Multivariate models for correlated count data

Rodrigues-Motta, Mariana; Pinheiro, Hildete Prisco; Martins, Eduardo Gomes; Araújo, Márcio S.; Reis, Sérgio Furtado dos

doi:10.1080/02664763.2013.789098

Cited by 10 publications

(5 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A specific example of a GLMM for multivariate count data was presented by Rodrigues‐Motta et al . (). GLMMs are computationally demanding, and many different algorithms have been proposed in the past three decades; see McCulloch () and Fong et al .…”

Section: Introductionmentioning

confidence: 97%

Multivariate Covariance Generalized Linear Models

Bonat

Jørgensen

2016

Journal of the Royal Statistical Society Series C: Applied Statistics

View full text Add to dashboard Cite

Summary. We propose a general framework for non-normal multivariate data analysis called multivariate covariance generalized linear models (McGLMs), designed to handle multivariate response variables, along with a wide range of temporal and spatial correlation structures defined in terms of a covariance link function combined with a matrix linear predictor involving known matrices. The method is motivated by three data examples that are not easily handled by existing methods. The first example concerns multivariate count data, the second involves response variables of mixed types, combined with repeated measures and longitudinal structures, and the third involves a spatio-temporal analysis of rainfall data. The models take non-normality into account in the conventional way by means of a variance function, and the mean structure is modelled by means of a link function and a linear predictor. The models are fitted using an efficient Newton scoring algorithm based on quasi-likelihood and Pearson estimating functions, using only second-moment assumptions. This provides a unified approach to a wide variety of different types of response variables and covariance structures, including multivariate extensions of repeated measures, time series, longitudinal, spatial and spatio-temporal structures.

show abstract

Section: Introductionmentioning

confidence: 97%

Multivariate Covariance Generalized Linear Models

Bonat

Jørgensen

2016

Journal of the Royal Statistical Society Series C: Applied Statistics

View full text Add to dashboard Cite

show abstract

“…For instance, Ver Hoef and Jansen (2007) use both mixture and hurdle models for zero-inflated seal haul out analyses. More recent zero-inflation research focuses on the joint modelling of multiple outcomes (Diao et al 2013;Feng and Dean 2012;Hatfield et al 2012;Rodrigues-Motta et al 2013). For instance Feng and Dean (2012) use a latent random risk term to link the spatial component across outcomes and Dean and Lundy (pers.comm., Lundy's thesis dissertation, in preparation, 2015) model juvenile delinquent behaviour using a longitudinal joint model.…”

Section: Discussionmentioning

confidence: 99%

Advances and Challenges in Parametric and Semi-parametric Analysis for Correlated Data

Sutradhar¹

2016

Lecture Notes in Statistics

View full text Add to dashboard Cite

“…More broadly, Bonat and Jørgensen have developed a general modeling framework for regression models with multivariate response variables from any exponential family distribution [14], implemented in the R package "mcglm" [15]. An example of the joint modeling of discrete consumption data is a study conducted by Rodrigues-Motta et al, where the authors analyzed the distribution of insect types consumed by a Brazilian marsupial species based on fecal samples [16]. Their dataset was small (37 animals) and counts were generally very low with many zeroes, leading them to propose a zero-inflated Poisson model with Gaussian random effects terms that was estimated via maximum likelihood with numerical integration.…”

Section: Introductionmentioning

confidence: 99%

Multivariate analysis of food consumption profiles in crisis settings

2023

View full text Add to dashboard Cite

Preventing malnutrition is one of the primary objectives of many humanitarian agencies, and household surveys are regularly employed to monitor food insecurity caused by political, economic, or environmental crises. Consumption frequencies for standard food groups are often collected to characterize the depth of food insecurity in a community and measure the impact of food assistance programs, producing a vector of bounded, correlated counts for each household. While aggregate indicators are typically used to summarize these results with a single statistic, they can be difficult to interpret and provide insufficient detail to judge the effectiveness of assistance programs. To address these limitations, we have developed a multivariate modeling framework for consumption frequency data. We introduce methods to update baseline models for the analysis of the smaller and more variable surveys typically collected in crisis settings, and we present an application of our approach to national consumption data collected in Yemen in 2014 and 2016 by the World Food Programme. The approach provides more nuanced and interpretable information about consumption changes in response to shocks and the effectiveness of humanitarian assistance.

show abstract

Multivariate models for correlated count data

Cited by 10 publications

References 34 publications

Multivariate Covariance Generalized Linear Models

Multivariate Covariance Generalized Linear Models

Advances and Challenges in Parametric and Semi-parametric Analysis for Correlated Data

Multivariate analysis of food consumption profiles in crisis settings

Contact Info

Product

Resources

About