The Copula Information Criteria

We derive two types of Akaike information criterion (AIC)‐like model‐selection formulae for the semiparametric pseudo‐maximum likelihood procedure. We first adapt the arguments leading to the original AIC formula, related to empirical estimation of a certain Kullback–Leibler information distance. This gives a significantly different formula compared with the AIC, which we name the copula information criterion. However, we show that such a model‐selection procedure cannot exist for copula models with densities that grow very fast near the edge of the unit cube. This problem affects most popular copula models. We then derive what we call the cross‐validation copula information criterion, which exists under weak conditions and is a first‐order approximation to exact cross validation. This formula is very similar to the standard AIC formula but has slightly different motivation. A brief illustration with real data is given.     

EXTENDED ANALYSIS FOR RANKED DATA

Standard analyses for ranked data or data which are analyzed using ranks are extended to give better comparisons. The new analysis allows detection of significant quadratic or dispersion effects.     

On nonparametric profile analysis of several multivariate samples

A unified development is offered for asymptotically distribution-free profile analysis of several multivariate samples. This includes as special cases procedures based on generalized U-statistics and also those based on linear rank statistics. Furthermore, it includes as special cases analysis of location profiles and also scalar profiles. Finally, asymptotic power and consistency properties are discussed for tests of hypotheses and subhypotheses of interest.     

A note on the universal optimality criterion for full rank models

The aim of this note is to suggest a revised formulation of the universal optimality criterion for full rank models as stated in Kiefer (1975). We have presented the relevant results with indications of some possible applications.     

Gamma failure‐time mixture models: yet another way to establish efficacy

Using a Yamaguchi‐type generalized gamma failure‐time mixture model, we analyse the data from a study of autologous and allogeneic bone marrow transplantation in the treatment of high‐risk refractory acute lymphoblastic leukaemia, focusing on the time to recurrence of disease. We develop maximum likelihood techniques for the joint estimation of the surviving fractions and the survivor functions. This includes an approximation to the derivative of the survivor function with respect to the shape parameter. We obtain the maximum likelihood estimates of the model parameters. We also compute the variance‐covariance matrix of the parameter estimators. The extended family of generalized gamma failure‐time mixture models is flexible enough to include many commonly used failure‐time distributions as special cases. Yet these models are not used in practice because of computational difficulties. We claim that we have overcome this problem. The proposed approximation to the derivative of the survivor function with respect to the shape parameter can be used in any statistical package. We also address the issue of lack of identifiability. We point out that there can be a substantial advantage to using the gamma failure‐time mixture models over nonparametric methods. Copyright © 2003 John Wiley & Sons, Ltd.     

On Kendall's Autocorrelations

This brief article extends the theory of sample Kendall's autocorrelations by providing their exact variances at lags higher than one under the null hypothesis of randomness, by introducing and investigating their weighted modifications, and by numerical demonstration of these results and their usefulness.     

Comparison of nonparametric analysis of variance methods: A vote for van der Waerden

ABSTRACT

For two-way layouts in a between-subjects analysis of variance design, the parametric F-test is compared with seven nonparametric methods: rank transform (RT), inverse normal transform (INT), aligned rank transform (ART), a combination of ART and INT, Puri & Sen's L statistic, Van der Waerden, and Akritas and Brunners ANOVA-type statistics (ATS). The type I error rates and the power are computed for 16 normal and nonnormal distributions, with and without homogeneity of variances, for balanced and unbalanced designs as well as for several models including the null and the full model. The aim of this study is to identify a method that is applicable without too much testing for all the attributes of the plot. The Van der Waerden test shows the overall best performance though there are some situations in which it is disappointing. The Puri & Sen's and the ATS tests show generally very low power. These two and the other methods cannot keep the type I error rate under control in too many situations. Especially in the case of lognormal distributions, the use of any of the rank-based procedures can be dangerous for cell sizes above 10. As already shown by many other authors, nonnormal distributions do not violate the parametric F-test, but unequal variances do, and heterogeneity of variances leads to an inflated error rate more or less also for the nonparametric methods. Finally, it should be noted that some procedures show rising error rates with increasing cell sizes, the ART, especially for discrete variables, and the RT, Puri & Sen, and the ATS in the cases of heteroscedasticity.     

The robustness of the likelihood ratio test,the nonparametric sum rank test,and f-ratio tests when the populations are from the negative binomial family

In this study, the robustness of power and significance level of several statistical testing methods was evaluated under the assumption that the test populations were from Poisson, negative binomial, or geometric distributions. The F-ratio test, with or without appropriate transformations, was shown to be both safe and robust for all distributions examined.     

Tables of the cramér-von mises distributions

Tables of the one- and two-sample unweighted Cramer-von Mises statistics are given, and compared with the limiting distribution. The two-sample statistic may be useful in (for example) clinical trials when a proportional hazards assumption (which leads to the use of the log-rank test) is unjustified: see, for example, Schumacher (1984). It is often possible to stop clinical trials early if the Cramer-von Mises test (rather than say, the log-rank test) is employed.     

Tail-Adaptive Location Rank Test for the Generalized Secant Hyperbolic Distribution

The generalized secant hyperbolic distribution (GSHD) was recently introduced as a modeling tool in data analysis. The GSHD is a unimodal distribution that is completely specified by location, scale, and shape parameters. It has also been shown elsewhere that the rank procedures of location are regular, robust, and asymptotically fully efficient. In this article, we study certain tail weight measures for the GSHD and introduce a tail-adaptive rank procedure of location based on those tail weight measures. We investigate the properties of the new adaptive rank procedure and compare it to some conventional estimators.