How to Read a Scientific Paper

• The technical literature is very dense, and is often written in a way that makes it very hard to understand.
• The way to master the technical literature is to become an active reader who uses critical thinking.
• As you read, keep these questions in mind:
  • Does the data really support the central hypothesis?
  • What are potential limitations of the methods and the analysis?
  • Could some of the results be artifacts, i.e., irrelevant results produced by the apparatus or methods?
  • Are the assumptions made by the authors reasonable, and supported by evidence?
  • Are there other interpretations of the data?
  • If there are other interpretations, have the authors ruled them out experimentally?
  • Have appropriate controls been done?
• Even articles published in prestigious journals can be wrong.
• Rarely, data have been fabricated, or contradictory results have been left out.
• In what sequence should you read a scientific paper?
  • Read the Abstract first, and see if the paper is worth reading.
  • If you understand the methods well enough, immediately go to the Results section, and examine the data.
  • If you are not sure about the methods, read the Materials and Methods section carefully.
  • Go back to the Introduction, and see if the authors have clearly delineated the significance of their hypothesis, as well as stating the hypothesis clearly.
  • Read the Discussion critically, and see if, as you weigh the evidence yourself, you agree with the statements made by the authors.
• It is useful to write down questions and concerns as you read the paper.
• Reading it this way will help you write scientific papers better yourself, and train you to be a peer reviewer of science.

For an amusing view of the problem of reading scientific papers, see: How to read a scientific article