A0429
Title: Overall marginalized models for longitudinal zero-inflated count data
Authors: Keunbaik Lee - Sungkyunkwan University (Korea, South) [presenting]
Abstract: To analyze longitudinal zero-inflated count data, the aim is to build on existing models by proposing marginalized zero-inflated Poisson (MZIP) models with random effects that explicitly capture the marginal effect of covariates, addressing the limitations found in previous methods. Additionally, a marginalized zero-inflated negative binomial (MZINB) model is presented to handle zero-inflated overdispersed data. Both models account for subject-specific heterogeneity in the random effects covariance structure. Simulation studies highlight the performance of the MZIP and MZINB models, comparing their inferences with homogeneous and heterogeneous random effects. Lastly, the proposed models are applied to analyze data on systemic lupus erythematosus.
