reverse might be true.

This question asked where probability and statistics questions should be posted, but went unanswered.

I really think that the only possible justification for posting a specifically statistical question to this forum would be if the question required involved mathematics with which a typical statistician would be unfamiliar. But in this case, I think the mathematical "essence" could be extracted from the post, to the point where using statistical terminology and notation would be unnecessary. For example, if someone had a regression query that was about linear algebra alone, they could simply post the equivalent linear algebra query (explain projections, or whatever) without mentioning the unnecessary context.

Finally, I get the impression from glancing around on the stats site (which I just joined, as it so happens$^{[1]}$) that it isn't as busy as mathSE. Perhaps they would want the extra traffic?

$[1]$. It was pointed out in the thread alluded to above that a good number of users on mathSE (I don't know how many) are also on statsSE, so migrating a question there would not likely deprive the OP of a chance at a great answer (given that the more statistically-inclined users here would probably join the stats forum).

$\endgroup$
  • $\begingroup$ I found some other odd tags: "covariance"? "Random Variables" is supposed to be "Questions about maps from a probability space to a measure space which are measurable" but from the looks of it that isn't happening. $\endgroup$ – user142299 May 9 '14 at 1:04
  • $\begingroup$ "Probability", for "basic questions about probability" - should that be here? How do we know whether those belong on the statsSE or here? $\endgroup$ – user142299 May 9 '14 at 1:04

You must log in to answer this question.