Abstract
The repeated measures concordance correlation coefficient was proposed for measuring agreement between two raters or two methods of measuring a response in the presence of repeated measurements (King et al., Citation2007). This paper proposes a class of repeated measures concordance correlation coefficients that are appropriate for both continuous and categorical data. We illustrate the methodology with examples comparing (1) 1-hour vs. 2-hour blood draws for measuring cortisol in an asthma clinical trial, (2) two measurements of percentage body fat, from skinfold calipers and dual energy x-ray absorptiometry, and (3) two binary measures of quality of health from an asthma clinical trial.
ACKNOWLEDGMENTS
The work of Tonya S. King and Vernon M. Chinchilli was partially supported via NIH grant 5U10HL051845. The authors wish to thank the Asthma Clinical Research Network (ACRN) for the use of the DICE and IMPACT data which provide excellent examples for this paper.
Notes
*Variance estimates are multiplied by 10 to better compare asymptotic versus empirical values.
*Variance estimates are multiplied by 10 to better compare asymptotic versus empirical values.