Reading a Science Paper for the First Time

For Young Learners

What This Is

A research paper is how scientists tell each other what they found. Almost all of them follow the same five part structure, which is the only reason a first read is possible at all.

Abstract. A summary of the whole thing, written last.

Introduction. Why the question was worth asking.

Methods. Exactly what they did.

Results. What they found, usually in tables and graphs.

Discussion. What they think it means, which is the part most likely to overstate the case.

This page teaches the method on that structure. A guided read of a named paper is not attached yet, because no paper could be retrieved and checked when the page was written, and walking you through a paper nobody verified would be worse than waiting.

Papers look impenetrable, and most of that is vocabulary rather than difficulty. The reasoning underneath is usually simpler than the sentences suggest.

Reading one yourself also breaks a habit. Health claims normally reach you through several layers of summary, and each layer loses a qualification. Going to the paper is how you find out what was actually shown.

How It Works

Read the parts in this order, not the printed one.

1. Title and abstract. Work out what the question was. If you cannot, stop and find an easier paper. There is no prize for struggling.

2. The last paragraph of the introduction. Authors state their aim there, almost without exception.

3. Methods, and only two things from it on a first read. How many participants, and what was compared with what.

4. Figures and tables. Read the captions, which are written to stand alone. Most of the finding is here.

5. The first paragraph of the discussion. The authors say what they think they showed.

6. The limitations paragraph, near the end of the discussion. This is the most valuable paragraph in the paper and the one most readers skip.

Then stop. You now know what the paper claims and what the authors admit it cannot support, and that is a complete first read.

How to Prepare

Have the four questions from How We Know What We Know next to you: compared with what, how many, who was left out, and who paid for it.

Keep a list of words you had to look up. It will be short within a few papers, because the same two hundred terms carry most of medical writing.

Start with a Cochrane plain language summary rather than with a journal article. It states the question, the number of studies and the confidence in the answer in ordinary English, and it is a gentler first object.

Use PubMed to find papers, and prefer open access ones so you can read the whole thing rather than the abstract alone.

Common Mistakes

Reading from the first word to the last. Nobody in the field does this.

Believing the abstract. An abstract is written to be persuasive and routinely omits the limitation that matters most.

Skipping the limitations paragraph, which is where the authors tell you what they cannot claim.

Treating one paper as settled science. One paper is one paper, and the field will contain others that disagree with it.

Confusing a result that is statistically significant with one that is large enough to matter to a patient. They are different questions.

Where to Find Out More

PubMed is the main free index of medical research and is searchable by anyone.

Cochrane publishes plain language summaries alongside its full reviews.

Section 12 of this site covers critical appraisal properly, including the checklists used at postgraduate level.