Testing for varying zero-inflation and dispersion in generalized Poisson regression models

Homogeneity of dispersion parameters and zero-inflation parameters is a standard assumption in zero-inflated generalized Poisson regression (ZIGPR) models. However, this assumption may be not appropriate in some situations. This work develops a score test for varying dispersion and/or zero-inflation parameter in the ZIGPR models, and corresponding test statistics are obtained. Two numerical examples are given to illustrate our methodology, and the properties of score test statistics are investigated through Monte Carlo simulations.     

Testing for departures from nominal dispersion in generalized nonlinear models with varying dispersion and/or additive random effects

This paper discusses the tests for departures from nominal dispersion in the framework of generalized nonlinear models with varying dispersion and/or additive random effects. We consider two classes of exponential family distributions. The first is discrete exponential family distributions, such as Poisson, binomial, and negative binomial distributions. The second is continuous exponential family distributions, such as normal, gamma, and inverse Gaussian distributions. Correspondingly, we develop a unifying approach and propose several tests for testing for departures from nominal dispersion in two classes of generalized nonlinear models. The score test statistics are constructed and expressed in simple, easy to use, matrix formulas, so that the tests can easily be implemented using existing statistical software. The properties of test statistics are investigated through Monte Carlo simulations.     

Asymptotic properties of the score test for varying dispersion in exponential family nonlinear models

Wei et al. [B.C. Wei, J.Q. Shi, W.K. Fung, and Y.Q. Hu, Testing for varying dispersion in exponential family nonlinear models, Ann. Inst. Statist. Math. 50 (1998), pp. 277–294.] developed the score diagnostics for varying dispersion in exponential family nonlinear models, such as the normal, inverse Gaussian, and gamma models, and investigated the powers of these tests through Monte Carlo simulations. In this paper, the asymptotic behaviours, including asymptotic chi-square and approximate powers under local alternatives of the score tests, are studied and examined by Monte Carlo simulations. The methods to estimate local powers of the score tests are illustrated with Grass yield data [P. McCullagh, and J.A. Nelder, Generalized Linear Models, Chapman and Hall, London (1989).].     

Score test for homogeneity of dispersion in generalized Poisson mixed models with excess zeros

In many applications, the clustered count data often contain excess zeros and the zero-inflated generalized Poisson mixed (ZIGPM) regression model may be suitable. However, dispersion in ZIGPM is often treated as fixed unknown parameter, and this assumption may be not appropriate in some situations. In this article, a score test for homogeneity of dispersion parameter in ZIGPM regression model is developed and corresponding test statistic is obtained. Sampling distribution and power of the score test statistic are investigated through Monte Carlo simulation. Finally, results from a biological example illustrate the usefulness of the diagnostic statistic.     

A semiparametric Bayesian approach to simplex regression model with heterogeneous dispersion

ABSTRACT

Simplex regression model is often employed to analyze continuous proportion data in many studies. In this paper, we extend the assumption of a constant dispersion parameter (homogeneity) to varying dispersion parameter (heterogeneity) in Simplex regression model, and present the B-spline to approximate the smoothing unknown function within the Bayesian framework. A hybrid algorithm combining the block Gibbs sampler and the Metropolis-Hastings algorithm is presented for sampling observations from the posterior distribution. The procedures for computing model comparison criteria such as conditional predictive ordinate statistic, deviance information criterion, and averaged mean squared error are presented. Also, we develop a computationally feasible Bayesian case-deletion influence measure based on the Kullback-Leibler divergence. Several simulation studies and a real example are employed to illustrate the proposed methodologies.     

Data Dispersion: Now You See It… Now You Don't

Poisson regression is the most well-known method for modeling count data. When data display over-dispersion, thereby violating the underlying equi-dispersion assumption of Poisson regression, the common solution is to use negative-binomial regression. We show, however, that count data that appear to be equi- or over-dispersed may actually stem from a mixture of populations with different dispersion levels. To detect and model such a mixture, we introduce a generalization of the Conway-Maxwell-Poisson (COM-Poisson) regression model that allows for group-level dispersion. We illustrate mixed dispersion effects and the proposed methodology via semi-authentic data.     

Analysis of Bivariate Survival Data using Shared Inverse Gaussian Frailty Model

In this article, we consider shared frailty model with inverse Gaussian distribution as frailty distribution and log-logistic distribution (LLD) as baseline distribution for bivariate survival times. We fit this model to three real-life bivariate survival data sets. The problem of analyzing and estimating parameters of shared inverse Gaussian frailty is the interest of this article and then compare the results with shared gamma frailty model under the same baseline for considered three data sets. Data are analyzed using Bayesian approach to the analysis of clustered survival data in which there is a dependence of failure time observations within the same group. The variance component estimation provides the estimated dispersion of the random effects. We carried out a test for frailty (or heterogeneity) using Bayes factor. Model comparison is made using information criteria and Bayes factor. We observed that the shared inverse Gaussian frailty model with LLD as baseline is the better fit for all three bivariate data sets.     

Bayesian shrinkage estimators for a measure of dispersion of an inverse gaussian distribution

The Bayesian shrinkage estimation for a measure of dispersion with known mean is studied for the inverse Gaussian distribution. An optimum choice of the shrinkage factor and the properties of the proposed Bayesian shrinkage estimators are being studied. It is shown that these estimators have smaller risk than the usual estimator of the reciprocal measure of dispersion.     

A score test for overdispersion in Poisson regression based on the generalized Poisson-2 model

Overdispersion is a common phenomenon in Poisson modeling. The generalized Poisson (GP) regression model accommodates both overdispersion and underdispersion in count data modeling, and is an increasingly popular platform for modeling overdispersed count data. The Poisson model is one of the special cases in the collection of models which may be specified by GP regression. Thus, we may derive a test of overdispersion which compares the equi-dispersion Poisson model within the context of the more general GP regression model. The score test has an advantage over the likelihood ratio test (LRT) and over the Wald test in that the score test only requires that the parameter of interest be estimated under the null hypothesis (the Poisson model). Herein, we propose a score test for overdispersion based on the GP model (specifically the GP-2 model) and compare the power of the test with the LRT and Wald tests. A simulation study indicates the proposed score test based on asymptotic standard normal distribution is more appropriate in practical applications.     

Variable selection for varying dispersion beta regression model

The beta regression models are commonly used by practitioners to model variables that assume values in the standard unit interval (0, 1). In this paper, we consider the issue of variable selection for beta regression models with varying dispersion (VBRM), in which both the mean and the dispersion depend upon predictor variables. Based on a penalized likelihood method, the consistency and the oracle property of the penalized estimators are established. Following the coordinate descent algorithm idea of generalized linear models, we develop new variable selection procedure for the VBRM, which can efficiently simultaneously estimate and select important variables in both mean model and dispersion model. Simulation studies and body fat data analysis are presented to illustrate the proposed methods.     

Poisson-mixed Inverse Gaussian Regression Model and Its Application

In this article, we have developed a Poisson-mixed inverse Gaussian (PMIG) distribution. The mixed inverse Gaussian distribution is a mixture of the inverse Gaussian distribution and its length-biased counterpart. A PMIG regression model is developed and the maximum likelihood estimation of the parameters is studied. A dataset dealing with the number of hospital stays among the elderly population is analyzed by using the PMIG and the PIG (Poisson-inverse Gaussian) regression models and it has been shown that the PMIG model fits the data better than the PIG model.     

Multivariate Dispersion Models Generated From Gaussian Copula

In this paper a class of multivariate dispersion models generated from the multivariate Gaussian copula is presented. Being a multivariate extension of Jørgensen's (1987a) dispersion models, this class of multivariate models is parametrized by marginal position, dispersion and dependence parameters, producing a large variety of multivariate discrete and continuous models including the multivariate normal as a special case. Properties of the multivariate distributions are investigated, some of which are similar to those of the multivariate normal distribution, which makes these models potentially useful for the analysis of correlated non-normal data in a way analogous to that of multivariate normal data. As an example, we illustrate an application of the models to the regression analysis of longitudinal data, and establish an asymptotic relationship between the likelihood equation and the generalized estimating equation of Liang & Zeger (1986).     

On testing inference in beta regressions

This article deals with testing inference in the class of beta regression models with varying dispersion. We focus on inference in small samples. We perform a numerical analysis in order to evaluate the sizes and powers of different tests. We consider the likelihood ratio test, two adjusted likelihood ratio tests proposed by Ferrari and Pinheiro [Improved likelihood inference in beta regression, J. Stat. Comput. Simul. 81 (2011), pp. 431–443], the score test, the Wald test and bootstrap versions of the likelihood ratio, score and Wald tests. We perform tests on the parameters that index the mean submodel and also on the parameters in the linear predictor of the precision submodel. Overall, the numerical evidence favours the bootstrap tests. It is also shown that the score test is considerably less size-distorted than the likelihood ratio and Wald tests. An application that uses real (not simulated) data is presented and discussed.     

Testing for linearity in generalized linear models using kernel smoothing

A test statistic proposed by Li (1999) for testing the adequacy of heteroscedastic nonlinear regression models using nonparametric kernel smoothers is applied to testing for linearity in generalized linear models. Simulation results for models with centered gamma and inverse Gaussian errors are presented to illustrate the performance of the resulting test compared with log-likelihood ratio tests for specific parametric alternatives. The test is applied to a data set of coronary heart disease status (Hosmer and Lemeshow, (1990).     

Statistical inference for the extended generalized inverse Gaussian model

This paper deals with the extended generalized inverse Gaussian (EGIG) distribution which has more than one turning point of the failure rate for certain values of the parameters. The EGIG model is a versatile model for analysing lifetime data and has one additional parameter, δ, than the GIG model's three parameters [B. Jorgensen, Statistical Properties of the Generalized Inverse Gaussian Distribution, Springer-Verlag, New York, 1982]. For the EGIG model, the maximum-likelihood estimation of the four parameters is discussed and a score test is developed for testing the importance of the additional parameter, δ. A non-central chi-square approximation to the power of the score test is provided. Simulation studies are carried out to examine the performance of the score test and the Wald confidence intervals. Finally, an example discussed by Jorgensen [5] is provided to illustrate that the EGIG model fits the data better than the GIG of Jorgensen [5]. Three other examples are presented and the power comparisons are displayed for each.     

Simultaneous statistical modelling of excess zeros,over/underdispersion,and multimodality with applications in hotel industry

We propose zero-inflated statistical models based on the generalized Hermite distribution for simultaneously modelling of excess zeros, over/underdispersion, and multimodality. These new models are parsimonious yet remarkably flexible allowing the covariates to be introduced directly through the mean, dispersion, and zero-inflated parameters. To accommodate the interval inequality constraint for the dispersion parameter, we present a new link function for the covariate-dependent dispersion regression model. We derive score tests for zero inflation in both covariate-free and covariate-dependent models. Both the score test and the likelihood-ratio test are conducted to examine the validity of zero inflation. The score test provides a useful tool when computing the likelihood-ratio statistic proves to be difficult. We analyse several hotel booking cancellation datasets extracted from two recently published real datasets from a resort hotel and a city hotel. These extracted cancellation datasets reveal complex features of excess zeros, over/underdispersion, and multimodality simultaneously making them difficult to analyse with existing approaches. The application of the proposed methods to the cancellation datasets illustrates the usefulness and flexibility of the models.     

A doubly restricted exponential dispersion model

In this paper, we propose and develop a doubly restricted exponential dispersion model, i.e. a varying dispersion generalized linear model with two sets of restrictions, a set of linear restrictions for the mean response, and at the same time, for another set of linear restrictions for the dispersion of the distribution. This model would be useful to consider several situations where it is necessary to control/analyze drug-doses, active effects in factorial experiments, mean-variance relationships, among other situations. A penalized likelihood function is proposed and developed in order to achieve the restricted parameters and to develop the inferential results. Several special cases from the literature are commented on. A simply restricted varying dispersion beta regression model is exemplified by means of real and simulated data. Satisfactory and promising results are found.     

Calibration for inverse gaussian regression

Inverse Gaussian regression models are useful for regression data where both variables are nonnegative and the variance of the dependent variable depends on the independent variable, Zero intercept inverse Gaussian regression models are presented with non-constant variance, constant ratio of variance to the mean and constant coefficient of variation, For purposes of calibration, the prediction band is used to give point and interval estimators for the independent variable, The results are illustrated with a real data set.     

Asymptotically refined score and GOF tests for inverse Gaussian models

ABSTRACT

The score test and the GOF test for the inverse Gaussian distribution, in particular the latter, are known to have large size distortion and hence unreliable power when referring to the asymptotic critical values. We show in this paper that with the appropriately bootstrapped critical values, these tests become second-order accurate, with size distortion being essentially eliminated and power more reliable. Two major generalizations of the score test are made: one is to allow the data to be right-censored, and the other is to allow the existence of covariate effects. A data mapping method is introduced for the bootstrap to be able to produce censored data that are conformable with the null model. Monte Carlo results clearly favour the proposed bootstrap tests. Real data illustrations are given.     

On Fisher’s dispersion test for integer-valued autoregressive Poisson models with applications

The integer-valued autoregressive (INAR) model has been widely used in diverse fields. Since the task of identifying the underlying distribution of time-series models is a crucial step for further inferences, we consider the goodness-of-fit test for the Poisson assumption on first-order INAR models. For a test, we employ Fisher’s dispersion test due to its simplicity and then derive its null limiting distribution. As an illustration, a simulation study and real data analysis are conducted for the counts of coal mining disasters, the monthly crime data set from New South Wales, and the annual numbers of worldwide earthquakes.