How to Read a Research Paper: A Guide for Students and Non-Experts
A practical method for reading a scientific paper: what each section is for, the order to read them in, and what to do with the parts you do not understand.
Research papers are not written to be read from the first word to the last. They follow a fixed structure so that readers can jump straight to the part they need. Once you know what each section is for, reading a paper becomes far quicker and much less intimidating.
What each section is for
- Abstract. A summary of the whole paper in one paragraph: the question, the method, the main result.
- Introduction. Why the question matters and what was known before. The last paragraph usually states exactly what this paper contributes.
- Methods. What the researchers did, in enough detail for someone else to repeat it.
- Results. What they found, mostly through figures and tables.
- Discussion or conclusion. What the authors think the results mean, and the limits of the study.
- References. The earlier work the paper builds on.
Papers in mathematics, computer science and the humanities vary this pattern, but the abstract, introduction and conclusion nearly always play the same roles.
Read in three passes
Reading in passes means you can stop as soon as you have what you need.
First pass: is this paper relevant?
Read the title, the abstract and the conclusion. Glance at the section headings and the figures. This takes five to ten minutes and tells you what the paper claims and whether it is worth more of your time. For many papers, this is all you need.
Second pass: what did they find?
Read the introduction, then go through the figures and tables carefully, reading their captions. The figures are the evidence; the surrounding text explains them. Then read the discussion. Skip details you cannot follow and make a note of them. After this pass you should be able to explain the paper's main point to someone else.
Third pass: do I believe it?
Read the methods and the full results. Ask how the study was done, how large it was, what it was compared with and whether the conclusions follow from the data. This is slow, and it is only worth doing for papers that matter to your work.
When you do not understand something
- Look up terms as you meet them, but only the ones that keep recurring.
- Find a review article. Reviews summarise a whole area and are written to be read by newcomers. Searching for the topic plus the word "review" usually finds one.
- Follow one reference. If the introduction cites a paper for a key idea, that paper often explains the idea more fully.
- Accept partial understanding. Experts skip methods they do not use themselves. You do not have to follow every equation to understand the finding.
Questions worth asking
- What exactly did the study measure, and in whom or what?
- How big was the effect, and how certain is it?
- What do the authors themselves list as limitations?
- Is this a preprint or a peer-reviewed paper? Our guide to the kinds of open access explains the difference.
- Who funded the work, and do the authors declare any conflicts of interest?
Getting the paper in the first place
If the paper you want to read is behind a paywall, paste its DOI, link or title into the finder on our home page to look for a free, legal copy. Our guide to reading research papers for free covers the other options.
Find a free copy of a paper
Paste a DOI, a link to the paper or its title, and we will look for a free, legal copy. No signup.
Frequently asked questions
How long does it take to read a research paper?
A first pass to decide whether a paper is relevant takes five to ten minutes. Reading one closely enough to judge its methods can take several hours, especially outside your own field.
Should I read the abstract or the conclusion first?
Read the abstract first, then skip to the conclusion. Together they tell you what the paper claims before you spend time on the detail.
Do I need to understand the statistics?
Not all of them. Focus on what was compared, how large the difference was and how confident the authors are. The discussion section usually puts the numbers into words.