RESEARCH GUIDE
A procedure for finding studies and checking records
This site helps locate studies and inspect bibliographic records. It does not turn a listing into an individual conclusion or a decision about taking medicine or supplements.
1. Split the question before searching
- Write one substance or subject.
- Write population, study design, and publication year as separate conditions.
- Keep dose, duration, outcome, and adverse event as separate search terms.
- These are search conditions, not actions to take.
Examples of search-question shapes
- Caffeine / population / study design / publication year
- Energy drinks / population / adverse event / source
- Artificial sweeteners / human / study design / publication year
- Sugar / animal / study design / source
- Supplement / population / dose / duration
- Medicine / population / outcome / source
2. Narrow the search results
- Start with a substance or subject in paper search.
- Check study design, species, publication year, and topic.
- Record the terms and result count whenever one condition changes.
- Open an individual record rather than deciding from a title alone.
3. Check an individual record in order
- Check study design and species.
- Check publication year, source, and whether the abstract is displayed.
- Check publication status, status-checked date, freshness, review stage, and record status.
- When a publication-status notice appears, open the displayed evidence link.
- Return to the source record and locate the information you need there.
4. Check provenance and freshness
- Source
- The bibliographic reference point for checking displayed information.
- Publication status
- One of active, corrected, retracted, expression of concern, updated, or unknown. It is not a summary of the paper.
- Status checked
- The date on which that status was recorded.
- Freshness
- One of current, due, stale, or unknown. It helps trace when the record was checked.
- Review stage and record status
- These show how the record is being handled, such as automated, pending review, or reviewed.
5. Questions to keep while reading
- Is this one study, or a record covering multiple studies?
- Is the subject human, animal, or not determined?
- Is it observational, a review, an undisplayed abstract, an older record, or an unconfirmed record?
- Which records remain when you change one condition in the same search question?
6. Open the update record
- Use Updates to check the data version, collection date, status-checked date, and build date.
- Keep the public paper/topic count separate from the annotated internal-ledger count.
- Check changes from the prior version, hashes, and the count of unconfirmed or failed records.
Decision boundary
Individual decisions about beginning, stopping, increasing, or combining medication or supplements are outside this guide. For an individual health decision, speak with an appropriately qualified professional.
The methodology describes the collection and classification procedure.