I think this site could be improved by the addition of a tag for a "general" question, or a "non-specific" question.

This tag could be used when the person asking the question is interested in a general answer not in help solving a specific statistical problem or troubleshooting an issue.

I think this would help people answer questions, as they wouldn't have to ask a poster for more details about the specifics of their problem when they have no specifics to give. Instead they could concentrate on a general answer or not answer if they think it would take too long.

I also think this would make the site more friendly to new people who might have broad open questions about statistics, but no particular use case in mind. I think having good answers to broader, less specific questions will make the site more useful, and more used. It may also result in more answered questions!

Finally, I think it would avoid frustration on both ends caused by a misfit of intentions and expectations!


As noted in the comments this is a grievance of mine, so all the examples are my own questions! Here are some examples:

  1. Do you lose information when you encode numerical columns with two values?
  2. Are large samples always needed for a Z-Test?
  3. Calculating the probability my observation, $Y_i$, is drawn from a random variable $X$? (Long string of comments and revisions before an answer is given, when what I was after was mdhak's comment, but as an authoritative answer).

I'll add more as I find them.

  • 2
    $\begingroup$ Hi Connor. I appreciate your concern for the betterment of the site. However, the post could be more robust if you could provide some examples of general queries, queries which didn't receive good answers owing to being asked for further details etc. Even before articulating any view, I would actually want to understand what your grievances are considering the fact that those phrases are vague, imo. $\endgroup$ Apr 19 at 5:42
  • 3
    $\begingroup$ With that being said, my preliminary take is this: when do we ask for more details or want the query to be more specific: exactly when we feel the present query is unanswerable due to dearth of information (data, methods applied, reproducible example, approach and where OP is stuck) or it is so broad that it cannot be answered without writing a chapter or so (there is no problem with a long answer - there are plenty, but certainly there is a line). There are many questions which I feel are quite general (I would still want to know what that means to you) that received precise answers... $\endgroup$ Apr 19 at 5:54
  • 3
    $\begingroup$ But if the community deems adding specific information and narrowing down could help it receive better usable as well as concise answers, then there is no con in asking for the same. I am unable to see the necessity of a new tag henceforth. Nevertheless, I feel I still need to comprehend your concern in full length once you provide with few examples. $\endgroup$ Apr 19 at 5:56
  • $\begingroup$ I understand your point, but I believe that in the cases I've provided, the query isn't unanswerable. My hunch is that becuase the answer is simple, it's assumed it must be XY, there must be more to the question, and therefore refuse to answer without specifics. A sensible practice, but unhelpful in the cases I've given. $\endgroup$
    – Connor
    Apr 19 at 7:05
  • 1
    $\begingroup$ This is an interesting post, but ultimately I am not convinced that anything different is needed or would be helpful. Broad general questions are already allowed so long as they satisfy the community, but most that are closed (or ignored) suffer that fate because they are too general, or at least lacking in focus. A tag with the purpose of flagging a naive question or questioner would serve no good purpose, as there are not different standards for different levels of user. This comes back to basics: CV is not intended as a help line, but as a repository of good questions and answers. $\endgroup$
    – Nick Cox
    Apr 19 at 13:03
  • 3
    $\begingroup$ (ctd) More positively put, if you have a good question, it can stand on its own two feet, It shouldn't need propping up or protecting with a special tag. $\endgroup$
    – Nick Cox
    Apr 19 at 13:04
  • 2
    $\begingroup$ As is I think standard on Meta, my downvote implies disagreement with the proposal, not that the question is of poor quality or lacking in interest. $\endgroup$
    – Nick Cox
    Apr 19 at 13:09
  • $\begingroup$ @NickCox What do you think is the solution then? I'm sure I'm not the only person who's experienced this, and it would be good to have best-practices for this case. A pre-amble could work, but what should it be, what should it include? I get the general stack exchange mantra of, "this is not a helpline", but on this site, especially, it does tend to become that. Asking for a lot of detail about the specific problem you're working on makes the site like a helpline. I think a general question, even if it's naive, is less like a helpline than what currently happens. $\endgroup$
    – Connor
    Apr 19 at 19:19
  • 1
    $\begingroup$ An asymmetry is built into SE by design and moderators and people with high reputation or long experience work with it just like newcomers. Anyone who asks a question has some rights, notably which answer to accept if there are several. Otherwise what happens to a question depends on the rest of the community. The best strategy you or anyone in a similar position has is to ask the best question you can. How a question is tagged won't, I think, make any detectable difference to its prospects of surviving or flourishing. That's a personal take, and its own survival depends on others agreeing. $\endgroup$
    – Nick Cox
    Apr 19 at 20:13
  • $\begingroup$ In other words, tags are largely pointless for the success of a question? I absolutely see your point there, however, I think there are isolated times where a tag does make a difference. For example, the soft-question tag on mathematics stack exchange, which does make a material difference to the success of a question. math.stackexchange.com/tags/soft-question/info $\endgroup$
    – Connor
    Apr 19 at 20:33
  • 1
    $\begingroup$ Each community sets and follows its own norms. Here on CV tags mostly indicate content, not intent or level. Speaking for myself I don't ever answer or not answer a question becaise of tags. I sometimes edit out useless tags. I respond much more to question titles. Some people I imagine do follow particular tags and some people put a lot of work into editing and maintaining tags. I respond much more to editing text, so people differ. A community can change its habits, but so far your proposal is meeting some scepticism and even more lack of interest. disappointing though that will be. $\endgroup$
    – Nick Cox
    Apr 19 at 23:53
  • 2
    $\begingroup$ Not the question, but people can and do add personal comments on being new to statistics, or very confused, or whatever. Such comments often get edited out, even ruthlessly, on the grounds that only your question matters. not who you are or what you are. But while they remain visible they can serve a helpful psychological purpose of signalling quite what level of answer is expected or requested. $\endgroup$
    – Nick Cox
    Apr 20 at 10:11

2 Answers 2


At the very outset, let me share my own quick assessment of your cited examples (your own posts) based on the original forms (without looking at any of the comments, at first):

$\bullet$ Are large samples always needed for a Z-Test?

There were improper usage of notations and terminologies and some misunderstandings. I feel a bit hesitant seeing words like good too.

$\bullet$ Do you lose information when you encode numerical columns with two values?

This seemed to be also not replete with proper usage of notations; in fact it was outright ambiguous which prompted it being closed. Finally you edited it to provide better structure and I was one of the reviewers who voted to reopen it.

$\bullet$ Calculating the probability my observation, $Y_i$, is drawn from a random variable $X$?

Again it is infested with the same problem: improper usage of notations and terminologies.

However, I do feel all those queries were not bad; they were sanguine and intended to be valid queries - unfortunately they were not adequately articulated, at least in the original form.

We don't know everything. That's why questions arise (that's basic). However, there is a certain extent of knowledge that we apparently take for granted as if we were well versed with those notions and concepts. So when we mould a query resorting to the language from that for granted knowledge that we acquired (might be a quick reading of a blog, site, video etc.), we are certain we have raised a legit proper question.

But things go awry when we see responses from people that implicitly attack that for granted evident knowledge. It might start to seem nitpicking to us: instead of answering in a general way to our evidently clear queries, their comments are vacillatory and moving back and forth.

However, then a mature way to react is to take a step back and try to read between the lines of the comments: what are they are trying to air and communicate? Then we can realize what we have been regurgitating might not be correct. The notations and handwavy concepts might need to be altered or outright restructured.

Then we would be able to conclude while the query could be genuine, the way it was raised proved to be its own impediment in getting a good concise answer.

Comments are absolutely necessary. When it gets long, chances are OP is not able to clearly provide ample explanation of what they are asking: the question could be sensed, but a proper useful answer warrants an explicitly clear post, even the confusion that evoked the question at the first place.

Users like Christian Hennig, Tim, Dave, utobi, Sextus Empiricus and all other respected users could have easily dissuaded from interacting and commenting for the sole purpose of eliciting further details and informing you of any misconception. They didn't. Why? Because they understand the value of this community and the site's goal: create a repository of high quality contents: be it an undergrad homework, a research based query or an industrial programming query.

Finally you were able to receive a good answer to one of your posts thanks to those interactions.

Seeing comments as response instead of an answer that solves everything can be indeed frustrating. But for the latter, you need the former in the light of the perspective added above.

One's knowledge can be limited, what I think is correct need not have to be. Those long comments help to remove those misconceptions.

At the risk of sounding like a broken record, I would assert comments are quintessential. You need to be patient.

[It takes days for me to sometimes understand something, takes hours for me to sometimes solve a problem which when seeing the hint/solution showed me how my approach was wrong. But instead of getting frustrated, I take a break and try to learn the necessary materials needed or other things.]

I appreciate you reacted rationally by asking a meta question rather than ranting at those users (I have witnessed many things during my not so long stay here).

To answer your titular question, I am of the conviction that such kind of tag is not necessary.

  • $\begingroup$ Thank you for such a complete response. I agree that comments are helpful, I also agree that my notation was (and still is) sloppy, and that I haven't been perfectly clear in the questions I've asked. However, I still think the idea of a "general", "non-specific" or "basic" tag has merit. The most recent question that you voted to re-open is a perfect example of this. In that post, all I want to know is if this could ever be a problem. But the tendency on this site is to ask for specifics, which might not even exist! $\endgroup$
    – Connor
    Apr 19 at 8:55
  • 1
    $\begingroup$ Of course my knowledge is also limited. We learn new things day by day. I appreciate you interacted rationally. Just take those as learning materials. $\endgroup$ Apr 19 at 9:00
  • $\begingroup$ I appreciate that a pre-amble could solve this, and I'm happy with that as a workaround. But I think a tag that makes it clear that the question is non-specific could be really helpful to avoid "why are you doing this" comments, when you've not done it at all yet! I think Cross Validated has great people on it, erudite statisticians with a lot to share! I think adding such a tag could help get to that knowledge faster in the instances I'm talking about. $\endgroup$
    – Connor
    Apr 19 at 9:01
  • 1
    $\begingroup$ Re specifics, it still applies that the premise might be warranting some additional info that OP is not familiar with. But it is only my opinion that a tag would be an overkill. But this is my take only. Wait for others to add their cents to the proposal. $\endgroup$ Apr 19 at 9:04
  • 2
    $\begingroup$ @Connor "why are you doing this" I ask that question a lot. And it is mainly because I try to understand the problem. Statistics starts with understanding as much of the process that is underlying the problem. It is about the numerical and data gathering issues surrounding some practical problem. Statistics is never general and always has an application in mind. --- A stereotypical question would be "I have this data <insert table or graph>, how do I analyse it?" Without any background given. $\endgroup$ Apr 20 at 10:03
  • 1
    $\begingroup$ When a question has a mathematical orientation then can I find myself answering it, without understanding the motivation, and asking afterwards why is this useful. (For instance recently here: "I am curious about the original problem where you are using this inequality.") $\endgroup$ Apr 20 at 10:06
  • $\begingroup$ @SextusEmpiricus What do you think of the most recent question: stats.stackexchange.com/questions/613150/… . Is that general or specific? It feels general to me. $\endgroup$
    – Connor
    Apr 20 at 17:30

First cent:

I had a short interaction in the comments about the question Do you lose information when you encode numerical columns with two values? and I didn't respond yet (instantly) to your reply because I am still confused about your question and what it means. (With or without 'general' as tag)

Instead they could concentrate on a general answer or not answer if they think it would take too long.

So my thought process is like, ok let me read that post a second time, or a third time again before answering. Often I might be juggling multiple questions at the same time and some end up on the floor and so forgotten if they are not shiny enough. I actually have a quite long backlog of open questions in which I had an interest to answer them, but which were in a way requiring some additional work to get it done and that I could not quickly fix. Questions being posed in a general way and with several unclear definitions, expressions, motivations, background, etc. are among them.

So currently my mind goes like 'not answer' (yet) because it takes too long. But that doesn't mean that you shouldn't try to clarify your question better. Adding a tag 'general' isn't changing people's responses.

While a question might be non-specific, the problem remains that you still need specific examples to easily explain the non-specific issue. Examples are maybe not neccesary, but they are still helpful. You can not replace that helpfulness with a tag 'general'.

Second cent:

Tags are mainly metadata that help to easily classify questions. The purpose of the classification can be variable, but functions like communicating how one would wish the question to be received or how it should be interpreted might be far stretched (one tag that has such a function is and it is a bit of an abuse of tags). #Dont-take-this-the-wrong-way #yes-that-is-an-ironic-hashtag #this-aswell-and-showing-how-tags-shouldnt-be-abused

  • $\begingroup$ Thank you for the answer! That does help me understand why the questions go unanswered. After some back-and-forth in the comments, I ended up partially answering my question, so I understand there's no one way of doing things. My question to you is, what is it that makes you think a question will take too long to answer? $\endgroup$
    – Connor
    Apr 20 at 17:32

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .