How to read a journal article without reading all of it

A reading list of twelve papers looks like twelve hours. It is not, because most of those papers do not deserve an hour and you can tell which within three minutes.

Journal articles are written in a fixed order for the benefit of reviewers, not readers. Reading them in that order — abstract, introduction, methods, results, discussion — means grinding through the methods of a paper you have not yet decided to care about.

The three-minute triage

Before reading anything properly, do this for every paper on the list.

1. The abstract. Not for the findings — for whether the question is the one you care about. Papers are frequently on your list because a word matched, not because the argument is relevant.

2. The last paragraph of the conclusion. This is where authors say what they think it means, in plain language, without hedging for reviewers. If you read one paragraph of a paper, read this one.

3. The figures and their captions. In empirical work the figures usually carry the finding, and captions are written to stand alone. You can often extract the entire result here without reading a word of the body.

Three minutes, and you now know whether this is a core paper you will read properly, a supporting paper you will cite for one point, or one you can drop.

Most reading lists sort into roughly two core papers, six supporting, and four you never needed. That is the hour saved.

Reading a core paper properly

Only for the two or three that matter. Still not front to back.

Methods last is the part that feels wrong and saves the most time. Methods are dense, and reading them before you know whether you care about the result is how an afternoon disappears. When you do read them, you are reading with a specific question — does this design support that claim? — which is faster and more critical than reading them cold.

The questions worth asking as you go

Four, and they apply to almost any empirical paper.

Who was studied, and how many? A striking finding from thirty undergraduates at one university is a different thing from the same finding across four countries. This is the single most useful thing to note, and the most commonly omitted from student summaries.

What was actually measured? Papers routinely measure a proxy — self-reported hours rather than hours, intention rather than behaviour — and then discuss the thing itself. The gap between the two is where a lot of over-claiming lives.

What do the authors say it does not show? Almost every discussion section contains a limitations paragraph. It is the most honest part of the paper and the part students skip. It is also where your critical evaluation comes from, ready made.

Who else cites it, and how? A paper cited two hundred times as a foundation is a different animal from one cited twice in passing. Sixty seconds on a citation count tells you how the field treats it.

Record it once, properly

The failure that costs the most time is not slow reading — it is reading something twice because the notes from the first pass were useless.

For every paper you keep, write down:

• The full reference, at the moment you have it open. Never later.
• One sentence on the argument, in your own words. If you cannot write it, you have not understood it yet, and that is worth knowing now.
• Any quotation you might use, in quotation marks, with the page number, in a different colour from your own words. This is the discipline that prevents the most common form of accidental plagiarism — see our guide on avoiding it.
• Your reaction. What it supports, what it contradicts, what it made you think of. In three weeks this is the note you will actually want, and it is the one nobody writes.

Our citation generator will fetch the full details from a DOI and keep the source in your browser, so a paper captured in week two is still there with its reference intact in week nine. That is what turns the referencing phase of an essay from an evening into half an hour.

How long it should take

Rough figures, so a reading plan has something to stand on:

A twelve-paper reading list is therefore about five hours, not twelve — and most of that lands on two or three papers.

Those figures also feed the reading phase in our assignment planner, which budgets three hours per thousand words for reading and notes. If your reading list is longer than that allows, better to find out at the start than in the last week.

One last thing worth saying: if a paper is genuinely impenetrable, it may not be you. Academic writing varies enormously in quality, and some of it is badly written by any measure — our readability checker will score a pasted passage if you want confirmation that a sentence really is as tangled as it felt.