A quick reference for reading and evaluating research literature. Written for master's students starting a thesis or research project, and for anyone newer to the process.
Not all studies answer a question with the same confidence. A case report can tell you something happened once. A randomized controlled trial can tell you an intervention probably caused an outcome, on average, in the people studied. Neither is “wrong”. They answer different kinds of questions, and a good literature review usually needs several types.
The standard shorthand for how much confidence a study design gives you, before you've looked at how well any one study was run.
Systematic reviews and meta-analyses
Best for describing the overall state of the evidence. Cite one when you mean "what do the studies add up to", not to back a single result.
Higher usually means stronger evidence for cause and effect. A well-run study lower down can still beat a poorly-run one higher up. Click a tier to see when to lean on it.
This ranks strength of evidence, not citation practice. For one finding, cite the original study it came from. Cite a systematic review or meta-analysis when you are describing the overall state of the evidence, not backing a single result.
What systematic reviews, trials, cohort studies, qualitative studies and more really are, and when each is the right tool.
Read moreWhat odds ratios, confidence intervals and p-values mean, and how to read AI summaries with a critical eye.
Read moreWhat kind of outcomes and measurements show up across more than 40 fields, from cardiology to computer science.
Read more