Journal of the American Statistical Association

Journal covers statistical science and its applications, theory, and methods in economic, social, physical, engineering and health sciences.

Articles from Vol. 85, No. 410, June

A Multivariate Generalization of Quantile-Quantile Plots
1. INTRODUCTION The univariate quantile--quantile (Q-Q) plot is among the most widely used graphical techniques for examining the distributional shape of univariate data. It may also be the most powerful. It was apparently first introduced by Hazen...
A Multivariate Signed-Rank Test for the One-Sample Location Problem
1. INTRODUCTION Assume that [Y.sub.1], ..., [Y.sub.[eta]] are iid as Y = ([Y.sub.1], ..., [Y.sub.p])', where Y is from an absolutely continuous p-variate population with median [theta] =([[theta].sub.1], ..., [[theta].sub.p] = (Med([Y.sub.1]), ...,...
An Application of the Seasonal Fractionally Differenced Model to the Monetary Aggregates
1. INTRODUCTION In this article I examine the monetary aggregates M1, M2, and M3, with an extension of the fractionally differenced model of Granger and Joyeux (1980), Porter-Hudak (1982), and Geweke and Porter-Hudak (1983), the seasonal fractionally...
A Simulation Study of the Analysis of Sets of 2 X 2 Contingency Tables under Cluster Sampling: Estimation of a Common Odds Ratio
1. INTRODUCTION AND NOTATION Previous simulation studies of the behavior of estimators of a common odds ratio in sets of contingency tables have been based on the assumption that data are independent. Hauck, Leahy, and Anderson (1980) and Hauck,...
Bootstrap Prediction Intervals for Autoregression
1. INTRODUCTION An important problem in the statistical analysis of time series is the following: Given a realization of a series up to time t, what can be said about the observation at time (t + k)? The methodology presented here is an application...
Bounded-Influence Rank Regression: A One-Step Estimator Based on Wilcoxon Scores
1. INTRODUCTION We consider a bounded-influence one-step rank estimator for the unknown parameter [beta], a p x 1 vector, in the linear regression model [y.sub.i] = [x.sub.i]'[beta] + [e.sub.i] where [y.sub.i] is the ith observation on the...
Breakdown Robustness of Tests
1. INTRODUCTION Many authors have proposed tests with desirable robustness properties, including Huber (1965), Rieder (1978, 1982), Schrader and Hettmansperger (1980), Beran (1981), Ronchetti (1982), and Hampel, Ronchetti, Rousseeuw, and Stahel...
Comparison of Linear Estimators Using Pitman's Measure of Closeness
1. INTRODUCTION Alternative methods to least squares estimation have received considerable attention by data analysts over the last several years. These pursuits have led to the development of such procedures as ridge regression, least absolute...
Components of Pearson's Phi-Squared Distance Measure for the K-Sample Problem
1. INTRODUCTION The problem of testing the equality of distributions for k [greater than or equal to] 2 populations occupies a fundamental role in the statistical literature. This article presents a general approach to this problem for continuous...
Confidence Curves in Nonlinear Regression
1. PRELUDE In this article we present a graphical alternative to Wald and likelihood-based confidence intervals for a component of the p x 1 parameter vector [theta] in the usual normal nonlinear regression model. We develop this graphical method...
Editors' Report for 1989
We are pleased to report that JASA has had another good year. Readers have probably noticed that several changes recommended in previous years by the editors, the ASA Publications Committee, and the ASA Board of Directors have been implemented in JASA...
Efficiencies of Interblock Rank Statistics for Repeated Measures Designs
1. INTRODUCTION A mature and elegant theory has been developed for linear models having a Gaussian error structure. This theory allows for testing of hypotheses involving main effects and interactions. It also includes multiple comparison procedures...
Estimating Fecundability from Data on Waiting Times to First Conception
1. INTRODUCTION An important problem in the demography of fertility is the estimation of the distribution of population fecundability or the estimation of the first few moments of the population distribution. Fecundability is the rate at which a...
Estimating the Electoral Consequences of Legislative Redistricting
1. INTRODUCTION State and national legislators in the United States are largely elected by plurality vote in individual geographic districts, whose boundaries are redrawn after every decennial census. In addition to ensuring equal populations in...
Evaluating Screening for the Early Detection and Treatment of Cancer without Using a Randomized Control Group
1. INTRODUCTION One approach to reducing the death rate from cancer is to increase the use of screening for the early detection and treatment of cancer, that is, examining asymptomatic people to find cancer earlier, followed by treatment for those...
Exact Inference for Contingency Tables with Ordered Categories
1. INTRODUCTION This article presents exact small-sample inferences for cross-classifications of ordinal variables. Let {[n.sub.ij]} denote cell counts in an r x c contingency table, and let {[m.sub.ij]} denote their expected values. Denote the...
Extension of the Stein Estimating Procedure through the Use of Estimating Functions
1. INTRODUCTION Often statisticians face the problem of estimating many parameters simultaneously. Interesting applications in the field of public health include (a) the simultaneous estimation of mortality or incidence rates of a disease for different...
Homophily and Social Distance in the Choice of Multiple Friends: An Analysis Based on Conditionally Symmetric Log-Bilinear Association Models
1. INTRODUCTION 1.1 An Overview of Literature and a Definition of Key Concepts This article is concerned with the analysis of friendship choice data from a sample of subjects having different numbers of friends. The analysis focuses on the modeling...
Inference from Coarse Data Via Multiple Imputation with Application to Age Heaping
1. INTRODUCTION 1.1 Deficient Data Many data sets are deficient in the sense that the variables are partially missing, rounded to only one digit, recorded with measurement errors, and so forth. The specific example that concerns us here involves...
Influence on Confidence Regions for Regression Coefficients in Generalized Linear Models
1. INTRODUCTION The aim of many influence diagnostics for linear regression and generalized linear models is to identify observations with disproportionately large effect on estimation of the regression coefficients. A related problem that has received...
Kernel Quantile Estimators
1. QUANTILE ESTIMATORS The estimation of population quantiles is of great interest when one is not prepared to assume a parametric form for the underlying distribution. In addition, quantiles often arise as the natural thing to estimate when the...
Modeling and Monitoring Biomedical Time Series
1. INTRODUCTION This article is concerned with the problems of modeling and monitoring time series that arise in clinical or research biomedical contexts, where the goal is the rapid on-line detection of events of clinical or biological interest,...
Modeling Time-Varying Dynamical Systems
1. INTRODUCTION Nonstationary time series models with deterministically varying paremeters were discussed by Farley, Hinich, and McGuire (1975), Melard and Kihem (1981), Bittanti (1986), and others. With respect to the random-coefficient approach,...
Models for Distributions on Permutations
1. INTRODUCTION Distributions on the set of permutations of the first k integers are used in nonparametric statistics (Mallows 1957), analysis of multicandidate elections (Upton and Brook 1975), and social choice theory (Luce 1959). In nonparametric...
Multinomial Runs Tests to Detect Clustering in Constrained Free Recall
1. INTRODUCTION To draw inferences about mental processes, psychologists often want to detect category structure in free recall. Typically, a list of randomly ordered words is presented to a subject. Unbeknownst to the subject, the experimenter...
Refining Bootstrap Simultaneous Confidence Sets
1. INTRODUCTION In constructing an ordinary confidence set, the first goal is controlling level. When constructing a simultaneous confidence set, the initial goals are two-fold: (a) controlling the overall level; (b) controlling the relative levels...
Sampling-Based Approaches to Calculating Marginal Densities
1. INTRODUCTION In relation to a collection of random variables, [U.ub.1], [U.sub.2], ..., [U.sub.k], suppose that either (a) for i = 1, ..., k, the conditional distributions [U.sub.1] | [U.sub.j] (j [not equal to] i) are available, perhaps having...
Sensitivity of Two-Sample Permutation Inferences in Observational Studies
1. INTRODUCTION AND NOTATION The use of permutation inferences in randomized experiments is sustained or warranted by the physical act of randomization, which creates the only probability distribution used in the inference (Fisher 1935, sec. 2)....
Signed-Rank Tests for Censored Matched Pairs
1. INTRODUCTION I consider the problem of testing whether ([X.sub.1i], [X.sub.2i]) has the same distribution as ([X.sub.2i], [X.sub.1i]) (i = 1, ..., n), where ([X.sub.1i] [X.sub.2i]) are iid nonnegative bivariate random vectors representing failure...
Sliding-Spans Diagnostics for Seasonal and Related Adjustments
1. INTRODUCTION The sliding-spans technique involves the comparison of the correlated seasonal adjustments of a given month's datum which are obtained by applying the adjustment procedure to a sequence of three or four overlapping spans of data,...
Small-Sample Confidence Intervals
1. INTRODUCTION The precise setting for this article is as follows. We have an iid sample [X.sub.1], ..., [X.sub.n] of m-dimensional observations, drawn from a population with distribution [F.sub.[eta]] involving an unknown p-dimensional parameter...
Statistical Densities, Cumulatives, Quantiles, and Power Obtained by S-System Differential Equations
1. INTRODUCTION Explicit or implicit formulas for densities, cumulatives, and quantiles of univariate statistical distributions are generally available. Tabulations can be made quite extensive (e.g., see Abramowitz and Stegun 1972; Miller 1981;...
Testing the Goodness of Fit of a Linear Model Via Nonparametric Regression Techniques
1. INTRODUCTION Diagnostic checks for departures from linearity are frequently employed by regression analysts when they fit a linear model to a set of data. It is important to detect such departures, since failure to do so can result in improper...
Testing the Mixture of Exponentials Hypothesis and Estimating the Mixing Distribution by the Method of Moments
1. INTRODUCTION The assumption that empirical distributions are mixtures of underlying exponential distributions is widely invoked in applied science and social science (Everitt and Hand 1981). In the context of duration analysis, the popularity...
Tests of Hypotheses in Overdispersed Poisson Regression and Other Quasi-Likelihood Models
1. INTRODUCTION Regression analysis of count data using maximum likelihood techniques based on the Poisson distribution (Frome, Kutner, and Beauchamp 1973) is increasingly used in toxicology (Stead, Hasselblad, Creason, and Claxton 1981), epidemiology...
The Accuracy of Approximate Intervals for a Binomial Parameter
1. INTRODUCTION Let X be the number of successes in n independent Bernoulli trials with common success probability p; that is, X is distributed as binomial with parameters (n, p). In this article, we are interested in constructing a confidence interval...
The Deterrent Effect of Capital Punishment: An Analysis of Daily Homicide Counts
1. INTRODUCTION Social scientists and legal scholars have been troubled for decades by whether capital punishment deters murder. Although researchers from several disciplines have used widely varying methodologies to test different aspects of the...
The Evaluation of Integrals of the Form [[Integral].Sub.-[Infinity].Sup.+[Infinity]] F(t)exp(-[T.Sup.2]) Dt: Application to Logistic-Normal Models
1. INTRODUCTION Logistic--normal models (Aitchinson and Shen 1980) arise naturally in a wide variety of statistical applications when outcomes are discrete. Two important areas of current interest in which such models appear are logistic regression...
The Maximal Smoothing Principle in Density Estimation
1. INTRODUCTION Nonparametric density estimates attempt to reconstruct the probability density from which a random sample has come, using the sample values and as few assumptions as possible about the density. These methods are smoothing operations...
The Relationship between the Length of the Base Period and Population Forecast Errors
1. INTRODUCTION The base period of a population forecast may be defined as the time period from which historical data are collected and statistical relationships are formulated to provide forecasts of future population values. A recent discussion...
The Relative Efficiency of Goodness-of-Fit Statistics in the Simple and Composite Hypothesis-Testing Problem
1. INTRODUCTION Let [X.sub.1], ..., [X.sub.n] be a sequence of iid random variables with an underlying continuous cdf F. We will consider the following hypotheses: [H.sub.0s] : G(x) = F(x; [[theta].sub.0]) versus [H.sub.1s]: G(x) [not equal to]...
Two-Sample Inference for Median Survival Times Based on One-Sample Procedures for Censored Survival Data
1. INTRODUCTION Nonparametric two-sample tests based on ranks, for example, median and Wilcoxon tests, have been extended by Gehan (1965), Efron (1967), Prentice (1978), and Brookmeyer and Crowley (1982a), among others, to accommodate censored survival...