A team titled their paper “When Ink Speaks”. It reported how handwritten feedback affected first-year students’ motivation to learn English. Three years later it had two citations, both self-citations. Researchers working on the same problem typed “written feedback motivation EFL” into their databases and never saw it.
The title and keywords are the only parts of your paper that are guaranteed to be read, and they carry the most weight for search engines. Their job is not to impress a committee; it is to let the person who needs your result find it.
How readers actually find papers
Most readers reach an article in one of three ways: a keyword search in Google Scholar or a database such as Scopus, Web of Science or PubMed; following a reference list; or seeing it shared. The second and third routes depend on someone having found you through the first.
Academic search tools typically give the most weight to words in the title, then the abstract and keywords. An important term that only appears on page eight is effectively invisible.
The four parts of a findable title
- What: the main variable, phenomenon or intervention.
- Who: the population.
- Where: the setting, when it matters for interpretation.
- How: the design. Reporting guidelines such as CONSORT and PRISMA recommend naming the design in the title.
A rewrite of the opening example: “Handwritten Feedback and First-Year Students’ Motivation to Learn English: A Controlled Trial at a Vietnamese University”. Longer, yes, but every phrase is something a potential reader might type.
Descriptive, declarative and question titles
| Type | Example | When it works |
|---|---|---|
| Descriptive | Sleep Duration and Exam Performance in Medical Students | A safe default for most papers |
| Declarative | Sleeping Under Six Hours Lowers Medical Students’ Exam Scores | Clear results from a strong design; favoured by some biomedical journals |
| Question | Does Short Sleep Lower Exam Scores? | Rarely; better for commentaries and reviews |
| Colon | Sleep and Exams: A Cohort Study of Medical Students | Very common; put keywords and design after the colon |
Declarative titles are appealing but risky with cross-sectional data. “Lowers” is a causal claim your design may not support, and reviewers will notice.
Titles that make papers disappear
- Puns, metaphors and quotations with no descriptive half. If you love the phrase, keep it before the colon and put the full description after.
- Uncommon abbreviations: nobody searches for your in-house acronym.
- Empty openers: “A study of”, “Some aspects of”, “An investigation into”. They use space and add nothing.
- Titles that are too broad: “Education and Technology” returns millions of results and buries you.
- Literal translation from another language, producing phrases that native speakers in the field never use.
Keywords: do not repeat the title
This is the most common mistake. If “motivation” and “written feedback” are already in the title, repeating them in the keyword field adds little. Use keywords for:
- Synonyms and variants: “feedback” and “corrective feedback”; “EFL” and “English as a foreign language”.
- One broader concept: “second language writing”.
- Setting or population if missing from the title.
- Controlled vocabulary such as MeSH in biomedicine or the ERIC thesaurus in education, where the journal asks for it.
A quick test: search each candidate keyword in Google Scholar. If the top ten results are papers you would like to sit beside, keep it. If they come from an unrelated field, the term is ambiguous.
The abstract is searchable too
The first two sentences of your abstract often appear directly in search results. Place key phrases there naturally, without stuffing. An abstract that repeats “learning motivation” six times irritates readers and gains nothing.
When to change the title mid-review
The title you submit is not fixed. After a major revision the emphasis often shifts: reviewers ask you to drop an analysis, or a secondary result turns out to be the main finding. In that case, propose a new title in your response letter and explain briefly why. Editors rarely object to a title that describes the paper more accurately. What you want to avoid is an old title promising something the final version no longer delivers, which frustrates readers who click through and quietly costs you citations.
In practice: a 30-minute title and keyword routine
- Write down five ways a colleague might search for your paper. Ask a real colleague too; they will phrase it differently.
- Run each phrase in Google Scholar and your field’s main database; note which ones return the closest papers.
- Draft three titles: descriptive, colon, and declarative if the design allows.
- Check each against what, who, where and how. Strike every word that carries no information.
- Check the journal’s length limits; many also ask for a short running title.
- Choose four to six keywords that do not duplicate the title, favouring synonyms and standard terms.
Next step: open the three most-cited papers in your exact niche, copy their titles and underline the searchable phrases. You will see the vocabulary your field uses to find each other, and which phrase your own title is still missing.
Câu hỏi thường gặp
How long should a research paper title be?
There is no single rule; many empirical papers use roughly 10 to 20 words. What matters more is that every word carries information and the title respects the journal’s limit.
Should keywords repeat words in the title?
They can, but it wastes space. Use the keyword field for synonyms, broader concepts and controlled vocabulary terms that are not already in the title.
Is it okay to use a question as a paper title?
It is usually a weak choice for empirical research because questions carry little searchable information. Question titles suit commentaries and reviews better.
How many keywords should a journal article have?
Most journals ask for three to six. Check the author guidelines and use the full allowance for terms with real search value.
How can I tell if my keywords are good?
Search each keyword in Google Scholar or your field’s database. If the top results are papers you would want to appear next to, the keyword works.