I am admittedly an undergraduate so I do not have very much experience yet in finding resources.

When I have an idea of something I would consider worthwhile to research (in my case for a proposal to a supervisor about an undergraduate research opportunity), assuming I do extensive research online to see what has been found out about the subject and end up concluding that what I want to look into has not been investigated yet.

How can I ensure that it actually hasn't, and it isn't just a case of me not finding the information that is already available somewhere?

  • 3
    You can ask on some of the SEs: those that might be relevant to you are bio.SE and cogsci.SE. Jun 23 '12 at 19:27
  • 1
    That is part of "extensive research online" of course ;)
    – Armatus
    Jun 23 '12 at 19:29
  • 15
    It's almost impossible to know for sure that a result is new. As suggested in the answers, ask experts and read the literature (a citation index is helpful). In spite of this, at least 3 times before I've worked hard on a result, only to discover it was previously proved, usually a decade or more earlier. Don't feel bad. This happens to most people. When it happens to you, be thankful for what you've learned in the process. Often the experience will help you later on in unexpected ways.
    – Dan C
    Aug 30 '12 at 14:16
  • I am now grappling with the same dilemma that, may be, what am about to do is already done. But my consolation is that I will bring to the fold something new or slightly different from what has been done. So don't lose heart do your best to bring something different just like prepairing the same food in a different way!
    – user14614
    Apr 27 '14 at 16:44

This is of course always hard to protect oneself from. However, I would say the best way to ensure that your research proposal hasn't been investigated before is to get in contact with an expert in the field and ask him or her this.

Of course, you should investigate the topic thoroughly before approaching the expert. One tip to do this is by using a citation index. This enables you to track how a certain paper has been cited, meaning you can follow the trail and more effectevly find out what has been done. For example, if you have located an old seminal paper in your field, finding out which papers have referenced this one should give you a fair chanse of finding what you're looking for.


As an undergraduate doing research, your primary goal is to learn and publishing is secondary. Of course, publishing your results and going through all the associated things for the first time is part of the learning. There is nothing fundamentally wrong with spending your undergraduate years rediscovering something. To bring up a particularly relevant example for today: Alan Turing was elected a fellow of King's College in 1935 (at the end of his undergrad) by proving the central limit theorem. He failed to notice that it was already proved in 1922. I am sure he learned a lot from it.

Of course, you have to be prepared to rediscover. If you are not it can be very discouraging. From a person experience, when I was an undergrad preparing my first first-author paper for submission I ended up finding previous work that had shown almost the same results 3 years prior. I had spent a lot of time rediscovering the ideas and writing my paper and it was a pretty devastating blow. My supervisor's words of encouragement: "Don't worry, at least we know we're on the right track".

He was right, and I was able to learn a lot from the experience (eventually leading to other publications). The particular paper in question sat in the back of my filesystem for a while, but I returned to it 2 years later with fresh eyes to discover that it was not as derivative as I had originally judged it and even this paper ended up being published (with some modifications) at a conference.

Another adviser takes the idea of rediscovery even further. When we move into new territory on a project, he will suggest some directions knowing that the relevant theorems have been proved already. He will tell me: "this result has been proven, but don't look it up, you will arrive at it on your own and then we will both be able to understand it better". This approach works very well in math-like fields since you have the advantage of knowing what you are going after as true, but still get to experience all the joy of (re-)discovery and the lessons you learn along the way. Something that is often not obvious from just reading somebody else's paper.

As an undergrad, don't worry about accidentally working on something that has been done. Concentrate on learning as much as you can!


My answer is probably hidden in the previous ones but I'll give it a go anyways:

1) never assume that you discovered a completely new field: if there's something that I learned in my research years is to be humble. Maybe soeone starts very boldly thinking that they're the best researchers in the world, but for sure everybody at some point realises that the field is populated by experts, who know a great deal of things in the area. So in the case of finding out others sharing your ideas, humbly try and learn from them; in the (exceptional) case of really finding out a new research avenue, humbly start collaborating with others because you won't make any difference to the field only working by yourself.

2) similar ideas do not prevent you from engaging on the same topic: this is very common, and I've seen it happening a lot of times. Researchers become depressed when they see that something similar to their ideas has been published already. I think that what happens there is that you find something related, and even if you did not do the work, tend to believe that you would have done exactly the same steps. It's only when you engage directly with the problem that you discover that you would tackle it (slightly or vastly) differently: that gives you a lot of potency in my opinion, and it's a very good exercise. Take an existing problem, which has been tackled already, and give it your spin.

  • "I think that what happens there is that you find something related, and even if you did not do the work, tend to believe that you would have done exactly the same steps." - it becomes really depressing when you realize you would have done exactly the same as what they did for their first step, before the other 9 steps that go a lot further than what you had even considered for now. And then, halfway through the paper, you discover they describe the result of those 10 steps as okay, but vastly inferior to their actual contribution ... ;) Apr 30 '17 at 14:47

Several years down the line, I'm now be able to answer this question myself: you can't exclude it, and this isn't limited to academic topics. People will disclose what research, products, startups or concepts they are working on at conferences, competitions, social events,... but you can never know whether they're working on something and intentionally not disclosing it. Internet research is the closest you can get, supplemented by asking people who know the field you're interested in.


Nowadays, people usually submit the preprints for the papers on arxiv. Hence, if you have an idea you can search the databases(arxiv and different research engines such as inspires in physics, for other topics I believe it would be different) by different combinations of words and see if someone else explored that particular idea. You shouldn't be bothered by the novelty of an idea. Suppose you have an idea and you work on it, the outcome would be in general very personal. Even if someone proved a known result and therefore can not publish this result since it has been published before, what one gained in the precess is more than just an idea. In the process of exploring an idea, one develops the techniques which are necessary of creating/observing/exploring ideas in the field, and in no time a new idea which hasn't been explored before comes in. In the end, don't bother with this problems, just go ahead with the research.

  • "Nowadays, people usually submit the preprints for the papers on arxiv." - that is very field-dependent. In my HCI-related CS subfield, the concept of publishing preprints is rather unheard of, and the only times I usually hear about arXiv at all is here on Academia SE. Apr 30 '17 at 14:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.