Search

Summary Multivariate one-sided hypotheses testing problems arise frequently in practice. Various tests have been developed. In practice, there are often missing values in multivariate data. In this case, standard testing procedures based on complete data may not be applicable or may perform poorly if the missing data are discarded. In this article, we propose several multiple imputation methods for multivariate one-sided testing problem with missing data. Some theoretical results are presented. The proposed methods are evaluated using simulations. A real data example is presented to illustrate the methods. [ABSTRACT FROM AUTHOR]

Summary In a recent paper, Gaugler and Akritas (unpublished manuscript) considered testing for no main effect in a two-factor mixed effects design when the traditional assumptions do not hold. Here we extend the nonparametric modeling to the random effects design and consider the problem of testing for no interaction effect. The new models for these designs allow for dependence among the random effects, heteroscedasticity in the error and interaction terms, and do not require normality. At a more systemic level, these models differ from the classical ones in that they do not consider the random interaction term as an additional, extraneous source of variability. The proposed test procedure applies to settings where the random factor in the case of the mixed model or at least one of the random factors in the case of the random effects model has many levels. The number of replications can be small and possibly unbalanced. Moreover, the model and test procedure are general enough to accommodate data missing at random (MAR), provided the missingness mechanism is the same for each level of the random effect. The limiting distribution of the test statistic is normal. Extensive simulations indicate that our test procedure, with or without missing data, maintains the nominal Type I error rate in all simulation settings. On the contrary, the standard procedures (the F-test of PROC GLM in SAS, and the ML and REML methods of PROC MIXED in SAS), as well as the exact F-test of (1998 in Statistical Tests for Mixed Linear Models), are extremely liberal in heteroscedastic settings, while under homoscedasticity and normality, the proposed test procedure is comparable to them. An analysis of a dataset from the Mussel Watch Project is presented. [ABSTRACT FROM AUTHOR]

Methods for performing multiple tests of paired proportions are described. A broadly applicable method using McNemar's exact test and the exact distributions of all test statistics is developed; the method controls the familywise error rate in the strong sense under minimal assumptions. A closed form (not simulation-based) algorithm for carrying out the method is provided. A bootstrap alternative is developed to account for correlation structures. Operating characteristics of these and other methods are evaluated via a simulation study. Applications to multiple comparisons of predictive models for disease classification and to postmarket surveillance of adverse events are given. [ABSTRACT FROM AUTHOR]

We consider the problem of evaluating a statistical hypothesis when some model characteristics are nonidentifiable from observed data. Such a scenario is common in meta-analysis for assessing publication bias and in longitudinal studies for evaluating a covariate effect when dropouts are likely to be nonignorable. One possible approach to this problem is to fix a minimal set of sensitivity parameters conditional upon which hypothesized parameters are identifiable. Here, we extend this idea and show how to evaluate the hypothesis of interest using an infimum statistic over the whole support of the sensitivity parameter. We characterize the limiting distribution of the statistic as a process in the sensitivity parameter, which involves a careful theoretical analysis of its behavior under model misspecification. In practice, we suggest a nonparametric bootstrap procedure to implement this infimum test as well as to construct confidence bands for simultaneous pointwise tests across all values of the sensitivity parameter, adjusting for multiple testing. The methodology's practical utility is illustrated in an analysis of a longitudinal psychiatric study. [ABSTRACT FROM AUTHOR]

We introduce a method of estimating disease prevalence from case–control family study data. Case–control family studies are performed to investigate the familial aggregation of disease; families are sampled via either a case or a control proband, and the resulting data contain information on disease status and covariates for the probands and their relatives. Here, we introduce estimators for overall prevalence and for covariate-stratum-specific (e.g., sex-specific) prevalence. These estimators combine the proportion of affected relatives of control probands with the proportion of affected relatives of case probands and are designed to yield approximately unbiased estimates of their population counterparts under certain commonly made assumptions. We also introduce corresponding confidence intervals designed to have good coverage properties even for small prevalences. Next, we describe simulation experiments where our estimators and intervals were applied to case–control family data sampled from fictional populations with various levels of familial aggregation. At all aggregation levels, the resulting estimates varied closely and symmetrically around their population counterparts, and the resulting intervals had good coverage properties, even for small sample sizes. Finally, we discuss the assumptions required for our estimators to be approximately unbiased, highlighting situations where an alternative estimator based only on relatives of control probands may perform better. [ABSTRACT FROM AUTHOR]

Pharmacovigilance systems aim at early detection of adverse effects of marketed drugs. They maintain large spontaneous reporting databases for which several automatic signaling methods have been developed. One limit of those methods is that the decision rules for the signal generation are based on arbitrary thresholds. In this article, we propose a new signal-generation procedure. The decision criterion is formulated in terms of a critical region for the P-values resulting from the reporting odds ratio method as well as from the Fisher's exact test. For the latter, we also study the use of mid-P-values. The critical region is defined by the false discovery rate, which can be estimated by adapting the P-values mixture model based procedures to one-sided tests. The methodology is mainly illustrated with the location-based estimator procedure. It is studied through a large simulation study and applied to the French pharmacovigilance database. [ABSTRACT FROM AUTHOR]

Regression models are often used to test for cause-effect relationships from data collected in randomized trials or experiments. This practice has deservedly come under heavy scrutiny, because commonly used models such as linear and logistic regression will often not capture the actual relationships between variables, and incorrectly specified models potentially lead to incorrect conclusions. In this article, we focus on hypothesis tests of whether the treatment given in a randomized trial has any effect on the mean of the primary outcome, within strata of baseline variables such as age, sex, and health status. Our primary concern is ensuring that such hypothesis tests have correct type I error for large samples. Our main result is that for a surprisingly large class of commonly used regression models, standard regression-based hypothesis tests (but using robust variance estimators) are guaranteed to have correct type I error for large samples, even when the models are incorrectly specified. To the best of our knowledge, this robustness of such model-based hypothesis tests to incorrectly specified models was previously unknown for Poisson regression models and for other commonly used models we consider. Our results have practical implications for understanding the reliability of commonly used, model-based tests for analyzing randomized trials. [ABSTRACT FROM AUTHOR]

Many assessment instruments used in the evaluation of toxicity, safety, pain, or disease progression consider multiple ordinal endpoints to fully capture the presence and severity of treatment effects. Contingency tables underlying these correlated responses are often sparse and imbalanced, rendering asymptotic results unreliable or model fitting prohibitively complex without overly simplistic assumptions on the marginal and joint distribution. Instead of a modeling approach, we look at stochastic order and marginal inhomogeneity as an expression or manifestation of a treatment effect under much weaker assumptions. Often, endpoints are grouped together into physiological domains or by the body function they describe. We derive tests based on these subgroups, which might supplement or replace the individual endpoint analysis because they are more powerful. The permutation or bootstrap distribution is used throughout to obtain global, subgroup, and individual significance levels as they naturally incorporate the correlation among endpoints. We provide a theorem that establishes a connection between marginal homogeneity and the stronger exchangeability assumption under the permutation approach. Multiplicity adjustments for the individual endpoints are obtained via stepdown procedures, while subgroup significance levels are adjusted via the full closed testing procedure. The proposed methodology is illustrated using a collection of 25 correlated ordinal endpoints, grouped into six domains, to evaluate toxicity of a chemical compound. [ABSTRACT FROM AUTHOR]

Marginal mean models of temporal processes in event time data analysis are gaining more attention for their milder assumptions than the traditional intensity models. Recent work on fully functional temporal process regression (TPR) offers great flexibility by allowing all the regression coefficients to be nonparametrically time varying. The existing estimation procedure, however, prevents successive goodness-of-fit test for covariate coefficients in comparing a sequence of nested models. This article proposes a partly functional TPR model in the line of marginal mean models. Some covariate effects are time independent while others are completely unspecified in time. This class of models is very rich, including the fully functional model and the semiparametric model as special cases. To estimate the parameters, we propose semiparametric profile estimating equations, which are solved via an iterative algorithm, starting at a consistent estimate from a fully functional model in the existing work. No smoothing is needed, in contrast to other varying-coefficient methods. The weak convergence of the resultant estimators are developed using the empirical process theory. Successive tests of time-varying effects and backward model selection procedure can then be carried out. The practical usefulness of the methodology is demonstrated through a simulation study and a real example of recurrent exacerbation among cystic fibrosis patients. [ABSTRACT FROM AUTHOR]

Fluorescence curves are useful for monitoring changes in photosynthesis activity. Various summary measures have been used to quantify differences among fluorescence curves corresponding to different treatments, but these approaches may forfeit valuable information. As each individual fluorescence curve is a functional observation, it is natural to consider a functional regression model. The proposed model consists of a nonparametric component capturing the general form of the curves and a semiparametric component describing the differences among treatments and allowing comparisons of treatments. Several graphical model-checking approaches are introduced. Both approximate, asymptotic confidence intervals as well as simulation-based confidence intervals are available. Analysis of data from a crop experiment using the proposed model shows that the salient features in the fluorescence curves are captured adequately. The proposed functional regression model is useful for analysis of high throughput fluorescence curve data from regular monitoring or screening of plant growth. [ABSTRACT FROM AUTHOR]

We provide a method for obtaining confidence intervals, point estimates, and p-values for the primary effect size parameter at the end of a two-arm group sequential clinical trial in which adaptive changes have been implemented along the way. The method is based on applying the adaptive hypothesis testing procedure of Müller and Schäfer (2001, Biometrics 57, 886–891) to a sequence of dual tests derived from the stage-wise adjusted confidence interval of Tsiatis, Rosner, and Mehta (1984, Biometrics 40, 797–803). In the nonadaptive setting this confidence interval is known to provide exact coverage. In the adaptive setting exact coverage is guaranteed provided the adaptation takes place at the penultimate stage. In general, however, all that can be claimed theoretically is that the coverage is guaranteed to be conservative. Nevertheless, extensive simulation experiments, supported by an empirical characterization of the conditional error function, demonstrate convincingly that for all practical purposes the coverage is exact and the point estimate is median unbiased. No procedure has previously been available for producing confidence intervals and point estimates with these desirable properties in an adaptive group sequential setting. The methodology is illustrated by an application to a clinical trial of deep brain stimulation for Parkinson's disease. [ABSTRACT FROM AUTHOR]

In this article, we consider comparing the areas under correlated receiver operating characteristic (ROC) curves of diagnostic biomarkers whose measurements are subject to a limit of detection (LOD), a source of measurement error from instruments' sensitivity in epidemiological studies. We propose and examine the likelihood ratio tests with operating characteristics that are easily obtained by classical maximum likelihood methodology. [ABSTRACT FROM AUTHOR]

A small literature discusses locally most powerful rank tests when only a fraction of treated subjects respond to treatment. The ranks used in these tests are very different from conventional ranks, being relatively flat for low responses and then rising steeply, and the associated tests are much more powerful than conventional rank tests when, indeed, only a small fraction of treated subjects exhibit dramatic responses. Because the tests are derived from considerations of local power, they do not yield a plausible family of models for effect, and therefore they do not yield confidence intervals for the magnitude of effect formed by inverting the tests. There is a similarity between these tests and another family of tests, originally motivated by different considerations involving peak performance in small subsets. Exploiting this similarity, a method for obtaining confidence statements is proposed. In the case of observational studies, sensitivity to unobserved bias from nonrandom assignment of treatments is also examined. Two examples are used as illustrations: (i) a study of smoking during pregnancy and its effects on birth weight, in which smokers are matched to six controls, and (ii) a matched pair study of damage to DNA among workers at aluminum production plants. [ABSTRACT FROM AUTHOR]

This article considers the problem of comparing several treatments (dose levels, interventions, etc.) with the best, where the best treatment is unknown and the treatments are ordered in some sense. Order relations among treatments often occur quite naturally in practice. They may be ordered according to increasing risks, such as tolerability or safety problems with increasing dose levels in a dose–response study, for example. We tackle the problem of constructing a lower confidence bound for the smallest index of all treatments being at most marginally less effective than the (best) treatment having the largest effect. Such a bound ensures at confidence level 1 −α that all treatments with lower indices are relevantly less effective than the best competitor. We derive a multiple testing strategy that results in sharp confidence bounds. The proposed lower confidence bound is compared with those derived from other testing strategies. We further derive closed-form expressions for power and sample size calculations. Finally, we investigate several real data sets to illustrate various applications of our methods. [ABSTRACT FROM AUTHOR]

Several goodness-of-fit tests of a lifetime distribution have been suggested in the literature; many take into account censoring and/or truncation of event times. In some contexts, a goodness-of-fit test for the truncation distribution is of interest. In particular, better estimates of the lifetime distribution can be obtained when knowledge of the truncation law is exploited. In cross-sectional sampling, for example, there are theoretical justifications for the assumption of a uniform truncation distribution, and several studies have used it to improve the efficiency of their survival estimates. The duality of lifetime and truncation in the absence of censoring enables methods for testing goodness of fit of the lifetime distribution to be used for testing goodness of fit of the truncation distribution. However, under random censoring, this duality does not hold and different tests are required. In this article, we introduce several goodness-of-fit tests for the truncation distribution and investigate their performance in the presence of censored event times using simulation. We demonstrate the use of our tests on two data sets. [ABSTRACT FROM AUTHOR]

The annual percent change (APC) is often used to measure trends in disease and mortality rates, and a common estimator of this parameter uses a linear model on the log of the age-standardized rates. Under the assumption of linearity on the log scale, which is equivalent to a constant change assumption, APC can be equivalently defined in three ways as transformations of either (1) the slope of the line that runs through the log of each rate, (2) the ratio of the last rate to the first rate in the series, or (3) the geometric mean of the proportional changes in the rates over the series. When the constant change assumption fails then the first definition cannot be applied as is, while the second and third definitions unambiguously define the same parameter regardless of whether the assumption holds. We call this parameter the percent change annualized (PCA) and propose two new estimators of it. The first, the two-point estimator, uses only the first and last rates, assuming nothing about the rates in between. This estimator requires fewer assumptions and is asymptotically unbiased as the size of the population gets large, but has more variability since it uses no information from the middle rates. The second estimator is an adaptive one and equals the linear model estimator with a high probability when the rates are not significantly different from linear on the log scale, but includes fewer points if there are significant departures from that linearity. For the two-point estimator we can use confidence intervals previously developed for ratios of directly standardized rates. For the adaptive estimator, we show through simulation that the bootstrap confidence intervals give appropriate coverage. [ABSTRACT FROM AUTHOR]

Nonparametric and parametric approaches have been proposed to estimate false discovery rate under the independent hypothesis testing assumption. The parametric approach has been shown to have better performance than the nonparametric approaches. In this article, we study the nonparametric approaches and quantify the underlying relations between parametric and nonparametric approaches. Our study reveals the conservative nature of the nonparametric approaches, and establishes the connections between the empirical Bayes method and p-value-based nonparametric methods. Based on our results, we advocate using the parametric approach, or directly modeling the test statistics using the empirical Bayes method. [ABSTRACT FROM AUTHOR]

For a continuous-scale diagnostic test, the most commonly used summary index of the receiver operating characteristic curve (ROC) is the area under the curve (AUC) that measures the accuracy of the diagnostic test. In this article, we propose an empirical likelihood (EL) approach for the inference on the AUC. First we define an EL ratio for the AUC and show that its limiting distribution is a scaled chi-square distribution. We then obtain an EL-based confidence interval for the AUC using the scaled chi-square distribution. This EL inference for the AUC can be extended to stratified samples, and the resulting limiting distribution is a weighted sum of independent chi-square distributions. Additionally we conduct simulation studies to compare the relative performance of the proposed EL-based interval with the existing normal approximation-based intervals and bootstrap intervals for the AUC. [ABSTRACT FROM AUTHOR]

The objective of this article is to develop a hypothesis-testing procedure to determine whether a common source outbreak has ended. We consider the case when neither the calendar date of exposure to the pathogen nor the exact incubation period distribution is known. The hypothesis-testing procedure is based on the spacings between ordered calendar dates of disease onset of the cases. A simulation study was performed to evaluate the robustness of the methods to various models for the incubation period of infectious diseases. We investigated the impact of multiple testing on the overall outbreak-wise type I error probability. We derive expressions for the outbreak-wise type I error probability and show that multiple testing has minimal effect on inflating that error probability. The results are discussed in the context of the 2001 U.S. anthrax outbreak. [ABSTRACT FROM AUTHOR]

This article develops methods for stratified analyses of additive or multiplicative causal effect on binary outcomes in randomized trials with noncompliance. The methods are based on a weighted estimating function for an unbiased estimating function under randomization in each stratum. When known weights are used, the derived estimator is a natural extension of the instrumental variable estimator for stratified analyses, and test-based confidence limits are solutions of a quadratic equation in the causal parameter. Optimal weights that maximize asymptotic efficiency incorporate variability in compliance aspects across strata. An assessment based on asymptotic relative efficiency shows that a substantial enhancement in efficiency can be gained by using optimal weights instead of conventional ones, which do not incorporate the variability in compliance aspects across strata. Application to a field trial for coronary heart disease is provided. [ABSTRACT FROM AUTHOR]

In many practical problems, a hypothesis testing involves a nuisance parameter which appears only under the alternative hypothesis.(1977,Biometrika64,247–254) proposed the maximum of the score statistics over the whole range of the nuisance parameter as a test statistic for this type of hypothesis testing.(1999,Biometrics55,883–886) studied two other simpler maximum test statistics, the maximum of the score statistics at two extreme points of the nuisance parameter, and the maximum of the score statistics at three points of the nuisance parameter including the two extreme points. In this article, we compare the powers of these three maximum-type statistics in the context of three genetic problems. [ABSTRACT FROM AUTHOR]

The false discovery rate (FDR) procedure has become a popular method for handling multiplicity in high-dimensional data. The definition of FDR has a natural Bayesian interpretation; it is the expected proportion of null hypotheses mistakenly rejected given a measure of evidence for their truth. In this article, we propose controlling the positive FDR using a Bayesian approach where the rejection rule is based on the posterior probabilities of the null hypotheses. Correspondence between Bayesian and frequentist measures of evidence in hypothesis testing has been studied in several contexts. Here we extend the comparison to multiple testing with control of the FDR and illustrate the procedure with an application to wavelet thresholding. The problem consists of recovering signal from noisy measurements. This involves extracting wavelet coefficients that result from true signal and can be formulated as a multiple hypotheses-testing problem. We use simulated examples to compare the performance of our approach to the(1995,Journal of the Royal Statistical Society, Series B57,289–300) procedure. We also illustrate the method with nuclear magnetic resonance spectral data from human brain. [ABSTRACT FROM AUTHOR]

The main purpose of a phase IIA trial of a new anticancer therapy is to determine whether the therapy has sufficient promise against a specific type of tumor to warrant its further development. The therapy will be rejected for further investigation if the true response rate is less than some uninteresting level and the test of hypothesis is powered at a specific target response rate. Two-stage designs are commonly used for this situation. However, in many situations investigators often express concern about uncertainty in targeting the alternative hypothesis to study power at the planning stage. In this article, motivated by a real example, we propose a strategy for adaptive two-stage designs that will use the information at the first stage of the study to either reject the therapy or continue testing with either an optimistic or a skeptic target response rate, while the type I error rate is controlled. We also introduce new optimal criteria to reduce the expected total sample size. [ABSTRACT FROM AUTHOR]

When the intraclass correlation coefficient or the equivalent version of the kappa agreement coefficient have been estimated from several independent studies or from a stratified study, we have the problem of comparing the kappa statistics and combining the information regarding the kappa statistics in a common kappa when the assumption of homogeneity of kappa coefficients holds. In this article, using the like-lihood score theory extended to uisance parameters (Tarone, 1988, Communications in Statistics-Theory and Methods 17 (5), 1549 -556) we present a efficient homogeneity test for comparing several independent kappa statistics and, also, give a modified homogeneity score method using a noniterative and consistent estimator as an alternative. We provide the sample size using the modified homogeneity score method and compare it with that using the goodness-of-fit method (GOF)(Donner, Eliasziw, and Klar, 1996, Biometrics 52, 176 -183). A simulation study for small and moderate sample sizes showed that the actual level of the homogeneity score test using the maximum likelihood estimators (MLEs) of parameters is satisfactorily close to the nominal and it is smaller than those of the modified homogeneity score and the goodness-of-fit tests. We investigated statistical properties of several noniterative estimators of a common kappa. The estimator (Donner et al.,1996) is essentially efficient and can be used as an alternative to the iterative MLE. An efficient interval estimation of a common kappa using the likelihood score method is presented. [ABSTRACT FROM AUTHOR]

A proposal to improve trend tests by using noninteger scores is examined. It is concluded that despite improved power such tests are usually inferior to the simpler integer scored approach. [ABSTRACT FROM AUTHOR]

Asymptotic Distribution of Score Statistics for Spatial Cluster Detection with Censored Data Cook et Lee (20007) ont étendu la statistique de scan de Kulldorff (1997) pour la détection de clusters à partir de données de survie. Leur approche était basée sur la statistique du score et ils on proposé une distribution de permutation pour le maximum des score tests. L’utilisation de la statistique du score permet d’introduire des variables explicatives. Cependant nous montrons que la distribution de permutation demande des hypothèses fortes d’indépendance entre les clusters potentiels et à la fois la variable de censure et les variables explicatives. Nous présentons une approche utilisant la distribution asymptotique du maximum des statistique de score qui ne demande pas ces hypothèses. Inference for Proportions in a 2 × 2 Contingency Table: HPD or not HPD? Les intervalles fondés sur la densité a posteriori la plus élevée sont courants en inférence bayésienne, mais ils ne sont pas invariants après transformations, comme noté par Agresti et Min (2005). Agresti & Min suggèrent de préférer l’emploi d’intervalles centraux ou ≪en queue≫ dans le contexte du risque relatif et celui de l’odds-ratio. Nous proposons une modification pour des résultats extrêmes, l’invariance étant assurée en remplaçant les intervalles centraux par des intervalles unilatéraux. Les a priori de bayes-Laplace pour le paramètre binomial apparaissent ici préférables aux à priori de Jeffrey, contrairement à la suggestion d’Agresti et Min, basée sur la couverture fréquentiste. [ABSTRACT FROM AUTHOR]