B1858
Title: Functional multiple indicators, multiple causes measurement error models
Authors: Carmen Tekwe - Indiana University - Bloomington (United States) [presenting]
Abstract: Energy expenditure is used by obesity researchers to approximate the amount of energy expended by the body to perform its routine functions. Since it is not directly observable, it can be viewed as a latent construct with multiple physical indirect measures such as respiratory quotient, volumetric oxygen consumption and volumetric carbon dioxide production. Metabolic rate assesses the body's ability to perform metabolic processes and is often approximated by heat production plus some error. Obesity development involves an imbalance between dietary energy intake and whole-body energy expenditure. The sparse functional multiple indicators are defined, as multiple cause measurement error (MIMIC ME) models by extending the linear MIMIC ME model to allow responses that are sparsely observed functional data. The mean curves are modeled using basis splines and functional principal components. A novel approach to identifying classical measurement error associated with approximating true metabolic rate by heat production based on functional principal components is also presented. The parameters are estimated using ME algorithm and a discussion of the model's identifiability is provided. The newly defined model is not a trivial extension of longitudinal or functional data methods due to the presence of the latent construct. Results from simulations and an application to study the relationship between metabolic rate and the multiple indicators of energy expenditure are provided.
