Evaluate the reliability and reproducibility of measurements.
Definition
The ICC measures consistency or agreement between multiple raters or repeated measurements. It is widely used in clinical research to validate inter-rater reliability or the reproducibility of a measurement instrument.
When to use it
Evaluate inter-rater reliability (agreement between judges)
Measure test-retest reproducibility of a tool
Validate a new measurement device
Requirements
Continuous variable measured by ≥ 2 raters or occasions
Choose ICC model according to design (1,1 / 2,1 / 3,1)
Absolute agreement or consistency depending on the objective
What StatsLab computes
ICC with 95% CI
One-Way, Two-Way Mixed, Two-Way Random models
Absolute agreement and consistency
Qualitative interpretation (poor → excellent)
Table of measurements per subject and rater
Worked example
Context : Two physiotherapists measure joint range of motion (°) in 20 patients.
Result : ICC(2,1) = 0.88, 95% CI [0.73; 0.95]
Interpretation : Excellent inter-rater reliability (ICC > 0.75). Both physiotherapists produce highly consistent measurements. The tool can be used interchangeably by both practitioners.