01 · Follow the citation
Find the original source.
Look for the study itself, its publication date and the organization behind the article. A page’s purpose matters: education, opinion and selling are different contexts. NCCIH encourages readers to ask who runs a health website, where its information comes from and whether it is current.
NCCIH: evaluating health information →02 · Identify the participants
Ask who was actually studied.
Research in cells or animals is not the same as a clinical study in people. For human studies, check the number and characteristics of participants and how long they were followed. These details affect how much the results can tell you about a particular health question.
NCCIH: questions about health stories →03 · Look for a comparison
What were the results compared with?
In a randomized, placebo-controlled trial, participants are assigned by chance to receive the intervention or a placebo. A placebo is an inactive comparison. Expectations can affect health outcomes, which is one reason a comparison group matters when interpreting a reported improvement.
NCCIH: placebo effects and trials →04 · Keep the limits in view
Look beyond the most exciting result.
Read the limitations, funding disclosures and size of the reported benefit. A difference that interests researchers may have little practical importance. One study rarely settles a health question; consider how it fits with other studies rather than treating it as a final answer.
NCCIH: understanding the full story →