The abstract is the most-read and least-read section of your paper. Most readers decide from it whether to read the rest; most writers draft it last, in a rush, and lose marks that a careful fifteen minutes would have saved. This guide shows you how to write an abstract that earns the time it asks for — the four moves it has to make, the word-count conventions you need to hit, and the difference between an abstract and the introduction it is not. You will see a worked example for an empirical paper, the mistakes that recur across disciplines, and a starting draft pattern you can revise into your own voice.
PaperDraft is a writing assistant, not a paper generator — the draft is your starting point, not your submission. You are responsible for editing, verifying sources, and following your school's academic integrity policy.
What an abstract is — and what it is not
An abstract is a single-paragraph, self-contained summary of a finished paper. It is usually 150–250 words, placed after the title page and before the introduction, and it makes four moves: it names the context and the problem, states what you did, reports what you found, and says why it matters. A reader who has read only your abstract should be able to answer three questions: what question does this paper address, how did the author address it, and what is the answer?
An abstract is not an introduction. The introduction sets up the paper at length, argues for the gap, previews structure, and leans on citations. The abstract does none of that — it does not cite, it does not preview sections, and it does not explain background a reader already has.
An abstract is not a table of contents. "This paper discusses X, then Y, then Z" is a structural description, not a summary. Tell the reader what you found, not what sections you wrote.
An abstract is not a teaser. Do not withhold results to build suspense. In academic writing, the abstract is the spoiler. If you found no effect, say so in the abstract. Readers are choosing whether to invest an hour in your paper — respect that decision.
Finally, an abstract is not the place for hedges you have not earned. Save "may suggest," "could indicate," and "further research is needed" for where they belong in the discussion; in the abstract, state what you actually found.
Before you write
You should only start the abstract after the paper is drafted to the point that nothing structural will change. Methods may be tightened, results may be reformatted, the discussion may be rewritten — but if the question, approach, or core finding is still in flux, you are not ready to write the abstract. Writing it too early is the single most common abstract mistake, because you end up summarizing a paper that no longer exists.
Gather four things before you draft:
- The research question in a single sentence, as it actually appears in your paper's introduction.
- The method in three to five key nouns — "randomized controlled trial, n = 240, pre-post design" or "close reading of three novels by Toni Morrison."
- The headline result — the effect size, the theme, the claim — in one sentence.
- The implication — what the result changes for theory, practice, or future work, in one sentence.
Check the submission requirements. Your target journal, conference, or instructor will specify a word count, sometimes a structured format (Background / Methods / Results / Conclusions as labeled subsections), and occasionally keywords or MeSH terms you must list below. Assume the specification is binding — reviewers and instructors both dock marks for abstracts that overshoot the word limit or ignore a structured-abstract requirement.
If you are submitting to a venue that uses IMRaD-style abstracts for empirical work, the four moves map cleanly onto Introduction, Methods, Results, and Discussion. If it is a structured abstract, each move gets an explicit heading. If it is a narrative abstract (common in the humanities), the moves are still there but they flow as continuous prose.
Step-by-step: how to write an abstract
1. Finish the paper first
Draft the abstract after the paper is stable. If you try to abstract a work-in-progress, you will either describe a version that no longer exists by submission or keep rewriting the abstract every time the methods shift. One full pass of the paper from introduction to conclusion, with all citations in place, is the minimum before you open the abstract document.
2. Identify the four moves
Every academic abstract makes four moves, even when the paper is theoretical rather than empirical. For an empirical paper: context/motivation → methods → results → implications. For a theoretical or humanities paper: problem/context → approach/argument → finding/claim → implications. Write these four on a blank page first. If you cannot articulate each in one sentence, the paper itself may have a clarity problem worth solving before you abstract it.
3. Write one sentence per move, then expand
Take your four sentences and expand each to two or three, staying proportional. A 200-word abstract for an empirical paper might split roughly 40/50/70/40 — a little less for context, a little more for methods and results, short implications. Do not let any single move eat more than half the word budget.
4. Match tense and register to each move
Use past tense for the context you described, the methods you applied, and the results you obtained ("We examined," "Participants completed," "Scores increased"). Use present tense for framing statements about what the paper does and what the results imply ("This paper argues," "The findings suggest"). Keep the register formal and the voice active where convention allows. Avoid first-person plural in disciplines that discourage it; check past issues of your target venue.
5. Strip every sentence that is not doing work
An abstract at 240 words that could be 210 is a 210-word abstract with 30 words of padding. Cut generic openers ("In recent years, there has been growing interest in..."), cut restated thesis sentences, cut transitions between the four moves. Every sentence should add a fact a reader did not already have.
6. Check against journal or instructor specs
Count words (not characters). Confirm the word count is at or under the limit. If the venue asks for keywords, list five to eight, ordered from most to least specific. If structured headings are required, use them verbatim from the style guide.
7. Read it cold the next day
Set it aside overnight. Read it the next morning, out loud, pretending you have never seen the paper. If you cannot answer what was done, what was found, and what it means from the abstract alone, revise. If the four moves are there and the prose does not trip you, you are finished.
Stuck at the start? PaperDraft scaffolds a starting abstract — the moves, the register, the structure — for you to revise. Start this paper — free.
Example abstract
Below is a 198-word empirical abstract with the moves annotated in brackets. Annotations do not appear in the final version — they are here to show which move each sentence performs.
[Context/motivation] Undergraduate students report high rates of academic procrastination, yet existing interventions focus primarily on time-management training with mixed results. [Motivation continued] Recent work has suggested that self-compassion, rather than self-discipline, may be the more tractable mechanism for reducing avoidance behavior. [Methods] We conducted a pre-registered randomized controlled trial (N = 240) comparing a four-week self-compassion intervention to an active control focused on implementation intentions. [Methods continued] Participants completed the Academic Procrastination Scale, the Self-Compassion Scale, and behavioral measures of study-session initiation at baseline, post-intervention, and four-week follow-up. [Results] Relative to the control, the self-compassion condition showed a reliable reduction in procrastination scores (d = 0.42, 95% CI [0.17, 0.67]) that was maintained at follow-up. [Results continued] The effect was mediated by changes in self-compassion scores rather than shifts in perceived time pressure. [Implications] These findings extend prior correlational work by establishing a causal pathway from self-compassion to reduced procrastination in an academic context. [Implications continued] Brief self-compassion protocols may be a lower-friction alternative to time-management training for student populations.
Notice what is not there: no citations, no section previews, no background the field already has, no hedging about limitations. Every sentence delivers a fact the reader did not have before the sentence.
Common mistakes
- Writing the abstract before the paper is finished. The abstract summarizes a stable paper; drafting it early means summarizing a draft that will change. Finish the paper first.
- Using the introduction as the abstract. The introduction argues for the importance of the problem; the abstract reports what you did and found. A paragraph that ends at "this paper will examine..." is an introduction paragraph, not an abstract.
- Omitting results. The most common structural failure. Students describe the problem and the method, then stop. A reader must know what you found. Honest null results belong in the abstract too — see our academic responsibility guide for why accurate reporting matters more than flattering framing.
- Citing sources. Abstracts do not cite. If a claim in your abstract needs a citation to be defensible, it is background — cut it or rework it into your own finding.
- Missing the word count. Overshooting a 150-word limit by 40 words looks like you did not read the instructions. Undershooting by 80 looks like you did not finish.
How PaperDraft helps you start
PaperDraft is a writing assistant, not a paper generator — the draft is your starting point, not your submission. For abstracts, PaperDraft scaffolds a starting version with the four-move structure in place, matched to whether your paper is empirical or theoretical, in the register your discipline expects. You then revise it into your own abstract, verify that every claim in it is supported by the paper, and match it to your venue's word count and formatting rules. The scaffolding is the quickest part; the rewriting and verification are the work that earns your name on the paper. Start scaffolding a research paper abstract at PaperDraft's research-paper workflow, and check our APA citation guide for the formatting conventions of the surrounding document.
Frequently asked questions
How long should my abstract be?
Most abstracts fall between 150 and 250 words. APA empirical abstracts are typically 150–250, humanities journals often cap at 150–200, and some STEM venues allow up to 300. Always check your specific venue or instructor's spec — the conventions vary more than students expect.
Do I include citations in my abstract?
No. Standard academic abstracts do not cite sources. If a claim in your abstract requires a citation to be defensible, rewrite it as a finding of your own paper or move it into the introduction.
Should I write the abstract first or last?
Last. An abstract summarizes a finished paper. Writing it before the paper is stable means you will either keep rewriting it as the paper changes, or submit an abstract that no longer matches the paper.
What is a structured abstract?
A structured abstract uses explicit subheadings — typically Background, Methods, Results, and Conclusions — rather than continuous prose. It is common in medical, health, and some empirical social-science journals. Check your target venue's style guide.
Do I have to disclose if I used an AI tool to draft my abstract?
Probably yes, depending on your program's policy. Many universities and journals now require disclosure when generative AI has been used in drafting. See our AI disclosure guide for program-specific conventions and how to word the disclosure statement honestly.