A0821
Title: Estimating marginal association in clustered data with informative subgroups induced by a given covariate
Authors: Samuel Anyaso-Samuel - National Cancer Institute (United States) [presenting]
Abstract: Informative cluster size (ICS) typically introduces bias in cluster-correlated data analyses. A complex form of informativeness where the number of observations corresponding to latent levels of a unit-level continuous covariate is studied within a cluster is associated with the response variable. This type of informativeness has not been explored in prior research. A novel test statistic designed to evaluate the effect of the covariate while accounting for informativeness is presented. The covariate induces a continuum of latent subgroups within the clusters, and the test statistic is formulated by aggregating values from an established statistic that accounts for informative subgroup sizes when comparing group-specific marginal distributions. Through simulations, the test is compared with four traditional methods commonly employed in cluster-correlated data analyses. Only the test maintains the size across all data-generating scenarios with informativeness. The proposed method is illustrated to test for marginal associations in periodontal data with this distinctive form of informativeness.
