Direct products: a powerful tool for the analysis of balanced data

This paper focuses attention on the use of direct (Kronecker) product techniques in the study of general balanced models. Direct products provide a powerful tool which facilitates the derivation of certain properties of these models. This is clearly demonstrated in determining the distribution of sums of squares in a balanced mixed-effects model under normality assumptions. Further evidence of its usefulness is also indicated.     

Approximate confidence bounds for ratios of variance components in two-way unbalanced crossed models with interactions

Consider the general unbalanced two-factor crossed components-of-variance model with interaction given by Y_ijk: = μ+A_i: +B_j: + C_ij: +E_ijk: (i = 1,2, … a; j = 1,…,b; k = 1,…,.n_ij:=0) A_i:,B_j:, C_ij: and E_ijk: are independent unobservable random variables. Also A_i:sim; N(0,σ² _A),B_j: ～ N(0,σ² _B), C_ij:～N(0,s² _C:) and E_ijk:～N(0,s² _E:). In this paper approximate confidence bounds are obtained for ρ_A: = ρ² _A/² and ρ_B: = ρ² _B:/ρ² (where σ² = σ² _A:+ σ² _B+σ² _Cσ² _E) for special cases of the above model. The balanced incomplete block model is studied as a special case.     

Estimation of variance components in the mixed effects models: A comparison between analysis of variance and spectral decomposition

The mixed effects models with two variance components are often used to analyze longitudinal data. For these models, we compare two approaches to estimating the variance components, the analysis of variance approach and the spectral decomposition approach. We establish a necessary and sufficient condition for the two approaches to yield identical estimates, and some sufficient conditions for the superiority of one approach over the other, under the mean squared error criterion. Applications of the methods to circular models and longitudinal data are discussed. Furthermore, simulation results indicate that better estimates of variance components do not necessarily imply higher power of the tests or shorter confidence intervals.     

The power function and an approximation for testing variance components in the presence of interaction in two-way random effects models

In this paper we consider a two-way layout random-effects model with interaction proportional to the product of two random components and derive the distributions of the sums of squares and the F-ratio under the usual normality assumption for the random effects. Approximations to these distributions by finite series Laguerre polynomial expansions are then developed. Some numerical results are given to illustrate the applications of the theory.     

Estimation of missing data in analysis of covariance: A least-squares approach

Abstract

A method is proposed for the estimation of missing data in analysis of covariance models. This is based on obtaining an estimate of the missing observation that minimizes the error sum of squares. Specific derivation of this estimate is carried out for the one-factor analysis of covariance, and numerical examples are given to show the nature of the estimates produced. Parameter estimates of the imputed data are then compared with those of the incomplete data.     

Additional comparisons of designs to estimate variance components in a two-way classification model

In this paper, the research of Muse and Anderson is extended to include additional comparisons of designs, featuring planned unbalance, for the estimation of variance components in a two-way cross classification model. Their results are extended to Include the following: (i) a small sample study of the original off-diagonal (OD) design and (ii) an asymptotic maximum likelihood investigation of three modifica-tions of the balanced diagonal rectangles (BD) design and one modification of the 01) design to permit the estimation of row, column, interaction and error variance components. Also a general iterative least.     

An application of linear feature selection to estimation of proportions

Let X be a random n-vector whose density function is given by a mixture of known multivariate normal density functions where the corresponding mixture proportions (a priori probabilities) are unknown. We present a numerically tractable method for obtaining estimates of the mixture proportions based on the linear feature selection technique of Guseman, Peters and Walker (1975).     

An extension of the median test to analysis of variance

In a k-way analysis of variance model, the major concern is testing for main effects and for the presence of interaction between the factors. When the assumptions of normality and equal variances are satisfied, the appropriate test to use is the usual F-test for ANOVA. However, when the normality assumption is not satisfied then a robust or nonparametric test is needed to conduct the analysis. In this paper a nonparametric method based on cell counts is proposed. Each cell is divided into L subcells based on predetermined outpoints and the resulting frequencies are laid out in a contingency table. Then the Pearson x² and tne likelihood ratio tests are performed. A comparison with the classical ANOVA F-test indicates that the proposed method is preferable when the data comes from a thick-tailed highly skewed distribution.     

A bayesian approach to prediction in analysis of variance settings

This paper takes the results of Lindley and smith ( 1972 ) one step further, by finding the predictive distribution of an observation y^* whose distribution is normal, and centred at A^* ₁θ₁ We then apply this distribution to the case of prediction based on data obtained in one and two wau ANOVA situations. For Example, it turns out that for two way ANOVA with interaction, the predictive mean, (which we would use as the predictor) is a weighted combination of sample main effects and interaction effects     

Population Marginal Means in the Linear Model: An Alternative to Least Squares Means

The parameter concept in the term least squares mean is defined and given the more meaningful name population marginal mean; and its estimation is discussed.     

On the analysis of multivariate repeated measures designs

The relationship between the mixed-model analysis and multivariate approach to a repeated measures design with multiple responses is presented. It is shown that by taking the trace of the appropriate submatrix of the hypothesis (error) sums of squares and crossproducts (SSCP) matrix obtained from the multivariate approach, one can get the hypothesis (error) SSCP matrix for the mixed-model analysis. Thus, when analyzing data from a multivariate repeated measures design, it is advantageous to use the multivariate approach because the result of the mixed-model analysis can also be obtained without additional computation.     

Choice of the primary analysis in longitudinal clinical trials

Missing data, and the bias they can cause, are an almost ever‐present concern in clinical trials. The last observation carried forward (LOCF) approach has been frequently utilized to handle missing data in clinical trials, and is often specified in conjunction with analysis of variance (LOCF ANOVA) for the primary analysis. Considerable advances in statistical methodology, and in our ability to implement these methods, have been made in recent years. Likelihood‐based, mixed‐effects model approaches implemented under the missing at random (MAR) framework are now easy to implement, and are commonly used to analyse clinical trial data. Furthermore, such approaches are more robust to the biases from missing data, and provide better control of Type I and Type II errors than LOCF ANOVA. Empirical research and analytic proof have demonstrated that the behaviour of LOCF is uncertain, and in many situations it has not been conservative. Using LOCF as a composite measure of safety, tolerability and efficacy can lead to erroneous conclusions regarding the effectiveness of a drug. This approach also violates the fundamental basis of statistics as it involves testing an outcome that is not a physical parameter of the population, but rather a quantity that can be influenced by investigator behaviour, trial design, etc. Practice should shift away from using LOCF ANOVA as the primary analysis and focus on likelihood‐based, mixed‐effects model approaches developed under the MAR framework, with missing not at random methods used to assess robustness of the primary analysis. Copyright © 2004 John Wiley & Sons, Ltd.     

On the use of cross-validation to assess performance in multivariate prediction

We describe a Monte Carlo investigation of a number of variants of cross-validation for the assessment of performance of predictive models, including different values of k in leave-k-out cross-validation, and implementation either in a one-deep or a two-deep fashion. We assume an underlying linear model that is being fitted using either ridge regression or partial least squares, and vary a number of design factors such as sample size n relative to number of variables p, and error variance. The investigation encompasses both the non-singular (i.e. n > p) and the singular (i.e. n p) cases. The latter is now common in areas such as chemometrics but has as yet received little rigorous investigation. Results of the experiments enable us to reach some definite conclusions and to make some practical recommendations.     

Estimation of parameters in heavy-tailed distribution when its second order tail parameter is known

Estimating parameters in heavy-tailed distribution plays a central role in extreme value theory. It is well known that classical estimators based on the first order asymptotics such as the Hill, rank-based and QQ estimators are seriously biased under finer second order regular variation framework. To reduce the bias, many authors proposed the so-called second order reduced bias estimators for both first and second order tail parameters. In this work, estimation of parameters in heavy-tailed distributions are studied under the second order regular variation framework when the second order parameter in the distribution tail is known. This is motivated in large part by a recent work by the authors showing that the second order tail parameter is known for a large class of popular random difference equations (for example, ARCH models). The focus is on least squares estimators that generalize rank-based and QQ estimators. Though other possible estimators are also briefly discussed, the least squares estimators are most simple to use and perform best for finite samples in Monte Carlo simulations.     

An application of graph theory to a computer storage problem in residual analysis

The r test procedure for adequacy of the multiple regression model is extended for moderate sample sizes by approximating the moments of the distribution of the test statistic with those of the distribution of the Durbin and Watson (1950) upper-bounding variable. The computation of these moments is simplified by the application of some elementary graph theory. Recommendations con-cerning the selection of near-neighbor pairs are suggested in order to reduce computer storage and execution time.     

An application of nonparametric regression to missing data in large market surveys

Non-response (or missing data) is often encountered in large-scale surveys. To enable the behavioural analysis of these data sets, statistical treatments are commonly applied to complete or remove these data. However, the correctness of such procedures critically depends on the nature of the underlying missingness generation process. Clearly, the efficacy of applying either case deletion or imputation procedures rests on the unknown missingness generation mechanism. The contribution of this paper is twofold. The study is the first to propose a simple sequential method to attempt to identify the form of missingness. Second, the effectiveness of the tests is assessed by generating (experimentally) nine missing data sets by imposed MCAR, MAR and NMAR processes, with data removed.     

A Population Approach to Analysis of Variance Models

Several authors have contributed to what can now be considered a rather complete theory for analysis of variance in cases with orthogonal factors. By using this theory on an assumed basic reference population, the orthogonality concept gives a natural definition of independence between factors in the population. By looking upon the treated units in designed experiments as a formal sample from a future population about which we want to make inference, a natural parametrization of expectations and variances connected to such experiments arises. This approach seems to throw light upon several controversial questions in the theory of mixed models. Also, it gives a framework for discussing the choice of conditioning in models     

Evaluation of an analysis approach used to account for extra-variation in clustered categorical responses

This article presents the results of a simulation study investigating the performance of an approach developed by Miller and Landis (1991) for the analysis of clustered categorical responses. Evaluation of this “two-step” approach, which utilizes the method of moments to estimate the extra-variation pardmeters and subsequently incorporates these parameters into estimating equations for modelling the marginal expectations, is carried out in an experimental setting involving a comparison between two groups of observations. We assume that data for both groups are collected from each cluster and responses are measured on a three-point ordinal scale. The performance of the estimators used in both “steps” of the analysisis investigated and comparisons are made to an alternative analysismethod that ignores the clustering. The results indicate that in the chosen setting the test for a difference between groups generally operatbs at the nominal α=0.05 for 10 or more clusters and hasincreasing power with both an increasing number of clusters and an inrreasing treatment effect. These results provide a striking contrasc to those obtained from an improper analysis that ignores clustering.     

On the level probabilities for useful partially ordered alternatives in the analysis of variance

In the analysis of variance, we often encounter situations in which we want to test the null hypothesis of homogeneity of the normal means against various partially ordered alternative hypotheses. We study likelihood ratio tests for three useful types of alternatives: d-star, bipartite and broom tree. Especially, we give computational formulas for the level probabilities of the alternative types. The results permit us to obtain critical values for practical use.