Abstract Examples for Research Papers: 6 Annotated Samples + How to Write One

Quick answer

An abstract is a 150–250 word summary of a paper that lets readers decide whether to read the full text. A good one makes five moves in order: background (why it matters), aim (the question), methods (what you did), results (what you found, with numbers) and conclusion (what it means). Structured abstracts label these parts; unstructured ones run them together as one paragraph.

What Is an Abstract? (The 5 Moves)

An abstract is the short summary at the top of a research paper, thesis or conference submission. Most readers only read the title and the abstract, and search engines and databases index it first, so it decides whether your paper gets read at all.

Almost every good research paper abstract makes the same five "moves", usually one or two sentences each:

1. Background – the problem and why it matters.
2. Aim – the research question, objective or hypothesis.
3. Methods – design, sample or data, and main measures.
4. Results – the main findings, with actual numbers.
5. Conclusion – what the findings mean and why someone should care.

Below you will find six original abstract examples across fields, each annotated move by move, followed by real published abstracts you can open as free PDFs.

Updated October 2026.

1. Background — Why it matters
2. Aim — Question or hypothesis
3. Methods — Design, sample, measures
4. Results — Key numbers
5. Conclusion — So what?

Structured vs Unstructured vs Descriptive vs Informative Abstracts

Abstracts differ in two ways: whether they use headings (structured vs unstructured) and whether they report findings (informative vs descriptive).

• Structured abstract: separate labelled parts such as Background, Methods, Results, Conclusions. Common in medicine, nursing and public health journals.
• Unstructured abstract: one paragraph with the same moves but no headings. Common in psychology (APA style), computer science, economics and most sciences.
• Informative abstract: reports the actual results and conclusions. This is what almost every empirical research paper needs.
• Descriptive abstract: describes what the paper covers, but not what it found. Short (often under 150 words) and common in humanities essays, review chapters and some conference programmes.

Always check the journal or course instructions first: they decide which type you use.

Example 1: Structured Abstract Example (Medicine, IMRaD)

Illustrative example written for this guide; not a real study.

Title: Effect of a Community Health Worker Text-Message Programme on Blood Pressure Control in Adults with Hypertension: A Cluster Randomised Trial

Background: Uncontrolled hypertension is a leading cause of stroke and heart disease in low- and middle-income countries, and many patients stop taking medicines within months of diagnosis.

Objective: To test whether weekly text-message reminders sent by community health workers improve blood pressure control.

Methods: We randomised 24 primary health centres in two districts to an intervention (weekly reminder messages plus monthly phone follow-up) or usual care. Adults aged 30–70 years with diagnosed hypertension were eligible. The primary outcome was the proportion of participants with blood pressure below 140/90 mm Hg at 12 months. Analysis was by intention to treat with adjustment for clustering.

Results: Of 1,412 participants enrolled, 1,306 (92%) completed follow-up. At 12 months, 48% of the intervention group and 36% of the usual-care group had controlled blood pressure (adjusted risk difference 11 percentage points; 95% CI 6 to 16). Mean systolic pressure fell by 4.2 mm Hg more in the intervention group. No adverse events were linked to the programme.

Conclusions: A low-cost text-message programme delivered by community health workers improved blood pressure control over one year. Health systems with existing community health worker networks could adopt it at scale.

Example 2: APA Abstract Example (Psychology)

Illustrative example written for this guide; not a real study.

In APA 7 student and professional papers, the abstract sits on its own page after the title page, with the bold, centred label Abstract, a single paragraph with no indent, and an optional Keywords: line underneath.

Abstract

Students often check their phones while studying, but it is unclear whether simply having a phone nearby reduces learning when it is not used. We tested whether phone location affects reading comprehension in undergraduates. In a preregistered experiment, 186 students were randomly assigned to study a 1,200-word text with their phone on the desk (face down, silent), in a bag, or in another room. Comprehension was measured with a 20-item test, and participants rated their self-reported distraction. Students whose phones were in another room scored higher (M = 14.1, SD = 2.9) than those with phones on the desk (M = 12.6, SD = 3.1), d = 0.50, while the bag condition fell between them. Self-reported distraction did not differ between groups, suggesting the effect may occur without students noticing it. These findings suggest that the mere presence of a smartphone can carry a small but meaningful cost for learning, and that advice to "put the phone away" should mean moving it out of the room. We discuss limits of a single-session laboratory task and directions for classroom field studies.

Keywords: smartphones, attention, reading comprehension, learning, distraction

Example 3: Computer Science Abstract Example

Illustrative example written for this guide; not a real paper.

Title: LiteDetect: A Lightweight Model for Detecting Crop Leaf Disease on Low-Cost Smartphones

Plant disease causes large yield losses for smallholder farmers, and deep learning models can detect disease from leaf photos. However, most accurate models are too large to run offline on the low-cost phones these farmers use. We present LiteDetect, a compact convolutional network that combines depthwise-separable convolutions with knowledge distillation from a larger teacher model. We train and evaluate it on a public dataset of 54,000 leaf images covering 38 crop–disease classes, and on a new field set of 2,100 photos taken by farmers under natural light. LiteDetect reaches 96.8% top-1 accuracy on the public test set and 88.4% on the field set, within 1.5 percentage points of the teacher model, while using 14 times fewer parameters (1.9 million) and running in 41 ms per image on a mid-range Android phone. An ablation study shows that distillation accounts for most of the accuracy gain. Our results show that accurate disease detection is possible fully offline on inexpensive hardware. We release the code and the field dataset to support further work.

Example 4: Economics Abstract Example

Illustrative example written for this guide; not a real study.

Title: Mobile Money and Household Resilience to Rainfall Shocks: Evidence from Rural Districts

Rural households in developing economies are exposed to weather shocks that reduce income and force cuts in food spending. This paper asks whether access to mobile money helps households smooth consumption after rainfall shocks. I combine four rounds of a household panel survey (n = 3,200 households, 2016–2022) with satellite rainfall data and exploit the staggered rollout of mobile money agents across villages in a difference-in-differences design. After a negative rainfall shock, households without mobile money access reduce food consumption by 9%, whereas households with access show no significant decline. The difference is explained mainly by larger and faster remittances from relatives in cities, not by savings. Effects are strongest for female-headed households and for villages more than 10 km from a bank branch. Event-study estimates show no pre-trends, and results are robust to alternative shock definitions. The findings suggest that expanding mobile money agent networks can serve as low-cost social protection against climate risk.

Example 5: Biology Abstract Example

Illustrative example written for this guide; not a real study.

Title: Urban Light Pollution Shifts the Daily Activity of a Common Garden Lizard

Artificial light at night is spreading rapidly, but its effects on reptiles are poorly understood compared with birds and insects. We asked whether night lighting changes the daily activity and body condition of a common garden lizard. Over two breeding seasons, we tagged 142 adult lizards at six urban sites with street lighting and six nearby dark sites, and recorded activity with miniature accelerometers. We also measured body condition and counted insect prey with light traps. Lizards at lit sites began activity 38 minutes earlier and stayed active 52 minutes later on average than lizards at dark sites, and 21% were recorded foraging after sunset, a behaviour never seen at dark sites. Insect prey was 2.4 times more abundant near lights. Despite this, body condition did not differ between site types, suggesting that extra foraging time may be offset by higher energy use or predation risk. Our results show that light pollution can reshape the daily routine of a diurnal reptile, and that management of urban lighting should consider reptiles as well as birds and insects.

Example 6: Descriptive Abstract Example (Humanities)

Illustrative example written for this guide; not a real article.

Title: Letters Home: Migration, Memory and Language Choice in Early Twentieth-Century Family Correspondence

This article examines a collection of 214 family letters exchanged between labour migrants and their relatives over three decades of the early twentieth century. It considers how writers moved between their home language and the language of the host country, and how these choices reflected changing ideas of belonging. The article first places the letters within the history of labour migration and the spread of basic literacy. It then analyses three themes: the use of religious and seasonal formulas to keep ties with home, the gradual entry of borrowed words for work and money, and the silences around illness and death. Drawing on approaches from historical sociolinguistics and memory studies, the article argues that private letters should be read not only as sources of facts but as places where migrants actively negotiated their identities. It concludes by discussing what such ordinary writing can add to histories that rely mainly on official records.

Real Published Abstracts to Study (Free PDFs)

The examples above are illustrative. To see how professional authors do it, open these real, open-access papers and read the abstract first, then compare it with the full text. Each one shows a different abstract style:

• NEJM (structured, clinical trial): labelled Background, Methods, Results, Conclusions; every result has a number.
• BMJ (structured, meta-analysis): BMJ-style headings such as Objective, Design, Data sources, Results, Conclusions.
• eLife (unstructured, modelling study): one flowing paragraph with a strong opening question.
• arXiv / NeurIPS (unstructured, computer science): the famous Transformer paper abstract states the problem, the new model, and benchmark results in under 200 words.

You can find thousands more: search any topic on Ocean of Papers and read the abstracts in the results before opening a PDF. For full papers by field, see our research paper samples.

• Remdesivir for 5 or 10 Days in Patients with Severe Covid-19 — Goldman, J. D., Lye, D. C. B., Hui, D. S., et al. (2020), New England Journal of Medicine / PMC. Structured abstract (RCT). A classic structured clinical abstract: Background, Methods, Results and Conclusions, with the primary outcome and adjusted comparison stated clearly.
• Dose-Response Associations Between Accelerometry Measured Physical Activity and Sedentary Time and All Cause Mortality: Systematic Review and Harmonised Meta-Analysis — Ekelund, U., Tarp, J., Steene-Johannessen, J., et al. (2019), BMJ / PMC. Structured abstract (meta-analysis). Shows the BMJ structured format (Objective, Design, Data sources, Eligibility criteria, Results, Conclusions) — a good model for systematic review abstracts.
• Publication Bias and the Canonization of False Facts — Nissen, S. B., Magidson, T., Gross, K., Bergstrom, C. T. (2016), eLife / PMC. Unstructured abstract. A single-paragraph abstract that opens with a big question ("Does this imply that many scientific facts are false?") and moves quickly to the model and its answer.
• Attention Is All You Need — Vaswani, A., Shazeer, N., Parmar, N., et al. (2017), NeurIPS 2017 / arXiv. Unstructured abstract (CS). A short computer science abstract: the problem with existing models, the new architecture, and benchmark results with exact BLEU scores and training cost.

Abstract Word Limits by Style and Paper Type

Word limits vary a lot, so the journal's "instructions for authors" page always wins. As general guidance:

• APA 7: university guides based on the APA Publication Manual (section 2.9) commonly cite a limit of 250 words for most papers.
• Most journals: usually ask for 150–300 words; medical journals with structured abstracts are often near the top of that range.
• Conferences: abstract submissions for talks or posters are often 250–500 words and may include a short results section even if the work is in progress.
• Theses and dissertations: often longer (up to one page); check your university's thesis rules.

If no limit is given, aim for 150–250 words.

To see where the abstract sits in a full paper, see our APA research paper example.

How to Write an Abstract: Step by Step

Write the abstract last, after the rest of the paper is finished. Then:

1. Read the instructions. Note the word limit, whether headings are required, and whether keywords are needed.
2. Pull one sentence from each section. Take the key point from your introduction, methods, results and discussion. This is your rough draft.
3. State the gap and the aim clearly. One sentence on what is unknown, one on what you did about it ("We tested whether…", "This paper examines…").
4. Give methods in one or two sentences. Design, sample size, setting and main measure. Skip details a reader does not need to judge the study.
5. Report the main results with numbers. Effect sizes, percentages, confidence intervals or accuracy — not "significant differences were found".
6. End with the meaning. What should readers think or do differently? Do not claim more than your data show.
7. Cut to the limit. Remove citations, abbreviations you do not reuse, and background that does not lead to your question.
8. Add keywords that someone would type into a database, including synonyms not in the title.
9. Check it stands alone. A reader who sees only the abstract should understand what you did and what you found.

For the full paper, see how to write a research paper. If you are writing a proposal rather than a finished study, your abstract describes planned methods and expected outcomes instead — see our research proposal example. For review papers, the literature review guide explains how to summarise search methods and themes.

Common Abstract Mistakes (and Fixes)

• No results. "Results will be discussed" tells the reader nothing. Fix: state the main finding with a number.
• Too much background. Half the abstract spent on context. Fix: one or two sentences of background at most.
• Missing the aim. The reader cannot tell what question was asked. Fix: add an explicit "We aimed to…" or "This paper asks…" sentence.
• New information. Claims that do not appear in the paper. Fix: everything in the abstract must be in the main text.
• Citations and undefined abbreviations. Fix: avoid references and spell out abbreviations on first use.
• Overclaiming. "Proves" or "cures" from one small study. Fix: match the strength of words to the strength of evidence.
• Ignoring the word limit. Many submission systems cut or reject long abstracts. Fix: count words before submitting.
• Vague methods. "A study was conducted." Fix: name the design, sample size and setting.

Frequently Asked Questions

What is a good abstract example for a research paper?

A good research paper abstract example states the problem, the aim, the methods, the main results with numbers, and the conclusion in 150–250 words. Examples 1–5 above show this in medicine, psychology, computer science, economics and biology. For real published examples, open the free PDFs in the "Real Published Abstracts" section.

How do I write an abstract for a research paper?

Write it last. Take one key sentence from each section of your paper (introduction, methods, results, discussion), add the main numbers, then cut it to the word limit and check that it makes sense on its own.

What is a structured abstract example?

A structured abstract uses labelled headings such as Background, Objective, Methods, Results and Conclusions. Example 1 above is a structured abstract example, and the NEJM and BMJ papers listed above are real ones.

What does an APA abstract example look like?

In APA 7, the abstract goes on its own page with the bold, centred label "Abstract", is written as one unindented paragraph, and is commonly limited to 250 words. An optional keywords line follows in italics. See Example 2 above.

How long should an abstract be?

Usually 150–250 words, but journals and conferences set their own limits, so check the instructions for authors. Structured medical abstracts and conference abstracts are often longer.

What is the difference between a descriptive and an informative abstract?

An informative abstract reports results and conclusions; a descriptive abstract only says what the paper covers. Most empirical research papers need an informative abstract, while humanities essays and reviews sometimes use a descriptive one (Example 6).

Should an abstract include citations?

Usually no. Most styles and journals ask you to avoid references in the abstract so it can stand alone. If you must mention a key earlier study, describe it in words.

Where can I find more abstract examples for research papers?

Search your topic on Ocean of Papers — each result shows the abstract, and many link to a free, legal PDF from PubMed Central, arXiv, bioRxiv, medRxiv or Unpaywall. Reading ten abstracts from your target journal is the fastest way to learn its style.

Related guides

• Research paper examples (annotated, free PDFs)
• How to write a research paper
• Research proposal example (annotated) + free PDFs
• Thesis and dissertation examples (free PDFs)
• How to write a literature review (with example PDFs)
• Quantitative research example (Chapters 1–5)
• APA research paper example (7th edition)

Search free research papers · Build a citation map from any search result · Browse all research guides