Answering off-topic and/or bad questions — what is our stance and recommendation for how to handle?

Question

I've recently observed an off-topic question (maybe -- the jury is still out) being answered and the answer accepted or up-voted; I've seen the same happen with 'bad' questions that ought to be improved before they're answered. (I don't want to focus on a specific question, or on named users, but on the general issue, so I'm not quoting -- and won't quote -- specifics).

Other Stack Exchange sites frown heavily on answering off-topic and/or bad questions (there's a pejorative term involved that I won't quote as there's been a recent move to ban it and anyway I don't think the motivation here is to increase rep but the negative impact is the same IMO).

As far as I can tell, the questions at issue are answered because somebody really wants to help the OP because they know or can easily find an/the answer, even if the question doesn't belong here or isn't good enough (yet). Or else the person who answers believes the question is on-topic and good-enough. As does anyone who upvotes an answer to such a question.

But the fact that there's an answer (that has been accepted or upvoted) leaves us with a question that:

• might lead others to assume the subject is on-topic (difficult to say otherwise if they can point to an answered question -- maybe easier if the question has been closed, but I'm not certain if people are comfortable closing an answered question with an accepted answer as I believe the person who answers will lose the rep they gained?
• might lead others to misjudge the quality of questions we work towards here.
• doesn't help the OP learn to ask on-topic and/or better questions.

In all situations, it detracts from the overall perception of quality of this site, as questions with accepted or up-voted answers cannot be deleted, even if they're subsequently closed. Poor questions could subsequently be improved, but there's little motivation to do so if the OP has done a hit-and-run with the question/answer and there's also the risk that post-improvement the answer doesn't match the question which also doesn't help the perception of the site.

Where do we want to strike the balance between being friendly and welcoming and cluttering the site with off-topic and sub-standard questions?

Answering but also voting to close at the same time might be an option, as would improving and then answering but both strategies have the risk of the question ending up unable to be deleted if closed. Encouraging people not to answer at all is more draconian but less welcoming -- encouraging people to provide helpful information in comments not answers might be more friendly (and mods could convert comments to answers to make this happen if they felt comfortable doing so).

For the avoidance of doubt, I do not favour naming/shaming people who answer the 'wrong questions'.

I don't have an answer, although I have a favoured position (comments not answers, plus vote to close or edit to improve as appropriate when you comment) but I might be an outlier, which is why I'm asking for opinions.

Answer 1

As the answerer-offender to (I think) the question that inspired this post, my thoughts:

• I agree with just about everything you said ;)
• However, it is far from clear what is on-topic vs off-topic. We have a little incomplete list of on-topic topics in the help pages. The question alluded to was in general about ancestry, even if the underlying question had a more present application.
• I completely agree for the reasons mentioned that bad questions shouldn't be answered. They should be closed, then deleted through either votes (which never happens) or the automated process
• Off-topic questions also shouldn't be answered. If a good question they should be migrated to the appropriate site. In cases where it is unclear if a question is on or off-topic I don't think there is a great solution, other than us making more clear guidelines about what is on vs off-topic. Until there is a consensus I've made the answer a community wiki – I certainly did not answer it for reputation.
• There are common questions about ancestry that are basic and potentially on the brink of off-topic or too broad. Having a question that we can direct users to, closing further duplicates, may be useful in the future. Making them community wikis could be useful. We can attempt to frame such questions with more of a genealogy slant so that they are on-topic, and can be directed to as duplicates when off-topic variants of the question are asked. Given the low number of questions asked here, I think we should do all we can to keep users, but obviously not at the cost of question quality.
• Also a somewhat related question on meta - Processing brief generic questions?

Answer 2

I started out by commenting on the answers that came before me about the on/off topic guidelines -- I've put that part of the answer below the line, so I can focus on answering the question.

I think this part of the question is spot-on:

But the fact that there's an answer (that has been accepted or upvoted) leaves us with a question that:

• might lead others to assume the subject is on-topic (difficult to say otherwise if they can point to an answered question -- maybe easier if the question has been closed, but I'm not certain if people are comfortable closing an answered question with an accepted answer as I believe the person who answers will lose the rep they gained?
• might lead others to misjudge the quality of questions we work towards here.
• doesn't help the OP learn to ask on-topic and/or better questions.

We wouldn't be here if we weren't willing to help others, but we have to look out for our own self-interest, too. I'm not as concerned about the effect on reputation points as I am about the overall quality of the site. If Genealogy.SE looks to be the same quality as any junky Ancestry message board or ordinary dump-and-run query site, there's no incentive for people to spend time writing quality answers here.

Since I participate in a chat on Twitter every week, and am always on the lookout for something to tweet about to promote the site, the idea of having questions which are worth sharing is always on my mind. I would love to see more self-answered questions from the more experienced users here, so we can have them as examples to show new users the potential of the site.

One thing that bugs me especially is seeing poor questions upvoted in an attempt to "encourage new users to participate." Upvoted questions get auto-tweeted, so they get broadcast outside the site. Maybe this is no big deal now, since the Twitter account only has around 100 followers, but all of the tweets are searchable. Once the question is upvoted, the tweets go out, bad form or no, and whatever shows up in the preview is (as far as I know) preserved in the ugly form for all time. One example: For the question How to confirm my research of Charles P. Cantrell? the author started out in by chatting instead of getting down to the question and the talky part didn't get edited out (I've just edited it now).

I'd rather the OP did the improving so that they start to understand what's expected rather than maybe feel railroaded, but recognise that isn't always going to happen.

I'd rather see us act quickly to clean up questions for several reasons. I think cosmetic edits (making the question more readable by breaking out facts into timelines, reducing chatty clutter) should be done as soon as possible, to increase the chance that others will read the question and answer it.

If we don't clean up the questions when we see them, they get shoved off the front page / top of the newest & active lists, and then we forget about fixing them.

I'd like to see the OP improve their own question, too, but what I'm usually asking them in the comments is to add more about their research effort. Asking them to do formatting cleanup as well is a bit much, especially for newbies who are unfamiliar with Markdown.

Right now I am trying to sell people on the idea that you get more out of your research if you can write a summary of your prior research and formulate a focus(s)ed research question. It's hard to show them how this site can work for that when it is cluttered with the same dump-and-run "I want to know anything" queries they can find on any other site.

I don't have an answer, although I have a favoured position (comments not answers, plus vote to close or edit to improve as appropriate when you comment) but I might be an outlier, which is why I'm asking for opinions.

I think this sums up my own position pretty well, although there have been cases where a question has been criticized for being "too broad" where I've come along afterwards and said, "Oh, no, I can answer this" like Determining what records are available in a particular locale?.

But that's the point -- if the question is closed on the grounds of being too broad, we have to stop and think about how to write a good answer that might genuinely help someone instead of just firing off an answer which is not well thought-out.

---

Re: Harry's comment "it is far from clear what is on-topic vs off-topic".

Here's the list from on-topic --

If your question is about:

• Starting your research or improving your methodology
• Finding a source or understanding how to use it
• Documenting or presenting what you’ve learned
• Breaking down brick-walls in your family tree
• Using technology to support your research

On one of the FHISO mailing lists, Elizabeth Shown Mills said that genealogy involves two basic tasks.

1. Identifying a person
2. Establishing their relationship to others

Our main goals as genealogists are to identify that the people named in a historical document or record are indeed the persons we are looking for, and not other with the same or similar names, and to establish the relationships between those people. We don't state this explicitly in the site guidelines -- they are implied in the 'breaking down brick walls" bullet point. Brick walls / roadblocks occur when we can't identify someone sufficiently, or can't find enough information about them to determine their relationships to parents. However, all of the bullet points are about doing research better in support of these goals.

This is from don't-ask --

You should only ask practical, answerable questions based on actual problems that you face.

So my question is -- what part of this is not clear? How can it be improved?

Using ESM's two tasks as a guide, I would say that all of the 'what do I call this person?' questions are off-topic because the relationship of the two people involved is already defined -- the person writing the question has already stated it when asking the question. (I reserve the right to change my mind after viewing Judy G. Russell's webinar on Kinship Determination.)

Similarly, all the "what ethnicity am I" questions are problematic. I don't see how any of the ethnicity questions of the "who am I" variety -- as opposed to the "how do I interpret this historical document?" kind -- can be used to help solve a problem -- not when we have an industry that can produce different ethnicity estimates for full-blood siblings as described in Judy G. Russell's "not soup yet" post .

Re: PolyGeo's guess about what question was the inspiration for this question --

If Blaine Bettinger and other DNA experts were here answering questions about the finer points of the reference populations used to generate ethnicity profiles, that might be sufficiently technical to warrant inclusion on a Stack Exchange site, but that's not likely to happen. Most experts have their own sites and make questions submitted to them the basis of posts on their own blogs.

So -- given that the real experts are likely to answer most questions on their own blogs and forums -- what is our purpose here?

It's no surprise that I favor the "help I'm stuck" kind of question where community members post a summary of their prior research -- the genealogy equivalent of "my code doesn't work, where did I go wrong?" kind of question -- the kind of question where it can help someone to have a friendly eye look over their work, which can be made into a case study for others to learn from.

Answer 3

I dislike closing questions if there are other options.

If it is a poorly written question, attempts should be made to get the original poster to improve it.

If it is an off-topic question, then likely there is another Stack Exchange site that might be more appropriate for it, and first thought should be given to migrate it.

But if the question cannot be improved, cannot be migrated anywhere, or maybe is a duplicate of another question, then yes it should be closed.

Answer 4

Most Stack Exchange sites frown heavily on answering off-topic and/or bad questions but most Stack Exchange sites receive more questions per day than our current 1.1 - see http://stackexchange.com/sites#questionsperday. The 15 or so sites that receive less questions are all Beta sites (except StackApps, which is a highly atypical site).

I agree with you that off-topic questions should be closed. In the case of Categorizing people of North African descent (i.e. Egypt) on Ethnicity surveys/questions for US documents?, which I am assuming is the trigger for your question, I was wavering but on the verge of closing it, when what I thought was a very good answer to an awkward question came in. My hesitation in voting to close was that I could not pinpoint an off-topic reason. I think I did the right thing by upvoting what I still think is a very good answer, and one which I think should NOT have been made into Community Wiki.

Where I think I disagree with you is in where to place the needle on bad enough to close for other reasons (too broad, unclear, etc). With our site getting so few questions I think we need to do whatever we can to turn any bad questions we receive into reasonable questions where we can. My current strategy is that if I think I can see a focussed on-topic question in a question that may be very poorly focussed, and that I think should be closed if left unedited, then I edit it, sometimes quite heavily.

I have been through the whole Beta period and graduation of another site (GIS) and the whole Beta period (so far) of G&FH SE. Both sites started with 40-45 questions per day when their Beta opened. The question volume on both sites dipped heavily after the first few days. GIS now (6 years in) averages 49 questions per day while G&FH (4 years in) averages 1.1. What I saw on GIS was pretty much a free for all where its users for the first year or two pretty much asked anything, few were closed and a LOT of bad questions were tolerated. Over the next few years those bad questions were gradually closed (some even deleted) so that they could not be used as precedents for more bad questions. GIS is now a site where only good questions stay open for more than a few hours. Imposing high quality standards works for that site now because it has a steady flow of questions coming in, but without that flow being established first I think the site would still be struggling.

In contrast on G&FH, we opted for high quality from the start, and I was part of that push because by then (2 years after GIS had started and when we were starting to raise our standards) I thought high quality attracted high quality. However, for the past few years I have been convinced that I was wrong in the approach that I thought at the start of G&FH would be right for it.

I believe that there is no reason why G&FH cannot be a high volume site, probably much higher than GIS, but I think the key to getting there is to get the questions flowing first, and then to start to try and discourage bad questions in the next (post-graduation) phase. I don't want bad questions on the site in perpetuity any more than you do but I think at this point in our site maturation we need to tolerate spending time editing and clarifying bad questions to try and make them into OK questions.

While it is true to say that "questions with accepted or up-voted answers cannot be deleted" by ordinary users, moderators can delete any questions worthy of deletion, and their answers, with a single click. This makes such deletions only a flag away for everyone else.

To assist the flow of questions in the Beta phase I also support being generous in our application of upvotes, and as we get to graduation to start changing towards a more suitable balance (perhaps 2:1) of upvotes:downvotes.