Online Program Home
My Program

Abstract Details

Activity Number: 421
Type: Topic Contributed
Date/Time: Tuesday, August 2, 2016 : 2:00 PM to 3:50 PM
Sponsor: Section on Nonparametric Statistics
Abstract #320811
Title: Estimation of Interpretable Growth Curves
Author(s): Jianhui Zhou* and Yin Zhang and Rashidul Haque and William A. Petri and Jennie Z. Ma
Companies: Healthy Birth, Growth and Development knowledge integration (HBGDki) Community and University of Virginia and International Centre for Diarrhoeal Diseas Research and University of Virginia and University of Virginia
Keywords: growth curve ; functional principal component analysis ; monotone ; penalization

During the critical period in child growth from early age to teenage years, it is important to distinguish normal and pathological patterns of growth curves, especially for children in developing countries. We study the height growth data from a cohort of preschool children in Dhaka, Bangladesh, who enrolled in the study during preschool years with varying follow-ups. We propose to estimate the growth curves using functional data analysis approaches. The main challenge of this study is that most subjects are not completely observed from 3 to 18 years old due to later enrolling or earlier dropping off, and the estimated growth curves from functional principal component analysis (fPCA) are not guaranteed to be monotone. Our goal is to obtain interpretable and meaningful growth curves that are monotonically increasing. Data transformation is used to guarantee monotonicity, and a penalized estimator is used for extrapolating growth curves beyond dropping off after fPCA estimates. The penalty function helps to borrow information from the cohort for subjects who had fewer observations. The developed method is applied to the growth curve modeling in a Bangladesh preschool cohort.

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

Back to the full JSM 2016 program

Copyright © American Statistical Association