You do not need to settle an entire scientific debate before deciding whether a headline accurately describes its source. You need a clear record of the claim and the evidence attached to it.
This is our reading checklist, not a validated clinical scoring instrument. It produces a better description of a report—not a treatment recommendation or a numerical “truth score.”
1. What exactly is the claim?
Write one sentence identifying the intervention, population and outcome. Avoid copying promotional adjectives. “Researchers measured a change in a specified group” is a claim you can investigate. “A revolutionary breakthrough changes everything” is not sufficiently specific.
Keep the original headline separately. Comparing the two makes the article's interpretive additions easier to see.
2. What is the closest available source?
Find the study, regulator document or attributable announcement. Record whether you read a full paper, abstract or release. Access to a summary should not be described as independent review of every method and result.
Several news articles citing the same release are not several independent studies. Count the underlying evidence, not the number of pages repeating it.
3. What was actually studied?
Record whether the evidence concerns cells, an animal model or people. For human research, record the population and number of participants when available. A broad disease label cannot replace the actual eligibility criteria.
4. What was the comparison?
Identify the comparison group or explain that none was supplied. Our comparator guide shows why improvement after an intervention and improvement caused by it are not identical claims.
5. What outcome changed, and when?
Write the measured endpoint and follow-up. Include counts and denominators where available. Do not substitute “cured” for an outcome that has a narrower definition.
Our absolute-benefit guide explains why a percentage without its baseline can be difficult to interpret.
6. What important information is missing?
Record harms, uncertainty and limits on applicability. “Not reported in the source reviewed” is an answer. It is not the same as “none occurred.”
A missing item is a reason to investigate or narrow the conclusion, not automatically a reason to declare the research invalid.
7. What would change the answer?
Name the next informative evidence: results from a stated study, additional follow-up, a comparison or a regulator decision. Do not replace this with an unsupported date when a treatment will become available.
The availability guide explains how to keep expected and observed milestones separate.
Save the record, not just the link
Use the free evidence-note tool to create a plain-text record on your own device. Do not enter medical records or personal health details. The tool organizes your notes; it does not verify a source or decide whether its claim is true.
You can also use the printable checklist on our resources page. The objective is modest and useful: make the claim small enough to check, preserve the source and leave uncertainty visible.