A statistic showing how consistent repeated measurements are for the same subject.
ICC quantifies reliability by comparing variance between subjects to total variance across repeated measures, raters, or testing sessions; values range from 0 (no agreement) to 1 (perfect agreement), with >0.75 often cited as good and >0.90 as excellent, though thresholds vary by field and application. Several ICC forms exist (e.g., 2,1 vs 3,k), reflecting different assumptions about raters and single- vs average-measure use, so results are only comparable when the same model and confidence interval are reported.
This guide was auto-drafted and is pending editorial review.