A BAYESIAN SIMULATION APPROACH TO INFERENCE ON A MULTI‐STATE LATENT FACTOR INTENSITY MODEL

The influence of economic conditions on the movement of a variable between states (for example a change in credit rating from A to B) can be modelled using a multi‐state latent factor intensity framework. Estimation of this type of model is, however, not straightforward, as transition probabilities are involved and the model contains a few highly analytically intractable distributions. In this paper, a Bayesian approach is adopted to manage the distributions. The innovation in the sampling algorithm used to obtain the posterior distributions of the model parameters includes a particle filter step and a Metropolis–Hastings step within a Gibbs sampler. The feasibility and accuracy of the proposed sampling algorithm is supported with a few simulated examples. The paper contains an application concerning what caused 1049 firms to change their credit ratings over a span of ten years.     

Nonlinear models for ground-level ozone forecasting

One of the main concerns in air pollution is excessive tropospheric ozone concentration. The aim of this work is to develop statistical models giving shortterm forecasts of future ground-level ozone concentrations. Since there are few physical insights about the dynamic relationship between ozone, precursor emissions and/or meteorological factors, a nonparametric and nonlinear approach seems promising in order to specify the forecast models. First, we apply four nonparametric procedures to forecast daily maximum 1-hour and maximum 8-hour averages of ozone concentrations in an urban area. Then, in order to improve the forecast performances, we combine the time series of the forecasts. This idea seems to give encouraging results. This work was supported by a MURST grant. The authors would like to thank two anonymous referees for their helpful comments.     

Maximum entropy sampling and optimal Bayesian experimental design

When Shannon entropy is used as a criterion in the optimal design of experiments, advantage can be taken of the classical identity representing the joint entropy of parameters and observations as the sum of the marginal entropy of the observations and the preposterior conditional entropy of the parameters. Following previous work in which this idea was used in spatial sampling, the method is applied to standard parameterized Bayesian optimal experimental design. Under suitable conditions, which include non-linear as well as linear regression models, it is shown in a few steps that maximizing the marginal entropy of the sample is equivalent to minimizing the preposterior entropy, the usual Bayesian criterion, thus avoiding the use of conditional distributions. It is shown using this marginal formulation that under normality assumptions every standard model which has a two-point prior distribution on the parameters gives an optimal design supported on a single point. Other results include a new asymptotic formula which applies as the error variance is large and bounds on support size.     

Bayesian methods for contingency tables using Gibbs sampling

Cell counts in contingency tables can be smoothed using loglinear models. Recently, sampling-based methods such as Markov chain Monte Carlo (MCMC) have been introduced, making it possible to sample from posterior distributions. The novelty of the approach presented here is that all conditional distributions can be specified directly, so that straight-forward Gibbs sampling is possible. Thus, the model is constructed in a way that makes burn-in and checking convergence a relatively minor issue. The emphasis of this paper is on smoothing cell counts in contingency tables, and not so much on estimation of regression parameters. Therefore, the prior distribution consists of two stages. We rely on a normal nonconjugate prior at the first stage, and a vague prior for hyperparameters at the second stage. The smoothed counts tend to compromise between the observed data and a log-linear model. The methods are demonstrated with a sparse data table taken from a multi-center clinical trial. The research for the first author was supported by Brain Pool program of the Korean Federation of Science and Technology Societies. The research for the second author was partially supported by KOSEF through Statistical Research Center for Complex Systems at Seoul National University.     

Distance sampling: estimating animal density

All too often we hear in the news statements like "there are, at most, a few hundred individuals left of this endangered species" or "there is small hope for the persistence of this population given that so few are left". How do scientists count animals to make such statements? Tiago Marques explains the concepts and pitfalls of distance sampling—one of the most widely used methods for estimating animal populations.     

Modelling two way contingency tables with recursive logits and odds ratios

In this work a new type of logits and odds ratios, which includes as special cases the continuation and the reverse-continuation logits and odds ratios, are defined. We prove that these logits and odds ratios define a parameterization of the joint probabilities of a two way contingency table. The problem of testing equality and inequality constraints on these logits and odds ratios is examined with particular regard to two new hypotheses of monotone dependence. Work partially supported by a MIUR2004 grant. Preliminary findings have been presented at SIS (Società Italiana di Statistica) Annual Meeting, Torino, 2006.     

The Moran coefficient for non-normal data

This paper summarizes findings that extend statistical distribution properties of the Moran coefficient index measuring spatial autocorrelation to non-normal random variables. Pitman–Koopmans theorem results are extended for the mean and the variance of this index. This summary includes a corollary to this theorem, as well as a new theorem (with its proof) and two conjectures implied by it. The first of these statements is supported by asymptotic heuristics; the second is supported by simulation experiment results. Mixture random variables that include heteroscedasticity or overdispersion also are explored. In addition, a simple asymptotic variance for the Moran coefficient is presented, assessed, and found to be very precise for sample sizes as small as 25–100. The principal conclusion is that independence and sample size are the most relevant properties for Pitman–Koopmans theorem results to be extended to non-normal random variables. The independent and identically distributed property reduces the necessary sample size for this extension, as do the properties of symmetry and normal approximation.     

Some remarks on an admissibility condition

Summary In this note we deal with some admissibility conditions proved by G. B. Tranquilli to be sufficient in the class of unbiased estimators of finite population parameters and with respect to (w.r.t.) a quadratic loss function. We show that the same conditions:i) are sufficient for the admissibility of an unbiased estimator with any loss function;ii) imply hyperadmissibility with reference to a particular (critical) population of the. From this fact we deduce that, for a fixed critical population, there is at most one estimator, in the class of all unbiased estimator of a finite population parameter, which satisfies Tranquilli condition. This research was partially supported by a M.U.R.S.T. grant ?Metodi inferenziali basati sul ricampionamento?.     

Tabulating weighted sup-norm functionals used in change-point analysis

We introduce a number of weighted partial sum processes of which certain sup-norm functionals may be used, for example, to detect changes in the mean of independent observations. Their limiting distributions are derived mathematically, simulated and then tabulated. With this information, a detailed numerical investigation of the power of these functionals is carried out.* ^*Research mainly done while at Carleton University, partially supported by an NSERC Canada Grant of Miklós Csörg? at Carleton University, Ottawa. Also supported by the Austrian Science Foundation (FWF) under grant SFB#010 (‘Adaptive Information Systems and Modelling in Economics and Management Science’) while at Technische Universität Wien. ? ^?Research partially supported by an NSERC Canada Grant of Miklós Csörg? at Carleton University, Ottawa.      

On the usability of the fluctuation test statistic to identify multiple cointegration break points

The fluctuation test suggested by Hansen and Johansen [Some tests for parameter constancy in cointegrated VAR models, Econometrics J. 2 (1999), pp. 306–333] intends to distinguish between the presence of zero and one break in cointegration relations. In this article, we provide evidence by Monte Carlo simulations that it also serves as a graphical device to detect even multiple break locations. It suffices to consider a simplified and easy-to-implement version of the original fluctuation test. Its break detection performance depends on the sign of change in cointegration parameters and the break height. The sign issue can be approached successfully by a backward application of the test statistic. If breaks are observable, the break locations are detected at the true location on average. We apply the graphical procedure to assess the cointegration of bond yields of Spain, Italy and Portugal with German yields for the period 1995–2013 which is surprisingly supported by the trace test. However, the recursive cointegration approach shows that a stable relationship with German yields is only present for sub-periods between the introduction of the Euro and the global financial crisis which is in line with expectations. The statistical robustness of these results is supported by a forward and backward application of the cointegration breakdown test by Andrews and Kim [Tests for cointegration breakdown over a short time period, J. Bus. Econom. Stat. 24 (2006), pp. 379–394].     

国有企业管理层在职消费的决定因素及经济后果

本文在明确定义在职消费经济性质的基础上,提出了一个有关在职消费的经验估计模型。借助于这一经验模型,本文对国有企业管理层在职消费的相对水平做出估计,并以此为基础对国有企业管理层在职消费的决定因素及其经济后果做出实证检验。研究表明：影响国有企业管理层在职消费动机的激励因素主要有管理层现金薪酬水平、管理层和员工之间的薪酬差距、管理层持股、管理层的平均年龄以及公司规模,在职消费的地域性激励以及自由现金流量假说仅得到有限证据的支持,最优雇佣契约理论则没有得到支持;此外,资本市场能够及时识别在职消费的事前非效率。同时,国有企业管理层在职消费还存在着事后的低效率。     

Analysis of a supersaturated design using Entropy Prior Complexity for binary responses via generalized linear models

A supersaturated design is a factorial design in which the number of effects to be estimated is greater than the available number of experimental runs. It is used in many experiments for screening purposes, i.e., for studying a large number of factors and then identifying the active ones. The goal with such a design is to identify just a few of the factors under consideration, that have dominant effects and to do this at minimum cost. While most of the literature on supersaturated designs has focused on the construction of designs and their optimality, the data analysis of such designs remains still at an early stage. In this paper, we incorporate the parameter model complexity into the supersaturated design analysis process, by assuming generalized linear models for a Bernoulli response, for analyzing main effects designs and discovering simultaneously the effects that are significant.     

Consistency of minimizing a penalized density power divergence estimator for mixing distribution

In this paper, we study the MDPDE (minimizing a density power divergence estimator), proposed by Basu et al. (Biometrika 85:549–559, 1998), for mixing distributions whose component densities are members of some known parametric family. As with the ordinary MDPDE, we also consider a penalized version of the estimator, and show that they are consistent in the sense of weak convergence. A simulation result is provided to illustrate the robustness. Finally, we apply the penalized method to analyzing the red blood cell SLC data presented in Roeder (J Am Stat Assoc 89:487–495, 1994). This research was supported (in part) by KOSEF through Statistical Research Center for Complex Systems at Seoul National University.     

George W. Snedecor,Pioneer Statistician, 1881–1974

Users of statistical packages need to be aware of the influence that outlying data points can have on their statistical analyses. Robust procedures provide formal methods to spot these outliers and reduce their influence. Although a few robust procedures are mentioned in this article, one is emphasized; it is motivated by maximum likelihood estimation to make it seem more natural. Use of this procedure in regression problems is considered in some detail, and an approximate error structure is stated for the robust estimates of the regression coefficients. A few examples are given. A suggestion of how these techniques should be implemented in practice is included.     

The Bayesian approach for highly reliable electro-explosive devices using one-shot device testing

Due to the high reliability and high testing cost of electro-explosive devices, even though an accelerated test is performed, one may observe very few failures or even no failures at all due to censoring. In this paper, we consider modelling the reliability of such devices by an exponential lifetime distribution in which the failure rate is assumed to be a function of some covariates and that the observed data are binary. The Bayesian approach, with three different prior settings, is used to develop inference on the failure rate, lifetime and the reliability under some settings. A Monte Carlo simulation study is carried out to show that this approach is quite useful and suitable for analysing data of the considered form, especially when the failure rates are very small. Finally, illustrative data are analysed using this approach.     

Share equations in econometrics: A story of repression,trustration and dead ends

Share equations play an important role in applied economic research, notably in marketing and demand analysis. Both market shares and budget shares have been used as dependent variables in econometric models which were partly motivated by microeconomic theory. However attempts of econometricians (and other statisticians) to treat share equations adequately led mostly to unsatisfactory approaches: Some researchers although admitting that shares satisfy a sum constraint simply repressed the fact that shares cannot be normally distributed. Some researchers looked in vain for a stochastic specification which at the same time is consistent and allows a flexible covariance structure. Last not least almost nobody has properly taken care of additional problems arising from dynamic share models. The paper discusses these three issues and proposes a possible way out of this dilemma which was first suggested by Aitchison (1982) and has been applied to econometric demand analysis by Considine and Mount (1984). Demandtheoretic implications as well as methods of estimation are discussed. An example using German import data illustrates some of the results. Research for this paper is financially supported by Deutsche Forschungsgemeinschaft, SFB 178 “Internationalisierung der Wirtschaft” at the University of Konstanz. Some results reported in this paper have been achieved jointly with Karl Ringwald whose permission for using unpublished joint work is gratefully acknowledged. Critical remarks by an anonymous referee helped to improve the exposition of the paper.     

Model-Based Estimation of Unemployment Rates in Small Areas of Portugal

The high level of unemployment is a major problem in most European countries nowadays. Hence, the demand for small area labor market statistics has rapidly increased over the past few years. The Portuguese Labour Force Survey is the main source of official statistics at the macro level. However, it was not designed to produce reliable design-based statistics at the micro level due to small sample sizes. The goal of this article is to analyze the performance of model-based small area estimators to estimate the unemployment rate at micro level. Our results showed that the temporal estimator is the most suitable.     

Change-of-variance problem for linear processes with long memory

In this paper we investigate the asymptotic properties of the test statistics for detecting change-points in the variance of infinite moving average sequences with long memory. This research is partly supported by NSFC Grants and SRF for ROCS, SEM.     

Empirical likelihood for the difference of quantiles under censorship

In this paper, we use a smoothed empirical likelihood method to investigate the difference of quantiles under censorship. An empirical log-likelihood ratio is derived and its asymptotic distribution is shown to be chi-squared. Approximate confidence regions based on this method are constructed. Simulation studies are used to compare the empirical likelihood and the normal approximation method in terms of its coverage accuracy. It is found that the empirical likelihood method provides a much better performance. The research is supported by NSFC (10231030) and RFDP.