A p-value for testing the equivalence of the variances of a bivariate normal distribution

A p-value is developed for testing the equivalence of the variances of a bivariate normal distribution. The unknown correlation coefficient is a nuisance parameter in the problem. If the correlation is known, the proposed p-value provides an exact test. For large samples, the p-value can be computed by replacing the unknown correlation by the sample correlation, and the resulting test is quite satisfactory. For small samples, it is proposed to compute the p-value by replacing the unknown correlation by a scalar multiple of the sample correlation. However, a single scalar is not satisfactory, and it is proposed to use different scalars depending on the magnitude of the sample correlation coefficient. In order to implement this approach, tables are obtained providing sub-intervals for the sample correlation coefficient, and the scalars to be used if the sample correlation coefficient belongs to a particular sub-interval. Once such tables are available, the proposed p-value is quite easy to compute since it has an explicit analytic expression. Numerical results on the type I error probability and power are reported on the performance of such a test, and the proposed p-value test is also compared to another test based on a rejection region. The results are illustrated with two examples: an example dealing with the comparability of two measuring devices, and an example dealing with the assessment of bioequivalence.     

Confidence intervals for the scale parameter of exponential distribution based on Type II doubly censored samples

This paper deals with the problem of interval estimation of the scale parameter in the two-parameter exponential distribution subject to Type II double censoring. Base on a Type II doubly censored sample, we construct a class of interval estimators of the scale parameter which are better than the shortest length affine equivariant interval both in coverage probability and in length. The procedure can be repeated to make further improvement. The extension of the method leads to a smoothly improved confidence interval which improves the interval length with probability one. All improved intervals belong to the class of scale equivariant intervals.     

Asymptotic power comparison of T2-type test and likelihood ratio test for a mean vector based on two-step monotone missing data

Abstract

In a 2-step monotone missing dataset drawn from a multivariate normal population, T²-type test statistic (similar to Hotelling’s T² test statistic) and likelihood ratio (LR) are often used for the test for a mean vector. In complete data, Hotelling’s T² test and LR test are equivalent, however T²-type test and LR test are not equivalent in the 2-step monotone missing dataset. Then we interest which statistic is reasonable with relation to power. In this paper, we derive asymptotic power function of both statistics under a local alternative and obtain an explicit form for difference in asymptotic power function. Furthermore, under several parameter settings, we compare LR and T²-type test numerically by using difference in empirical power and in asymptotic power function. Summarizing obtained results, we recommend applying LR test for testing a mean vector.     

The superharmonic condition for simultaneous estimation of means in exponential familles

The mean vector associated with several independent variates from the exponential subclass of Hudson (1978) is estimated under weighted squared error loss. In particular, the formal Bayes and “Stein-like” estimators of the mean vector are given. Conditions are also given under which these estimators dominate any of the “natural estimators”. Our conditions for dominance are motivated by a result of Stein (1981), who treated the N_p (θ, I) case with p ≥ 3. Stein showed that formal Bayes estimators dominate the usual estimator if the marginal density of the data is superharmonic. Our present exponential class generalization entails an elliptic differential inequality in some natural variables. Actually, we assume that each component of the data vector has a probability density function which satisfies a certain differential equation. While the densities of Hudson (1978) are particular solutions of this equation, other solutions are not of the exponential class if certain parameters are unknown. Our approach allows for the possibility of extending the parametric Stein-theory to useful nonexponential cases, but the problem of nuisance parameters is not treated here.     

Sequential and two-stage procedures for selecting the better exponential population covering the case of unknown and unequal scale parameters

The problems of selecting the larger location parameter of two exponential distributions are discussed. When the scale parameters are the same but unknown, we consider the procedure of Desu et al. (1977) in detail, and study some of its exact and asymptotic properties. We indicate how this procedure can be modified along the lines of Mukhopadhyay (1979, 1980) to achieve first-order asymptotic efficiency. We then propose a sequential procedure for this set-up and show that it is asymptotically second-order efficient according to Ghosh and Mukhopadhyay (1981). In case the scale parameters are completely unknown and unequal, we propose a two-stage procedure that guarantees the probability of correct selection to exceed the prescribed nominal level in the preference zone. We do not need any new tables to implement this particular procedure other than those in Krishnaiah and Armitage (1964), Gupta and Sobel (1962), Guttman and Milton (1969). We also propose a sequential method in this case and derive some of its asymptotic properties.     

Nonparametric tests for scale shift at an unknown time point

A class of linear rank tests is suggested for testing a shift in scale at an unknown time point in a sequence of independent observations. The tests,based on inverse normal scores and on ordered exponential scores,are shown to be asymptotically as efficient as their distribution-oriented competitors. Critical values and powers for these two rank tests are also discussed.     

Goodness of Fit Statistics for the Exponential Distribution When the Data Are Grouped

Abstract

In many industrial and biological experiments, the recorded data consist of the number of observations falling in an interval. In this paper, we develop two test statistics to test whether the grouped observations come from an exponential distribution. Following the procedure of Damianou and Kemp (Damianou, C., Kemp, A. W. (1990 Damianou, C. and Kemp, A. W. 1990. New goodness of statistics for discrete and continuous data. American Journal of Mathematical and Management Sciences, 10: 275–307. [Taylor & Francis Online] , [Google Scholar]). New goodness of statistics for discrete and continuous data. American Journal of Mathematical and Management Sciences 10:275–307.), Kolmogrov–Smirnov type statistics are developed with the maximum likelihood estimator of the scale parameter substituted for the true unknown scale. The asymptotic theory for both the statistics is studied and power studies carried out via simulations.     

Efficiency of reduced logistic regression models

One feature of the usual polychotomous logistic regression model for categorical outcomes is that a covariate must be included in all the regression equations. If a covariate is not important in all of them, the procedure will estimate unnecessary parameters. More flexible approaches allow different subsets of covariates in different regressions. One alternative uses individualized regressions which express the polychotomous model as a series of dichotomous models. Another uses a model in which a reduced set of parameters is simultaneously estimated for all the regressions. Large-sample efficiencies of these procedures were compared in a variety of circumstances in which there was a common baseline category for the outcome and the covariates were normally distributed. For a correctly specified model, the reduced estimates were over 100% efficient for nonzero slope parameters and up to 500% efficient when the baseline frequency and the effect of interest were small. The individualized estimates could have efficiencies less than 50% when the effect of interest was large, but were also up to 130% efficient when the baseline frequency was large and the effect of interest was small. Efficiency was usually enhanced by correlation among the covariates. For an underspecified reduced model, asymptotic bias in the reduced estimates was approximately proportional to the magnitude of the omitted parameter and to the reciprocal of the baseline frequency.     

Efficiency of nonparametric tests for scale shift at an unknown time point

Recently Hsieh considered three nonparametric tests for the scale-change problem and showed them to be asymptotically as efficient as their parametric competitors. Here the class of so called sum-type statistics is studied which contains the statistics of Hsieh. It is proved that max-type statistics can be constructed that are at least as efficient in the sense of Bahadur as the sum-type statistics.     

Efficiency of nonparametric tests for scale shift at an unknown time point

Recently Hsieh considered three nonparametric tests for the scalechange problem and showed them to be asymptotically as efficient as their parametric competitors. Here the class of so called sum-type statistics is studied which contains the statistics of Hsieh. It is proved that max-type statistics can be constructed that are at least as efficient in the sense of Bahadur as the sum-type statistics.     

The use of an identity in Anderson for multivariate multiple testing

Consider the model where there are I$I$ independent multivariate normal treatment populations with p×1$p\times 1$ mean vectors _μi${\mathit{\mu }}_i$, i=1,…,I$i=1,\dots ,I$, and covariance matrix Σ$\Sigma$. Independently the (I+1)$(I+1)$st population corresponds to a control and it too is multivariate normal with mean vector _μI+1${\mathit{\mu }}_{I+1}$ and covariance matrix Σ$\Sigma$. Now consider the following two multiple testing problems.     

The Decline and Fall of Type II Error Rates

For general linear models with normally distributed random errors, the probability of a Type II error decreases exponentially as a function of sample size. This potentially rapid decline reemphasizes the importance of performing power calculations.     

Nonparametric repeated significance tests for one-way anova with adoptation to multiple comparisons

Based on two-sample rank order statistics, a repeated significance testing procedure for a multi-sample location problem is considered. The asymptotic distribution theory of the proposed tests is given under the null hypothesis as well as under local alternatives. A Bahadur efficiency result of the repeated significance test relative to the terminal test based solely on the target sample size is presented. In the adaptation of the proposed tests to multiple comparisons, an asymptotically equivalent test statistic in terms of the rank estimators of the location parameters is derived from which the Scheffé method of multiple comparisons can be obtained in a convinient way.     

On an asymptotic relative efficiency concept based on expected volumes of confidence regions

ABSTRACT

The paper deals with an asymptotic relative efficiency concept for confidence regions of multidimensional parameters that is based on the expected volumes of the confidence regions. Under standard conditions the asymptotic relative efficiencies of confidence regions are seen to be certain powers of the ratio of the limits of the expected volumes. These limits are explicitly derived for confidence regions associated with certain plugin estimators, likelihood ratio tests and Wald tests. Under regularity conditions, the asymptotic relative efficiency of each of these procedures with respect to each one of its competitors is equal to 1. The results are applied to multivariate normal distributions and multinomial distributions in a fairly general setting.     

The costs of using incomplete response data for the exponential regression model

This paper investigates the asymptotic and small sample costs of using incomplete response data, Situations are identified where the information loss is substantial, Moreover, the small sample properties of the estimators are even worse than suggested by their asymptotic counterparts. These results provide the practitioner with guidance as to the severity of the costs he can incur, This is especially helpful when he cars choose the type of incomplete data that he observes.     

Locally Most Powerful Rank Tests for Comparison of Two Failure Rates Based on Multiple Type-II Censored Data

This article deals with the locally most powerful rank tests for testing the hypothesis that two failure rates are equal against the alternative that one failure rate is greater than the other, when the combined ordered sample is multiple Type-II censored. A modified version of the Dupa? and Hájek (1969 Dupa? , V. , Hájek , J. (1969). Asymptotic normality of simple linear rank statistics under alternatives II. Ann. Math. Statist. 40:1992–2017.[Crossref] , [Google Scholar]) theorem is used to establish their asymptotic normality under fixed alternative since the scores generating functions associated with these rank test statistics have a finite number of jump discontinuities. The modified version that leads to a simpler centering constant, is proved by Dupa? (1970 Dupa? , V. ( 1970 ). A contribution to the asymptotic normality of simple linear rank statistics. In: Puri, M. L., ed. Nonparametric Techniques in Statistical Inference. Cambridge: Cambridge University Press . [Google Scholar]) using the results of Hájek (1968 Hájek , J. ( 1968 ). Asymptotic normality of simple linear rank statistics under alternatives . Ann. Math. Statist. 39 : 325 – 346 .[Crossref] , [Google Scholar]). The Pitman AREs of these rank tests based on censored data relative to the corresponding tests based on complete data are obtained under some Lehmann-type alternative distributions such that their failure rates dominate the failure rates of the respective null distributions. The AREs are computed numerically for single (left or right) and double censored data, and the extent of loss due to these censoring schemes is discussed. The rank tests considered here include among them the Mann-Whiney-Wilcoxon (MWW) test, the Savage test, and the linear combination of these two tests. In the case of all the tests, except the MWW test, it is found that the loss of efficiency due to left censoring is considerably less than that due to right censoring. In the case of finite samples, Monte Carlo simulation results showing the empirical levels and empirical powers against some Lehmann alternatives are presented.