Empirical phi-divergence test statistics for testing simple and composite null hypotheses

The main purpose of this paper is to introduce first a new family of empirical test statistics for testing a simple null hypothesis when the vector of parameters of interest is defined through a specific set of unbiased estimating functions. This family of test statistics is based on a distance between two probability vectors, with the first probability vector obtained by maximizing the empirical likelihood (EL) on the vector of parameters, and the second vector defined from the fixed vector of parameters under the simple null hypothesis. The distance considered for this purpose is the phi-divergence measure. The asymptotic distribution is then derived for this family of test statistics. The proposed methodology is illustrated through the well-known data of Newcomb's measurements on the passage time for light. A simulation study is carried out to compare its performance with that of the EL ratio test when confidence intervals are constructed based on the respective statistics for small sample sizes. The results suggest that the ‘empirical modified likelihood ratio test statistic’ provides a competitive alternative to the EL ratio test statistic, and is also more robust than the EL ratio test statistic in the presence of contamination in the data. Finally, we propose empirical phi-divergence test statistics for testing a composite null hypothesis and present some asymptotic as well as simulation results for evaluating the performance of these test procedures.     

A Generalization of Ahmad's Class of Mann–Whitney–Wilcoxon Statistics

In the spirit of the recent work of Ahmad (1996) this paper introduces another class of Mann–Whitney–Wilcoxon test statistics. The test statistic compares the r th and s th powers of the tail probabilities of the underlying probability distributions. The choice of r + s = 4 improves the Pitman efficiency for uniform, exponential, lognormal and normal distributions and keeps the same efficiency as the Mann–Whitney–Wilcoxon test for logistic and double exponential distributions. The two-sample test is modified for the one-sample problem with symmetric underlying distribution.     

Testing stationarity and trend stationarity against the unit root hypothesis

In this paper we propose a family of relativel simple nonparametrics tests for a unit root in a univariate time series. Almost all the tests proposed in the literature test the unit root hypothesis against the alternative that the time series involved is stationarity or trend stationary. In this paper we take the (trend) stationarity hypothesis as the null and the unit root hypothesis as the alternative. The order differnce with most of the tests proposed in the literature is that in all four cases the asymptotic null distribution is of a well-known type, namely standard Cauchy. In the first instance we propose four Cauchy tests of the stationarity hypothesis against the unit root hypothesis. Under H1 these four test statistics involved, divided by the sample size n, converge weakly to a non-central Cauchy distribution, to one, and to the product of two normal variates, respectively. Hence, the absolute values of these test statistics converge in probability to infinity 9at order n). The tests involved are therefore consistent against the unit root hypothesis. Moreover, the small sample performance of these test are compared by Monte Carlo simulations. Furthermore, we propose two additional Cauchy tests of the trend stationarity hypothesis against the alternative of a unit root with drift.     

Simulation on probability points for testing of lognormal or weibull distribution with a small sample

Two extensive computer simulated tables of percentage points of the asymptotic test statistics for testing lognormal or Weibull population proposed by Pereira (1978) are discussed. Special attention is given to small sample cases. Some of the most commonly used 16 symmetrical probability points are reported. These points are 0.001, 0.005, 0.01. 002. 0.025. 0.05. 0.10.0.15, 0.85, 0.90, 0.95, 0.975, 0.98, 0.99, 0.995 and 0.999. These empirical Sumulated results can be used to test hypotheses for these two particular populations and are adequate when using a normal approximation.     

On the joint distribution of the random variables ‘number of inversions’ and ‘number of outstanding variables’ in a randomly arranged sequence

The bivariate probability distribution of the random variables [number of inversions] and [number of outstanding variables] in a sequence of n i.i.d. random variables is derived. As an application, the null covariance between the test statistics proposed by Mann and Brunk, respectively, for the ‘trend in location’ problem is obtained. It is shown that these test statistics are asymptotically uncorrelated under the null hypothesis.     

Measurements of separation among probability densities or random variables

Distance between two probability densities or two random variables is a well established concept in statistics. The present paper considers generalizations of distances to separation measurements for three or more elements in a function space. Geometric intuition and examples from hypothesis testing suggest lower and upper bounds for such measurements in terms of pairwise distances; but also in L_p spaces some useful non-pairwise separation measurements always lie within these bounds. Examples of such separation measurements are the Bayes probability of correct classification among several arbitrary distributions, and the expected range among several random variables.     

Results of exploratory data analysis in the broken stick model

The broken stick model is a model of the abundance of species in a habitat, and it has been widely extended. In this paper, we present results from exploratory data analysis of this model. To obtain some of the statistics, we formulate the broken stick model as a probability distribution function based on the same model, and we provide an expression for the cumulative distribution function, which is needed to obtain the results from exploratory data analysis. The inequalities we present are useful in ecological studies that apply broken stick models. These results are also useful for testing the goodness of fit of the broken stick model as an alternative to the chi square test, which has often been the main test used. Therefore, these results may be used in several alternative and complementary ways for testing the goodness of fit of the broken stick model.     

Divergence-based estimation and testing with misclassified data

The well-known chi-squared goodness-of-fit test for a multinomial distribution is generally biased when the observations are subject to misclassification. In Pardo and Zografos (2000) the problem was considered using a double sampling scheme and ø-divergence test statistics. A new problem appears if the null hypothesis is not simple because it is necessary to give estimators for the unknown parameters. In this paper the minimum ø-divergence estimators are considered and some of their properties are established. The proposed ø-divergence test statistics are obtained by calculating ø-divergences between probability density functions and by replacing parameters by their minimum ø-divergence estimators in the derived expressions. Asymptotic distributions of the new test statistics are also obtained. The testing procedure is illustrated with an example.     

Asymptotic behaviour and statistical applications of divergence measures in multinomial populations: a unified study

Divergence measures play an important role in statistical theory, especially in large sample theories of estimation and testing. The underlying reason is that they are indices of statistical distance between probability distributions P and Q; the smaller these indices are the harder it is to discriminate between P and Q. Many divergence measures have been proposed since the publication of the paper of Kullback and Leibler (1951). Renyi (1961) gave the first generalization of Kullback-Leibler divergence, Jeffreys (1946) defined the J-divergences, Burbea and Rao (1982) introduced the R-divergences, Sharma and Mittal (1977) the (r,s)-divergences, Csiszar (1967) the ϕ-divergences, Taneja (1989) the generalized J-divergences and the generalized R-divergences and so on. In order to do a unified study of their statistical properties, here we propose a generalized divergence, called (h,ϕ)-divergence, which include as particular cases the above mentioned divergence measures. Under different assumptions, it is shown that the asymptotic distributions of the (h,ϕ)-divergence statistics are either normal or chi square. The chi square and the likelihood ratio test statistics are particular cases of the (h,ϕ)-divergence test statistics considered. From the previous results, asymptotic distributions of entropy statistics are derived too. Applications to testing statistical hypothesis in multinomial populations are given. The Pitman and Bahadur efficiencies of tests of goodness of fit and independence based on these statistics are obtained. To finish, apendices with the asymptotic variances of many well known divergence and entropy statistics are presented. The research in this paper was supported in part by DGICYT Grants N. PB91-0387 and N. PB91-0155. Their financial support is gratefully acknowledged.     

Diagnostic tests as residual analysis

Many applied workers are strongly oriented to residual analysis for assessing model adequacy. Formal test statistics of adequacy however are frequently derived from likelihood theory, particularly through Lagrange Multipliers. In contraGt, the present paper derives the formal statistics by concentrating Upon the distribution of residuals. It is shown that most existing tests can be derived in this way from a few elementary principles of specification analysis. One advantage of this alternative methodology is that it highlights some difficulties in existing approaches and simultaneously indicates a resolution of them; a good example being testing for heteroscedasticity in simultaneous equations. Other issues such as independence and robustness of diagnostic tests are also easily explored within the proposed framework.     

Comments on “bootstrapping time series models”

Many applied workers are strongly oriented to residual analysis for assessing model adequacy. Formal test statistics of adequacy however are frequently derived from likelihood theory, particularly through Lagrange Multipliers. In contraGt, the present paper derives the formal statistics by concentrating Upon the distribution of residuals. It is shown that most existing tests can be derived in this way from a few elementary principles of specification analysis. One advantage of this alternative methodology is that it highlights some difficulties in existing approaches and simultaneously indicates a resolution of them; a good example being testing for heteroscedasticity in simultaneous equations. Other issues such as independence and robustness of diagnostic tests are also easily explored within the proposed framework.     

Percentage points for directional anderson-darling goodness-of-fit tests

This paper discusses the problem of assessing the asymptotic distribution when parameters of the hypothesized distribution are estimated from a sample, pointing out a common mistake included in the paper by Sinclair, Spurr, and Ahmad (1990) which introduced two modifications of the Anderson-Darling goodness-of-fit test statistic. Their two test statistics modify the popular Anderson-Darling test statistic to be sensitive to departures of the fitted distribution from the true distribution in one or the other of the tails. This paper uses these new test statistics to develop tests of fit for the normal and exponential distributions. Easy to use formulas are given so the reader can perform these tests at any sample size without consulting exhaustive tables of percentage points. Finally a power study is given to demonstrate the test statistics’ viability against a broad range of alternatives.     

Two-way ANOVA when the distribution of the error terms is skew t

In this article, we assume that the distribution of the error terms is skew t in two-way analysis of variance (ANOVA). Skew t distribution is very flexible for modeling the symmetric and the skew datasets, since it reduces to the well-known normal, skew normal, and Student's t distributions. We obtain the estimators of the model parameters by using the maximum likelihood (ML) and the modified maximum likelihood (MML) methodologies. We also propose new test statistics based on these estimators for testing the equality of the treatment and the block means and also the interaction effect. The efficiencies of the ML and the MML estimators and the power values of the test statistics based on them are compared with the corresponding normal theory results via Monte Carlo simulation study. Simulation results show that the proposed methodologies are more preferable. We also show that the test statistics based on the ML estimators are more powerful than the test statistics based on the MML estimators as expected. However, power values of the test statistics based on the MML estimators are very close to the corresponding test statistics based on the ML estimators. At the end of the study, a real life example is given to show the implementation of the proposed methodologies.     

Minimum Scoring Rule Inference

Proper scoring rules are devices for encouraging honest assessment of probability distributions. Just like log‐likelihood, which is a special case, a proper scoring rule can be applied to supply an unbiased estimating equation for any statistical model, and the theory of such equations can be applied to understand the properties of the associated estimator. In this paper, we discuss some novel applications of scoring rules to parametric inference. In particular, we focus on scoring rule test statistics, and we propose suitable adjustments to allow reference to the usual asymptotic chi‐squared distribution. We further explore robustness and interval estimation properties, by both theory and simulations.     

Testing hypotheses about covariance matrices using bootstrap methods

The usual chi-squared approximation to test statistics based on normal theory for testing covariance structures of multivariate populations is very sensitive to the normality assumption. Two general bootstrap procedures are developed in this paper to obtain approximately valid critical values for these test statistics when the data are not normally distributed. The first is based on separate sampling from individual samples, and the second is based on sampling from pooled samples. Although the second method requires more assumptions, its small sample properties are better.     

Power studies of tests for uniformity,II

Results from a power study of six statistics for testing that a sample is from a uniform distribution on the unit interval (0,1) are reported. The test statistics are all well-known and each of them was originally proposed because they should have high power against some alternative distributions. The tests considered are the Pearson probability product test, the Neyman smooth test, the Sukhatme test, the Durbin-Kolmogorov test, the Kuiper test, and the Sherman test. Results are given for each of these tests against each of four classes of alternatives. Also, the most powerful test against each member of the first three alternatives is obtained, and the powers of these tests are given for the same sample sizes as for the six general "omnibus" test statistics. These values constitute a "power envelope" against which all tests can be compared. The Neyman smooth tests with 2nd and 4th degree polynomials are found to have good power and are recommended as general tests for uniformity.     

A New Goodness-of-Fit Test for Censored Data with an Application in Monitoring Processes

In this article, we propose a new goodness-of-fit test for Type I or Type II censored samples from a completely specified distribution. This test is a generalization of Michael's test for censored data, which is based on the empirical distribution and a variance stabilizing transformation. Using Monte Carlo methods, the distributions of the test statistics are analyzed under the null hypothesis. Tables of quantiles of these statistics are also provided. The power of the proposed test is studied and compared to that of other well-known tests also using simulation. The proposed test is more powerful in most of the considered cases. Acceptance regions for the PP, QQ, and Michael's stabilized probability plots are derived, which enable one to visualize which data contribute to the decision of rejecting the null hypothesis. Finally, an application in quality control is presented as illustration.     

Detecting parameter shift in garch models

This paper applies recent theories of testing for parameter constancy to the conditional variance in a GARCH model. The supremum Lagrange multiplier test for conditional Gaussian GARCH models and its robustified variants are discussed. The asymptotic null distribution of the test statistics are derived from the weak convergence of the scores, and the critical values from the hitting probability of squared Bessel process.

Monte Carlo studies on the finite sample size and power performance of the supremum LM tests are conducted. Applications of these tests to S&P 500 indicate that the hypothesis of stable conditional variance parameters can be rejected.     

A NOTE ON A SEQUENTIAL GROUP t-TEST

A sequential probability ratio test (SPET) of the mean of a normal distribution with unknown variance, based on an independent sequence of groups of observations, is investigated and its efficiency compared with that of the WAGE sequential t-test, which is based on an invariantly sufficient sequence of test statistics.     

A review of model selection procedures relevant to the weibull distribution

A number of goodness-of-fit and model selection procedures related to the Weibull distribution are reviewed. These procedures include probability plotting, correlation type goodness-of-fit tests, and chi-square goodness-of-fit tests. Also the Kolmogorow-Smirniv, Kuiper, and Cramer-Von Mises test statistics for completely specified hypothesis based on censored data are reviewed, and these test statistics based on complete samples for the unspecified parameters case are considered. Goodness-of-fit tests based on sample spacings, and a goodness-of-fit test for the Weibull process, is also discussed.

Model selection procedures for selecting between a Weibull and gamma model, a Weibull and lognormal model, and for selecting from among all three models are considered. Also tests of exponential versus Weibull and Weibull versus generalized gamma are mentioned.