Writing an abstract
An abstract is a short summary of a longer piece of work, usually between 150 and 300 words. It appears at the front of journal articles, conference proposals and some assessment tasks, and it is usually the first part of your work that most people will read.
Readers use abstracts to decide whether to read further, editors use them to decide whether to send a manuscript out for review, and conference committees use them to accept or reject proposals. Your abstract carries a lot of weight for its length.
Because an abstract summarises work you have already done, it is often best written last. Some writers also draft one early as a thinking tool, which is covered further down this page.
What goes into an abstract
Most abstracts move through the same five stages, whatever the discipline (Hyland, 2000). Not every abstract includes all five, and the balance shifts depending on your field, but the sequence is generally consistent.
| Move | What it does |
|---|---|
| Background | Establishes the topic and why it matters. |
| Gap or problem | Identifies what is missing, unresolved or contested. |
| Aim | States what your study or paper sets out to do. |
| Method | Explains how you did it. |
| Findings or argument | Reports what you found or what you argue. |
| Implications | Says what the outcomes mean and why the reader should care. |
Annotated examples
Abstracts look different across disciplines. An abstract reporting an experiment has to account for a method and a set of results, while one introducing an argument-based article needs to signal a position and the evidence behind it. Compare these two, and read the abstracts published in your own field alongside them.
Physical inactivity among university students is associated with poorer mental health outcomes, yet participation in campus sport remains low [Background]. Existing research has focused largely on facility access, leaving motivational barriers underexplored [Gap]. This study examined why first-year students who intend to exercise do not go on to do so [Aim]. A survey of 412 first-year students at an Australian university was followed by semi-structured interviews with 18 participants [Method]. Intention was found not to predict participation, and perceived social risk emerged as the strongest reported barrier [Findings]. Campus programs that reduce the social exposure of beginners may therefore be more effective than those that expand facility access [Implications].
Note how much work the method sentence does in very little space: sample size, setting, and both data sources. Readers in empirical fields scan for exactly this.
Depictions of automation in interwar British fiction are conventionally read as expressions of anxiety about unemployment [Background]. This reading rests, however, on a small group of canonical texts [Gap]. This article examines the treatment of machinery in twelve lesser-known novels published between 1919 and 1939 [Aim and method combined]. It argues that these works treat automation less as an economic threat than as a challenge to established ideas of craftsmanship and masculine identity [Argument]. Recognising this shift complicates the standard account of interwar technological pessimism, and suggests the anxiety was as much about status as about work [Implications].
Here the aim and method fold into a single sentence, and "findings" become an argument. The underlying sequence is the same.
Australian crime fiction has been read predominantly through the lens of national identity [Background]. The genre's treatment of urban space has attracted far less attention [Gap]. This article surveys twenty novels published since 2000, grouped by setting, and considers how each constructs the relationship between crime and place [Aim and scope]. Directions for further work on regional and suburban settings are outlined in the closing section [Signposting, not findings].
This one describes what the article covers and how it is organised, without reporting findings. Descriptive abstracts are common in the humanities and in review pieces, but check what your field expects, as most empirical journals now require the informative form.
Background. Medication non-adherence is common among older adults managing multiple chronic conditions, and pharmacist-led review has been proposed as a low-cost response. Evidence from Australian primary care settings remains limited [Gap].
Objective. To evaluate whether a pharmacist-led medication review improves adherence among patients aged over 65 taking five or more medications.
Methods. A randomised controlled trial was conducted across six general practices in regional Victoria. Two hundred and forty patients were allocated to review or usual care, with adherence measured at six and twelve months by pill count and self-report.
Results. Adherence was higher in the intervention group at six months, but the difference was no longer significant at twelve.
Conclusions. Pharmacist-led review produces short-term gains that are not sustained without follow-up, suggesting review should be scheduled repeatedly rather than delivered once [Implications].
The headings look like they map onto the moves, but they do not quite: the gap sits inside Background, and the implications inside Conclusions. Check your target journal's required headings, since they vary and are usually non-negotiable.
Note-taking in lectures is increasingly mediated by AI transcription and summarisation tools [Background]. Existing studies of student note-taking predate the widespread availability of these tools [Gap]. This paper reports on a study, currently in data collection, examining how undergraduate students integrate AI-generated summaries into their own notes [Aim]. Screen recordings and think-aloud protocols are being collected from 24 students across two faculties, with analysis to be completed in October [Method]. Preliminary observations suggest students treat the summary as a substitute for attention during the lecture rather than a supplement after it [Provisional findings]. The presentation will set out the full analysis and consider what it means for how lecturers structure material [Implications].
Proposals are written before the work is finished, so you are describing what you will have done by the time you present.
What to leave out
View
Structured and unstructured abstracts
Some journals, particularly in health and medical fields, require a structured abstract with explicit subheadings such as Background, Methods, Results and Conclusions. Others expect an unstructured abstract, written as a single continuous paragraph.
Conference proposals are often shorter, and are usually written before the work is complete. In that case you are describing what you will have done by the time you present, which requires some forward thinking.
Using an abstract to plan your writing
Although abstracts are usually finalised last, drafting one early can help you work out what you are actually arguing. If you cannot summarise your paper in 250 words, that is often a sign the argument has not settled yet.
One approach is to answer a set of questions and build the answers into a draft. Brown's 'Eight Questions' is a well-known example.
Write for 30 minutes on the subject of your paper.
| Question | Word limit |
|---|---|
| 1 Who are the intended readers? List three to five of them by name. | 50 words |
| 2 What did you do? | 50 words |
| 3 Why did you do it? | 50 words |
| 4 What happened when you did it? | 50 words |
| 5 What do the results mean in theory? | 50 words |
| 6 What do the results mean in practice? | 50 words |
| 7 What is the key benefit for readers? | 25 words |
| 8 What remains unresolved? | No limit |
Adapted from Brown (1994/95, p. 6, as cited in Murray, 2013, p. 131).
The answers give you the core details for an abstract and they also help you structure the paper itself. Murray (2013) suggests identifying the keywords in each answer, building those keywords into your section headings, and then reflecting on them through topic sentences in each section. Define and explain the terms as you go.
Before you submit
- Does the abstract stand alone, making sense to someone who has not read the paper?
- Does it match the finished paper, rather than the paper you planned to write?
- Are the findings given more space than the background?
- Is it within the word limit set by the journal, conference or unit?
- Have you used the keywords a reader would search for?
References
Hyland, K. (2000). Disciplinary discourses: Social interactions in academic writing. Longman.
Murray, R. (2013). Writing for academic journals (3rd ed.). Open University Press.