A mixed-effect model for positive responses augmented by zeros

Author(s)

Rodrigues-Motta, Mariana ^[1] ; Galvis Soto, Diana Milena ^[1] ; Lachos, Victor H. ^[1] ; Vilca, Filidor ^[1] ; Baltar, Valeria Troncoso ^[2] ; Verly Junior, Eliseu ^[3] ; Fisberg, Regina Mara ^[4] ; Lobo Marchioni, Dirce Maria ^[4]

Affiliation

^[1] Univ Estadual Campinas, Dept Stat, Sao Paulo - Brazil

^[2] Univ Fed Fluminense, Dept Epidemiol & Biostat, Inst Community Hlth, Rio De Janeiro - Brazil

^[3] Univ Estado Rio De Janeiro, Inst Social Med, Dept Epidemiol, Rio De Janeiro - Brazil

^[4] Univ Sao Paulo, Sch Publ Hlth, Dept Nutr, Sao Paulo - Brazil

• Document type: Journal article
• Source: STATISTICS IN MEDICINE; v. 34, n. 10, p. 1761-1778, MAY 10 2015.
• Web of Science Citations: 3

Abstract

In this research article, we propose a class of models for positive and zero responses by means of a zero-augmented mixed regression model. Under this class, we are particularly interested in studying positive responses whose distribution accommodates skewness. At the same time, responses can be zero, and therefore, we justify the use of a zero-augmented mixture model. We model the mean of the positive response in a logarithmic scale and the mixture probability in a logit scale, both as a function of fixed and random effects. Moreover, the random effects link the two random components through their joint distribution and incorporate within-subject correlation because of the repeated measurements and between-subject heterogeneity. A Markov chain Monte Carlo algorithm is tailored to obtain Bayesian posterior distributions of the unknown quantities of interest, and Bayesian case-deletion influence diagnostics based on the q-divergence measure is performed. We apply the proposed method to a dataset from a 24hour dietary recall study conducted in the city of SAo Paulo and present a simulation study to evaluate the performance of the proposed methods. Copyright (c) 2015John Wiley \& Sons, Ltd. (AU)

• FAPESP's process: 14/02211-1 - XXVII International Biometric Conference
• Grantee: Mariana Rodrigues Motta
• Support Opportunities: Research Grants - Meeting - Abroad
• FAPESP's process: 14/02938-9 - Estimation and diagnostics for censored mixed effects models using scale mixtures of skew-normal distributions
• Grantee: Víctor Hugo Lachos Dávila
• Support Opportunities: Regular Research Grants
• FAPESP's process: 11/10628-1 - Food intake, serum biomarkers of folate, vitamins B6 and B12, polymorphism of gene MTHFR and the effects on homocysteine levels in the ISA-Capital study: disentangling multiple relations with structural equation models
• Grantee: Valéria Troncoso Baltar
• Support Opportunities: Scholarships in Brazil - Post-Doctoral
• FAPESP's process: 09/15831-0 - Dietary factors, homocystein, MTHFR gene polymorphisms, and cardiovascular risk in adults and the elderly: a population-based study - ISA - Capital
• Grantee: Regina Mara Fisberg
• Support Opportunities: Regular Research Grants