Begin With Your Purpose
The same paper is read differently for a class, a review, a method, or a research decision.
Decision framework
How deeply should you read?
What do you need from this paper today?
Reading depth depends on discipline, paper type, and purpose. A short conceptual paper and a complex empirical study require different forms of attention.
Use Three Passes Instead of One Long Struggle
Each pass has a different question.
The three-pass workflow
- 1
First pass: orientation
Read the title, abstract, introduction, headings, figures or tables, conclusion, and reference pattern. Identify the paper type and central question.
- 2
Second pass: reconstruction
Read the main sections carefully. Track how the authors move from the question to the method, evidence, result, and interpretation.
- 3
Third pass: evaluation
Reconstruct the study or argument as if you had to reproduce, review, or extend it. Examine assumptions, alternatives, limitations, and unanswered questions.
Build an Argument Map
A useful note should preserve the paper's intellectual structure.
| Element | Question to answer | One-sentence output |
|---|---|---|
| Problem | What gap, contradiction, or practical difficulty motivates the paper? | The paper addresses… |
| Question | What exactly is being asked or tested? | The central question is… |
| Method | How is the question investigated? | The authors use… |
| Evidence | What observations, data, proof, sources, or analysis support the result? | The main evidence is… |
| Result | What is found, proved, classified, or argued? | The paper concludes… |
| Limitation | Where should the conclusion not be extended? | The result is limited by… |
| Use | Why does this paper matter for my work? | I will use this paper for… |
Local checklist
Questions to answer before highlighting
Progress is stored only in this browser.
Read the Method According to the Discipline
Do not judge every paper with the same checklist.
| Paper type | What to inspect closely |
|---|---|
| Mathematical or theoretical | Definitions, assumptions, logical dependencies, proof strategy, counterexamples, and scope of the theorem. |
| Quantitative empirical | Research design, sample, variables, measurement, comparison, uncertainty, analysis choices, and practical size of the result. |
| Qualitative | Research context, participant selection, data collection, coding or interpretation, reflexivity, and transferability. |
| Review article | Search scope, inclusion criteria, type of synthesis, source quality, missing literature, and whether conclusions match the reviewed evidence. |
| Computational or modelling | Data provenance, model assumptions, baseline comparisons, validation, reproducibility, and sensitivity to parameter choices. |
A Worked Example of Paper Mapping
The example is hypothetical so the method is visible without borrowing a real study's claims.
Handle Confusion Precisely
A difficult section becomes manageable when the difficulty is named.
Vague and useful reading notes
| Topic | Vague note | Useful note |
|---|---|---|
| Technical difficulty | The method is too difficult. | I do not understand why the authors chose this estimator or what assumption it requires. |
| Result | The findings are interesting. | The reported result supports X under conditions Y, but does not address Z. |
| Relevance | This paper is useful for my thesis. | This paper provides a definition and method I can compare with two later studies. |
| Criticism | The sample is small. | The sample limits precision and may not represent the population named in my research question. |
Turn Reading Into a Reusable Note
A PDF becomes research material only after its function is recorded.
The one-page research note
- 1
Step 1
Record complete bibliographic details and the stable DOI or publisher URL.
- 2
Step 2
Write the research question and main contribution in your own words.
- 3
Step 3
Describe the method at the level needed for your purpose.
- 4
Step 4
Record the evidence that directly supports the main claim.
- 5
Step 5
Write at least one stated or inferred limitation and label which it is.
- 6
Step 6
Separate the authors' conclusions from your interpretation.
- 7
Step 7
Add one sentence explaining how the paper will be used in your assignment, review, or research.
Use AI as a Question Generator, Not a Substitute Reader
The source paper remains the authority for what the authors actually claim.
Local checklist
Before citing the paper
Progress is stored only in this browser.
Questions about research reading
Must I understand every equation before using a paper?
Not always. The required depth depends on how you will use the paper. You should not rely on a technical result that you cannot interpret responsibly, but background reading may not require reproducing every derivation.
How long should reading one paper take?
There is no universal time. A first pass may take minutes, while a method you plan to reproduce may require several sessions. Decide the purpose before deciding the depth.
Can I cite a paper after reading only the abstract?
An abstract can help screening, but it is usually not enough for an important claim. Read the relevant method, result, discussion, and limitation sections before citing the paper as evidence.