22

What are the best practices for writing a research paper title? I am looking for some general guidelines.

  • Does one go with short and sweet?
  • Or a bit longer and informative?
  • Should one phrase it as a question or rather as a statement?
  • Should one use humour and play on words?
  • What should one avoid?
  • Are titles written different for different areas of research? (i.e. Computer Science, Maths, Neuroscience etc.)
  • possible duplicate: academia.stackexchange.com/questions/9892/… – Dirk Jan 3 '17 at 14:10
  • 14
    @MikeyMike I disagree. These days I get about 100 papers my newsfeed every day. I decide on what papers I take a closer look on the grounds of 1. Title, 2. Authors, 3. Abstract (in descending order). – Dirk Jan 3 '17 at 16:17
  • 2
    Two of the best review articles I ever read had humorous titles. The content was well written and the humor retained through the body of the work. Much more pleasant than most while still being excellent information. If you can be clever and engaging, go for it. You are writing for humans, after all. – The Nate Jan 3 '17 at 17:36
  • 5
    My favourite title is 'Altruism: What's in it for me?' Short, snappy, but effectively conveys the essential paradox. – Strawberry Jan 3 '17 at 18:28
  • 1
    Plays on words may be a bad idea. And other "literary" flourishes. Because: perhaps many of the readers of the paper do not have English as their first language. – GEdgar Jan 4 '17 at 15:19
16

Follow the conventions of your target journal or conference. The most reliable predictor of how a title will be perceived is if other papers at the same venue have had similar titles.

Specifically, your question Are titles written different for different areas of research? can be answered with a clear yes. For example, compare the following programming languages and neuro-imaging venues.

  • 3
    This is generally sound advice, but needs a bit of caution. I have come across some venues in which the majority of titles are, to my eyes, very poorly written. – user2390246 Jan 3 '17 at 18:10
  • 1
    @user2390246 Well, "poorly written" can mean at least two things: language issues or being boring. Language issues should be avoided, of course. Boredom is a bit more tricky. One can try out a more fancy title. There are certainly benefits of that, for example, the chance of becoming an eye-catcher. But by doing so, one also risks to displease one of the reviewers (although it usually won't be a dealbreaker). – lighthouse keeper Jan 3 '17 at 19:26
  • 2
    @user2390246 And yet - even if you consider them poorly written, going against the grain without knowing why those titles are like that is dangerous. If most titles are in a certain why, you can expect that at least many PC members actually like them this way, even if you don't. – xLeitix Jan 3 '17 at 22:02
  • 1
    "compare the following programming languages and neuro-imaging venues" - and then there is this paper, to appear at POPL'17 (dl.acm.org/citation.cfm?id=3009897). To be fair, that title is fun and insanely impractical, given that it's impossible to google for. – xLeitix Jan 3 '17 at 22:09
  • 1
    @xLeitix Actually (and perhaps with the help of the filter-bubble), googling "do be do be do" gives a preprint of that paper as result #7. ;) – lighthouse keeper Jan 4 '17 at 11:06
12

A scientific basis for choosing the right title of your paper

For every person who reads the whole of a scientific paper, about 500 read only the title . One way to improve this statistic could be to make the title declarative by including what the paper says, not just what it covers. Gustavii, B. (2008). How to Write and Illustrate Scientific Papers (2nd ed.). Cambridge: Cambridge University Press.

I always been told to use declarative rather then neutral/descriptive title but seems that the following article shows the opposite: Jamali, H.R. and Nikzad, M. (2011). Article title type and its relation with the number of downloads and citations, Scientometrics, 88 (2):653-661

I'll use it to answer to your question with the support of real data.

Q: Does one go with short and sweet? Or a bit longer and informative?

A: Articles with longer titles are downloaded slightly less than the articles with shorter titles.

Q: Should one phrase it as a question or rather as a statement?

A: There are differences between articles with different types of titles in terms of downloads and citations, especially articles with question titles tend to be downloaded more but cited less than the others.

Q: Should one use humour and play on words?

A: some authors hypothesize that "humorous titles communicate a non-serious subject matter, and as found previously in experimental settings harm the credibility of the paper"Journal of Information Science, 34 (5) 2008, pp. 680–687

Q: What should one avoid?

In my opinion: one should avoid abbreviations, and to put the keywords at the end of the titles.

Q: Are the titles written different for different areas of research?

I think common sense is not specifically related to any discipline.

  • 1
    "In my opinion: abbreviations, put the keywords at the end of the titles." To me this seems ambiguous: Should they put the keywords at the end, or avoid doing that? – Weckar E. Jan 4 '17 at 9:40
  • 1
    In the statement quoted from the Journal of Information Science (paper available via SemanticScholar), the authors seem to misrepresent the mentioned findings, which were about the use of humour in textbooks, and students' assessments of such uses. – lighthouse keeper Jan 4 '17 at 10:46
  • @lighthousekeeper is a possible interpreetation on the data on a previous finding "as found previously in experimental settings harm the credibility of the paper" so is sill valid in this context. – G M Jan 4 '17 at 13:47
  • @GM Thanks, after your edit, your answer now portrays the original statement accurately. Still, the issue with the original statement remains, as it seems to misrepresent the previous findings from experimental settings. – lighthouse keeper Jan 4 '17 at 14:04
5

(Disclaimer: Quite subjective point of view, based on my experience more than any kind of reference or "authority". Ph.D student here, Computer Science, ML & NLP)

  • Are titles written different for different areas of research? (i.e. Computer Science, Maths, Neuroscience etc.)

I've no evidence about it, but I'm pretty sure it will be way different. Thus, my background might influence me a lot.

  • Should one phrase it as a question or rather as a statement?
  • Should one use humour and play on words?

I would say without hesitation, to me, humour has to be banned, and you should use a statement. You publish to give answer, not question. Question might be present at the end of your work to introduce what question does your work open, what are the next step. Still, I will be statement like "we may wonder what would...." (no "?" involved, you are wondering, not asking).

  • Does one go with short and sweet?
  • Or a bit longer and informative?

Here's the trap: it has to be short and informative. You have to tease future reader by telling him "I managed (or kind of) to do this, that way, it rocks right?".
This part will probably be the trickiest as you need to make people feel close to your research by showing what you work on, while not being too specific otherwise they will say "wow what's that?".

As a thumb rule, for a random paper (random = no hype = didn't read about it previously, no one told me to read it, I don't know/notice who wrote it), I will be interested if: the goal interests me or if the method interests me. Lets say if I'm interested about cars: Making faster car or Using cars for food delivery.

Personaly what I see most is:

  • [WHAT I DID] <separator > [HOW DID I DO IT ], usually to present a solution to a general problem

    • (1) is the high level goal like "Understanding human voice";
    • (2) is "using", "with", ": ";
    • (3) is the method, "Super Hugely Deep Neural Network" or approach "using semantic representations
  • [WHAT I USED] [WHAT FOR], using well known techniques for a context specific use or something not initially studied

    • (1) Precisely name the technique
    • (2) Describe your context/need.
  • [MY PRODUCT]: [DESCRIPTION], introduce a named product/innovation.

  • [WHAT I DID], when something is really new. Usually impressive work.

    • e.g. "Efficient way of doing this" where the paper explains what this is, and how awesome it is.

Hope it helps, do not hesitate to comment, suggest edits etc. I find this discussion interesting and my really new to academic world, give me feedback!

pltrdy

3

As with all communication, the key is to consider your aims and your audience. Your aim, I presume, is to get your paper read. Certain people will read it no matter what the title - mainly, those who work in very similar areas. Your most important audience is therefore people who know a little about your subject area (those in exactly your field will read it no matter what) who are scanning through a list of paper titles (either from a journal or a keyword search). We come across far too many papers to read them all, you want people to pause when they see your title. Think about what might make you stop to read further if you were scanning down a list of papers.

On that basis:

  • Be as clear as possible. Minimise jargon and use simple phrasing. If I can't understand the title at first glance, I'm not going to read it again.

  • You should aim to get attention within the first few words. That might sound dramatic, but I suspect that lots of people won't even read the whole title if it doesn't look relevant or interesting. So try to put your most important key word/phrase in the first few words.

  • Personally I don't see much disadvantage with having a long title per se, but you need to make sure that the first phrase can stand on its own. So quite a few titles follow something along the lines of "[Interesting discovery]: using [method] to address [question]".

  • Being creative/humorous/unusual might make people stop and look twice if you can pull it off well. But realistically, drawing people in this way may not be especially constructive: if they wouldn't have read your abstract but for the intriguing title, chances are that they won't end up reading any further than that. If you're going down this route, you certainly need to make sure that you don't sacrifice informativeness.

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.