Evidence-appraisal glossary
Imprecision
Imprecision describes how much random error clouds an effect estimate, usually shown by a wide confidence interval. When the interval is wide enough to span meaningfully different decisions, certainty in the evidence is rated down for imprecision.
Also called: random error, lack of precision.
An estimate can be unbiased yet still uninformative if too few patients or events sit behind it, leaving a confidence interval that stretches from helpful to harmful. In evidence appraisal, imprecision asks whether the interval crosses a decision threshold and whether the total sample meets the optimal information size needed to detect the effect of interest. It is distinct from a non-significant result: a precise interval tightly around no effect is informative, while a wide interval around a large effect is not.
This is a plain-language methodology definition for reading research. It is general education, not medical advice.