Influential observations in the analysis of variance

The effect of influentia lob servations on t h e parameter estimates of ordinary l e a s t squares regression models has received considerable attentio n fn the last decade. However, very little attention has been given t o the problem of in fluent ia lobserva- tions in the analysis of variance . The purpose of t h i s paper is t o show by way of examples that influential observations can alter the conclusions of tests of hypotheses in the analysis of variance . Regression diagnostics for identif y in g both extreme points and outliers can be used to reveal potential data and design problems.     

Defective data in the analysis of variance

Several kinds of defective data are considered, the first being that of data completely missing. General expressions for block designs are given for the variances of treatment contrasts. Sometimes even a moderate loss of data will seriously increase an important variance. A similar defect arises when some plots receive the wrong treatment. Again the consequences can be serious though, if the correct form of analysis used, variances of contrasts may be much better estimated than if defective data are regarded as missing. Finally, the problem of the mixed-up plots is considered, the total value for certain plots being known but not the individual data. Formulae are presented for dealing with such a case. Again the correct form of analysis can sometimes retain useful information that would be lost if affected plots are regarded as missing. Finally some considerations are set out for dealing with defective data in general.     

Robust analysis of variance

SUMMARY For the c -sample location problem with equal and unequal variances, we compare the classical F -test and its robustified version-the Welch test-with some nonparametric counterparts defined for two-sided and one-sided ordered alternatives, such as trend and umbrella alternatives. A new rank test for long-tailed distributions is proposed. The comparison is referred to level alpha and power beta of the tests, and is carried out via Monte Carlo simulation, assuming short-, medium- and long-tailed as well as asymmetric distributions. It turns out that the Welch test is the best one in the case of unequal variances but in the case of equal variances special non-parametric tests are to prefer.     

Robust analysis of variance

The classes of R- and M-estimates contain practical robust alternatives to least squares estimation in linear models. These estimates form the basis for a robust analysis of variance. This inference procedure is described and its versatility demonstrated.     

Bayesian analysis of quasi-life tables

Quasi-life tables, in which the data arise from many concurrent, independent, discrete-time renewal processes, were defined by Baxter (1994, Biometrika 81:567–577), who outlined some methods for estimation. The processes are not observed individually; only the total numbers of renewals at each time point are observed. Crowder and Stephens (2003, Lifetime Data Anal 9:345–355) implemented a formal estimating-equation approach that invokes large-sample theory. However, these asymptotic methods fail to yield sensible estimates for smaller samples. In this paper, we implement a Bayesian analysis based on MCMC computation that works equally well for large and small sample sizes. We give three simulated examples, studying the Bayesian results, the impact of changing prior specification, and empirical properties of the Bayesian estimators of the lifetime distribution parameters. We also study the Baxter (1994, Biometrika 81:567–577) data, and uncover structure that has not been commented upon previously.     

Statistical evidence in contingency tables analysis

The likelihood ratio is used for measuring the strength of statistical evidence. The probability of observing strong misleading evidence along with that of observing weak evidence evaluate the performance of this measure. When the corresponding likelihood function is expressed in terms of a parametric statistical model that fails, the likelihood ratio retains its evidential value if the likelihood function is robust [Royall, R., Tsou, T.S., 2003. Interpreting statistical evidence by using imperfect models: robust adjusted likelihood functions. J. Roy. Statist. Soc. Ser. B 65, 391–404]. In this paper, we extend the theory of Royall and Tsou [2003. Interpreting statistical evidence by using imperfect models: robust adjusted likelihood functions. J. Roy. Statist. Soc., Ser. B 65, 391–404] to the case when the assumed working model is a characteristic model for two-way contingency tables (the model of independence, association and correlation models). We observe that association and correlation models are not equivalent in terms of statistical evidence. The association models are bounded by the maximum of the bump function while the correlation models are not.     

On the estimation of variance components in stability analysis

Chow and Shao (1989, 1991) indicated that the presence of batch-to-batch variation has an impact on the determination of drug shelf-life in stability studies. In this paper, we propose two unbiased estmators for batch-to-batch variation. The proposed estimators are compared in terms of their corresponding variances. An example concerning a stability study is discussed to illustrate the use of the proposed estimators.     

Counterexamples in unbalanced two-way analysis of variance

Recommended methods for analyzing unbalanced two-way data may be classified into two major categories:the parametric interpretation approach and the model comparison approach. Each approach has its advantages and its drawbacks. The main drawback of the parametric interpretation approach is non-orthogonality.For the model comparison approach the main drawback is the dependence of the hypothesis tested on the cell sizes. In this paper we provide examples to illustrate these drawbacks.     

Quantile dispersion graphs for analysis of variance estimates of variance components

SUMMARY The exact distribution of an analysis of variance estimator of a variance component is obtained by determining its quantiles on the basis of R. B. Davies' algorithm. A plot of these quantiles provides useful information concerning the efficiency of the estimator, including the extent to which it can be negative. Furthermore, the variability in the values of each quantile is assessed by varying the values of the variance components for the model under consideration. The maximum and minimum of such quantile values can then be determined. A plot of the maxima and minima for various selected quantiles produces the so-called 'quantile dispersion graphs'. These graphs can be used to provide a comprehensive picture of the quality of estimation obtained with a particular design. They also provide an effective graphical tool for comparing designs on the basis of their estimation capabilities.     

A note on the greenwood variance estimator in cohort life tables: Isolating the effect of one source of bias

A modification of the Greenwood variance estimator is defined and shown to be free of bias whenever its constitu­ent interval estimators are conditionally unbiased, given the sample size at the start of the interval. Using the modified estimator as a standard of comparison, the original Greenwood estimator is seen to have an intrinsic positive bias.Under­estimation of variances through the use of Greenwood's formula must be due to bias in the constituent interval estimators and/or, with fixed interval bounds, due to disregarding the random character of the total number of life table intervals to exhaustion of ttje sample. Some easy to prove properties of the modified and the original Greenwood estimators are stated that apply in the absence of censoring. A suggest­ion is made for reducing the bias of the interval variance estimators.     

Inference in smoothing spline analysis of variance

Summary. Smoothing spline analysis of variance decomposes a multivariate function into additive components. This decomposition not only provides an efficient way to model a multivariate function but also leads to meaningful inference by testing whether a certain component equals 0. No formal procedure is yet available to test such a hypothesis. We propose an asymptotic method based on the likelihood ratio to test whether a functional component is 0. This test allows us to choose an optimal model and to compare groups of curves. We first develop the general theory by exploiting the connection between mixed effects models and smoothing splines. We then apply this to compare two groups of curves and to select an optimal model in a two-dimensional problem. A small simulation is used to assess the finite sample performance of the likelihood ratio test.     

Tests of dimensionality in multivariate analysis of variance

Procedures for determining the dimensionality of the hyperplane containing the mean vectors in the m-group multivariate analysis of variance set-up are compared in a computer study.     

Some multiple decision problems in analysis of variance

In most practical situations to which the analysis of variance tests are applied, they do not supply the information that the experimenter aims at. If, for example, in one-way ANOVA the hypothesis is rejected in actual application of the F-test, the resulting conclusion that the true means θ₁,…,θ_k are not all equal, would by itself usually be insufficient to satisfy the experimenter. In fact his problems would begin at this stage. The experimenter may desire to select the “best” population or a subset of the “good” populations; he may like to rank the populations in order of “goodness” or he may like to draw some other inferences about the parameters of interest.

The extensive literature on selection and ranking procedures depends heavily on the use of independence between populations (block, treatments, etc.) in the analysis of variance. In practical applications, it is desirable to drop this assumption or independence and consider cases more general than the normal.

In the present paper, we derive a method to construct optimal (in some sense) selection procedures to select a nonempty subset of the k populations containing the best population as ranked in terms of θ_i’s which control the size of the selected subset and which maximizes the minimum average probability of selecting the best. We also consider the usual selection procedures in one-way ANOVA based on the generalized least squares estimates and apply the method to two-way layout case. Some examples are discussed and some results on comparisons with other procedures are also obtained.     

Testing for autocorreiation in nested analysis of variance

This paper considers the problem in which repeated measures are taken on each subject in an experimental design and the model errors follow a first order autoregressive process. The appropriate form of the Durbin-Watson test statistic for autocorrelated errors is presented. Critical values for the test statistic are reported.     

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.     

Central limit theorems for four new types of U-designs

Orthogonal array (OA)-based Latin hypercube designs, also called U-designs, have been popularly adopted in designing a computer experiment. Nested U-designs, sliced U-designs, strong OA-based U-designs and correlation controlled U-designs are four types of extensions of U-designs for different applications in computer experiments. Their elaborate multi-layer structure or multi-dimensional uniformity, which makes them desirable for different applications, brings difficulty in analysing the related statistical properties. In this paper, we derive central limit theorems for these four types of designs by introducing a newly constructed discrete function. It is shown that the means of the four samples generated from these four types of designs asymptotically follow the same normal distribution. These results are useful in assessing the confidence intervals of the gross mean. Two examples are presented to illustrate the closeness of the simulated density plots to the corresponding normal distributions.     

Multivariate analysis of variance for functional data

Functional data are being observed frequently in many scientific fields, and therefore most of the standard statistical methods are being adapted for functional data. The multivariate analysis of variance problem for functional data is considered. It seems to be of practical interest similarly as the one-way analysis of variance for such data. For the MANOVA problem for multivariate functional data, we propose permutation tests based on a basis function representation and tests based on random projections. Their performance is examined in comprehensive simulation studies, which provide an idea of the size control and power of the tests and identify differences between them. The simulation experiments are based on artificial data and real labeled multivariate time series data found in the literature. The results suggest that the studied testing procedures can detect small differences between vectors of curves even with small sample sizes. Illustrative real data examples of the use of the proposed testing procedures in practice are also presented.     

A nonparametric method for analysis of variance

A nonparametric method for analyzing analysis of variance models is introduced which is highly resistant to outliers, computationally simple, and comprehensible to anyone with a rudimentary knowledge of classical analysis of variance. The methodology is based on Mood's median test and is highly useful as an exploratory technique.