Reading path: Underwater images and evidence →
Find the problem before the conclusion
Start with the abstract, but treat it as a map rather than a substitute for the paper. Look for a sentence describing a gap, uncertainty, or problem. Then read the introduction's final paragraphs, where authors often state their objective, question, or prediction more precisely. Separate that question from the broader subject. “Coastal warming and plankton” is a topic; “Does the timing of seasonal warming coincide with a change in plankton abundance at these stations?” is a researchable question.
A useful question usually identifies what is being compared, measured, or described. It may ask whether two variables covary, how a community differs between places, or whether an organism's distribution changes with depth. If the introduction presents several aims, list them separately. A paper may have a primary question and secondary analyses, and confusing them can make its results seem more sweeping than intended.
Translate the question into evidence
Next, inspect the methods and ask what observations would count as an answer. Suppose the stated question concerns whether a small drifting animal is more abundant in colder water. Relevant evidence might include repeated counts of that animal and water-temperature measurements collected at the same locations and times. A methods section describing only a genetic comparison of preserved specimens might address species identity, but it would not directly test the proposed abundance-temperature relationship.
This check is not a judgment that the methods are good or bad overall. It tests alignment: do the sampling units, measurements, and analysis correspond to the question? If the question concerns change through a season, a single visit may describe conditions on that date but cannot by itself show a seasonal pattern. If the question concerns differences among sites, repeated measurements at one site do not provide that comparison.
Watch for a shift in scale
Research questions can narrow as a study proceeds. An introduction may discuss ocean-wide change while the actual work samples one bay, one season, and a limited set of organisms. Write down both the broad motivation and the narrower question the study can answer. This prevents a common reading error: treating evidence from a local case as if it directly settled a global issue.
A practical note might read: “Question: Is the observed animal count associated with temperature across three sampled stations in spring? Evidence: paired station measurements and counts; scope: these stations and dates.” This compact restatement forces the reader to name the comparison and the boundaries. If you cannot fill in those details, return to the objectives and methods rather than guessing from the title.
Use a one-sentence test
After reading, try writing one sentence that begins, “The authors ask whether…” Then write a second: “They address this by measuring…” Compare the two. If the second sentence does not supply evidence relevant to the first, the paper may be answering a narrower question, using an indirect measure, or pursuing a different aim than the abstract led you to expect.
Finish by checking the results and discussion for the same question. Results should present the observations or analyses that bear on it; discussion should explain what the authors think those results mean. A clear question makes the rest of the paper easier to evaluate: you can tell which figures matter, which caveats limit the answer, and what remains unknown.
Have a correction or a source to suggest? Send a note →