Saddlepoint approximations to exact tests and improved likelihood ratio tests for the gamma distribution

For the most common one-sample and two-sample tests in the gamma distribution we derive the log likelihood ratio tests and the improved versions obtained by a Bartlett adjustment. For most of these tests an exact test exists and we give the saddlepoint approximation to the latter. The tests are compared with previously published tests and a small simulation study is included.     

RAO'S SCORE TESTS IN SURVIVAL ANALYSIS: EXAMPLES AND BARTLETT TYPE ADJUSTMENTS

Score method in hypothesis testing is one of Professor C. R. Rao's great contributions to statistics. It provides a simple and unified way to test some simple and composite hypotheses in many statistical problems. Some popular tests in statistical practice derived with the help of intuitions can be shown as score tests under some statistical models. The subject-years test and log-rank test in survival analysis are two of the examples. In this paper, we first introduce these two examples. After formulating these two tests as score tests, we then review some recent results on the Bartlett type adjustments for these tests.     

Impact of surviving time on tests for carcinogenicity

In tumorigenicity experiments, each animal begins in a tumor-free state and then either develops a tumor or dies before developing a tumor. Animals that develop a tumor either die from the tumor or from other competing causes. All surviving animals are sacrificed at the end of the experiment, normally two years. The two most commonly used statistical tests are the logrank test for comparing hazards of death from rapidly lethal tumors and the Hoel-Walburg test for comparing prevalences of nonlethal tumors. However, the data obtained from a carcinogenicity experiment generally contains a mixture of fatal and incidental tumors. Peto et al.(1980)suggested combining the fatal and incidental tests for a comparison of tumor onset distributions.

Extensive simulations show that the trend test for tumor onset using the Peto procedure has the proper size, under the simulation constraints, when each group has identical mortality patterns, and the test with continuity correction tends to be conservative. When the animals n the dosed groups have reduced survival rates, the type I error rate is likely to exceed the nominal level. The continuity correction is recommended for a small reduction in survival time among the dosed groups to ensure the proper size. However, when there is a large reduction in survival times in the dosed groups, the onset test does not have the proper size.     

Bartlett corrections in beta regression models

We consider the issue of performing accurate small-sample testing inference in beta regression models, which are useful for modeling continuous variates that assume values in (0,1), such as rates and proportions. We derive the Bartlett correction to the likelihood ratio test statistic and also consider a bootstrap Bartlett correction. Using Monte Carlo simulations we compare the finite sample performances of the two corrected tests to that of the standard likelihood ratio test and also to its variant that employs Skovgaard's adjustment; the latter is already available in the literature. The numerical evidence favors the corrected tests we propose. We also present an empirical application.     

Tests for Detecting Overdispersion in the Positive Poisson Regression Model

This article derives score tests for extra-Poisson variation in the positive or truncated-at-zero Poisson regression model against truncated-at-zero negative binomial family alternatives. It also develops size-corrected tests of overdispersion that are expected to improve their small-sample properties. Further, small-sample performance of the tests is investigated by means of Monte Carlo experiments. As an illustration, the proposed tests are applied to a model of strikes in U.S. manufacturing. The proposed tests have an interpretation as conditional moment tests and require only the positive Poisson model to be estimated. It is shown that most of the tests for overdispersion in the regular Poisson model given in the econometric and statistical literature can be obtained as special cases of the tests developed in this article. Monte Carlo experiments indicate that the size correction, based on the asymptotic expansions of the score function, is effective in improving the accuracy of the size and power of the tests in small samples.     

Lagrance-multiplier tersts for weak exogeneity: a synthesis

This paper unifies two seemingly separate approaches to test weak exogeneity in dynamic regression models with Lagrange-mulptiplier statistics. The first class of tests focuses on the orthogonality between innovations and conditioning variables, and thus is related to the Durbin-Wu-Hausman specification test. The second approach has been developed more recently in the context of context of cointegration and error correction models, ad concentrates on the question whether the conditioning variables display error correction behaviour. It is shown that the vital difference between the two approaches stems from the choice of the parmeters of interest. A new test is derived, which encompasses both its predecessors. The test is applied to an error correction model of the demand for money in Switzerland.     

Lagrance-multiplier tersts for weak exogeneity: a synthesis

This paper unifies two seemingly separate approaches to test weak exogeneity in dynamic regression models with Lagrange-mulptiplier statistics. The first class of tests focuses on the orthogonality between innovations and conditioning variables, and thus is related to the Durbin-Wu-Hausman specification test. The second approach has been developed more recently in the context of context of cointegration and error correction models, ad concentrates on the question whether the conditioning variables display error correction behaviour. It is shown that the vital difference between the two approaches stems from the choice of the parmeters of interest. A new test is derived, which encompasses both its predecessors. The test is applied to an error correction model of the demand for money in Switzerland.     

Power transformations to induce normality and their applications

Summary.  Random variables which are positive linear combinations of positive independent random variables can have heavily right-skewed finite sample distributions even though they might be asymptotically normally distributed. We provide a simple method of determining an appropriate power transformation to improve the normal approximation in small samples. Our method contains the Wilson–Hilferty cube root transformation for χ ² random variables as a special case. We also provide some important examples, including test statistics of goodness-of-fit and tail index estimators, where such power transformations can be applied. In particular, we study the small sample behaviour of two goodness-of-fit tests for time series models which have been proposed recently in the literature. Both tests are generalizations of the popular Box–Ljung–Pierce portmanteau test, one in the time domain and the other in the frequency domain. A power transformation with a finite sample mean and variance correction is proposed, which ameliorates the small sample effect. It is found that the corrected versions of the tests have markedly better size properties. The correction is also found to result in an overall increase in power which can be significant under certain alternatives. Furthermore, the corrected tests also have better power than the Box–Ljung–Pierce portmanteau test, unlike the uncorrected versions.     

A note on asymptotic expansions for the power of perturbed tests

For simplicity or tractability reasons one sometimes uses modified test statistics, which differ from the original ones up to O_p(a_n) terms with a_n→0. In this note, some technical conditions are provided under which a corresponding expansion for the powers of such perturbed tests holds. The necessity of some of these conditions is discussed and illustrated by examples. An application to invariant testing multivariate normality is presented.     

The Appropriateness of Some Common Procedures for Testing the Equality of Two Independent Binomial Populations

For testing the equality of two independent binomial populations the Fisher exact test and the chi-squared test with Yates's continuity correction are often suggested for small and intermediate size samples. The use of these tests is inappropriate in that they are extremely conservative. In this article we demonstrate that, even for small samples, the uncorrected chi-squared test (i.e., the Pearson chi-squared test) and the two-independent-sample t test are robust in that their actual significance levels are usually close to or smaller than the nominal levels. We encourage the use of these latter two tests.     

On the theory of modified randomization tests for nonparametric hypotheses

A general randomization test for nonparametric hypotheses which is a modification of permutation tests in proposed. The exact level of the test is derived and under mild gegularity conditions, a general result on the consistency of the power function is obtained. Applications to several testing problems are considered. Asymptotic expansions of the power of this test are derived with respect to contiguous alternatives thus test are derived with respect to contiguous alternatives thus enabling us to make deficiency comparisons with permutation tests. The paper concludes with some Monte Carlo simulations verifying the theoretical results derived.     

Bootstrap methods for heteroskedastic regression models: evidence on estimation and testing

This paper uses Monte Carlo simulation analysis to study the finite-sample behavior of bootstrap estimators and tests in the linear heteroskedastic model. We consider four different bootstrapping schemes, three of them specifically tailored to handle heteroskedasticity. Our results show that weighted bootstrap methods can be successfully used to estimate the variances of the least squares estimators of the linear parameters both under normality and under nonnormality. Simulation results are also given comparing the size and power of the bootstrapped Breusch-Pagan test with that of the original test and of Bartlett and Edgeworth-corrected tests. The bootstrap test was found to be robust against unfavorable regression designs.     

The rank von neumann test as a test for autocorrelation in regression models

The rank Von Neumann test, which performs extremely well as a test for serial correlation in raw data, is here compared with the Durbin-Watson and Geary tests as a test for autocorrelation in regression residuals. The test convincingly outperforms the Geary test but it is less robust than the Durbin-Watson test     

IMPROVED SCORE TESTS FOR VON MISES REGRESSION MODELS

We present, in matrix notation, a finite-sample correction formula to improve score tests in von Mises regression models with concentration covariates. The formula only requires simple operations on matrices and can be used to obtain analytically closed-form corrections for score test statistics in a variety of special von Mises models. The paper also provides a numerical comparison of the size of two score test statistics with bootstrap-based critical values.     

Large sample power of absolute moment tests

Sample kurtosis is a member of the large class of absolute moment tests of normality. We compare kurtosis to other absolute moment tests to determine which are the most powerful at detecting long‐tailed symmetric departures from normality for large samples. The large sample power of the tests is calculated using Geary's (1947) approximations of the moments of the test statistics. Using the system of Gram-Charlier symmetric distributions as alternatives, the most power is obtained using a moment in the range 2.5 ‐ 3.5.     

Evaluating the Accuracy of Small P‐Values In Genetic Association Studies Using Edgeworth Expansions

The asymptotic distributions of many classical test statistics are normal. The resulting approximations are often accurate for commonly used significance levels, 0.05 or 0.01. In genome‐wide association studies, however, the significance level can be as low as 1×10⁻⁷, and the accuracy of the p‐values can be challenging. We study the accuracies of these small p‐values are using two‐term Edgeworth expansions for three commonly used test statistics in GWAS. These tests have nuisance parameters not defined under the null hypothesis but estimable. We derive results for this general form of testing statistics using Edgeworth expansions, and find that the commonly used score test, maximin efficiency robust test and the chi‐squared test are second order accurate in the presence of the nuisance parameter, justifying the use of the p‐values obtained from these tests in the genome‐wide association studies.     

Monte carlo sampling approach to testing nonnested hypothesis: monte carlo results

Alternative ways of using Monte Carlo methods to implement a Cox-type test for separate families of hypotheses are considered. Monte Carlo experiments are designed to compare the finite sample performances of Pesaran and Pesaran's test, a RESET test, and two Monte Carlo hypothesis test procedures. One of the Monte Carlo tests is based on the distribution of the log-likelihood ratio and the other is based on an asymptotically pivotal statistic. The Monte Carlo results provide strong evidence that the size of the Pesaran and Pesaran test is generally incorrect, except for very large sample sizes. The RESET test has lower power than the other tests. The two Monte Carlo tests perform equally well for all sample sizes and are both clearly preferred to the Pesaran and Pesaran test, even in large samples. Since the Monte Carlo test based on the log-likelihood ratio is the simplest to calculate, we recommend using it.     

Testing homogeneity of control and treatment populations — local optimality and related issues

Durairajan and Raman (1996 a, b) studied the robustness of Locally most powerful invariant (LMPI) tests for compound normal model in control and treatment populations. In the present paper, the Locally most powerful (LMP) tests are constructed for no contamination in normal mixture model through testing the parameter of mixture of distributions and the mixing proportion. The expected performance of LMP tests are compared using Efron's Statistical Curvature on the lines of Sen Gupta and Pal (1991). The Locally most powerful similar (LMPS) tests for the equality of control and treatment populations in the presence of nuisance parameters are also constructed. Further, the null and non-null distributions of the test statistics are derived and some power computations are made. Received: September 1, 1999; revised version: August 31, 2000     

Power of depth-based nonparametric tests for multivariate locations

Li and Liu [New nonparametric tests of multivariate locations and scales. Statist Sci. 2004;19(4):686–696] introduced two tests for a difference in locations of two multivariate distributions based on the concept of data depth. Using the simplicial depth [Liu RY. On a notion of data depth based on random simplices. Ann Stat. 1990;18(1):405–414], they studied the performance of these tests for symmetric distributions, namely, the normal and the Cauchy, in a simulation study. However, to the best of our knowledge, the performance of these tests for skewed distributions has not been studied in the current literature. This paper is a contribution in that direction and examines the performance of these depth-based tests in an extensive simulation study involving ten distributions belonging to five well-known families of multivariate skewed distributions. The study includes a comparison of the performance of these tests for four popular affine-invariant depth functions. Conclusions and recommendations are offered.     

Some Second-Order Asymptotics for Extreme Value Linear Regression Models

ABSTRACT

In this article we derive finite-sample corrections in matrix notation for likelihood ratio and score statistics in extreme-value linear regression models. We consider three corrected score tests that perform better than the usual score test. We also derive general formulae for second-order biases of maximum likelihood estimates of the linear parameters. Some simulations are performed to compare the likelihood ratio and score statistics with their modified versions and to illustrate the bias correction.