Pooled Autoregressive Models for Categorical Data

• In Event: AERA Poster Session 4
  In Poster Session: Poster Session (San Diego)

Abstract

Time series capture time dependent intra-individual variation within a single participant. When data are collected from more than one subject, methods developed for single subject intra-individual relationship may not fully work and laws governing inter-individual relationship may not apply to intraindividual relationship, especially when outcomes are categorical or ordinal data. These data are usually collected by the Likert table. This article aims to investigate the performance of four estimation methods for pooling time series data focusing on categorical outcomes and to address related issues through an AR(1) model. In this article, models for pooling time series were formulated, estimation methods were derived, simulation studies were conducted, results were summarized and compared.

Authors

• Zhenqiu Laura Lu, University of Georgia

• Zhiyong Zhang