Logo image
A Bayesian approach to functional mixed-effects modeling for longitudinal data with binomial outcomes
Journal article   Peer reviewed

A Bayesian approach to functional mixed-effects modeling for longitudinal data with binomial outcomes

Stephanie Kliethermes and Jacob Oleson
Statistics in medicine, Vol.33(18), pp.3130-3146
08/15/2014
DOI: 10.1002/sim.6166
PMCID: PMC4107023
PMID: 24723495

View Online

Abstract

Longitudinal growth patterns are routinely seen in medical studies where individual growth and population growth are followed up over a period of time. Many current methods for modeling growth presuppose a parametric relationship between the outcome and time (e.g., linear and quadratic); however, these relationships may not accurately capture growth over time. Functional mixed-effects (FME) models provide flexibility in handling longitudinal data with nonparametric temporal trends. Although FME methods are well developed for continuous, normally distributed outcome measures, nonparametric methods for handling categorical outcomes are limited. We consider the situation with binomially distributed longitudinal outcomes. Although percent correct data can be modeled assuming normality, estimates outside the parameter space are possible, and thus, estimated curves can be unrealistic. We propose a binomial FME model using Bayesian methodology to account for growth curves with binomial (percentage) outcomes. The usefulness of our methods is demonstrated using a longitudinal study of speech perception outcomes from cochlear implant users where we successfully model both the population and individual growth trajectories. Simulation studies also advocate the usefulness of the binomial model particularly when outcomes occur near the boundary of the probability parameter space and in situations with a small number of trials.
Biostatistics Markov Chains Humans Linear Models Models, Statistical Algorithms Computer Simulation Bayes Theorem Statistics, Nonparametric Cochlear Implants Longitudinal Studies Monte Carlo Method Speech Perception

Details

Metrics

Logo image