Online Program Home
My Program

Abstract Details

Activity Number: 614 - Statistical Methods for Longitudinal and Other Dependent Data
Type: Contributed
Date/Time: Thursday, August 1, 2019 : 8:30 AM to 10:20 AM
Sponsor: Section on Nonparametric Statistics
Abstract #304203 Presentation
Title: Modeling Longitudinal Data with Interval Censored Anchoring Events
Author(s): Chenghao Chu* and Ying Zhang and Wanzhu Tu
Companies: Vertex Pharmaceuticals and University of Nebraska Medical Center and Indiana University
Keywords: Empirical Process; Interval Censor; Longitudinal Data; nonparametrics

The time scale in many longitudinal data are often anchored by unobserved random events with unknown distribution. In many situations, the anchoring event can only be known to fall into an observable interval (i.e., interval censored). In this cases, analysts will not be able to use the traditional models to describe the temporal effect. In practice, ad hoc or strong assumptions on the anchoring events are usually assumed, which are unverifiable and prone to biased estimation and invalid inference. In this research, we proposed a two-stage method to model longitudinal data with interval censored anchoring events. We first obtain a robust nonparametric estimate of the anchoring events distribution, and then obtain the parameter estimates as stochastic functional of the estimated distribution. Algorithmically efficient algorithms for computing the parameter estimates were provided. For the purpose of large-sample statistical inference, we studied the asymptotic properties of the proposed functional estimator using empirical process theory. Finite sample performance of the proposed method was examined through simulation study. Applications were illustrated using real data analysis.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2019 program