How to write a strong introduction into a research paper?

  • How does one write a strong (good) introduction into a research paper? Some introductions make me really curious about the rest of the paper while others do not. Although it is relatively easy to say which introductions are good and which are not, I find it difficult to distill what makes the difference. There is a previous question about writing introductions (How to write a Ph.D. thesis Introduction chapter?) but it is about Ph.D theses.

    write the paper first, then tack on a beginning and end. that way you will know what its about.

    Here's a bit of meta-advice on this point. One major way I've learned how to improve introductions is by thinking hard about negative referee reports. In my experience, when a paper gets a referee report that I disagree with, the explanation is often that the introduction needed to be clearer about something.

    @Noah's advice is spot on (although one might aspire to write a good introduction earlier in the day than this) and really points to the importance of the question: the difference between an average introduction and a good one is often the difference between your paper being grokked or not by some fairly random referee who has sufficient subject-level expertise but is not closely clued in to your particular perspective.

  • Suresh

    Suresh Correct answer

    7 years ago

    This is very area specific. I'll start with the caveat that I write papers in computer science, so YMMV.

    The way I think about introductions (which is not to say they are GOOD introductions) is that they tell the story of the paper in brief. Every paper has a story to tell, starting with

    • Here's a fascinating question
    • Here's what people have tried to do (in brief: not a full related work section, but a high level assesment)
    • here's the key challenge preventing further progress
    • Voila: here's our complete/partial/intermediate/awesome solution
    • (additionally) and here's how it works.

    The intro is typically the "hook" to read the rest of the paper, so you have to provide a birds-eye view that draws the reader in without drowning them in details.

    The thing that separates a good intro from a bad one is knowing where that right level of detail is, so you're not either totally vacuous or mired in details. Getting this right is an art and depends on your field, your results, the problem, and your understanding of the target audience.

    A few pure theory papers begin with "Let X be a..." instead of "Hey look at this shiny rock I found."

    This is similar to Simon Peyton Jones's approach

License under CC-BY-SA with attribution


Content dated before 6/26/2020 9:53 AM