Writing a clearer title and abstract

The short answer

A clear title tells readers exactly what the paper is about. A clear abstract tells them what you did, what you found, and what it means, in a form that stands on its own. Both must be accurate to the full paper, and both are written for someone deciding whether to read further: an editor screening submissions, a researcher scanning search results, or a database matching search terms. The ICMJE Recommendations (external link) point out that abstracts are the only substantive portion of an article indexed in many electronic databases, which is why their accuracy and completeness matter so much.

What a title must do

A research title has three jobs.

• Be accurate. It describes what the study actually examined and does not imply a stronger finding, a broader population, or a causal relationship the design cannot support.
• Be specific. It names the topic precisely and, where relevant, the population, setting, intervention, or design. “Stress in nurses” is a topic; a specific title says which nurses, which aspect of stress, and how it was studied.
• Be findable. It contains the words a researcher would type into a database. The ICMJE describes the title as a distilled description of the complete article that, together with the abstract, should make electronic retrieval both sensitive and specific.

Common title patterns

• Descriptive: states the topic and approach without the result, typically naming the exposure or intervention, the outcome, the population, and the design.
• Declarative: states the main finding. This can be effective, but it commits the title to a claim, so make sure the evidence fully supports it.
• Question: poses the research question. Occasionally useful, but often less informative than a descriptive title.
• Two-part: a topic followed by a colon and a subtitle naming the design or setting, such as “…: a cross-sectional study.” The ICMJE notes that reporting guidelines recommend, and some journals require, study design information in the title, particularly for randomized trials and systematic reviews and meta-analyses.

Avoiding hype

Words such as novel, first, breakthrough, and paradigm-shifting rarely add information, and they invite skepticism from editors and reviewers. Also avoid unexplained abbreviations, filler openings such as “A study of…” or “Investigations into…”, and wordplay that hides the topic from search tools. Check your target journal’s title length limit and whether it also wants a short running title.

Structured and unstructured abstracts

A structured abstract uses labeled headings, typically Background or Objective, Methods, Results, and Conclusions, mirroring the IMRaD structure of the paper. Many journals in medicine and the health sciences require one, and the ICMJE states that original research, systematic reviews, and meta-analyses require structured abstracts. An unstructured abstract is a single paragraph without headings. It is common in the humanities, many social sciences, and other fields, but it still needs to cover the same ground in a logical order.

Word limits vary between journals, and between article types in the same journal, so check the instructions for authors rather than reusing an abstract written for another journal. Journals may also specify the headings to use, ask for a trial registration number at the end of the abstract, or restrict citations and abbreviations within it.

What each part of the abstract should state

Whatever the format, an abstract for an empirical study should let a reader answer these questions without opening the paper:

• Background or objective: What gap or question does the study address, and what was its aim? One or two sentences are usually enough.
• Methods: What was the design? Who or what was studied, in what setting and period? How were the main outcomes measured and analyzed?
• Results: What were the main findings, with the key numbers your field expects, such as effect sizes and confidence intervals, rather than only a statement that results were significant?
• Conclusions: What do the findings mean, stated no more strongly than the results allow, and what is the main implication or limitation?

Reviews, theoretical papers, and humanities articles follow a different logic, but the principle is the same: state the question, the approach, the argument or findings, and the contribution.

Consistency with the paper

The abstract must agree with the full text. Every number should match the results section, every claim should appear in the paper, and the conclusion should match the discussion. The ICMJE asks authors and editors to make sure information is consistent in both places during revision and review. Inconsistencies often appear when the paper is revised and the abstract is not, so recheck the abstract after every round of changes, including after peer review.

Reporting guidelines for abstracts

For some study designs, reporting guidelines specify what an abstract should contain. The EQUATOR Network (external link) maintains a searchable library of reporting guidelines. Two widely used examples for abstracts:

• CONSORT for Abstracts (external link) covers the reporting of randomized trials in journal and conference abstracts.
• PRISMA 2020 for Abstracts (external link) covers systematic reviews; its checklist is contained in the main PRISMA 2020 statement.

Check whether your target journal requires a particular guideline. If one applies to your design, draft the abstract with its checklist in front of you rather than checking against it afterward.

Keywords and discoverability

Keywords help databases and journal systems classify your paper and help readers find it. Choose terms a researcher in your field would actually search for, including synonyms and alternative spellings that are not already in your title. Where your field has a controlled vocabulary, use it: in biomedicine, the National Library of Medicine’s Medical Subject Headings (MeSH) (external link) are used for indexing, cataloging, and searching health information. Some journals ask for keywords from their own list or classification scheme, and most limit how many you can supply.

Avoid keywords so broad they distinguish nothing, such as “health” or “education” on their own. Because many databases index the title, abstract, and keywords rather than the full text, make sure your study’s main terms appear naturally in the title and abstract, without repeating them artificially.

Plain-language summaries and graphical abstracts

Some journals ask for additional summaries alongside the abstract. A plain-language summary explains the study and its significance for non-specialist readers, such as patients, practitioners, or policymakers, without technical terms. A graphical abstract is a single image that summarizes the main finding or process. Each journal sets its own requirements for length, format, and image specifications, so follow its instructions for authors. The same accuracy rule applies to both: neither should claim more than the paper shows. If you are preparing summaries for wider audiences, research communication support covers this kind of writing.

A final check before submission

Titles and abstracts are short, but they are often the first part of a paper that editors, reviewers, and readers see, and sometimes the only part. Draft them early, revise them last, and check them alongside the rest of the pre-submission checklist before you submit.