How to summarise a paper: the five-sentence template
Summarising a paper means turning it into something you can recall without reopening it and explain to someone else in two minutes. The hard part is not shortening; it is deciding what to drop. The template below makes most of that decision for you.
The most common mistake first
Copying the abstract and rewording it is not a summary. The abstract is the text the authors wrote to present their study; it rarely includes the limitations and it rounds the numbers. Your summary answers a reader’s question, usually your future self’s: is this study useful to me?
The five-sentence template
Write these five sentences for every paper. The order matters; the length does not:
- 1. Question: what did the study want to find out? (“Does X affect Y?”, one sentence.)
- 2. Method: with whom, with how much data, in what design? Give numbers: how many participants, how many years, which measurement.
- 3. Main finding: a single result, with a number where possible. Not “a significant increase” but “a twelve percent increase”.
- 4. Limitation: the most important limit the authors themselves state. Small sample, one country, short duration, self-reported measures.
- 5. What it means for you: how does this finding change your question? If it does not, write that too; “irrelevant” is a result.
An example
Applied to a clinical psychology paper, the template reads like this: “The study asks what predicted anxiety in the general population during the first weeks of the pandemic. Online survey of 1,210 people in China, single time point. About a third reported moderate-to-severe anxiety; being a student and having physical symptoms raised it. Limit: self-reported, one country, no follow-up. For me: a reference for early-phase measurement, not usable for long-term outcomes.”
Five sentences, eighty words. The paper itself is twelve pages.
Which section gives you what
The template’s sentences map onto the paper’s sections: the question sits in the last paragraph of the introduction, the method in the first paragraph of methods and in the tables, the main finding in the first figure of the results, the limitation at the end of the discussion. Knowing the structure makes summarising faster; the IMRaD structure is covered in a separate post.
Common mistakes
- No numbers. “Found to be effective” tells you nothing a month later. Effect size, sample, duration: at least one belongs in the summary.
- Mixing interpretation with findings. What the authors argue in the discussion and what the data shows can differ. Data goes in sentence three, interpretation in sentence five.
- Skipping the limitation. Most people never write sentence four; it is the most common reason a paper gets cited for something it did not show.
- Same length for every paper. Five sentences is too little for a paper at the centre of your field and too much for a peripheral source. The template is a floor, not a ceiling.
Why Pofolia’s summaries share this skeleton
Every paper summary in Pofolia follows the first four sentences in the same order: question, method, finding, limit. That is not a coincidence; it is the structure we give the model when generating them, because what a reader is looking for is the same every time. We do not write the fifth sentence: “what it means for you” is the one only you can write.
And an honest warning: a ready-made summary is good for choosing a paper, not for citing one. Numbers and conditions need checking against the source; where AI summaries go wrong is covered in a separate post.