Search

This paper considers the problem of using observational data in the presence of selection bias to identify causal effects in the framework of linear structural equation models. We propose a criterion for testing whether or not observed statistical dependencies among variables are generated by conditioning on a common response variable. When the answer is affirmative, we further provide formulations for recovering the covariance matrix of the whole population from that of the selected population. The results of this paper provide guidance for reliable causal inference, based on the recovered covariance matrix obtained from the statistical information with selection bias. [ABSTRACT FROM AUTHOR]

This paper presents a class of Mantel & Haenszel(1959) type statistics for testing independence between pairs of failure times. This class is extended to include a weight function in order to enhance test power. Large sample distributions of these statistics are given under the null hypothesis of independence. The paper also presents a simplified test, the covariance test, which is asymptotically equivalent to the Mantel-Haenszel-type test. In order to account for heterogeneity among subjects, both tests are generalised to include subject-specific covariates. In a simulation study, the weighted Mantel-Haenszel and covariance tests are shown to lead to an increase in power for detecting dependence, compared to unweighted tests and to a score test (Commenges & Andersen, 1995) that is based on a random effects model. An example is used to illustrate the methodology. [ABSTRACT FROM AUTHOR]

Tang et al. (2003) considered a regression model with missing response, where the missingness mechanism depends on the value of the response variable and hence is nonignorable. They proposed three pseudolikelihood estimators, based on different treatments of the probability distribution of the completely observed covariates. The first assumes the distribution of the covariate to be known, the second estimates this distribution parametrically, and the third estimates the distribution nonparametrically. While it is not hard to show that the second estimator is more efficient than the first, Tang et al. (2003) only conjectured that the third estimator is more efficient than the first two. In this paper, we investigate the asymptotic behaviour of the third estimator by deriving a closed-form representation of its asymptotic variance. We then prove that the third estimator is more efficient than the other two. Our result can be straightforwardly applied to missingness mechanisms that are more general than that in Tang et al. (2003). [ABSTRACT FROM AUTHOR]

Graphical Gaussian models with edge and vertex symmetries were introduced by Højsgaard & Lauritzen (2008), who gave an algorithm for computing the maximum likelihood estimate of the precision matrix for such models. In this paper, we take a Bayesian approach to its estimation. We consider only models with symmetry constraints and which thus form a natural exponential family with the precision matrix as the canonical parameter. We identify the Diaconis-Ylvisaker conjugate prior for these models, develop a scheme to sample from the prior and posterior distributions, and thus obtain estimates of the posterior mean of the precision and covariance matrices. Such a sampling scheme is essential for model selection in coloured graphical Gaussian models. In order to verify the precision of our estimates, we derive an analytic expression for the expected value of the precision matrix when the graph underlying our model is a tree, a complete graph on three vertices, or a decomposable graph on four vertices with various symmetries, and we compare our estimates with the posterior mean of the precision matrix and the expected mean of the coloured graphical Gaussian model, that is, of the covariance matrix. We also verify the accuracy of our estimates on simulated data. [ABSTRACT FROM AUTHOR]

We consider the estimation of the semiparametric proportional hazards model with an unspecified baseline hazard function where the effect of a continuous covariate is assumed to be monotone. Previouswork on nonparametric maximum likelihood estimation for isotonic proportional hazard regression with right-censored data is computationally intensive, lacks theoretical justification, and may be prohibitive in large samples. In this paper, partial likelihood estimation is studied. An iterative quadratic programming method is considered, which has performed well with likelihoods for isotonic parametric regression models. However, the iterative quadratic programming method for the partial likelihood cannot be implemented using standard pool-adjacent-violators techniques, increasing the computational burden and numerical instability. The iterative convex minorant algorithm which uses pool-adjacent-violators techniques has also been shown to perform well in related parametric likelihood set-ups, but evidences computational difficulties under the proportional hazards model.An alternative pseudo-iterative convex minorant algorithm is proposed which exploits the pool-adjacent-violators techniques, is theoretically justified, and exhibits computational stability.Aseparate estimator of the baseline hazard function is provided. The algorithms are extended to models with time-dependent covariates. Simulation studies demonstrate that the pseudo-iterative convex minorant algorithm may yield orders-of-magnitude reduction in computing time relative to the iterative quadratic programming method and the iterative convex minorant algorithm, with moderate reductions in the bias and variance of the estimators. Analysis of data from a recent HIV prevention study illustrates the practical utility of the isotonic methodology in estimating nonlinear, monotonic covariate effects. [ABSTRACT FROM AUTHOR]

Propensity score matching is widely used to control covariates when analysing the effects of a nonrandomized binary treatment. However, it requires several arbitrary decisions, such as how many matched subjects to use and how to choose them. In this paper a simple least squares estimator is proposed, where the treatment, and possibly the response variable, is replaced by the propensity score residual. The proposed estimator controls covariates semiparametrically if the propensity score function is correctly specified. Furthermore, it is numerically stable and relatively easy to use, compared with alternatives such as matching, regression imputation, weighting, and doubly robust estimators. The proposed estimator also has a simple valid asymptotic variance estimator that works well in small samples. The least squares estimator is extended to multiple treatments and noncontinuously distributed responses. A simulation study demonstrates that it has lower mean squared error than its competitors. [ABSTRACT FROM AUTHOR]

In this paper, we consider clustered right-censored time-to-event data. Such data can be analysed either using a marginal model if one is interested in population effects or using so-called frailty models if one is interested in covariate effects on the individual level and in estimation of correlation. The Cox frailty model has been studied extensively in the last decade or so and estimation techniques and large sample results are now available. It is, however, difficult to deal with time-changing covariate effects when using the Cox model. An appealing alternative model is the Aalen additive hazards model, in which it is easy to work with time dynamics. In this paper, we describe an innovative approach to estimation in the Aalen additive gamma frailty hazards model. We give the large sample properties of the estimators and investigate their small sample properties by Monte Carlo simulation. A real example is provided for illustration. [ABSTRACT FROM PUBLISHER]

Distributional theory regarding principal components is less well developed for correlation matrices than it is for covariance matrices. The intent of this paper is to reduce this disparity. Methods are proposed that enable investigators to fit and to make inferences about flexible principal components models for correlation matrices. The models allow arbitrary eigenvalue multiplicities and allow the distinct eigenvalues to be modelled parametrically or nonparametrically. Local parameterisations and implicit functions are used to construct full-rank unconstrained parameterisations. First-order asymptotic distributions are obtained directly from the theory of estimating functions. Second-order accurate distributions for making inferences under normality are obtained directly from likelihood theory. Simulation studies show that the Bartlett correction is effective in controlling the size of the tests and that first-order approximations to nonnull distributions are reasonably accurate. The methods are illustrated on a dataset. [ABSTRACT FROM AUTHOR]

Random-effects models are frequently used to synthesize information from different studies in metaanalysis. While likelihood-based inference is attractive both in terms of limiting properties and of implementation, its application in random-effects meta-analysis may result in misleading conclusions, especially when the number of studies is small to moderate. The current paper shows how methodology that reduces the asymptotic bias of the maximum likelihood estimator of the variance component can also substantially improve inference about the mean effect size. The results are derived for the more general framework of random-effects meta-regression, which allows the mean effect size to vary with study-specific covariates. [ABSTRACT FROM AUTHOR]

This paper considers the construction of an upper confidence bound on the range of a set of p (≥2) location parameters following a sequential range test. Although details are given only for the one-way analysis of variance model with known variance, generalisation to other situations is pointed out. This upper confidence bound is useful in assessing the equivalence of all the location parameters. An illustrative example is given. [ABSTRACT FROM AUTHOR]

This paper concerns regression methodology for assessing relationships between multidimensional response variables and covariates that are correlated within a network. To address analytical challenges associated with the integration of network topology into the regression analysis, we propose a hybrid quadratic inference method that uses both prior and data-driven correlations among network nodes. A Godambe information-based tuning strategy is developed to allocate weights between the prior and data-driven network structures, so the estimator is efficient. The proposed method is conceptually simple and computationally fast, and has appealing large-sample properties. It is evaluated by simulation, and its application is illustrated using neuroimaging data from an association study of the effects of iron deficiency on auditory recognition memory in infants. [ABSTRACT FROM AUTHOR]

The sample covariance matrix, which is well known to be highly nonrobust, plays a central role in many classical multivariate statistical methods. A popular way of making such multivariate methods more robust is to replace the sample covariance matrix with some robust scatter matrix. The aim of this paper is to point out that multivariate methods often require that certain properties of the covariance matrix hold also for the robust scatter matrix in order for the corresponding robust plug-in method to be a valid approach, but that not all scatter matrices possess the desired properties. Plug-in methods for independent components analysis, observational regression and graphical modelling are considered in more detail. For each case, it is shown that replacing the sample covariance matrix with a symmetrized robust scatter matrix yields a valid robust multivariate procedure. [ABSTRACT FROM AUTHOR]

The use of high-throughput data to study the changing behaviour of biological pathways has focused mainly on examining the changes in the means of pathway genes. In this paper, we propose instead to test for changes in the co-regulated and unregulated variability of pathway genes. We assume that the eigenvalues of previously defined pathways capture biologically relevant quantities, and we develop a test for biologically meaningful changes in the eigenvalues between classes. This test reflects important and often ignored aspects of pathway behaviour and provides a useful complement to traditional pathway analyses. [ABSTRACT FROM AUTHOR]

Linear discriminant analysis has been widely used to characterize or separate multiple classes via linear combinations of features. However, the high dimensionality of features from modern biological experiments defies traditional discriminant analysis techniques. Possible interfeature correlations present additional challenges and are often underused in modelling. In this paper, by incorporating possible interfeature correlations, we propose a covariance-enhanced discriminant analysis method that simultaneously and consistently selects informative features and identifies the corresponding discriminable classes. Under mild regularity conditions, we show that the method can achieve consistent parameter estimation and model selection, and can attain an asymptotically optimal misclassification rate. Extensive simulations have verified the utility of the method, which we apply to a renal transplantation trial. [ABSTRACT FROM PUBLISHER]

Many statistical analysis procedures require a good estimator for a high-dimensional covariance matrix or its inverse, the precision matrix. When the precision matrix is banded, the Cholesky-based method often yields a good estimator of the precision matrix. One important aspect of this method is determination of the band size of the precision matrix. In practice, crossvalidation is commonly used; however, we show that crossvalidation not only is computationally intensive but can be very unstable. In this paper, we propose a new hypothesis testing procedure to determine the band size in high dimensions. Our proposed test statistic is shown to be asymptotically normal under the null hypothesis, and its theoretical power is studied. Numerical examples demonstrate the effectiveness of our testing procedure. [ABSTRACT FROM PUBLISHER]

In the modelling of longitudinal data from several groups, appropriate handling of the dependence structure is of central importance. Standard methods include specifying a single covariance matrix for all groups or independently estimating the covariance matrix for each group without regard to the others, but when these model assumptions are incorrect, these techniques can lead to biased mean effects or loss of efficiency, respectively. Thus, it is desirable to develop methods for simultaneously estimating the covariance matrix for each group that will borrow strength across groups in a way that is ultimately informed by the data. In addition, for several groups with covariance matrices of even medium dimension, it is difficult to manually select a single best parametric model among the huge number of possibilities given by incorporating structural zeros and/or commonality of individual parameters across groups. In this paper we develop a family of nonparametric priors using the matrix stick-breaking process of Dunson et al. (2008) that seeks to accomplish this task by parameterizing the covariance matrices in terms of their modified Cholesky decompositions (Pourahmadi, 1999). We establish some theoretical properties of these priors, examine their effectiveness via a simulation study, and illustrate the priors using data from a longitudinal clinical trial. [ABSTRACT FROM PUBLISHER]

Inferences related to the second-order properties of functional data, as expressed by covariance structure, can become unreliable when the data are non-Gaussian or contain unusual observations. In the functional setting, it is often difficult to identify atypical observations, as their distinguishing characteristics can be manifold but subtle. In this paper, we introduce the notion of a dispersion operator, investigate its use in probing the second-order structure of functional data, and develop a test for comparing the second-order characteristics of two functional samples that is resistant to atypical observations and departures from normality. The proposed test is a regularized M-test based on a spectrally truncated version of the Hilbert–Schmidt norm of a score operator defined via the dispersion operator. We derive the asymptotic distribution of the test statistic, investigate the behaviour of the test in a simulation study and illustrate the method on a structural biology dataset. [ABSTRACT FROM PUBLISHER]

The mean residual life provides the remaining life expectancy of a subject who has survived to a certain time-point. When covariates are present, regression models are needed to study the association between the mean residual life function and potential regression covariates. In this paper, we propose a flexible class of semiparametric mean residual life models where some effects may be time-varying and some may be constant over time. In the presence of right censoring, we use the inverse probability of censoring weighting approach and develop inference procedures for estimating the model parameters. In addition, we provide graphical and numerical methods for model checking and tests for examining whether or not the covariate effects vary with time. Asymptotic and finite sample properties of the proposed estimators are established and the approach is applied to real life datasets collected from clinical trials. [ABSTRACT FROM PUBLISHER]

Covariate measurement error and missing responses are typical features in longitudinal data analysis. There has been extensive research on either covariate measurement error or missing responses, but relatively little work has been done to address both simultaneously. In this paper, we propose a simple method for the marginal analysis of longitudinal data with time-varying covariates, some of which are measured with error, while the response is subject to missingness. Our method has a number of appealing properties: assumptions on the model are minimal, with none needed about the distribution of the mismeasured covariate; implementation is straightforward and its applicability is broad. We provide both theoretical justification and numerical results. [ABSTRACT FROM PUBLISHER]

The present paper deals with sequential designs intended to balance the allocations of two competing treatments in the presence of prognostic factors. After giving a theoretical framework on the optimality of balanced designs that can arise when covariates are taken into account, we propose a new family of covariate-adaptive randomized designs that represents higher order approximation to balance treatments, both globally and also across covariates. We derive the theoretical properties of the suggested designs in terms of loss of precision and predictability. The performance of this proposal is illustrated through a simulation study and compared with those of other procedures suggested in the literature. [ABSTRACT FROM PUBLISHER]

Gneiting (2002) proposed a nonseparable covariance model for spatial-temporal data. In the present paper we show that in certain circumstances his model possesses a counterintuitive dimple. In some cases, the magnitude of the dimple can be nontrivial. [ABSTRACT FROM PUBLISHER]

In standard regression analyses of clustered data, one typically assumes that the expected value of the response is independent of cluster size. However, this is often false. For example, in studies of surgical interventions, investigators have frequently found surgery volume and outcomes to be related to the skill level of the surgeons. This paper examines the effect of ignoring response-dependent, informative, cluster sizes on standard analytical methods such as mixed-effects models and conditional likelihood methods using analytic calculations, simulation studies and an example from a study of periodontal disease. We consider the case in which cluster sizes and responses share random effects which we assume to be independent of the covariates. Our focus is on maximum likelihood methods that ignore informative cluster sizes, and we show that they exhibit little bias in estimating covariate effects that are uncorrelated with the random effects associated with cluster sizes. However, estimation of covariate effects that are associated with the random effects can be biased. In particular, for models with random intercepts only, ignoring informative cluster sizes can yield biased estimators of the intercept but little bias in estimation of all covariate effects. [ABSTRACT FROM PUBLISHER]

Gaussian graphical models explore dependence relationships between random variables, through the estimation of the corresponding inverse covariance matrices. In this paper we develop an estimator for such models appropriate for data from several graphical models that share the same variables and some of the dependence structure. In this setting, estimating a single graphical model would mask the underlying heterogeneity, while estimating separate models for each category does not take advantage of the common structure. We propose a method that jointly estimates the graphical models corresponding to the different categories present in the data, aiming to preserve the common structure, while allowing for differences between the categories. This is achieved through a hierarchical penalty that targets the removal of common zeros in the inverse covariance matrices across categories. We establish the asymptotic consistency and sparsity of the proposed estimator in the high-dimensional case, and illustrate its performance on a number of simulated networks. An application to learning semantic connections between terms from webpages collected from computer science departments is included. [ABSTRACT FROM PUBLISHER]

In this paper we propose a new regression interpretation of the Cholesky factor of the covariance matrix, as opposed to the well-known regression interpretation of the Cholesky factor of the inverse covariance, which leads to a new class of regularized covariance estimators suitable for high-dimensional problems. Regularizing the Cholesky factor of the covariance via this regression interpretation always results in a positive definite estimator. In particular, one can obtain a positive definite banded estimator of the covariance matrix at the same computational cost as the popular banded estimator of Bickel & Levina (2008b), which is not guaranteed to be positive definite. We also establish theoretical connections between banding Cholesky factors of the covariance matrix and its inverse and constrained maximum likelihood estimation under the banding constraint, and compare the numerical performance of several methods in simulations and on a sonar data example. [ABSTRACT FROM PUBLISHER]

This paper develops smooth-threshold estimating equations that can automatically eliminate irrelevant parameters by setting them as zero. The resulting estimator enjoys the oracle property in the sense of Fan & Li (2001), even in estimators for which the covariance assumption of Wang & Leng (2007) is violated, such as the Buckley–James estimator. Furthermore, the estimator can be obtained without solving a convex optimization problem. A bic-type criterion for tuning parameter selection is also proposed. It is shown that the criterion achieves consistent model selection. A numerical study confirms the performance of the method. [ABSTRACT FROM PUBLISHER]

When used for modelling longitudinal data generalised estimating equations specify a working structure for the within-subject covariance matrices, aiming to produce efficient parameter estimators. However, misspecification of the working covariance structure may lead to a large loss of efficiency of the estimators of the mean parameters. In this paper we propose an approach for joint modelling of the mean and covariance structures of longitudinal data within the framework of generalised estimating equations. The resulting estimators for the mean and covariance parameters are shown to be consistent and asymptotically Normally distributed. Real data analysis and simulation studies show that the proposed approach yields e?cient estimators for both the mean and covariance parameters. [ABSTRACT FROM PUBLISHER]

Zhu & Hastie (2003) presented a general criterion for finding discriminant directions. To optimise their criterion, iterative methods are needed unless each class has a Gaussian distribution with a common covariance matrix. In this short paper, we present a slightly more general case where iterative methods can also be avoided. [ABSTRACT FROM PUBLISHER]

In the statistical literature, life expectancy is usually characterised by the mean residual life function. Regression models are thus needed to study the association between the mean residual life functions and their covariates. In this paper, we consider a linear mean residual life model and develop inference procedures in the presence of potential censoring. The new model and inference procedures are applied to the Stanford heart transplant data. Semiparametric efficiency calculations and information bounds are also considered. [ABSTRACT FROM AUTHOR]

In many applications, researchers are interested in estimating the mean of a multivariate normal random vector whose components are subject to order restrictions. Various authors have demonstrated that the likelihood-based methodology may perform poorly under certain conditions for such problems. The problem is much harder when the underlying covariance matrix is nondiagonal. In this paper a simple iterative algorithm is introduced that can be used for estimating the mean of a multivariate normal population when the components are subject to any order restriction. The proposed methodology is illustrated through an application to human reproductive hormone data. [ABSTRACT FROM AUTHOR]

For submodels of an exponential family, we consider likelihood ratio tests for hypotheses that render some parameters nonidentifiable. First, we establish the asymptotic equivalence between the likelihood ratio test and the score test. Secondly, the score-test representation is used to derive the asymptotic distribution of the likelihood ratio test. These results are derived for general submodels of an exponential family without assuming compactness of the parameter space. We then exemplify the results on a class of multivariate normal models, where null hypotheses concerning the covariance structure lead to loss of identifiability of a parameter. Our motivating problem throughout the paper is to test a random intercepts model against an alternative covariance structure allowing for serial correlation. [ABSTRACT FROM AUTHOR]

We study the profile‐kernel and backfitting methods in partially linear models for clustered/longitudinal data. For independent data, despite the potential root‐n inconsistency of the backfitting estimator noted by Rice (1986), the two estimators have the same asymptotic variance matrix, as shown by Opsomer & Ruppert (1999). In this paper, theoretical comparisons of the two estimators for multivariate responses are investigated. We show that, for correlated data, backfitting often produces a larger asymptotic variance than the profile‐kernel method; that is, for clustered data, in addition to its bias problem, the backfitting estimator does not have the same asymptotic efficiency as the profile‐kernel estimator. Consequently, the common practice of using the backfitting method to compute profile‐kernel estimates is no longer advised. We illustrate this in detail by following Zeger & Diggle (1994) and Lin & Carroll (2001) with a working independence covariance structure for nonparametric estimation and a correlated covariance structure for parametric estimation. Numerical performance of the two estimators is investigated through a simulation study. Their application to an ophthalmology dataset is also described. [ABSTRACT FROM AUTHOR]

Traditionally the neighbourhood size k in the k‐nearest‐neighbour algorithm is either fixed at the first nearest neighbour or is selected on the basis of a crossvalidation study. In this paper we present an alternative approach that develops the k‐nearest‐neighbour algorithm using likelihood‐based inference. Our method takes the form of a generalised linear regression on a set of k‐nearest‐neighbour autocovariates. By defining the k‐nearest‐neighbour algorithm in this way we are able to extend the method to accommodate the original predictor variables as possible linear effects as well as allowing for the inclusion of multiple nearest‐neighbour terms. The choice of the final model proceeds via a stepwise regression procedure. It is shown that our method incorporates a conventional generalised linear model and a conventional k‐nearest‐neighbour algorithm as special cases. Empirical results suggest that the method out‐performs the standard k‐nearest‐neighbour method in terms of misclassification rate on a wide variety of datasets. [ABSTRACT FROM PUBLISHER]

In many biomedical studies, longitudinal count data comprise repeated responses and a set of multidimensional covariates for a large number of individuals. When the response variable in such models is subject to overdispersion, the overdispersion parameter influences the marginal variance. In such cases, the overdispersion parameter plays a significant role in efficient estimation of the regression parameters. This raises the need for joint estimation of the regression parameters and the overdispersion parameter, the longitudinal correlations being nuisance parameters. In this paper, we develop a generalised estimating equations approach based on a general autocorrelation structure for the repeated overdispersed data. The asymptotic properties of the estimators of the main parameters are discussed, and the estimation methodology is illustrated by analysing data on epileptic seizure counts. [ABSTRACT FROM PUBLISHER]

Logistic regression is often used to find a linear combination of covariates which best discriminates between two groups or populations. The ROC, receiver operating characteristic, curve is a good way of assessing the performance of the resulting score, but using the same data both to fit the score and to calculate its ROC leads to an over‐optimistic estimate of the performance which the score would give if it were to be validated on a sample of future cases. The paper studies the extent of this overestimation, and suggests a shrinkage correction for the ROC curve itself and for the area under the curve. The correction is consistent with Efron's formula for the bias in the error rate of a binary prediction rule. Two medical examples are discussed. [ABSTRACT FROM PUBLISHER]

Grenander & Miller (1994) describe a model for representing amorphous two-dimensional objects with no obvious landmark. Each object is represented by n vertices around its perimeter, and is described by deforming an n-sided regular polygon using edge transformations. A multivariate normal distribution with a block circulant covariance matrix is used to model these edge transformations. The purpose of this paper is to describe in detail the statistical properties of this multivariate model and the eigenstructure of the covariance matrix. Various special cases of the model are considered, including articulated models and conditional Markov random field models. We consider maximum likelihood based inference and the model is applied to some datasets to explore shape variability. [ABSTRACT FROM PUBLISHER]

The positive-definiteness constraint is the most awkward stumbling block in modelling the covariance matrix. Pourahmadi's (1999) unconstrained parameterisation models covariance using covariates in a similar manner to mean modelling in generalised linear models. The new covariance parameters have statistical interpretation as the regression coefficients and logarithms of prediction error variances corresponding to regressing a response on its predecessors. In this paper, the maximum likelihood estimators of the parameters of a generalised linear model for the covariance matrix, their consistency and their asymptotic normality are studied when the observations are normally distributed. These results along with the likelihood ratio test and penalised likelihood criteria such as BIC for model and variable selection are illustrated using a real dataset. [ABSTRACT FROM PUBLISHER]

The canonical parameter of a covariance selection model is the inverse covariance matrix [sum ]-1 whose zero pattern gives the conditional independence structure characterising the model. In this paper we consider the upper triangular matrix Φ obtained by the Cholesky decomposition [sum ]-1 = ΦTΦ. This provides an interesting alternative parameterisation of decomposable models since its upper triangle has the same zero structure as [sum ]-1 and its elements have an interpretation as parameters of certain conditional distributions. For a distribution for [sum ], the strong hyper-Markov property is shown to be characterised by the mutual independence of the rows of Φ. This is further used to generalise to the hyper inverse Wishart distribution some well-known properties of the inverse Wishart distribution. In particular we show that a hyper inverse Wishart matrix can be decomposed into independent normal and chi-squared random variables, and we describe a family of transformations under which the family of hyper inverse Wishart distributions is closed. [ABSTRACT FROM PUBLISHER]

In previous work we have studied a nonparametric additive time-varying regression model for longitudinal data recorded at irregular intervals. The model allows the influence of each covariate to vary separately with time. For small datasets, however, only a limited number of covariates may be handled in this way. In this paper, we introduced a semiparametric regression model for longitudinal data. The influence of some of the covariates varies nonparametrically with time while the effect of the remaining covariates are constant. No smoothing is necessary in the estimation of the parametric terms of the model. Asymptotics are derived using martingale techniques for the cumulative regression functions, which are much easier to estimate and study than the regression functions themselves. The approach is applied to longitudinal data from the Copenhagen Study Group for Liver Diseases (Schlichting et al., 1983). [ABSTRACT FROM PUBLISHER]

When a study shows an association between a treatment and a response, before concluding that there is a causal relationship it is useful to assess whether or not an unobserved variable, U, might explain the observed association. Sensitivity analysis clarifies the properties U must have in terms of its relationship to the response and its imbalance in the groups being compared. A substantial literature has investigated the imbalance or association between U and group assignment. This paper develops a dual sensitivity analysis for matched pairs, focusing on the strength of the relationship between U and the response required to reduce an observed association to non-significance. A third simultaneous form of sensitivity analysis models both relationships between U and treatment assignment and U and response. The simultaneous form allows one to compare results from the sensitivity analysis to subject matter knowledge about both relationships. The methods are illustrated by several examples. [ABSTRACT FROM AUTHOR]

In this paper we use posterior covariance matrices to study the strong consistency of Bayes estimators in stochastic regression models under various assumptions on the stochastic regressors. The random errors are assumed to form a martingale difference sequence. Several results are obtained using a recursion satisfied by the sequence of posterior covariance matrices. These results suggest that the posterior covariance matrix is a useful tool in studying strong consistency problems in stochastic regression models. Three examples from sequential design and adaptive control are discussed. [ABSTRACT FROM AUTHOR]

This paper presents methods to handle missing covariates when the quasi-likelihood equations for the complete data are available. Our suggestion is to replace the functions of the missing data appearing in the quasi-likelihood equation with their conditional means given the observed data or with unbiased predictors, so that the resulting equation is unbiased. We focus on two models. One is a random effects model for count data, where random effects are treated as missing covariates. The other is the overdispersed binomial regression model with partially missing covariates. We also investigate the efficiency of the proposed estimates relative to the maximum likelihood estimators. [ABSTRACT FROM AUTHOR]

This paper considers the use of an auxiliary variable X* to estimate quantiles of a test statistic X; X* may be an asymptotic expansion of X, or a simplified version which ignores some of the covariance structure. The proposed estimator involves three stages. First a large sample is drawn, and X* is evaluated. Then a first subsample is drawn, X and X* are evaluated and the conditional distribution of X given X* is estimated. Then a second subsample is drawn with weighting which is optimised for this conditional distribution, and for a particular quantile. From this subsample, an importance sampling estimator of the quantile is obtained. The resulting estimator is shown to have substantially lower mean squared error than the conventional estimator, and to be reasonably robust both to errors in the model for the conditional distribution and to the quantile assumed for the optimisation. An example in genetics is given. [ABSTRACT FROM AUTHOR]

The nonparametric estimation of covariance lies at the heart of functional data analysis, whether for curve or surface-valued data. The case of a two-dimensional domain poses both statistical and computational challenges, which are typically alleviated by assuming separability. However, separability is often questionable, sometimes even demonstrably inadequate. We propose a framework for the analysis of covariance operators of random surfaces that generalizes separability while retaining its major advantages. Our approach is based on the expansion of the covariance into a series of separable terms. The expansion is valid for any covariance over a two-dimensional domain. Leveraging the key notion of the partial inner product, we generalize the power iteration method to general Hilbert spaces, and show how the aforementioned expansion can be efficiently constructed in practice at the level of the surface observations. Truncation of the expansion and retention of the leading terms automatically induces a nonparametric estimator of the covariance, whose parsimony is dictated by the truncation level. The resulting estimator can be calculated, stored and manipulated with little computational overhead relative to separability. Consistency and rates of convergence are derived under mild regularity assumptions, illustrating the trade-off between bias and variance regulated by the truncation level. The merits and practical performance of the proposed methodology are demonstrated in a comprehensive simulation study. [ABSTRACT FROM AUTHOR]

A centred Gaussian model that is Markov with respect to an undirected graph G is characterised by the parameter set of its precision matrices which is the cone M+(G) of positive definite matrices with entries corresponding to the missing edges of G constrained to be equal to zero. In a Bayesian framework, the conjugate family for the precision parameter is the distribution with Wishart density with respect to the Lebesgue measure restricted to M+(G). We call this distribution the G-Wishart. When G is nondecomposable, the normalising constant of the G-Wishart cannot be computed in closed form. In this paper, we give a simple Monte Carlo method for computing this normalising constant. The main feature of our method is that the sampling distribution is exact and consists of a product of independent univariate standard normal and chi-squared distributions that can be read off the graph G. Computing this normalising constant is necessary for obtaining the posterior distribution of G or the marginal likelihood of the corresponding graphical Gaussian model. Our method also gives a way of sampling from the posterior distribution of the precision matrix. [ABSTRACT FROM AUTHOR]

Model misspecification is discussed for the analysis of censored survival data with the Cox regression model. The maximum partial likelihood estimator for covariate effects has an asymptotic normal distribution under model misspecification (Lin & Wei, 1989). Situations where the assumed functional form for covariates is wrong are considered in a general framework in which we also derive the limit of Breslow's (1972) estimator. Furthermore, we give explicit expressions for the asymptotic variance of the maximum partial likelihood estimator. Numerical values for the errors of the maximum partial likelihood estimator and of the usual variance estimator are obtained for functional misspecification of a single relevant covariate and looking at the effect of a first factor, representing treatment or something similar, for functional wrong incorporation of auxiliary covariate measurements. In the latter situations we analyse the impact of dependence between the covariates. [ABSTRACT FROM PUBLISHER]

Liang & Zeger (1986) introduced methodology for the analysis of longitudinal data that provides an alternative to likelihood-based inference. They considered modelling the marginal means of the response follow-up measures, and proposed the use of unbiased estimating functions to handle inference. Here we wish to do the same for point or jump processes. We consider parametric models for the marginal means, and possibly the covariance structures, of processes that allow covariates. Inference is performed with unbiased estimating functions and robust variance estimates are provided. The optimal linear estimating function is presented in general. The special case of mixed Poisson processes is discussed in further detail with an asymptotic efficiency study and simulations. [ABSTRACT FROM AUTHOR]

SUMMARY Deletion diagnostics are proposed for generalised estimating equations. The diagnostics consider leverage and residuals to measure the influence of a subset of observations on the estimated regression parameters and on the estimated values of the linear predictor. Computational formulae are provided which correspond to the influence of a single observation and of an entire cluster of correlated observations. Additionally, diagnostics are given which approximate the effect of deletion of an arbitrary subset of observations under a model with general covariance structure and arbitrary link function, extending Proposition 3 of Christensen, Pearson & Johnson (1992). The proposed measures are applied to medical data. [ABSTRACT FROM AUTHOR]

SUMMARY The problem of estimating the number of binary responses that can be attributed to high levels of a continuous covariate is studied. The distribution of the covariate in the population is combined with the relative odds of response to produce a function, areas under which represent numbers of responses attributable to covariate values in that range. A combination of parametric and nonparametric smoothing is proposed. Consistent estimates of variance are developed as well as confidence bands for the attributable response function based on the bootstrap. The methods are illustrated on data relating fatal motor accident involvement to alcohol impairment of the driver. [ABSTRACT FROM AUTHOR]

Recent scientific and technological innovations have produced an abundance of potential markers that are being investigated for their use in disease screening and diagnosis. In evaluating these markers, it is often necessary to account for covariates associated with the marker of interest. Covariates may include subject characteristics, expertise of the test operator, test procedures or aspects of specimen handling. In this paper, we propose the covariate-adjusted receiver operating characteristic curve, a measure of covariate-adjusted classification accuracy. Nonparametric and semiparametric estimators are proposed, asymptotic distribution theory is provided and finite sample performance is investigated. For illustration we characterize the age-adjusted discriminatory accuracy of prostate-specific antigen as a biomarker for prostate cancer. [ABSTRACT FROM AUTHOR]

We consider nonparametric estimation of a covariance function on the unit square, given a sample of discretely observed fragments of functional data. When each sample path is observed only on a subinterval of length |$\delta<1$|⁠, one has no statistical information on the unknown covariance outside a |$\delta$| -band around the diagonal. The problem seems unidentifiable without parametric assumptions, but we show that nonparametric estimation is feasible under suitable smoothness and rank conditions on the unknown covariance. This remains true even when the observations are discrete, and we give precise deterministic conditions on how fine the observation grid needs to be relative to the rank and fragment length for identifiability to hold true. We show that our conditions translate the estimation problem to a low-rank matrix completion problem, construct a nonparametric estimator in this vein, and study its asymptotic properties. We illustrate the numerical performance of our method on real and simulated data. [ABSTRACT FROM AUTHOR]