Levene tests of homogeneity of variance for general block and treatment designs

This article develops a weighted least squares version of Levene's test of homogeneity of variance for a general design, available both for univariate and multivariate situations. When the design is balanced, the univariate and two common multivariate test statistics turn out to be proportional to the corresponding ordinary least squares test statistics obtained from an analysis of variance of the absolute values of the standardized mean-based residuals from the original analysis of the data. The constant of proportionality is simply a design-dependent multiplier (which does not necessarily tend to unity). Explicit results are presented for randomized block and Latin square designs and are illustrated for factorial treatment designs and split-plot experiments. The distribution of the univariate test statistic is close to a standard F-distribution, although it can be slightly underdispersed. For a complex design, the test assesses homogeneity of variance across blocks, treatments, or treatment factors and offers an objective interpretation of residual plots.     

Nonparametric tests for homogeneity of species assemblages: a data depth approach

Testing homogeneity of species assemblages has important applications in ecology. Due to the unique structure of abundance data often collected in ecological studies, most classical statistical tests cannot be applied directly. In this article, we propose two novel nonparametric tests for comparing species assemblages based on the concept of data depth. They can be considered as a natural generalization of the Kolmogorov-Smirnov and the Cramér-von Mises tests (KS and CM) in this species assemblage comparison context. Our simulation studies show that the proposed test is more powerful than other existing methods under various settings. A real example is used to demonstrate how the proposed method is applied to compare species assemblages using plant community data from a highly diverse tropical forest at Barro Colorado Island, Panama.     

Chop-Lump Tests for Vaccine Trials

Summary This article proposes new tests to compare the vaccine and placebo groups in randomized vaccine trials when a small fraction of volunteers become infected. A simple approach that is consistent with the intent‐to‐treat principle is to assign a score, say W, equal to 0 for the uninfecteds and some postinfection outcome X > 0 for the infecteds. One can then test the equality of this skewed distribution of W between the two groups. This burden of illness (BOI) test was introduced by Chang, Guess, and Heyse (1994, Statistics in Medicine 13 , 1807–1814). If infections are rare, the massive number of 0s in each group tends to dilute the vaccine effect and this test can have poor power, particularly if the X's are not close to zero. Comparing X in just the infecteds is no longer a comparison of randomized groups and can produce misleading conclusions. Gilbert, Bosch, and Hudgens (2003, Biometrics 59 , 531–541) and Hudgens, Hoering, and Self (2003, Statistics in Medicine 22 , 2281–2298) introduced tests of the equality of X in a subgroup—the principal stratum of those “doomed” to be infected under either randomization assignment. This can be more powerful than the BOI approach, but requires unexaminable assumptions. We suggest new “chop‐lump” Wilcoxon and t‐tests (CLW and CLT) that can be more powerful than the BOI tests in certain situations. When the number of volunteers in each group are equal, the chop‐lump tests remove an equal number of zeros from both groups and then perform a test on the remaining W's, which are mostly >0. A permutation approach provides a null distribution. We show that under local alternatives, the CLW test is always more powerful than the usual Wilcoxon test provided the true vaccine and placebo infection rates are the same. We also identify the crucial role of the “gap” between 0 and the X's on power for the t‐tests. The chop‐lump tests are compared to established tests via simulation for planned HIV and malaria vaccine trials. A reanalysis of the first phase III HIV vaccine trial is used to illustrate the method.     

Exact log-rank tests for unequal follow-up

The asymptotic log-rank and generalized Wilcoxon tests are the standard procedures for comparing samples of possibly censored survival times. For comparison of samples of very different sizes, an exact test is available that is based on a complete permutation of log-rank or Wilcoxon scores. While the asymptotic tests do not keep their nominal sizes if sample sizes differ substantially, the exact complete permutation test requires equal follow-up of the samples. Therefore, we have developed and present two new exact tests also suitable for unequal follow-up. The first of these is an exact analogue of the asymptotic log-rank test and conditions on observed risk sets, whereas the second approach permutes survival times while conditioning on the realized follow-up in each group. In an empirical study, we compare the new procedures with the asymptotic log-rank test, the exact complete permutation test, and an earlier proposed approach that equalizes the follow-up distributions using artificial censoring. Results confirm highly satisfactory performance of the exact procedure conditioning on realized follow-up, particularly in case of unequal follow-up. The advantage of this test over other options of analysis is finally exemplified in the analysis of a breast cancer study.     

Tests for multivariate recurrent events in the presence of a terminal event

In studies involving diseases associated with high rates of mortality, trials are frequently conducted to evaluate the effects of therapeutic interventions on recurrent event processes terminated by death. In this setting, cumulative mean functions form a natural basis for inference for questions of a health economic nature, and Ghosh and Lin (2000) recently proposed a relevant class of test statistics. Trials of patients with cancer metastatic to bone, however, involve multiple types of skeletal complications, each of which may be repeatedly experienced by patients over their lifetime. Traditionally the distinction between the various types of events is ignored and univariate analyses are conducted based on a composite recurrent event. However, when the events have different impacts on patients' quality of life, or when they incur different costs, it can be important to gain insight into the relative frequency of the specific types of events and treatment effects thereon. This may be achieved by conducting separate marginal analyses with each analysis focusing on one type of recurrent event. Global inferences regarding treatment benefit can then be achieved by carrying out multiplicity adjusted marginal tests, more formal multiple testing procedures, or by constructing global test statistics. We describe methods for testing for differences in mean functions between treatment groups which accommodate the fact that each particular event process is ultimately terminated by death. The methods are illustrated by application to a motivating study designed to examine the effect of bisphosphonate therapy on the incidence of skeletal complications among patients with breast cancer metastatic to bone. We find that there is a consistent trend towards a reduction in the cumulative mean for all four types of skeletal complications with bisphosphonate therapy; there is a significant reduction in the need for radiation therapy for the treatment of bone. The global test suggests that bisphosphonate therapy significantly reduces the overall number of skeletal complications.     

Statistical tests for clonality

Cancer investigators frequently conduct studies to examine tumor samples from pairs of apparently independent primary tumors with a view to determine whether they share a "clonal" origin. The genetic fingerprints of the tumors are compared using a panel of markers, often representing loss of heterozygosity (LOH) at distinct genetic loci. In this article we evaluate candidate significance tests for this purpose. The relevant information is derived from the observed correlation of the tumors with respect to the occurrence of LOH at individual loci, a phenomenon that can be evaluated using Fisher's exact test. Information is also available from the extent to which losses at the same locus occur on the same parental allele. Data from these combined sources of information can be evaluated using a simple adaptation of Fisher's exact test. The test statistic is the total number of loci at which concordant mutations occur on the same parental allele, with higher values providing more evidence in favor of a clonal origin for the two tumors. The test is shown to have high power for detecting clonality for plausible models of the alternative (clonal) hypothesis, and for reasonable numbers of informative loci, preferably located on distinct chromosomal arms. The method is illustrated using studies to identify clonality in contralateral breast cancer. Interpretation of the results of these tests requires caution due to simplifying assumptions regarding the possible variability in mutation probabilities between loci, and possible imbalances in the mutation probabilities between parental alleles. Nonetheless, we conclude that the method represents a simple, powerful strategy for distinguishing independent tumors from those of clonal origin.     

Application of projection pursuit based robust principal component analysis to ECG enhancement

The paper presents an application of principal component analysis (PCA) to ECG processing. For this purpose the ECG beats are time-aligned and stored in the columns of an auxiliary matrix. The matrix, considered as a set of multidimensional variables, undergoes PCA. Reconstruction of the respective columns on the basis of a low dimensional principal subspace leads to the enhancement of the stored ECG beats. A few modifications of this classical approach to ECG signal filtering by means of a multivariate analysis are introduced. The first one is based on replacing the classical PCA by its robust extension. The second consists in replacing the analysis of the whole synchronized beats by the analysis of shorter signal segments. This creates the background for the third modification, which introduces the concept of variable dimensions of the subspaces corresponding to different parts of ECG beats. The experiments performed show that introduction of the respective modifications significantly improves the classical approach to ECG processing by application of principal component analysis.     

Comparison of independent samples of high-dimensional data by pairwise distance measures

Pairwise distance or association measures of sample elements are often used as a basis for hierarchical cluster analyses. They can also be used in tests for the comparison of pre-defined subgroups of the total sample. Usually this is done with permutation tests In this paper, we compare such a procedure with alternative tests for high-dimensional data based on spherically distributed scores in simulation experiments and with real data. The tests based on the pairwise distance or similarity measures perform quite well in this comparison. As the number of possible permutations is small in very small samples, this might restrict the use of the test. Therefore, we propose an exact parametric small sample version of the test using randomly rotated samples.     

Simultaneous sign tests for marginal homogeneity of square contingency tables

For testing the nullhypothesis of marginal homogeneity of 1x1-contingency tables it is proposed to perform I simultaneous sign tests and to combine them according to the BONFFERRONI or the ?INáK inequality.     

An adaptive test for a subset of coefficients in a multivariate regression model

An adaptive multivariate test is proposed for a subset of regression coefficients in a linear model. This adaptive method uses the studentized deleted residuals to calculate an appropriate weight for each observation. The weights are then used to compute Wilk's lambda for the weighted model. The adaptive test is performed by permuting the independent variables corresponding to those parameters that are assumed to equal zero in the null hypothesis. The permuted variables are then weighted to obtain a permutation test statistic that is used to estimate the p-value. An example is presented of a multivariate regression that uses systolic and diastolic blood pressure as dependent variables with age and body mass index as independent variables. The simulation results show that the adaptive test maintains its size for the three multivariate error distributions that were used in the study. For normal error models the power of the adaptive test nearly equaled that of the non-adaptive test. For models that used non-normal errors the adaptive test was considerably more powerful than the traditional non-adaptive test.     

Variable selection in high-dimensional multivariate binary data with application to the analysis of microbial community DNA fingerprints

In order to understand the relevance of microbial communities on crop productivity, the identification and characterization of the rhizosphere soil microbial community is necessary. Characteristic profiles of the microbial communities are obtained by denaturing gradient gel electrophoresis (DGGE) of polymerase chain reaction (PCR) amplified 16S rDNA from soil extracted DNA. These characteristic profiles, commonly called community DNA fingerprints, can be represented in the form of high-dimensional binary vectors. We address the problem of modeling and variable selection in high-dimensional multivariate binary data and present an application of our methodology in the context of a controlled agricultural experiment.     

Assessing the agronomic and environmental effects of the application of cattle manure compost on soil by multivariate methods

Multivariate analysis was used for interpreting data from a pot experiment using samples of three Spanish soils. Samples of soil fertilized with compost were compared with untreated control samples. We also compared the effect of adding the compost to soil with a controlled moisture content of 50% of its water holding capacity (WHC), and to a near-saturated soil (95% WHC). Hierarchical cluster analysis (HCA) and principal component analysis (PCA) were used; they perfectly differentiated sample groups both as a function of the treatment applied and by sampling date. The compost samples were characterized by higher pH, electrical conductivity (EC), organic matter (OM) content and cation exchange capacity (CEC), together with nutrient concentrations than the control pots. The pots with a soil–compost mixture at 95% WHC presented lower values of EC, CEC, inorganic N, K, Na and B than the mixtures at 50% WHC. Multivariate methods may therefore be useful for the analysis and interpretation of a large number of data in soil research.     

A class of permutation tests for stratified survival data

We propose a class of permutation tests for stratified survival data. The tests are derived using the framework of Fay and Shih (1998, Journal of the American Statistical Association 93, 387-396), which creates tests by permuting scores based on a functional of estimated distribution functions. Here the estimated distribution function for each possibly right-, left-, or interval-censored observation is based on a shrinkage estimator similar to the nonparametric empirical estimator of Ghosh, Lahiri, and Tiwari (1989, Communications in Statistics--Theory and Methods 18, 121-146), and permutation is carried out within strata. The proposed test with a weighted Mann-Whitney functional is similar to the permutation form of the stratified log-rank test when there is a large strata effect or the sample size in each stratum is large and is similar to the permutation form of the ordinary log-rank test when there is little strata effect. Thus, the proposed test unifies the advantages of both the stratified and ordinary log-rank tests. By changing the functional, we may obtain a stratified Prentice-Wilcoxon test or a difference in means test with similar unifying properties. We show through simulations the advantage of the proposed test over existing tests for uncensored and right-censored data.