Reading by the numbers
- 3–6 hours typical time to deep-read a single empirical paper end-to-end (Vitae Researcher Survey, 2024).
- 30–60 minutes typical time using the three-pass method for the same paper, retaining 85% of useful content.
- 40 to 60 peer-reviewed sources expected for a UK master’s literature review — impossible without a triage system.
- 250 to 280 words/minute average academic reading speed; 400+ words/minute achievable for first-pass surveying.
- 72% of papers a researcher initially identifies turn out to be not directly relevant — making the first-pass triage decision the most valuable.
The three-pass method (Keshav, 2007)
Originally designed for computer science papers but now widely used across disciplines. Each pass has a different goal:
| Pass | Time | What you read | Decision after |
|---|---|---|---|
| 1 — Survey | 5–10 min | Title, abstract, introduction, headings, conclusion | Stop, or proceed to pass 2? |
| 2 — Skim | ~1 hour | All figures, tables, captions, opening/closing of each section | Cite as background, or proceed to pass 3? |
| 3 — Deep | 3–5 hours | Every paragraph, every equation; annotate; identify weaknesses | Cite in detail; possibly critique |
Pass 1 — Survey (the triage decision)
Goal: decide whether this paper is worth more time. Ask:
- What is the central claim?
- Does it actually relate to my research question?
- Is the methodology appropriate to my needs?
- Is the journal / venue credible?
If the paper passes, schedule pass 2. If not, log it in your reference manager with a one-line note (“not relevant: focuses on US K-12, my project is UK undergraduate”).
Pass 2 — Skim (build working understanding)
Read selectively for argument structure:
- First and last sentence of each paragraph (topic and conclusion)
- Every figure caption — figures often contain the paper’s “punchline”
- Every table — especially the rows for variables you care about
- Methodology section — to evaluate rigour
Pass 3 — Deep read (only for central papers)
Reserve for the 5–15 papers actually central to your literature review. Annotate aggressively. Look for: (a) cracks in the argument; (b) limitations the authors acknowledge; (c) connections to other sources you have read.
SQ3R — for textbook-style sources
For monographs, textbook chapters and longer reports, the SQ3R method (Robinson, 1946; updated 2024) is more appropriate than three-pass:
- Survey — chapter title, headings, subheadings, summary, questions at chapter end
- Question — turn each heading into a question (“What are the three forms of capital?”)
- Read — read actively, looking for answers to your questions
- Recite — close the book; explain the answer in your own words
- Review — within 24 hours, then weekly
Drowning in reading for your literature review?
PhD-qualified researchers can build your annotated bibliography or full literature review chapter from your reading list.
Worked examples by paper type
Example 1 — empirical study (Bloom et al. 2015 on remote work)
| Pass | What you extract in this paper |
|---|---|
| 1 (8 min) | RQ: does WFH increase productivity? Method: RCT at Ctrip, n=249. Result: +13% productivity. Decision: highly relevant for any remote-work paper, proceed. |
| 2 (45 min) | Productivity gains driven by (a) reduced break time and (b) quieter environment. Effect size, sample, and limitations all mapped to synthesis matrix. |
| 3 (3 hours) | Identify methodological constraints (call-centre setting limits generalisability), engagement with theory, and contribution claim. |
Example 2 — theoretical paper
Theoretical papers (philosophy, theory pieces) often do not have figures or empirical methodology. Adapt three-pass: pass 1 = abstract + section headings + final two paragraphs (where authors usually summarise contribution). Pass 2 = read the argument structure, identify premises and conclusion. Pass 3 = challenge each premise.
Example 3 — systematic review
Systematic reviews often run 8,000–15,000 words. Useful shortcut: read the PRISMA flow diagram (always Figure 1 or 2) — it tells you the search scope, exclusion criteria and final included studies in one figure. Often this is enough for pass 1.
Annotation strategies that work
| Symbol | Meaning |
|---|---|
| ! | Important / disagreement / surprising |
| ? | Unclear, look up later |
| → | Connects to another source / argument |
| QUOTE | Mark direct quote candidates with page number |
| DEF | Definition worth noting |
Use a reference manager from day 1
Three free or low-cost options:
- Zotero (free, open-source) — most-used in social sciences and humanities; browser-clipper saves PDFs and metadata in seconds
- Mendeley (free, Elsevier-owned) — strong PDF annotation; integrates with Word
- EndNote (£100, often institution-licensed) — most-used in life sciences
Tag every paper as you read with: #read-pass-1, #read-pass-2, #read-pass-3, plus topic tags. This makes literature review building trivial later.
Need help synthesising what you’ve read?
Our literature-review specialists turn your reading list into a viva-ready chapter.
References
- Keshav, S. (2007) How to Read a Paper. ACM SIGCOMM Computer Communication Review, 37(3), pp. 83–84.
- Robinson, F. P. (1946) Effective Study. New York: Harper & Row.
- Vitae (2024) Researcher Reading Practices Survey 2024. Cambridge: CRAC.
- Booth, A., Sutton, A. and Papaioannou, D. (2021) Systematic Approaches to a Successful Literature Review. 3rd edn. London: Sage.
- Page, M. J. et al. (2021) “The PRISMA 2020 statement”, BMJ, 372, n71.
- Bloom, N. et al. (2015) “Does working from home work?”, Quarterly Journal of Economics, 130(1), pp. 165–218.
- Dunlosky, J. et al. (2013) “Improving students’ learning with effective learning techniques”, Psychological Science in the Public Interest, 14(1), pp. 4–58.
From reading list to first-class lit review
PhD-qualified specialists synthesise your sources into a viva-ready literature review chapter. Free Turnitin pre-check.
