Flexible Bent-Cable Models for Mixture Longitudinal Data

• Activity Number: 64
• Type: Topic Contributed
• Date/Time: Sunday, August 1, 2010 : 4:00 PM to 5:50 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Shahedul Ahsan Khan*+ and Grace Chiu and Joel A. Dubin
• Companies: University of Waterloo and CSIRO and University of Waterloo
• Keywords: Bayesian inference ; Bent-cable regression ; Changepoint data ; Gradual and abrupt transitions ; Longitudinal data

Abstract

Continuous changepoint data may exhibit 1 of 2 types of transitions: gradual or abrupt. Modeling the trend for such data is challenging in the presence of discontinuous derivatives. Further complications arise when we have (1) longitudinal data, and (2) samples which come from two potential populations: one with a gradual transition, and the other abrupt. Bent-cable regression is an appealing statistical tool to model such data due to the model's flexibility and greatly interpretable regression coefficients. We extend bent-cable methodology for longitudinal data to account for both gradual and abrupt transitions. We describe explicitly the computationally intensive Bayesian implementation; and demonstrate our methodology by a simulation study, and with 2 applications: (1) assessing the transition to early hypothermia in a rat model, and (2) understanding CFC-11 trends monitored globally.