Abstract
A short summary of a publication, often describing the question, methods, findings and conclusions. It may not contain all the information in the full article.
Glossary
Research, explained.
Plain-language definitions of common research terms. Their meaning and limitations depend on the individual study.
These are general explanations, not assessments of particular publications.
Clinical trial
A study in which researchers examine an intervention, such as a medicine, procedure or support programme, in people.
Cohort study
A study that follows or compares groups of people to investigate how exposures and outcomes are associated.
Confidence interval
A range showing how uncertain a numerical estimate is. A wider interval generally means the estimate is less precise.
Correlation
When two things vary together. Correlation alone does not show that one causes the other.
Meta-analysis
A statistical method for combining results from several studies. Its reliability depends on the included evidence and how studies differ.
Observational study
A study in which researchers observe what happens without assigning the exposure or treatment being examined.
Peer review
Evaluation of a manuscript by other specialists before publication. It does not guarantee that every finding is correct.
Randomised controlled trial
A study in which participants are assigned by chance to comparison groups. This can help reduce certain types of bias.
Sample
The people, records or other observations included in a study. Who is included matters when interpreting who results apply to.
Systematic review
A review that uses a defined method to search for, select and assess studies addressing a specific question.
Statistical significance
A statement about how compatible data are with a statistical model or hypothesis. It does not necessarily mean a finding is important in everyday life.
For more detail about any result, read the study's own methods and limitations.