Glossary Term
Test-Retest Reliability
An OMERACT measurement property that assesses the stability and consistency of an instrument’s scores over time when no change in the construct being measured is expected to have occurred. It evaluates whether repeated measurements under similar conditions produce sufficiently similar scores.
Test-retest reliability should be assessed during a period when the target domain is expected to remain stable, while other potential sources of variability are kept as consistent as possible. Agreement between repeated scores provides information about the amount of measurement error or day-to-day variability associated with the instrument.
Relationship to Reliability: Reliability is the broader concept describing the consistency of measurements. Test-retest reliability specifically evaluates consistency over time in stable individuals, whereas inter-method reliability evaluates consistency across relevant sources of measurement variability, such as different raters, technicians, or machines.
Examples
A pain instrument is administered to patients on two occasions one week apart during a period when their pain is expected to remain stable. If patients receive similar scores at both assessments, this provides evidence of good test-retest reliability. The variability between the two measurements can also help determine how much change must occur before a change in score is likely to represent true change rather than normal measurement variability.