The FAQ says that I should ask only practical, answerable questions based on actual problems.

That seems very subjective and I want to avoid posting questions that will be down-voted or closed.

Is there a test than I can apply to help me catch questions that are not appropriate?

  • I hope we hear many different voices on this topic. Personally, I'm hopeful we will all do more with the questions posted by the community to the site--up vote, comment, etc.
    – GeneJ
    Commented Nov 10, 2012 at 17:17
  • 1
    The community needs to work out the issue of what makes for good genealogical query on StackExchange. I'm hoping to see evidence of that in the questions that are asked and voted up.
    – GeneJ
    Commented Nov 10, 2012 at 17:53

5 Answers 5


Before posting a new question, ask:

Can the question be answered with a generic search engine?

Paste your question title into Google (or Bing or Yahoo or ...) and run a basic search. Do not worry about the prepositions and articles, most search engines ignore them.

  • If you get nothing useful at all, then your title is probably a poor statement of what you want to know. Rewrite it and search again.

  • If you find a complete answer, then there is no need to use Stack Exchange at all.

  • If you find some related material but nothing that deals with the particular information you are seeking, then this may be a question for Stack Exchange.

Is your "question" really a question?

If you have used the expressions "interested in ..." or "find out about ...", then you have probably described an area in which you work but not framed a question with sufficient focus to enable others to assist you.

Is the significance of your question clear?

Remember that you are asking other family historians to put aside their own work to help you. Be sure that the question is worth their effort. Sometimes the way you ask a question can hide its true value. Who was Fred Smith's father? looks like a question that should be answered by reference to one record in a repository. Which of three John Smiths (b 1820-1830 in Devon) was the father of Fred (b 1847)? is far more likely to attract interest and effort.

Does your question have multiple independent (although related) threads?

Allow readers to focus on one task at a time by separating out questions that can be investigated independently. You might want to know about an illegitimate birth, migration to the New World and conscripted military service in one branch of your tree, but I might not have the time or interest to deal with all of them. Split your request into separate well-focused questions. If the same background material is necessary for several questions then copy and paste it into each. Mention in the body of each question that there are other questions on the same person or family (and add hyperlinks between them), but ensure that the title focuses on one and only one at a time.

Have you demonstrated your own effort to find an answer?

Stack Exchange is not meant to be the first place to which you turn every time you want some information. You should not write a question until you have (at least) exhausted the obvious sources of a possible answer. A question that appears to be trying to get someone else to do all the work will not attract (favourable) attention. At the very least, you should show how you gathered what you already know; and for extra credit, describe a search that failed to move your understanding forward.

Where is the answer likely to be found?

If people reading your question are likely to believe that the answer "should" be available in a standard reference (such as the help file for your software) then go back to the previous point and prove them wrong before you post. Do not ask How can I do X? I use A. because the answer See pages p-q of the manual! wastes others' time and makes you look silly. On the other hand I have seen users of software B and C do X. It seems that I cannot do it with A because of .... Is there a way around this? might generate very useful responses.

Does the question contain all of the necessary background and context?

Think about English crime stories on television where the suspect is cautioned You do not have to say anything. But it may harm your defence if you do not mention when questioned something which you later rely on in court. The one thing that you do not want to write when editing a question is "Didn't I say that ..." Before you post, have someone else read your question and note carefully what clarification they seek. If they ask something you know but did not write, add it in.

Is the task you are setting feasible?

A question might satisfy all of the criteria listed above but still attract no interest. Which records in the secret Vatican Archives would disprove the claim that my 6xgreat grandfather was the Supreme Prior of the Rosicrucians? I have read THE DA VINCI CODE from cover to cover but cannot find an online index for the relevant documents. However fascinating this may seem, I would not respond because there is no way in which I could access any useful information. Think carefully whether anyone will have the time and resources to deal with your question.

How many different answers could your question have?

  • If there are three or more possible answers, this is probably not a suitable question. Requests for personal preferences (such as Which .... do you use?) or extended lists (Where could I find ...) are generally not appropriate.

  • If your question has two possible answers and you expect to choose between them by counting the number of responses of each type, then it needs more work. Stack Exchange operates on the quality of the answers not the quantity. Which is better, ... or ...? questions are almost always inappropriate.

  • If your question can have no definitive answer, but will generate a really lively discussion; do not post it. Look in the top menu for Chat and take it there. Stack Exchange encourages robust exchanges of strongly-held opinions, but not in the Q&A pages.

Is your question about genealogy and family history or Genealogy & Family History?

There is a separate section of the site for questions about the site and how it operates. It is called the Meta. Do not post Meta questions on the Main Q&A page.

Has the question been asked (and answered) on SE before?

Do not post any question until you have run all the key words you have used through the search box at the top of the Questions page. Read each hit that it identifies (both questions and answers).

  • If there is a partial answer to what you want then edit your question to acknowledge that before you post.
  • If there is another question that appears to be a duplicate (or closely related) but you believe it is not, then make explicit reference to that. Question 12345 has similar content but is not the same because ...

Do you know the answer to your question already?

Stack Exchange makes provision both for asking questions to which you know the answer and for answering your own question. That does not mean that every such question is appropriate. Until you have some experience in framing good questions, you may be wise to avoid them.

If you are tempted to post a Dorothy Dixer, consider two more questions Why are you posting? and Does the form of the question reveal that you know the answer? A clumsily phrased question may undo whatever you are attempting to achieve.

  • 1
    I wish I could vote for this multiple times.
    – user104
    Commented Nov 10, 2012 at 9:52
  • Thank you for posting your insights about what makes a good question.
    – GeneJ
    Commented Nov 10, 2012 at 17:15
  • 1
    These are good rules, but I hope they are not so exclusive that few questions are left for anyone to answer. We currently have about 50 active people asking questions, trying to keep the questions per day going, and I'm pretty sure every one of them/us are finding it tougher and tougher to think up new questions. Let's just make sure our rules don't eliminate 90% of the questions people have or people won't come here because we won't answer their questions.
    – lkessler
    Commented Nov 10, 2012 at 18:45
  • @lkessler, I think these rules only eliminate the very worst 2% of questions, where no effort has been put into investigating a problem before asking, no effort has been put into constructing a reasonably-bounded question, and no effort is put into improving the question if it is needed. [continued]
    – user104
    Commented Nov 10, 2012 at 19:40
  • @lkessler, Nothing guarantees that people will get an answer to any question. I'm more likely to answer something that I can deal with quickly using 'general genealogy knowledge' or specialised knowledge of a region or topic, or else a meaty problem that will be an rewarding diversion to spend some time investigating. As well as having a concern about keeping the volume of questions up, I'm also worried that we'll get swamped in one-liners - "looking for information on X Smith, circa 1800, London" - if we're not careful and we get too many of those, people won't come here to answer questions.
    – user104
    Commented Nov 10, 2012 at 19:43
  • @ColeValleyGirl - These are a great set of guidelines, but I also think they're just that, guidelines for the perfect question. If someone even just does half or 1/3rd of what's listed, then I'd say they've done a pretty darn good job of asking a question. ;) In short, I wouldn't suggest using this as a rulebook for closing posts, but more as a guideline for users who wish to ask the best questions possible. I also think Fortiter intends this to be guidelines for askers, not something to use as ammunition against them.
    – jmort253
    Commented Nov 10, 2012 at 20:21
  • @jmort253, I had definitely understood them as guidelines for questioners not as a rule book for closing posts. My reference to the worst 2% was aimed at reassuring lkessler that I didn't think they would eliminate 90% of questions.
    – user104
    Commented Nov 10, 2012 at 20:37
  • @ColeValleyGirl - You're absolutely right. There's no magic bullet that will take care of the problem. It will always take work to ask great questions, both from askers as well as from those who can leave nice comments asking for clarification. :)
    – jmort253
    Commented Nov 10, 2012 at 20:50
  • "Who was Fred Smith's father? looks like a question that should be answered by reference to one record in a repository. Which of three John Smiths (b 1820-1830 in Devon) was the father of Fred (b 1847)? is far more likely to attract interest and effort." As a general rule, the US does not have repositories the contain the kind of direct evidence you are writing about.
    – GeneJ
    Commented Nov 10, 2012 at 21:25
  • I agree this is a good and well thought out checklist. It would apply to any SE subject though. My fear is that there's a whole class of questions in our field which cannot be given black-and-white answers. Typically requests for help searching for someone - the best answers require iterative and collaborative searches. I raised this separately under: meta.genealogy.stackexchange.com/q/1312/108.
    – ACProctor
    Commented Nov 15, 2012 at 22:16

Is there a test than I can apply …

Yes, there is a test (for lack of a better word) which was enumerated as a series "Guidelines for Great Subjective Questions." It's become one of our most often-referenced blog posts:

Good Subjective, Bad Subjective

You may also want to review your site's 'how to ask' guidance, adapted from Google’s tips for getting help.

How to Ask


Here's my simple test of a good question.

Pretend you are the guru expert everyone respects who knows absolutely everything about genealogy and family history.

Now read the question.

  1. Is it something you can answer or give direction for?
  2. Is it worthy of you answering?

You must answer "Yes" to both questions.


I think @Fortiter has the crux of what should go into a 'how to ask good questions' page referenced off the faq page. That is how stack overflow did it. However I think some of the sections of his answer should be meta questions of their own that rate discussion. Here are some examples:

"Is the significance of your question clear? ... Remember that you are asking other family historians to put aside their own work to help you."

'Significance' is subjective so if we are to include it then we need to define. It does not need to be significant to everyone.

New section on requesting info about an individual

There is a different meta on this this but this 'how to' should give explicit guidance on what is and what is not appropriate. I personally think there is still disagreement on this and I think it's the most critical issue to resolve if we want the site to thrive. Maybe wording this section will help crystallize consensus. @Fortiter's answer should be edited so with the specifics of whether the answer is "no" or "yes" to the meta question "should-we-be-asking-specific-questions-about-helps-with-specific-ancestors-on-this-site". This site has exhausted all the 'research methods' questions I can think to ask (although there are obviously more) but I still have hundreds of questions about individuals to ask if they are appropriate to ask on this site.

"Do you know the answer to your question already?"

Isn't this counter to SE instructions, especially during beta? Rather than experience posting questions, shouldn't the criteria be the question meets the rest of this 'how to ask a good question'. And even if it doesn't, comments (not closing) should help the OP improve it which gets them involved in more of the process. We are still trying to determine what are the good questions and it should be up to the masses by voting not the few who already gotten the approval of other early adopters. If our hit rate was growing I'd be more willing to accept this criteria, but it's not so I think we need to broaden. However the "Dorothy Dixer" is a valid part of this section and probably should be a section of it's own (as opposed to being part of 'answer your own') since it applies to questions you don't answer as well (or you don't answer right away). Questions should be asked because they are specific genealogical problems of interest to others (either all genealogists or many other descendants of a particular distant ancestor).

New section on context

I think another section should be added with respect to adding context. It touches on the subjective nature of 'significance' above but I think this new section should focus more on advice to include the research you already did, properly referenced if you have the references, etc (or whatever the group thinks is the right context to add). My point is advice on context is useful.

Overall I think @Fortiter's answer is a very good one and should be the basis of the how to ask good questions page. I just think we should be very careful this early in discouraging people who want to participate.

  • A few points: (1): We're in day 32, too early to compare against other sites. (2) Asking questions you know the answer to is okay, but it has to be in the same Q&A format as other posts. Don't make it obvious in the question that you already know the answer. ;) (3) As for growth, keep in mind that, for many cultures, the holidays are quickly approaching. This may explain the reduction in traffic. (4) Great content ranks higher in Google, meaning that eventually more people will find this site. It takes time.
    – jmort253
    Commented Nov 10, 2012 at 20:30

The list that Fortiter provides certainly frames a standard of excellence, but there are also notions we might ask ourselves about what we want to be.

In that context then, what questions are not a good fit for the new forum.

While I'm confident Internet searches will return the names/surnames/places mentioned in our Q&A posts, do we want to be cousin-connection based?

That is, will we allow questions principally written to find cousins or for which the "expert answer" would likely come from someone who has constructed a genealogy about that family?

(a) There are many 20th-century styled query forums that support cousin connections. Some have slick, up to date interfaces, but have often been long standing so they feature valuable archives. These forums/list sites often have categories for surnames. For example, RootsWeb/Ancestry.com hosts 32000 mailing lists, many of which are surname or location specific. They host many message boards, too (some are associated with a list). The RootsWeb/Ancestry.com Preston board has more than 2500 posts; its associated mailing list goes back to 1998. GenForum's surname page Preston dates back to 1997; has over 3500 messages.

Many posts to these forums/lists are "about people." The posts may take the form of a question or express a problem, but they don't have to. A post might just as easily read, I descend of ancestor so and so, born such and such. Does anyone connect or know anything about him/her? as it would, "I'd like to share such and such about my ancestor, so and so." Or even, I'm might be [or even am not] related to so and so, but does anyone know anything about him/her?

(b) More modern/complex "cousin connection" platforms exist in the form of tree sharing and/or collaborative tree sites. These platforms benefit from their associated database presentations "about" people. In other words, you can search names, dates, etc., and then the view pedigrees and/or descendants. Many of these sites have associated message boards/forums. Users may also be able to directly message the author of the tree (or those who follow that tree). The communities that form around these tree oriented sites may be highly engaged.

  • 1
    I'm very sure (which probably means I'm in the minority <G>) that I don't think this should be (1) a lookup exchange (2) a 'cousin-bait' site (3) hospitable to 'fishing expeditions'. I do think we ought to encourage queries about individual ancestors where the questioner doesn't know how to proceed, either because they're breaking new ground in their genealogical experience or because they've tried everything they can think of and are well and truly stuck at a point that they need expert help, or just a fresh set of eyes.
    – user104
    Commented Nov 12, 2012 at 14:49

You must log in to answer this question.

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