A study that watches what happens naturally, without the researcher assigning treatments.
In an observational study, researchers measure exposures and outcomes as they occur in real-world settings rather than randomising participants to an intervention. This makes it useful for studying rare outcomes, long-term effects, or situations where randomisation is impractical or unethical, but it is more vulnerable to confounding and bias, so causal claims should be made cautiously. Common subtypes include cohort, case-control, and cross-sectional designs, each with different strengths for inferring association versus causation.
This guide was auto-drafted and is pending editorial review.