Coefficient of accuracy and concordance correlation coefficient: new statistics for methods comparison.
Lawrence Lin, Lynn D. Torbeck
Abstract
Lawrence Lin, Lynn D. Torbeck
Abstract
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.
OpenAlex reports 107 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
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.
Key concepts: Concordance correlation coefficient, Statistic, Correlation coefficient, Statistics, Concordance, Cohen's kappa, Accuracy and precision, Correlation