Follow your curiosity
How to find the original study behind a science headline
A news story leaves you with a question. A title, name or link can help you start finding the publication behind it and save it for another reading break.
A quick way to choose
- Start with the story's clues: link, title, authors, journal and date.
- Check the record you find against those clues before treating the search as complete.
- Keep the news link and the scientific publication link separately.
You are scrolling when a story says a study found something surprising. You want to know where it came from, but the link opens another news story telling the same account. Your first goal can be small: locate the publication behind that claim.
This guide is about that search. Once you have the right publication, you can continue with our guide to reading a research paper. Here, we will connect the story to its source, without expecting you to work through the entire investigation in one break.
1. Save the clues before opening more tabs
Copy the news link and the sentence that caught your attention. Look for an article title, surname, institution, journal name or date within the story. Record what appears as written: translating a title at the start may take you further from its original record.
Separate what you know from what you are assuming. The news date is not necessarily the study date, and a university name may belong to someone interviewed. These are useful search clues; you still need to check that they lead to the publication being cited.
2. Follow the link and identify where it leads
Open the link beside the claim. Have you reached a journal article, a university press release or another news story? If it is a press release, look there for the publication reference. You can record the trail in one line: news story, press release, scientific publication.
Avoid choosing a paper solely because it covers the same topic. Check whether the names and finding match what the story describes. If there is no reference, keep that gap as an open question and search using the clues you have.
3. Search the title or use the DOI
If you have the title, try Crossref Metadata Search, which searches bibliographic information. Crossref recommends it for matching references to DOIs. Compare the title, authors and journal in the result; matching words can belong to a different publication. The Crossref documentation describes these options.
A DOI identifies a publication. If you have one, such as 10.3390/healthcare12232488, put https://doi.org/ before it to open its registered destination. The DOI Foundation explains this resolution method. Copy the whole identifier without accidentally including a sentence's closing period. Finding a DOI identifies a publication; it does not assess the quality of its findings.
4. For a biomedical topic, try PubMed
You can paste the title into PubMed. It also accepts citation details such as an author, journal and year. Its guide to finding a specific citation explains this approach. You do not need to begin by learning special search syntax.
If the record does not appear, return to your clues: do you have the paper's title or the news headline? Try a researcher's name and the topic. A missing result in that search is not enough to conclude that the publication does not exist.
5. Practice with a record you can cross-check
Try our reading about how long habit formation takes. Its source is a review by Singh and colleagues, published in Healthcare in 2024. The PubMed record shows its citation.
The title and review label matter: if someone describes this source as a single new experiment, you would need to look again at which publication they mean. This exercise stops at identifying it; we are not proposing a habit-formation deadline or evaluating its findings. Use these details to identify it:
- Title: Time to Form a Habit: A Systematic Review and Meta-Analysis of Health Behaviour Habit Formation and Its Determinants.
- First author: Ben Singh.
- Publication: Healthcare, December 9, 2024.
- DOI: 10.3390/healthcare12232488.
- Listed publication type: review.
6. Distinguish the record from the full text
The PubMed help on full text explains links to publishers and PubMed Central when available. You can also check library access. The record and abstract may be accessible while the full article has different access conditions.
Record what you could open. Finding the reference, reading the abstract and reading the full text describe different stages. If you only have the abstract, leave questions that require methods, tables or details missing from it for later.
7. Check that it is the source of this claim
Return to the sentence in the story. Does the publication study the same people, animals or materials? Is the story about a review or an experiment? Is it describing a finding in the paper or an interviewee's interpretation? These questions guide your check; they do not produce an automatic trust score.
If the story draws on a review, that review may be the publication you needed. Its references let you follow the trail to included studies later. If several candidates remain and you cannot distinguish them, save the possibilities without presenting one as confirmed.
Save a question for your next break
Finish with a four-line note: what the story said, which publication you found, which details match and what still needs checking. Add both links. If you have not found a convincing reference, your note can say so; that is a useful outcome too.
PaperFlow offers another starting point: the open reading library collects short readings linked to their sources. Choose a question, open the reference and practice recognizing its record. Then return to the reading guide to explore the text. You can continue the route another day.
How we prepared this guide
An editorial guide by PaperFlow using Crossref, DOI Foundation and NLM documentation consulted on October 31, 2026, with AI assistance and editorial review. The example uses bibliographic details checked in PubMed. The steps and note format are editorial suggestions, without a promise to locate every source or certify a study's quality.