Longitudinal data model selection

Authors:
Rahman Azari;Lexin Li;Chih-Ling Tsai
Affiliations:
Department of Statistics, University of California, Davis, CA 95616, USA;School of Medicine, University of California, Davis, CA 95616, USA;Graduate School of Management, University of California, Davis, CA 95616, USA and Guanghua School of Management, Peking University, Beijing 100871, P.R. China
Venue:
Computational Statistics & Data Analysis
Year:
2006

Citing 0
Cited 4

Bootstrap variants of the Akaike information criterion for mixed model selection

Computational Statistics & Data Analysis
Information methods for model selection in linear mixed effects models with application to HCV data

Computational Statistics & Data Analysis
Conditional Akaike information criterion for generalized linear mixed models

Computational Statistics & Data Analysis
The weighted average information criterion for multivariate regression model selection

Signal Processing

Quantified Score

Hi-index	0.03

Visualization

Abstract

In longitudinal data with correlated errors, we apply the likelihood and residual likelihood approaches to obtain the corrected Akaike information criterion (AICc) and the residual information criterion (RIC), respectively. Simulation studies show that AICc outperforms the Akaike information criterion (AIC) when the numbers of subjects and repeated observations are small, and RIC is superior to the Bayesian information criterion (BIC) when the signal-to-noise ratio is moderate to large. We illustrate the practical use of these selection criteria with an empirical example for modeling the serum cholesterol measured at six time occasions.