How to Read Research Papers Quickly for a Research Proposal

Source: Keshav, S. (2007). How to read a paper. ACM SIGCOMM Computer Communication Review, 37(3).

How long does it take to read a research paper?

A research paper takes about 5 to 10 minutes for a first-pass read, up to an hour for a careful second pass, and about 4 to 5 hours for a full critical read by a beginner, according to Keshav’s three-pass method (2007). An experienced reader needs about an hour for that same full critical read.

Why does a full read take hours when the words take minutes? It is the difference between reading a recipe and cooking it. Brysbaert’s meta-analysis of 190 studies puts the average adult silent reading rate in English at 238 words per minute for non-fiction (Journal of Memory and Language, 2019), so 1,000 words of a paper takes a little over four minutes to read once. The hours go on checking methods, rereading results and following references.

Speed-reading courses rarely help with research papers: a review by Rayner and colleagues found “a trade-off between speed and accuracy” (Psychological Science in the Public Interest, 2016). Reading research papers quickly comes from choosing what to read deeply, not from moving your eyes faster.

What is the three-pass method for reading research papers?

The three-pass method is a way of reading a research paper in up to three increasingly careful readings, set out by S. Keshav of the University of Waterloo in ACM SIGCOMM Computer Communication Review in 2007. Each pass has a fixed scope and time, and the reader stops after any pass once the paper has given what they need.

The first pass, of about five to ten minutes, has four steps in Keshav’s paper:

1. Carefully read the title, abstract and introduction.
2. Read the section and sub-section headings, but ignore everything else.
3. Read the conclusions.
4. Glance over the references, noting the ones you have already read.

At the end of the first pass, Keshav’s method asks you to answer the “five Cs”: Category (what type of paper is this?), Context (which other papers is it related to?), Correctness (do the assumptions appear valid?), Contributions (what are the paper’s main contributions?) and Clarity (is the paper well written?).

How do you read 100 papers for a research proposal?

To read 100 papers for a research proposal, give all 100 a first pass, give a second pass only to the papers your research question, method or argument relies on, and save a full third pass for the handful your proposal builds on or argues against. With Keshav’s timings, 100 first passes at 5 to 10 minutes each take roughly 8 to 17 hours in total.

Here is the arithmetic for a library of 100 papers, using Keshav’s 2007 timings for a beginner. The 30 and 5 below are illustrative counts, not a rule:

Where do the 100 papers come from? Keshav also describes a literature survey that fits a proposal well: find three to five recent papers in the area, give each a first pass, read their related-work sections, then look for citations and author names that keep recurring. Those recurring papers are the field’s key papers. Keshav then recommends two passes through the papers you collect. For the citation-following step in detail, see snowballing in a literature review; if your search returns hundreds of hits, screening abstracts comes before any of the three passes.

Which parts of a research paper should you read first?

Read the abstract first, then the introduction and the conclusion, and only then decide whether the methods and results deserve your time. Three university reading guides agree on that order, with small differences in what comes next.

• Vanderbilt University English Language Center (Scholarly Reading Guide: Reading Efficiently): when skimming, “Read specific sections of articles only: Abstract, Introduction, and Conclusion”.
• A University of Cambridge library guide (skim, scan and scrutinise): read the abstract, then the aims, skip the literature review, go to the conclusion for the main findings, then the figures and, if time and need allow, the methods.
• Caltech’s writing centre (reading a research article): many STEM professionals read “selectively and out of order”, starting with the abstract and reading in full only once a paper’s relevance is clear.

Do not skip the middle of every paper, though. Hubbard and Dunbar surveyed 260 students and researchers in biological sciences and found that inexperienced readers “found the methods and results sections of research papers the most difficult to read” and undervalued the results section and critical interpretation of data (PLOS ONE, 2017). For papers that reach a second pass, give the methods and results the extra time, because that is where a proposal’s argument gets tested.

How do you decide which papers to read in full for a research proposal?

Read a paper in full when your proposal depends on it: when it answers part of your research question, supplies the method you plan to use, reports a finding that conflicts with another paper, or is a recent review of your topic. Every other paper can stop at the first or second pass.

Ask what you would lose if the paper disappeared from your proposal. Picture the proposal’s literature review as a bridge: a few papers are load-bearing pillars, many are deck planks, some are railings. Pillars get a third pass, planks a second, railings a first. Keshav’s five Cs (category, context, correctness, contributions and clarity) help sort them, because the context and contributions answers tell you where a paper sits in that bridge.

Conflicting findings deserve a second pass even when the papers look minor, because a place where two studies disagree, or where no study covers your population, is often where a proposal’s research gap sits. How to find a research gap covers what to do next, and how to critically appraise a research paper gives a checklist for the third pass.

How should you take notes so you don’t reread research papers?

Take notes in a literature review matrix: one row per paper and the same columns for every paper, filled in at the end of each pass. Typical columns for a proposal are the research question, method, sample and setting, main finding, stated limitation and the page number for each entry, so you can return to the passage instead of rereading the paper.

Each pass leaves a record. After a first pass, a row may hold only Keshav’s five Cs (category, context, correctness, contributions, clarity) and a one-line contribution. After a second pass, the method, sample, finding and limitation columns fill in. Newcastle University’s Academic Skills Kit makes the same point in its guide to effective reading strategies for extended work: check a text’s relevance from the abstract, introduction, conclusion and reference list, and re-read it fully only if it is useful.

Litrevu’s free literature review matrix gives you the columns to start from. The rows that disagree and the empty cells then become the outline of your proposal’s literature review, covered in how to write a literature review for a research proposal.

Which reading method is best for a research proposal?

The three-pass method is the best fit for most research proposals, because it is the only one of these methods that attaches a time to each depth of reading, which gives clear stopping rules for a library of dozens of papers.

How do I read papers for a research proposal with an AI assistant for writing backed by sources?

With an AI assistant for writing backed by sources, you add the papers you found, ask all of them the same questions, check each answer against the passage it cites, and decide yourself which papers get a second or third pass. Litrevu is an AI assistant for writing backed by sources. It reads every paper, report and PDF you add, answers your questions with a citation to the exact passage behind each claim, compares sources side by side, and checks your own draft against them.

On your own proposal papers, the workflow looks like this:

1. Add your papers. Upload the PDFs from your search, paste a PDF link, a DOI or an arXiv link, paste a paper’s text, or let Litrevu find open-access papers on your topic and choose which to add.
2. Ask every paper the same first-pass questions. For example: “What is each study’s method and sample?”, “What is the main finding?” and “What limitation do the authors state?” Answers are cited only from your papers, and when your papers do not cover a question, Litrevu says so.
3. Open the citations before you trust them. Each citation opens the passage highlighted on the page of the PDF it came from, with the paper’s title, authors and year. Reading that passage is your check; the citation check confirms that a citation points to a passage in your papers, not that the passage fully supports the sentence.
4. Line the papers up in a comparison table. Compare studies side by side, with citations, to see where they agree and where they differ. The papers behind those disagreements, and any population no paper covers, are the ones to move into your second and third passes.
5. Check your own paragraph. When you write a paragraph of your proposal’s literature review, paste it in and each claim comes back as Supported, Partly supported, Contradicted or Not in your papers. “Not in your papers” means your library does not cover that claim, which is a cue to search wider, not proof that a gap exists in the literature.

The reading decisions stay yours. Litrevu finds, lines up and checks the evidence in your papers; you decide which papers matter, what the gap is and how to argue it, and you write the proposal. Many universities allow AI for reading and searching sources but not AI-written text in submitted work, so check your institution’s policy and declare how you used any AI tool.

Here is the shape of a cited answer, shown as a format illustration with placeholders rather than real output:

Frequently asked questions

What should you do next with the papers for your proposal?