When I ask a reference-request question, before asking I generally Google + Google Scholar a bunch of words I put in the question. It sounds a bit tedious to list all queries I have made, and it might be counterproductive as results for a given queries depend on the location, time, etc., I only look at the few first pages of results, and I may miss some interesting results.

However, some commentators asked me provide evidence of my efforts so far.

How can I provide evidence of my efforts when asking reference questions for which I have found no useful information so far?

3 Answers 3


I think that the push-back you are getting may have to do with the fact that you ask a lot of questions of this type. Of the 41 questions currently marked as on the site as a whole, 19 were asked by you over the last few months, and you've only accepted an answer on one of them. These are also nearly half of the questions that you have asked.

So far as I can see, this site is generally an excellent source for informed opinion, and people will often provide references voluntarily if they have them readily available. Moreover, many of your questions (e.g., on visit weekend weather, or on ESL vs. conference acceptance) draw good answers that are not references.

Insisting on references only when good informed opinion is available can make one wonder about the motivation. Are you asking the site to "do your homework" on literature searches? This can feel especially dubious given that you work in large-scale ML / data-mining, and a lot of your questions are for information where, if a study exists, it would likely be generated by one of your colleagues in the field.

So: is there a reason that you really need to tag so many of your questions as? Are these questions out of curiosity, or are you trying to use the answers to formulate research questions or related work sections of your own?

  • Thanks for your answer. I often insist on references to avoid having my questions closed as "primarily opinion-based" and because I prefer data-supported claims than educated guesses. I used to say clearly I am interested in both with a preference for references if any are available (e.g. academia.stackexchange.com/posts/27493/revisions). However a couple of close votes forced me to ask for references only, which isn't that bad as I'm afraid that many personal opinions answers deter future reference answers. Questions are out of curiosity, some stem from my interest in open science. Commented Dec 3, 2014 at 17:33
  • Regarding the number of answers I have accepted, to my defense very few of my questions got answers with references: if they do, I'm glad to accept them as an answer (apparently I had forgotten to do so for two answers). Also, I upvote all answers. Commented Dec 3, 2014 at 17:36
  • @FranckDernoncourt I think, then, that you may have over-corrected in your attempt to prevent questions being closed. It's also possible that the specific things you tend to wonder about are likely to simply not have an answer at present (e.g., percentage of papers previously rejected). You might also ask for the answer to the same data-driven questions without making it a reference request, as long as you make it clear it's not a "poll" question.
    – jakebeal
    Commented Dec 3, 2014 at 17:42
  • I agree with your answer. However, I think it is important to note that, almost by definition, asking a lot of questions is not a bad habit on a Q&A site. So the fact that @FranckDernoncourt was asking a lot of questions should not lead to a push-back, even though it might be happening.
    – xLeitix
    Commented Dec 4, 2014 at 17:08
  • 1
    @xLeitix Absolutely agreed. It is also the case that tweaking the way that one presents questions can encourage more productive dialogue.
    – jakebeal
    Commented Dec 5, 2014 at 1:30

I am not sure that it is useful to provide evidence about how you have searched for a topic. Even if a Google/Pubmed/Arxiv search turns up references, it doesn't really provide any expert insight. I think the value of making a reference request on AC.SE is that experts, or at least others with experience, can help guide you and refine the search. I would hope that reference request type questions are receiving better answers than just the first relevant hit in some search engine.


It's true that questions on SE sites are supposed to show at least a minimum of effort.

However, I think in the case of a question, the way to show that kind of effort is to

  • write a well-defined, specific question
  • give context for the question (i.e. explain the motivation or inspiration for the question)
  • explain why you think that a reference on the subject of your question might exist

This last point is, I think, where the pushback might be coming from - a couple of your questions are about things that I'd be very surprised if anyone had actually studied.

I don't think it's necessary or helpful to list Google Scholar search terms.

Regarding your comment on opinion-based questions, please don't abuse like that. It's perfectly valid to ask "why," "how," "how often," etc questions here without insisting on references. Use and insist on supporting citations for questions where you really need the supporting citations.

  • Thanks for your answer. The main reason why I specified reference-request in opinion-based questions is epitomized by the answer to the question: I want to do research but I'm too old for a PHD. There are 9 answers, from what I can see (I didn't follow all links) none of them provide any survey to back up their claims (except for my answer). I don't want the same thing to happen for my questions. I am lucky to be in an environment where I can get many personal opinions from academics, I am really much more looking for y research/study/survey. Commented Dec 5, 2014 at 4:29
  • Given the way to show that kind of effort, the context for the question has so far been always "I'm curious". And I think that a reference on the subject of my question might exist because I candidly think my questions point to interesting aspects of the academic worlds. But I've read other people suspecting me to use Academia StackExchange to crowd source my literature review for my personal research, so I take your point and will try to clarify that. It's hard to ask questions… I thought that any course 101 on research was making clear that formulating the question is the key. Commented Dec 5, 2014 at 4:32
  • 3
    @FranckDernoncourt There are other kinds of answers besides for "research/study/survey" and "personal opinion," e.g., "answers supported by experience," "answers supported by a logical explanation of some mechanism," "answers supported by an off-the-cuff study I just did on a small sample." If you are willing to accept some of those (as it seems you sometimes are), don't insist on only answers supported by research studies. You can clarify in your post what kind of answers you are looking for, and explicitly say you're not interested in unsupported opinions.
    – ff524
    Commented Dec 5, 2014 at 4:35
  • Good point, sounds like a good way to open it. One small issue I have with this though is that answers formulated on a public forum can be somehow distorted (of course studies aren't unbiased either). I'll think about it but thanks for the great suggestion! Commented Dec 5, 2014 at 4:38

You must log in to answer this question.

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