Check a scientific claim before relying on it

What does this claim actually establish, and how well does it fit my decision?

Write the exact claim, trace it to its original evidence, and examine the method, comparison and limits. Keep the strength of the conclusion proportional to what was actually observed.

Evidence checked October 15, 2026. No hands-on testing.

Start with a checkable statement

A headline such as “scientists prove it works” hides the important questions. What was studied, compared with what, in whom, for how long, and with which outcome? Write one sentence that preserves those limits. Keep an estimate or an association from silently becoming a promise.

Separate an observation from its explanation

Understanding Science 101 describes science as testing ideas against evidence and revising explanations when new evidence appears. Your checking task follows that principle: identify the observation, the proposed explanation and what other explanation might fit. A confident statement and a recent publication date are not substitutes for that work.

Use a visible checking process

1. Save the claim and its context. Record the wording, publisher, URL and date. Note whether the statement came from a paper, press release, advertisement, commentary or social post.
2. Find the underlying source. Follow the citation to the original report or study. If you can only see a summary, mark that limit. Several articles repeating one press release are not several independent investigations.
3. Identify the method and comparison. What was measured, and how? Was there a relevant comparison group or a before-and-after observation? Look for how the groups or periods differed. Write down what the design can answer rather than attaching a universal quality score.
4. Check the people, setting and time. A result in one group, device or environment may not answer another question. For health research, MedlinePlus suggests checking whether participants were people or animals, who they were, study size, duration, type and funding.
5. Read the numbers in context. Find the denominator, absolute amounts, uncertainty and the outcome being counted. Do not assume a large relative percentage means a large absolute change.
6. Look for limitations and independent checking. Ask what evidence could disagree, whether the source reports uncertainty, and whether independent work addresses the same question. A review can help, but its relevance and methods still matter.
7. Choose a proportionate conclusion. Record supported, tentative or unresolved, with a reason. If the consequence is substantial or personal, get help suited to that decision instead of forcing the evidence to produce a yes/no answer.

Read a percentage without losing its denominator

Hypothetical arithmetic: 8 events per 100 in one group and 4 per 100 in another differ by 4 percentage points. Relative to 8, the reduction is 4 ÷ 8 = 50 percent. Both descriptions use the same numbers. Neither calculation tells you whether the groups were comparable, whether the difference is reliable, or whether the result applies to you.

A compact claim-checking record

Question	Record this	A reason to pause
Exact proposition	Population, intervention/exposure, comparison, outcome and period	The headline changes what the original source measured.
Evidence and method	Original source and relevant section/table	Only publicity or an inaccessible summary is available.
Numbers and uncertainty	Counts, denominator, estimate and stated limitations	Only a relative percentage is given.
Relevance and incentives	Who benefits, funding and fit to the decision	The source does not explain a material difference or conflict.

Decision map: decide what the evidence supports

Text alternative: First trace the evidence, then examine the method and fit. Use that examination to write a qualified conclusion rather than treating a checklist as a truth detector.

Use the resources at the right level

Berkeley provides a no-login conceptual overview. MedlinePlus gives practical questions for health-information websites, including identity, funding, dates and sources. Cochrane Evidence Essentials is a further learning option about trials and reviews, but its modules require a free account. This catalog reviewed the course description, not the completed lessons.

Avoid these common errors

• Treating the institution, publication date or number of search results as proof of a particular claim.
• Using a review from a different population or outcome without noting the mismatch.
• Reporting a hypothetical example as research data.
• Turning incomplete evidence into certainty because a decision feels urgent. For personal health choices, discuss relevant information with an appropriate clinician.

Your next useful action

Choose one claim and complete the four-row record above. End with a sentence beginning “This source supports…” and another beginning “It does not establish…”. If the original evidence is unavailable, preserve that gap.

Resources and their limitations

Sources and review notes

The resource evaluations above use the following original pages. Public websites were checked on October 15, 2026 in America/Denver (October 16 UTC). A successful capture does not establish every linked page or individual result. No hands-on testing occurred.

• Berkeley — Understanding Science 101
• MedlinePlus — Evaluating Health Information
• Cochrane Evidence Essentials

Related guides

Health literacy and evaluating advice; practical calculation; corroboration, disagreement and corrections.

Examples are teaching aids; no course completion, field navigation or practical intervention was tested.

Corrections: email director@victimsguide.com with the guide title, the point to correct and an original source. Please omit private medical, policy and identity records.