On the Ratio of Two Folded Normal Distributions

A class of distributions associated with the ratio of two folded normal random variables is introduced which strictly includes the half standard Cauchy distribution. The properties of this class of distributions are studied, along with a graph of the possible shapes of its density functions. The salient features of this class are mathematical tractability and statistical applicability. Utility of the class of distributions is demonstrated by presenting four applications.     

Conditional Generalized Liouville Distributions on the Simplex

Liouville and generalized Liouville distributions on the simplex have been proposed for modeling compositional data and have been shown to be free from the extreme independence structure that characterizes the Dirichlet class. In this article, generalized Liouville distributions are shown to be rich enough to distinguish some lesser modes of independence as well. Unfortunately, it is noted that the applicability of the Liouville family will be limited, owing to the lack of invariance with respect to the chosen fill-up value. As an alternative, a new family of simplex distributions is proposed, one that admits invariance with respect to choice of fill-up value, as well as the ability to differentiate among many forms of independence.     

Linear Conditional Expectation for Discretized Distributions

Many statistical methods for continuous distributions assume a linear conditional expectation. Components of multivariate distributions are often measured on a discrete ordinal scale based on a discretization of an underlying continuous latent variable. The results in this paper show that common examples of discretized bivariate and trivariate distributions will have a linear conditional expectation. Examples and simulations are provided to illustrate the results.     

Approximate Marginal Posterior Distributions Using Asymptotic Modes

This article gives asymptotic expansions for marginal posterior distributions with asymptotic modes of order n ^?2, and shows their validity. In addition, by using the asymptotic expansion, an approximate central posterior credible interval is derived.     

On the Characterization of a Mixture of Generalized Power Function Distributions by Conditional Expectation of Order Statistics

In the present article, we give some theorems to characterize the mixture of two generalized power function distributions based on conditional expectation of order statistics.     

Composite Marginal Likelihoods to the Normal Bradley-Terry Model

Inference in generalized linear mixed models with crossed random effects is often made cumbersome by the high-dimensional intractable integrals involved in the marginal likelihood. This article presents two inferential approaches based on the marginal composite likelihood for the normal Bradley-Terry model. The two approaches are illustrated by a simulation study to evaluate their performance. Thereafter, the asymptotic variances of the estimated variance component are compared.     

Bivariate Continuous Distributions with Specified Conditional Hazard Functions

Analysis of covariance (ANCOVA) is the standard procedure for comparing several treatments when the response variable depends on one or more covariates. We consider the problem of testing the equality of treatment effects when the variances are not assumed to be equal. It is well known that classical F test is not robust with respect to the assumption of equal variances and may lead to misleading conclusions if the variances are not equal. Ananda (1998 Ananda , M. M. A. ( 1998 ). Bayesian and non-Bayesian solutions to analysis of covariance models under heteroscedasticity . J. Econometrics 86 : 177 – 192 .[Crossref], [Web of Science ®] , [Google Scholar]) developed a generalized F test for testing the equality of treatment effects. However, simulation studies show that the actual size of this test can be much higher than the nominal level when the sample sizes are small, particularly when the number of treatments is large. In this article, we develop a test using the parametric bootstrap approach of Krishnamoorthy et al. (2007 Krishnamoorthy , K. , Lu , F. , Mathew , T. ( 2007 ). A parametric bootstrap approach for ANOVA with unequal variances: Fixed and random models . Computat. Statist. Data Anal. 51 : 5731 – 5742 .[Crossref], [Web of Science ®] , [Google Scholar]). Our simulations show that the actual size of our proposed test is close to the nominal level, irrespective of the number of treatments and sample sizes. Our simulations also indicate that our proposed PB test is more robust, with respect to the assumption of normality, than the generalized F test. Therefore, our proposed PB test provides a satisfactory alternative to the generalized F test.     

On the Performance of Sequential Regression Multiple Imputation Methods with Non Normal Error Distributions

Sequential regression multiple imputation has emerged as a popular approach for handling incomplete data with complex features. In this approach, imputations for each missing variable are produced based on a regression model using other variables as predictors in a cyclic manner. Normality assumption is frequently imposed for the error distributions in the conditional regression models for continuous variables, despite that it rarely holds in real scenarios. We use a simulation study to investigate the performance of several sequential regression imputation methods when the error distribution is flat or heavy tailed. The methods evaluated include the sequential normal imputation and its several extensions which adjust for non normal error terms. The results show that all methods perform well for estimating the marginal mean and proportion, as well as the regression coefficient when the error distribution is flat or moderately heavy tailed. When the error distribution is strongly heavy tailed, all methods retain their good performances for the mean and the adjusted methods have robust performances for the proportion; but all methods can have poor performances for the regression coefficient because they cannot accommodate the extreme values well. We caution against the mechanical use of sequential regression imputation without model checking and diagnostics.     

Discriminating Between the Von Mises and Wrapped Normal Distributions

In the analysis of circular data, we encounter two models which lay claim to the title of normal distribution: the von Mises and wrapped normal distributions. In this paper, we consider whether it is possible to distinguish between them in a practical situation. More specifically, given data from a circular distribution which may be either von Mises or wrapped normal, we enquire as to how large a sample is required in order to tell them apart.     

Likelihood Inference for Multivariate Extreme Value Distributions Whose Spectral Vectors have known Conditional Distributions

Multivariate extreme value statistical analysis is concerned with observations on several variables which are thought to possess some degree of tail dependence. The main approaches to inference for multivariate extremes consist in approximating either the distribution of block component‐wise maxima or the distribution of the exceedances over a high threshold. Although the expressions of the asymptotic density functions of these distributions may be characterized, they cannot be computed in general. In this paper, we study the case where the spectral random vector of the multivariate max‐stable distribution has known conditional distributions. The asymptotic density functions of the multivariate extreme value distributions may then be written through univariate integrals that are easily computed or simulated. The asymptotic properties of two likelihood estimators are presented, and the utility of the method is examined via simulation.     

On the Product of Inverse Wishart and Normal Distributions with Applications to Discriminant Analysis and Portfolio Theory

Abstract. In this article we analyse the product of the inverse Wishart matrix and a normal vector. We derive the explicit joint distribution of the components of the product. Furthermore, we suggest several exact tests of general linear hypothesis about the elements of the product. We illustrate the developed techniques on examples from discriminant analysis and from portfolio theory.     

Non-Normal Bivariate Distributions with Normal Marginals

The purpose of this note is to indicate that Fieller's Theorem can be expressed in the matrix formulation of the general linear model. The practical consequence is that one general computer program which can estimate the parameters and test the validity of a pertinent model, can also compute confidence limits for the ratios of any linear combinations of the parameters.     

A Nonparametric Comparison of Conditional Distributions with Nonnegligible Cure Fractions

Survival data with nonnegligible cure fractions are commonly encountered in clinical cancer clinical research. Recently, several authors (e.g. Kuk and Chen, Biometrika 79 (1992) 531; Maller and Zhou, Journal of Applied Probability, 30 (1993) 602; Peng and Dear, Biometrics, 56 (2000) 237; Sy and Taylor, Biometrics 56 (2000) 227) have proposed to use semiparametric cure models to analyze such data. Much of the existing work has been emphasized on cure detections and regression techniques. In contrast, this project focuses on the hypothesis testing in the presence of a cure fraction. Specifically, our interest lies in detecting whether there exists survival differences among noncured patients between treatment arms. For this purpose, we investigate the use of a modified Cramér-von Mises statistic for two-sample survival comparisons within the framework of cure models. Such a test has been studied by Tamura et al., (Statistics in Medicine 19, 2000, 2169) using bootstrap procedure. We will focus on developing asymptotic theory and convergent algorithms in this paper. We show that the limiting distributions of the Cramér-von Mises statistic under the null hypothesis can be represented by stochastic integrals and a weighted noncentral chi-squares. Both representations lead to concrete numerical schemes for computing the limiting distributions. The algorithms can be easily implemented for data analysis and significantly reduce computing time compared to the bootstrap approach. For illustrative purposes, we apply the proposed test to a published clinical trial.     

Normal Inverse Gaussian Distributions and Stochastic Volatility Modelling

The normal inverse Gaussian distribution is defined as a variance-mean mixture of a normal distribution with the inverse Gaussian as the mixing distribution. The distribution determines an homogeneous Lévy process, and this process is representable through subordination of Brownian motion by the inverse Gaussian process. The canonical, Lévy type, decomposition of the process is determined. As a preparation for developments in the latter part of the paper the connection of the normal inverse Gaussian distribution to the classes of generalized hyperbolic and inverse Gaussian distributions is briefly reviewed. Then a discussion is begun of the potential of the normal inverse Gaussian distribution and Lévy process for modelling and analysing statistical data, with particular reference to extensive sets of observations from turbulence and from finance. These areas of application imply a need for extending the inverse Gaussian Lévy process so as to accommodate certain, frequently observed, temporal dependence structures. Some extensions, of the stochastic volatility type, are constructed via an observation-driven approach to state space modelling. At the end of the paper generalizations to multivariate settings are indicated.     

Estimation of a Copula when a Covariate Affects only Marginal Distributions

This paper is concerned with studying the dependence structure between two random variables Y₁ and Y₂ in the presence of a covariate X, which affects both marginal distributions but not the dependence structure. This is reflected in the property that the conditional copula of Y₁ and Y₂ given X, does not depend on the value of X. This latter independence often appears as a simplifying assumption in pair‐copula constructions. We introduce a general estimator for the copula in this specific setting and establish its consistency. Moreover, we consider some special cases, such as parametric or nonparametric location‐scale models for the effect of the covariate X on the marginals of Y₁ and Y₂ and show that in these cases, weak convergence of the estimator, at ‐rate, holds. The theoretical results are illustrated by simulations and a real data example.     

Improved Tolerance Factors for Multivariate Normal Distributions

The Q_os and Qm are two leading estimators of the probability of misclassification which are based on the asymptotic expansion of the the expected value of the Error Rate, P_i. The estimators are, however, not suitable for estimating the Error rates for certain ranges of the parameters p , n₁, n₂ and ß.We investigate the regions in which they produce unacceptable estimates , and show that the Q_os is, in general, better than the Q_m in producing acceptable estimates     

On the Distributions of Certain Spacings

A characterization of the uniform distribution based on distributions of spacings is presented which extends the existing result in this direction. Also, a result on the distribution of spacings for distributions close to the uniform one is discussed.     

On Conditional and Partial Correlation

A graphical technique, similar in spirit to probability plotting, can be used to judge whether a Poisson model is appropriate for an observed frequency distribution. This “Poissonness plot” can equally be applied to truncated Poisson situations. It provides a type of robustness for detecting isolated discrepancies in otherwise well-behaved frequency distributions.     

A Unified Method for Checking Compatibility and Uniqueness for Finite Discrete Conditional Distributions

Checking compatibility for two given conditional distributions and identifying the corresponding unique compatible marginal distributions are important problems in mathematical statistics, especially in Bayesian inferences. In this article, we develop a unified method to check the compatibility and uniqueness for two finite discrete conditional distributions. By formulating the compatibility problem into a system of linear equations subject to constraints, it can be reduced to a quadratic optimization problem with box constraints. We also extend the proposed method from two-dimensional cases to higher-dimensional cases. Finally, we show that our method can be easily applied to checking compatibility and uniqueness for a regression function and a conditional distribution. Several numerical examples are used to illustrate the proposed method. Some comparisons with existing methods are also presented.