How to Write an Abstract (With Examples)
How to write an abstract for a research paper, dissertation, conference, or proposal. Covers the 4 core elements, both formats, and annotated examples.
An abstract is a self-contained summary of a research paper, thesis, dissertation, conference presentation, or grant proposal. It gives readers enough information to decide whether to read the full work, and it appears before the paper itself, often in databases and search results where it may be the only part a reader ever sees.
Most abstracts are 150 to 250 words. A strong abstract is not a teaser or an introduction; it is a miniature version of the full paper that stands entirely on its own.
This guide covers the four core elements every abstract needs, the two main format types, how to write each type, and annotated examples for research papers, dissertations, conference proposals, and grant proposals.
The 4 Core Elements of Every Abstract
Regardless of discipline or document type, every abstract answers the same four questions in roughly this order:
| Element | The question it answers | What to include |
|---|---|---|
| Objective | Why does this research exist? | The research question, the gap it addresses, or the problem it solves |
| Methods | How was it done? | Study design, participants or data, analytical approach |
| Results | What was found? | Key findings or outcomes, with specific detail |
| Conclusion | So what? | The main implication, contribution, or takeaway |
These four elements are sometimes called IMRaD (Introduction, Methods, Results, and Discussion), though the abstract version compresses each to 1 to 3 sentences. In some disciplines and documents the order shifts slightly (proposals put conclusions last because results do not exist yet), but the four questions must always be answered.
Two Abstract Formats
Format 1: Paragraph abstract (most common)
A paragraph abstract is a single block of uninterrupted prose, usually 150 to 250 words. It is the default format for most journals in the social sciences, humanities, and many STEM fields.
It reads as a continuous argument, not as a list. Each sentence flows into the next. There are no headings or bullet points.
Format 2: Structured abstract
A structured abstract divides the four elements into labeled sections, usually bolded or set as headings. It is required by most medical, clinical, and many life science journals, and increasingly by social science journals as well.
Common heading labels for structured abstracts:
- Background / Context (or Introduction)
- Objective / Purpose
- Methods (or Design and Setting)
- Results (or Findings)
- Conclusions (or Implications)
Structured abstracts tend to run slightly longer (200 to 350 words) because each section needs a full sentence or two. Check the submission guidelines for your specific journal, conference, or institution. Format requirements are non-negotiable.
How to Write an Abstract: Step by Step
Step 1: Write it last
The abstract summarizes a paper that already exists. Writing it before the paper is finished almost always produces a vague or inaccurate abstract, because the exact findings and framing are not yet fixed. Write the abstract after the rest of the document is complete and revised.
Step 2: Start with the objective
The first sentence should establish the research question or problem clearly. Avoid broad scene-setting ("Many researchers have studied X...") and get to the specific gap or question within the first sentence or two.
Weak opening: "Mental health is an important issue for many people in society."
Strong opening: "Doctoral students experience anxiety and depression at rates significantly higher than age-matched peers, yet few interventions have been evaluated in this population."
Step 3: Summarize methods in one to two sentences
Name the study design, the sample or data source, and the key analytical method. Be specific but concise. A reader should be able to assess the reliability of your findings from this section.
Step 4: State the key results with specificity
This is the most commonly weak section of student abstracts. Vague results statements ("Results were significant" or "Themes were identified") tell a reader nothing. Name the specific finding.
Weak: "The results showed that students were affected by academic pressure."
Strong: "Students who received biweekly advisor check-ins showed a 31% reduction in self-reported burnout symptoms over one semester compared to a waitlist control group."
Step 5: State the conclusion and implication
End by telling the reader what the finding means for the field, for practice, or for future research. This is one or two sentences maximum.
Step 6: Check length and remove every unnecessary word
Count the words. If you are over the target length, cut:
- Filler phrases ("It is important to note that...")
- Redundant language ("past history," "current existing data")
- Background you already covered in the objective
- Hedging beyond what is scientifically necessary
Every sentence in an abstract should add specific information. If removing a sentence leaves the abstract complete, the sentence should not be there.
Annotated Example: Research Paper Abstract
The following is an example abstract for a psychology research paper, annotated to show each element.
Doctoral students represent a population at elevated risk for psychological distress, yet evidence-based interventions developed specifically for this group remain scarce. [OBJECTIVE: states the gap and why the study is needed]
This study examined whether structured peer mentoring groups reduced burnout symptoms in doctoral students across two research universities. Participants (N = 148, years 2–4 of their programs) were randomly assigned to an 8-week peer mentoring intervention or a waitlist control condition. Burnout was assessed at baseline, post-intervention, and 12-week follow-up using the Maslach Burnout Inventory (Student Survey). [METHODS: design, sample, instrument, procedure]
Students in the intervention condition showed significantly lower emotional exhaustion at post-intervention (d = 0.54, 95% CI [0.28, 0.79]) and at 12-week follow-up (d = 0.41). No significant difference in cynicism or academic efficacy was observed. [RESULTS: specific, with effect sizes and confidence intervals]
These findings suggest that structured peer mentoring is a feasible and modestly effective approach to reducing emotional exhaustion in doctoral students, with implications for graduate school mental health programming. [CONCLUSION: implication for practice]
(Word count: 172)
Annotated Example: Dissertation Abstract
A dissertation abstract covers the same four elements but often runs slightly longer (200 to 350 words) and must stand completely independent of the rest of the document, since it will be indexed in ProQuest Dissertations and Theses and may be read in isolation.
Despite decades of investment in doctoral education reform, approximately 40 to 50 percent of students who begin PhD programs in the United States do not complete them. Existing research identifies advisor relationship quality as a key predictor of completion, yet the mechanisms linking advisor behavior to student persistence have not been fully specified. [OBJECTIVE]
Drawing on survey data from 412 doctoral students across 17 STEM programs and 22 semi-structured interviews with completers and non-completers, this dissertation examines how specific advisor behaviors (feedback frequency, autonomy support, and professional socialization) affect student persistence decisions at the dissertation stage. [METHODS]
Hierarchical regression analyses indicate that feedback frequency and autonomy support independently predicted persistence intent after controlling for program-level factors and student characteristics. Interview data revealed three patterns of advisor-student relationships: collaborative partnerships associated with high persistence, transactional relationships with moderate persistence, and isolated relationships associated with attrition. Students in isolated relationships described the absence of clear completion milestones as the primary driver of departure decisions. [RESULTS]
These findings extend the literature on doctoral socialization by specifying the advisor behaviors most predictive of persistence and suggesting a model of relationship types. Practical implications for program design, advisor training, and institutional support structures are discussed. [CONCLUSION]
(Word count: 218)
Writing Abstracts for Different Purposes
| Abstract type | Typical length | Key differences from a paper abstract |
|---|---|---|
| Research paper | 150-250 words | Reports results that already exist |
| Dissertation / thesis | 200-350 words | Must be fully self-contained; indexed separately |
| Conference presentation | 150-300 words | Reviewers assess whether the work is worth presenting; results may be preliminary |
| Grant proposal | 200-300 words | No results yet; emphasize gap, approach, and expected contribution |
| Poster session | 100-200 words | Shorter; emphasis on visual clarity and inviting conversation |
Abstract for a Conference Proposal
Conference abstracts are peer-reviewed before the conference takes place, which means the abstract is what gets your work accepted. If you are submitting before the project is complete, state what findings you expect to present and what you will have confirmed by the conference date.
The structure is the same four elements, but the results section becomes: "This paper will present X, based on Y data, collected through Z period."
Abstract for a Grant Proposal
Grant proposal abstracts appear before a funding application and are often the only section that every reviewer reads before scoring. They must convey:
- The problem and its significance
- The specific aims of the proposed work
- The approach (methods)
- The expected outcome and its value to the field
Unlike a paper abstract, you are describing work you plan to do, not work you have done. Use future tense where appropriate ("This study will examine...") and be specific about what you will produce.
Common Mistakes in Abstracts
Including citations. Abstracts must be self-contained. Citations require the reader to go elsewhere to understand the abstract, which defeats the purpose. Exception: if your abstract references a specific dataset or established instrument by name (e.g., "the Patient Health Questionnaire-9"), you do not need a citation for that.
Writing the abstract before the paper is finished. This almost always produces a mismatch between the abstract and the final paper, because conclusions evolve during writing.
Vague results statements. "Results were significant" or "Several themes emerged" give the reader no useful information. State the specific finding, including effect sizes, percentages, or key themes where possible.
Treating the abstract as an introduction. An abstract does not set up questions; it answers them. It does not preview what is coming; it reports what was done and found.
Exceeding the word limit. Abstracts with word limits are submitted to systems that count words exactly. An abstract that is 260 words when 250 is the limit will be rejected or truncated.
Defining acronyms. The abstract cannot assume the reader has seen any other part of the paper. Acronyms should either be spelled out or avoided entirely. Avoid technical jargon that a general reader in your field would not know.
Frequently Asked Questions
What are the 5 components of an abstract? Most frameworks describe 4 components: objective (why), methods (how), results (what), and conclusion (so what). Some frameworks add "background" as a fifth element, placed before the objective, to provide brief context for the research question. In practice, the background is usually folded into the objective section unless the journal's structured abstract format requires a separate heading.
How do I start my abstract? Start with the research problem or gap, stated specifically. Avoid broad openings like "This paper examines..." or "X is an important topic..." The first sentence should orient the reader to the specific context of your work. A reliable template: "[Population/Context] faces [problem], yet [gap in current knowledge]. This study [did X]."
What is an example of an abstract? See the two annotated examples above for a research paper (172 words) and a dissertation (218 words), with each element labeled. The key principle both illustrate: specific findings stated with measurable detail, not general statements that could describe any paper in the field.
What are the four C's of abstract writing? The four C's are: Clear (avoid jargon and ambiguity), Concise (no unnecessary words), Complete (all four elements answered), and Correct (factually accurate, with no overclaiming). A useful self-check: does each sentence add a specific piece of information that is not duplicated elsewhere in the abstract?
How long should an abstract be? Standard range is 150 to 250 words for most papers. Dissertation abstracts run 200 to 350 words. Structured abstracts in medical journals may run up to 400 words. Conference and proposal abstracts vary widely by organization. Always check the submission guidelines.
For writing the full paper these abstracts summarize, see the companion guides on how to write a literature review and how to write a PhD dissertation.
Amos Oppong
View profile →Keep reading.
- ● research