Pass one: decide whether the paper deserves more time
Read the title, abstract, publication type, and enough of the methods to understand the design. Look at the main tables or figures. Ask whether the population, exposure or intervention, outcome, and time frame actually match the question you are researching.
This pass prevents a common graduate-school problem: spending forty minutes carefully reading a paper that will never be used.
Pass two: reconstruct the study
Write down the research question in one sentence. Then identify who was studied, how participants or data were selected, what was measured, what comparison was made, and which outcome was primary.
If you cannot reconstruct those elements, the abstract alone is not enough for a confident interpretation.
Pass three: interrogate the result
Look at effect sizes, confidence intervals, absolute numbers where available, missing data, and whether the analysis matches the design. Check whether the main conclusion corresponds to the actual primary outcome rather than a secondary or subgroup result.
Pass four: place it in the literature
For a literature review, the paper becomes one node in a larger map. Record how it agrees or conflicts with related work, what design limitation it addresses or introduces, and what unresolved question remains.
Purdue’s literature-review guidance emphasizes synthesis rather than a sequence of isolated summaries. Efficient reading should therefore produce structured notes that can later be compared across papers.
Read differently for different goals
- Background: prioritize reviews and conceptual papers.
- Methods: read procedures, instruments, and analysis closely.
- Evidence synthesis: extract comparable design and result fields.
- Replication: methods, supplements, code, and protocol become central.
Efficiency does not mean reading less carefully. It means matching depth to purpose.
Let the question determine the reading order
Reading linearly from the first sentence to the references is not always the most efficient way to evaluate a paper. If the goal is to decide whether the paper is relevant, start with the title, abstract, population, exposure or intervention, and outcomes. If the goal is to evaluate a specific claim, move quickly to the methods and the result that generated that claim. The introduction can then supply theory and prior literature after relevance is established.
Figures and tables are useful early because they reveal what was actually measured and reported. They can also expose a mismatch between a broad conclusion and a narrow outcome. After that first pass, return to the methods to understand how participants were selected, variables were defined, missing data were handled, and the analysis was constructed.
Write a five-line extraction before taking long notes
- Research question.
- Population and setting.
- Design and comparison.
- Main result with uncertainty.
- Largest limitation for the claim you care about.
If those five lines cannot be written accurately, the paper probably has not been understood well enough to summarize confidently.
