Should new users be allowed to provide questions similar to others?

Question

Recently on CrossValidated I have found that new members may not have studied all the rules prior to asking their first question. I find that moderators and other users will quickly vote to close either because they think it is a duplicate or statement of the question is not clear or the question is off topic. Such decisions are a matter of judgment. I have seen cases where these new members become very frustrated. I understand why the rules must be followed. However I think in such cases a comment to the new member suggesting what is wrong with the question should be done first. They should wait a reasonable amount of time for a an edit to the question or a comment from the new member before taking action.

On two occasions I did things that I thought were reasonable but got moderators upset with me. Situation 1: A new user asked his first question that I told him was not complete and looked like a self study question. He commented that he couldn't edit the question because he thought he was limited in the question length. So I agreed to edit it. The question was put on hold. I don't remember whether it was before or after I made the comment. So I edited the question adding the information he gave me in comments. I also added the self help tag. I then found a second question with similar problems so I edited that one too. Two moderators criticized me for this for different reasons. On the second question I was told that I should have given the OP the chance to edit the question himself (not knowing the discussion that went on with the first question). The other moderator didn't think I should have edited a question that was already put on hold.

Situation 2: A new user asked a simple question. I had a satisfactory answer but I also knew that there were many situations (possible for different modeling methods) where this problem came up. The user was worried that she did something wrong. I thought that it would be best for me to give the answer and point the OP to a search method that using the key words "negative R square" would yield many hits. I thought it would be a difficult task to see if I could consider any of them an exact duplicate. I thought that the OP should look through a few of them to see how commonly the question comes up with a variety of models. The OP and others liked my question and I got 2 up votes for it. A moderator suggested that I should not have answered the question but rather should have searched for a duplicate and recommend it for close/hold. He assumed that there would be an exact duplicate and that I was cluttering the site with repeat questions and an answer that was short (not canonical) and not as informative as a longer supposed duplicate.

Am I wrong in either situation and if you think so could you explain why?

Answer 1

Should new users be allowed to provide questions similar to others?

Depends on what you mean with "allowed" and "similar". A new user can post similar questions, but if it is similar to the point some users consider it a duplicate (aka, a borderline duplicate) it should be closed. Duplicate questions are ok to ask (as long it is not recurrent from the same user which is not doing any previous search within the site before posting; remembering if a user post too many questions that get closed he/she will receive a temporary question ban, etc; that is why it is not ok asking too many duplicates in a row), because there are many ways asking the same question, but it is important that answers don't be split across the site. So pointing to the canonical post (which not always is the older one) will help users finding the best answer.

But what if "similar" means it is not a duplicate, and it is closed anyway? No problem, the "On hold" notice/message will tell the OP to edit his/her question and explain why it is not a duplicate. Edits to questions with "On hold" status will be sent right away to the Reopen queue, where it can be voted to reopen. Moreover, a user with more than 3k reputation can also vote to reopen a question without any need of an edit being made, and the question will also be sent to the Reopen queue.

Recently on CrossValidated I have found that new members may not have studied all the rules prior to asking their first question.

Unfortunately, nowadays this is a big problem, but not exclusive to SE. Are you helping moderating the community leaving comments to such OP's to visit the tour page and the help center, for example?

Additionally, I think the actual problem is that many new users don't study any of the rules at all (not even the tour).

I find that moderators and other users will quickly vote to close either because they think it is a duplicate or statement of the question is not clear or the question is off topic.

Really? I find it is still a slow process here in CrossValidated, with experienced users refraining to rapidly closing questions to the point they will collect poor answers and OP will not be satisfied. You probably know that a question with a positive score answer is considered answered, so the community bot will not bump this question to the active page; it will be buried with that poor answer. On the other hand, if a question is quickly and correctly put "On hold" the OP will have a chance to edit the question to make it more accurate and chances of receiving a better answer will increase. At the same time, we keep our site the most clean as possible.

What we could try keeping improving, is to give a chance for a new user to improve his/her question before doing any downvote; and if possible, leaving comments with specific feedback, beyond what the closing messages already say.

Such decisions are a matter of judgment.

Yes, that is why closing, reopening, approving edits are mostly processes which are submitted to peer review (I say mostly, because mods and users with gold badges have the hammer vote; but I am confident to say in most cases this is not a problem. In this respect, I think we have a minor problem with obsolete comments, but this is another discussion).

I have seen cases where these new members become very frustrated.

No comments to this sentence, because you did not bring any example. It is difficult to say something.

...I think in such cases a comment to the new member suggesting what is wrong with the question should be done first.

I think it could be done at the same time. That will help. In fact, we have carried some important efforts to incentivize users leaving comments. See the work of Silverfish and gung here in meta: How best to use the review queue?.

They should wait a reasonable amount of time for a an edit to the question or a comment from the new member before taking action.

Well, I think they shouldn't if they don't want to. We are all volunteers here and time can be limited. We are already curating the site and answering questions. OPs need to make their part, i.e., helping users to help them. Reading the rules in the help center and taking the tour is one thing they can do. Another action they can take is to post the question and be ready to quickly provide replies in comments and edit their question. After all, it is in the OP's best interest to improve their own question.

On two occasions I did things that I thought were reasonable but got moderators upset with me.

Again, difficult to comment without real examples. I will make some general comments though.

Situation 1: A new user asked his first question that I told him was not complete and looked like a self study question. He commented that he couldn't edit the question because he thought he was limited in the question length. So I agreed to edit it. The question was put on hold. I don't remember whether it was before or after I made the comment. So I edited the question adding the information he gave me in comments. I also added the self help tag. I then found a second question with similar problems so I edited that one too. Two moderators criticized me for this for different reasons. On the second question I was told that I should have given the OP the chance to edit the question himself (not knowing the discussion that went on with the first question). The other moderator didn't think I should have edited a question that was already put on hold.

There is a very specific agreed internal policy for the self-study tag. Did you follow it? I don't agree having this tag on our site (despite I agree the specific closing reason and its respective guidelines to OPs is very helpful; but tag and closing reason are different things), but the majority of users agree with such policy, therefore it should be respected.

I suggest you to narrow down your question (with examples) to emphasize better what points you don't agree with the feedback received and how they are different from what the policy says.

Situation 2: A new user asked a simple question. I had a satisfactory answer but I also knew that there were many situations (possible for different modeling methods) where this problem came up.

It could be considered a "Too broad" question; hence, putting it "On hold" and asking the OP to narrow it down to which "situation". But if you felt it was not the best option, then, keep your opinion (like you said, it can be a subjective decision).

The user was worried that she did something wrong. I thought that it would be best for me to give the answer and point the OP to a search method that using the keywords "negative R square" would yield many hits. I thought it would be a difficult task to see if I could consider any of them an exact duplicate. I thought that the OP should look through a few of them to see how commonly the question comes up with a variety of models. The OP and others liked my question and I got 2 up votes for it. A moderator suggested that I should not have answered the question but rather should have searched for a duplicate and recommend it for close/hold. He assumed that there would be an exact duplicate and that I was cluttering the site with repeat questions and an answer that was short (not canonical) and not as informative as a longer supposed duplicate.

We should always give a deep thought to what moderators say. They are experienced users when it comes to moderation, they curate the site, they know better than most users the overall content we have within the site, because they need to read a lot.

With that being said, users have the right to disagree. If you think the best option was to answer that question, go ahead. Perhaps, you did some previous search and did not find a duplicate question. Always try to use your best judgement.

Answer 2

Leaving aside the specific incidents you mention (though gave some of my reasoning on one of those in comments), I want to address the title question about whether new users should be able to ask duplicate questions.

The policy against duplicates is there for a good reason. Indeed, it's one of the major attractions of the site for me. That linked article, by one of the Stackoverflow founders, Joel Spolsky, points out something that he and Jeff Atwood (the other one) had noticed about forums like Usenet (which I also answered questions on back in the day):

• newbies would frequently ask the same beginner questions

• old timers got tired of it

I've been answering stats questions online in various fora since at least 1992 (before even the Mosaic browser existed), and CrossValidated is the first forum I've been a regular on that made it really easy to find previous versions of the same question, and which tried to solve the problem by making it easy to get new versions of questions pointing to old ones which already had good answers. $^\dagger$

As Joel pointed out, the problem with forums that don't do this is that you end up answering the same "standard" questions over and over -- and the more nearly beginner-level the questions are, the more of them you get (because there will always be more beginners than experts). So as a site grows, you spend more time answering questions you've already answered before. The answers tend to become perfunctory, since there's no benefit in putting effort into really good answers -- the effort will disappear into history under the flood of new questions and in a matter of days or weeks is forgotten. Effort is only briefly recognized, and we'd only be helping one person at a time -- one question, one answer, one person helped. After writing a few thousand answers on the internet, it starts to get annoying.

With CrossValidated, by contrast, it's worth writing a good answer, because your good answers remain easy to find even when they're old, and new questions on the same topic will tend to point to older (and hopefully more canonical) versions of the same question. If you want to contribute an answer to the site, it is best done at the same place as other answers, so that the best answers are all in the same place (instead of spread all over). That way, once you find any good answers, all the other good answers are there, and can be compared (and, hopefully, the best ones have the most votes).

Here, some of my answers are seen by thousands of people. [And while I was typing this answer yesterday, I got an upvote on an answer that was almost a thousand days old... that sort of thing almost never happens elsewhere.]

[In addition, similar questions that are not exact duplicates can be linked (both via links mentioned in comments and in the "Related" column in the sidebar) to help users find additional useful information related to their problem.]

As a result, it's actually important to avoid answering duplicate questions, so that good answers are not all over the place but concentrated into as few locations as they reasonably can be.

This different way of organizing a site requires a different kind of effort -- in place of answering every question, we end up spending time in so-called curation -- closing duplicates, improving old questions and answers, finding previously missed duplicates and closing them/linking them to answered versions of the same question, and so on. But the end result is far better -- new users, instead of waiting to get a question answered, can find their question already answered and typically in better fashion than they'd have got if the site allowed duplicates.

This benefit happens even if they have never heard of StackExchange, because good answers here show up very high on search engines like Google.

Indeed I have noticed more and more over the last four years or so that searches for something stats-related that I type into Google have links to answers here -- and often they're quite useful. That doesn't happen with my old answers on other forums (at least hardly ever) -- even when I type a search specifically designed to locate an old answer I'm looking for.

But let's imagine that we could marshal some argument for only allowing users with low reputation / recently-created accounts to ask duplicate questions (which are still otherwise deliberately not allowed, following network policy), in spite of the fact that new users will be the main source of duplicates.

If we did that, then any user who desired to get around the policy could post duplicate questions by simply creating another account, so duplicates would proliferate even more. We might as well toss out the entire SE model .... and in that case, I'd be saying my goodbyes. I don't need to participate in still more forums that encourage us to ignore previous good answers. I've done that already.

---

$^\dagger$ why do we keep old duplicates around long after the person that posted them has moved on? Because they ask their questions-with-the-same-answers in somewhat different ways, and so different versions of such questions are found by searches. They widen the net of searches that can be pointed to the same good answers -- they do the job of funneling people to where they can get their questions answered without anyone writing a new answer.