Exact Distribution Theory Research Articles

The Offset Normal Shape Distribution for Dynamic Shape Analysis

This article deals with the statistical analysis of landmark data observed at different temporal instants. Statistical analysis of dynamic shapes is a problem with significant challenges due to the difficulty in providing a description of the shape changes over time, across subjects and over groups of subjects. There are several modeling strategies, which can be used for dynamic shape analysis. Here, we use the exact distribution theory for the shape of planar correlated Gaussian configurations and derive the induced offset-normal shape distribution. Various properties of this distribution are investigated, and some special cases discussed. This work is a natural progression of what has been proposed in Mardia and Dryden, Dryden and Mardia, Mardia and Walder, and Kume and Welling. Supplemental materials for this article are available online.

Robust inference with order constraint in microarray study

Gene classification can involve complex order-restricted inference. Examining gene expression pattern across groups with order-restriction makes standard statistical inference ineffective and thus, requires different methods. For this problem, Roy’s union-intersection principle has some merit. The M-estimator adjusting for outlier arrays in a microarray study produces a robust test statistic with distribution-insensitive clustering of genes. The Mestimator in conjunction with a union-intersection principle provides a nonstandard robust procedure. By exact permutation distribution theory, a conditionally distribution-free test based on the proposed test statistic generates corresponding p-values in a small sample size setup. We apply a false discovery rate (FDR) as a multiple testing procedure to p-values in simulated data and real microarray data. FDR procedure for proposed test statistics controls the FDR at all levels of α and π0 (the proportion of true null); however, the FDR procedure for test statistics based upon normal theory (ANOVA) fails to control FDR.

Joint distributions of numbers of runs of specified lengths on directed trees

In this paper, we consider exact joint distributions of numbers of success runs of specified lengths on a Markov directed tree under \(\ell \)-overlapping enumeration schemes. Methods for deriving the probability generating functions of the joint distribution are presented. We extend the exact distribution theory of runs based on sequences to runs based on directed trees. Some numerical results for the run statistics are given in order to illustrate the computational aspects and the feasibility of our theoretical results. The reliability systems and lifetime distributions are considered. Our general results are applied to reliability systems and lifetime distributions. Finally, we discuss parametric estimation problems based on the lifetimes of the systems.

Exact extreme value, product, and ratio distributions under non-standard assumptions

The exact distributions of many functions of random vectors are derived in the literature mainly for the case of a Gaussian vector distribution or under the assumption that the vector follows a spherical or an elliptically contoured distribution. Numerous standard statistical applications are given for these cases. Deriving analogous results, if the sample distribution comes from a large family of probability laws, needs to make use of new analytical tools from the area of exact distribution theory. The present paper provides the application of such tools suitable for deriving the exact cumulative distribution functions and density functions of extreme values, products, and ratios in \(l_{2,p}\)-symmetrically distributed populations. Accompanying simulation studies are presented in cases of power-exponentially distributed populations and for different sample sizes. As an application, well-known results on the increasing failure rate properties of extremes from Gaussian samples are extended to \(p\)-power exponential sample distributions.

A Note on Multiple Comparison Procedures for Analysis of Ranked Data

Ranked samples typically arise in sensory evaluation studies, in which a group of paneliests independently rank the same set of samples. In this setting, Christensen and colleagues (Journal of Food Science 71:S132-S143, 2006) have proposed a nonparametric procedure for multiple comparisons of the samples, based on the sums of the ranks of the samples, summed across the panelists' rankings. Christensen and colleagues utilized a simulation procedure in order to estimate critical differences between rank sums, at various alpha levels and combinations of numbers of panelists and samples. In this note, exact distribution theory relating to the rank sums is derived, using probability generating functions; and, critical values can be determined through computer algebra.

INCONSISTENT VAR REGRESSION WITH COMMON EXPLOSIVE ROOTS

Nielsen (Working paper, University of Oxford, 2009) shows that vector autoregression is inconsistent when there are common explosive roots with geometric multiplicity greater than unity. This paper discusses that result, provides a coexplosive system extension and an illustrative example that helps to explain the finding, gives a consistent instrumental variable procedure, and reports some simulations. Some exact limit distribution theory is derived and a useful new reverse martingale central limit theorem is proved.

Distributions of numbers of runs and scans on directed acyclic graphs with generation

In this paper, we introduce a class of a directed acyclic graph on the assumption that the collection of random variables indexed by the vertices has a Markov property. We present a flexible approach for the study of the exact distributions of runs and scans on the directed acyclic graph by extending the method of conditional probability generating functions. The results presented here provide a wide framework for developing the exact distribution theory of runs and scans on the graphical models. We also show that our theoretical results can easily be carried out through some computer algebra systems and give some numerical examples in order to demonstrate the feasibility of our theoretical results. As applications, two special reliability systems are considered, which are closely related to our general results. Finally, we address the parameter estimation in the distributions of runs and scans.

Distributions of simple patterns in some kinds of exchangeable sequences

In this article, we consider the distributions of simple patterns in some types of sequences of infinite exchangeable multi-state trials. The distributions on exchangeable multi-state trials are considered in terms of an extension of de Finetti's theorem. As an application of partially exchangeable sequences, distributions on a Markov exchangeable sequence are studied. Furthermore, we propose a new type of partially exchangeable sequence and examine its properties. In addition, we discuss the distribution theory in the case of the finite exchangeable sequences. The results presented here provide a wide framework for developing the exact distribution theory of simple patterns. Finally, some examples are given in order to illustrate our theoretical results.

Inconsistent VAR Regression with Common Explosive Roots

Nielsen (2009) shows that vector autoregression is inconsistent when there are common explosive roots with geometric multiplicity greater than unity. This paper discusses that result, provides a co-explosive system extension and an illustrative example that helps to explain the finding, gives a consistent instrumental variable procedure, and reports some simulations. Some exact limit distribution theory is derived and a useful new reverse martingale central limit theorem is proved.

Two New Zealand pioneer econometricians

Two distinguished New Zealanders pioneered some of the foundations of modern econometrics. Alec Aitken, one of the most famous and well-documented mental arithmeticians of all time, contributed the matrix formulation and projection geometry of linear regression, generalized least squares (GLS) estimation, algorithms for Hodrick Prescott (HP) style data smoothing (six decades before their use in economics), and statistical estimation theory leading to the Cramér Rao bound. Rex Bergstrom constructed and estimated by limited information maximum likelihood (LIML) the largest empirical structural model in the early 1950s, opened up the field of exact distribution theory, developed cyclical growth models in economic theory, and spent nearly 40 years of his life developing the theory of continuous time econometric modeling and its empirical application. We provide an overview of their lives, discuss some of their accomplishments, and develop some new econometric theory that connects with their foundational work.

Two New Zealand Pioneer Econometricians

Two distinguished New Zealanders pioneered some of the foundations of modern econometrics. Alec Aitken, one of the most famous and well-documented mental arithmeticians of all time, contributed the matrix formulation and projection geometry of linear regression, generalized least squares (GLS) estimation, algorithms for Hodrick Prescott (HP) style data smoothing (six decades before their use in economics), and statistical estimation theory leading to the Cramer Rao bound. Rex Bergstrom constructed and estimated by limited information maximum likelihood (LIML) the largest empirical structural model in the early 1950s, opened up the field of exact distribution theory, developed cyclical growth models in economic theory, and spent nearly 40 years of his life developing the theory of continuous time econometric modeling and its empirical application. We provide an overview of their lives, discuss some of their accomplishments, and develop some new econometric theory that connects with their foundational work.

EXACT DISTRIBUTION THEORY IN STRUCTURAL ESTIMATION WITH AN IDENTITY

Some exact distribution theory is developed for structural equation models with and without identities. The theory includes LIML, IV, and OLS. We relate the new results to earlier studies in the literature, including the pioneering work of Bergstrom (1962). General IV exact distribution formulas for a structural equation model without an identity are shown to apply also to models with an identity by specializing along a certain asymptotic parameter sequence. Some of the new exact results are obtained by means of a uniform asymptotic expansion. An interesting consequence of the new theory is that the uniform asymptotic approximation provides the exact distribution of the OLS estimator in the model considered by Bergstrom (1962). This example appears to be the first instance in the statistical literature of a uniform approximation delivering an exact expression for a probability density.

Assessing the magnitude of the concentration parameter in a simultaneous equations model

Poskitt and Skeels (2003) provide a new approximation to the sampling distribution of the IV estimator in a simultaneous equations model. This approximation is appropriate when the concentration parameter associated with the reduced form model is small and a basic purpose of this paper is to provide the practitioner with a method of ascertaining when the concentration parameter is small, and hence when the use of the Poskitt and Skeels (2003) approximation is appropriate. Existing procedures tend to focus on the notion of correlation and hypothesis testing. Approaching the problem from a different perspective leads us to advocate a different statistic for use in this problem. We provide exact and approximate distribution theory for the proposed statistic and show that it satisfies various optimality criteria not satisfied by some of its competitors. Rather than adopting a testing approach we suggest the use of p-values as a calibration device.

Distributions of Runs and Scans on Higher-Order Markov Trees

In this article, we consider the distributions of the number of success runs of specified length and scans on a higher-order Markov tree under three different enumeration schemes (the “non overlapping”, the “at least”, and the “overlapping” scheme). Recursive formulae for the evaluation of their probability generating functions are established. We provide a proper framework for extending the exact distribution theory of runs and scans from based on sequences to based on directed trees. Some numerical results for the run and scan statistics are given in order to illustrate the computational aspects and the feasibility of our theoretical results. Finally, two special reliability systems are considered, which are closely related to our general results.

Exact distribution theory for belief net responses

Bayesian belief networks provide estimates of conditional probabilities, called query responses. The precision of these estimates is assessed using posterior distributions. This paper discusses two claims and a conjecture by Kleiter (1996) concerning the exact posterior distribution of queries. The two claims provide conditions where a query has an exact beta distribution. The first claim is clarified by the following generalization. Assuming a BDe prior and complete data, a query has the same distribution under equivalent network structures. If the query can be represented as a network parameter under an equivalent structure, it must then have a beta distribution. Kleiter's second claim is contradicted by a counter-example. His conjecture, concerning finite mixtures of beta distributions, is also disproved.

Multiple comparisons of several homoscedastic multivariate populations

The limiting joint distribution of correlated Hotelling’s T 2 statistics associated with multiple comparisons with a control in multivariate one-way layout model is a multivariate central nonsingular chi-square distribution with one-factorial correlation matrix, which has the distribution function expressed in a closed form as an integral of a product of noncentral chi-square distribution functions with respect to a central chi-square density function. For pairwise comparisons, it is a multivariate central singular chi-square distribution whose distribution function is generally intricate. To overcome the complexity of the (exact or asymptotic) distribution theory of $$T^2_{\rm max}$$ -type statistics appeared in simultaneous confidence intervals of mean vectors, improved Bonferroni-type inequalities are applied to construct asymptotically conservative simultaneous confidence intervals for pairwise comparisons as well as comparisons with a control.

The Impact of the Bootstrap on Statistical Algorithms and Theory

Bootstrap ideas yield remarkably effective algorithms for realizing certain programs in statistics. These include the construction of (possibly simultaneous) confidences sets and tests in classical models for which exact or asymptotic distribution theory is intractable. Success of the bootstrap, in the sense of doing what is expected under a probability model for data, is not universal. Modifications to Efron's definition of the bootstrap are needed to make the idea work for modern procedures that are not classically regular.

The density of the sufficient statistics for a Gaussian AR(1) model in terms of generalized functions

This paper derives the exact joint distribution of the minimal sufficient statistics in the first-order AR(1) model with Gaussian errors and zero start-up value. The results are fundamental to an exact distribution theory for the statistics that are typically of interest in this model.

Distributions of Runs and Consecutive Systems on Directed Trees

In this paper we study exact distributions of runs on directed trees. On the assumption that the collection of random variables indexed by the vertices of a directed tree has a directed Markov distribution, the exact distribution theory of runs is extended from based on random sequences to based on directed trees. The distribution of the number of success runs of a specified length on a directed tree along the direction is derived. A consecutive-k-out-of-n:F system on a directed tree is introduced and investigated. By assuming that the lifetimes of the components are independent and identically distributed, we give the exact distribution of the lifetime of the consecutive system. The results are not only theoretical but also suitable for computation.

Tests of Bivariate Symmetry with a One‐sided Alternative in the Analysis of Variance

AbstractWe consider a two‐factor experiment in which the factors have the same number of levels with a natural ordering among levels. We test the hypothesis that the effects of the two treatments are symmetric against a one‐sided alternative using the likelihood ratio criteria. Test of the one‐sided alternative as a null hypothesis against no restriction has also been studied. Exact distribution theory under the null hypothesis is derived and is shown to be a weighted mixture of chi‐square distributions. An example is used to illustrate the procedure.