B1443
Title: A coefficient to measure agreement between two continuous variables based on a L1 norm
Authors: Ronny Vallejos - Universidad Tecnica Federico Santa Maria (Chile) [presenting]
Felipe Osorio - Universidad Tecnica Federico Santa Maria (Chile)
Clemente Ferrer - USM (Chile)
Abstract: The examination of agreement between two variables has gained significance over recent decades across various fields and is a frequent subject of study. An instance of this arises when two instruments measure experimental units within a study, prompting the interest in gauging the extent of agreement between these instruments. For data that is measured on a continuous scale, multiple approaches have been formulated to evaluate agreement among outcomes. A novel coefficient of concordance is introduced based on the L1 norm. The coefficient remains independent of additional parameters for estimation and is straightforward to compute. Explicit formulations for the coefficient are provided in the context of bivariate normal random vectors and elliptically contoured distributions. Furthermore, its relationship is explored with Lin's coefficient. The estimation procedure employs the delta method in cases where the sequence means are identical. To deepen the comprehension of the practical performance of the coefficient, numerical experiments are conducted.
