How to read a scientific research paper

Recently, I was lead author of a University of Bath online resource called Evaluating Scientific Research Literature. Aimed at helping science undergraduates read academic papers more critically, we involved postgraduates in designing it and undergraduates in testing it. I interviewed the authors of academic papers to reveal their motivations for writing them, and to offer guidance to students on how best to read them. Our project team learnt a great deal, and I gained insights that will help me write the second edition of my book on academic writing.

Our key insight? There is no single good way to read a scientific research paper. It depends why you’re reading it, your level of familiarity with the content, and your reading preferences.

If you’re a first-year undergraduate writing an essay, and you are using the paper as one of many information sources, you might focus on the introduction and conclusion, perhaps studying any figures or tables in the results section and reading the discussion. If you’re a final-year student designing your own research project, you might pay particular attention to the method section, to inform your practice. The emphasis of your reading changes depending on your level of knowledge and what you’re seeking to do.

One thing we advise all students to do if they’re thinking of reading a paper is to skim it first. This means going through the paper at speed, to gain an overview — a mental map if you will. Pay attention to the beginnings and ends – of the paper as a whole (title, abstract, introduction and conclusion), the sections (key paragraphs near the beginning and end of sections) and of the paragraphs themselves. Briefly study any figures and tables. As you go through, you can note important parts that you will need to read in greater detail.

Something we all agreed on: do not simply read a paper once slowly all the way through. That is very inefficient. You end up reading everything at a similar speed, whether it is relevant or not. Much better is to decide why you’re reading the paper before or after skimming it, and then read it to meet that purpose. You might not read the whole paper but just the parts that are relevant to your task. You may wish to read those important parts two or three times, while annotating and/or making notes, to make sure you’ve got everything you need. That way, you will meet your purpose and most likely gain a good grasp of the paper.

A final word: as an undergraduate it is unrealistic to expect to understand everything you read. One of the project’s professors talked about still not understanding some papers and he’s been researching for more than 20 years!