• Research papers

How to read a paper: almost never from beginning to end

Nobody who reads papers professionally reads them front to back, and yet that is exactly how students are left to do it — starting at the abstract, grinding through the methods, and arriving at the discussion exhausted and unsure what the paper actually claimed. A paper is a reference document with a narrative pasted over it, and it should be attacked, not read.

6 min readSubjects

The three-pass method

  1. 1

    Pass one — five minutes, decide if it matters

    Title, abstract, section headings, figures, conclusion. That's it. You are answering one question: is this paper relevant to me? Most papers die here, and they should. Being ruthless at this stage is the whole skill.

  2. 2

    Pass two — an hour, understand the claim

    Read the introduction and discussion properly. Look hard at the figures — in most papers the figures *are* the result and the prose is commentary. Skim the methods. You should now be able to say what they claim and roughly how they got there.

  3. 3

    Pass three — only if you must reproduce or attack it

    Now read the methods line by line, check the statistics, and try to find where it breaks. Reserve this for papers that matter — a handful per project, not per week.

Read the paper the authors didn't write

  • What's the sample, really? Buried in the methods, and frequently much less impressive than the abstract implies.
  • **What did they *not* measure?** Absences are where the weaknesses hide.
  • How do the limitations they admit differ from the ones you can see? The gap between those two is often the most interesting thing in the paper.
  • Does the conclusion actually follow from the result? Discussions routinely over-claim, and noticing it is exactly the critical faculty your degree is for.

Common questions

How do I read a research paper efficiently?

Three passes. Five minutes on title, abstract, headings, figures and conclusion to decide if it matters. An hour on introduction, discussion and figures to understand the claim. A full read only for papers you must reproduce or attack.

Should I read the methods section?

Skim it on pass two, read it properly only on pass three. But always check the sample — it's frequently much less impressive than the abstract suggests.

How do I read papers critically?

Ask what they didn't measure, whether the conclusion actually follows from the result, and how the limitations they admit differ from the ones you can see. The gap is usually the most interesting part.

Try it on your own material

Upload your notes, slides or lecture recordings and get a tutor that answers only from them — and says so when they don't cover it.

$12/month, or $8/month billed yearly · cancel in two clicks