PubMed · 1998 · 107 citations · 0 references
PharmacyBiostatisticsMethod ValidationMetrologyStatisticsComparative MeasurementConcordance AnalysisMeasurement PrecisionMethod ComparisonTest ReliabilityTest ValidityAccuracyAgreement StatisticsConcordance AssessmentCorrelation Of Test ResultsMethod Precision AssessmentPearson Correlation CoefficientStatistical Comparison Of Measurement ResultsTest Statistic
Lin introduced the Coefficient of Accuracy, C(a), to assess accuracy in methods comparison, addressing the limitation of the traditional Coefficient of Determination, r², which only captures precision. The study shows that combining C(a) with r² yields the Concordance Correlation Coefficient, rc, which simultaneously quantifies accuracy and precision.
A new statistic, the Coefficient of Accuracy, C(a), has been developed by Lin for methods comparison. When an old measurement method is compared to a new measurement method or if the same method is compared in two laboratories, the Coefficient of Determination, r2, is typically used to measure the relationship. However, r2 only measures the precision of the relationship. The newly developed statistic, C(a), measures the accuracy of the relationship. When these two statistics are combined together, they form a single statistic for both accuracy and precision called the Concordance Correlation Coefficient, rc.