Using Adjacent-category Logits Procedure for Estimating Receiver Operating Characteristic Surface

This article presents a semiparametric method for estimating receiver operating characteristic surface under density ratio model. The construction of the proposed method is based on the adjacent-category logit model and the empirical likelihood approach. A bootstrap approach for the VUS estimator inference is presented. In a simulation study, the proposed estimator is compared with the existing parametric and nonparametric estimators in terms of bias, standard error, and mean square error. Finally, a real data example and some discussions on the proposed method are provided.     

On the separability of assembly preferences

Analyses of assembly elections often assume that voters have well-defined preferences over candidates, even though preferences over assemblies are the natural analytic starting point. This candidate-based approach is usually justified by an assumption that preferences over assemblies are separable. We show, however, that if preferences over assemblies are themselves derived from underlying preferences over legislative or economic outcomes, then preferences over assemblies will not in general be separable. Received: 23 June 1997/Accepted: 3 March 1998     

Solution of problem 5/SP08: a birth–death process with polynomial transitions Proposed by Randall J. Swift, California State Polytechnic University, Pomona, USA

Problem Section

Solution of problem 5/SP08: a birth–death process with polynomial transitions Proposed by Randall J. Swift, California State Polytechnic University, Pomona, USA     

Making sense of political contention: Approaches, substantive issues, and interpretation in social movement analysis

This review essay discusses and compares three books that cover aspects of political contention and social movements. The books are Freedom is an Endless Meeting: Democracy in American Social Movements by Francesca Polletta (Chicago: The University of Chicago, 2002); Making Sense of Social Movements by Nick Crossley (Buckingham: Open University Press, 2002); and Methods of Social Movement Research, edited by Bert Klandermans and Suzanne Staggenborg (Minneapolis: The University of Minnesota Press, 2002).     

Factors associated with long-term benzodiazepine use among elderly women and men in Quebec

The goal of this study was to compare factors associated with long-term benzodiazepine use by elderly women and men (n = 1701) who participated in the Quebec Health Survey (QHS). Data from the 1998 QHS were linked with data from the administrative files of the Régie de l'assurance maladie du Québec. Results showed that elderly women were more at risk than men for long-term benzodiazepine use. Results of the multivariate logistic regression did not show a significant difference between women and men on any of the risk factors studied. Other factors such as elderly and physician attitudes deserve further study to explain differences in long-term benzodiazepine use between elderly women and men.     

Optimal designs based on exact confidence regions for parameter estimation of a nonlinear regression model

This paper is concerned with the proposal of optimality criteria, referred to as X  - and X^′$X^{\prime }$-optimality criteria, and the construction of X  - and X^′$X^{\prime }$-optimal designs, for nonlinear regression models. These optimal designs aim at improving the estimation of parameters of this class of models. The principle of these criteria is the minimization, with respect to the design, of the expected volume of a particular exact parametric confidence region. In this paper we give detailed definitions, properties, and computation methods of X  - and X^′$X^{\prime }$-optimal designs. We also compare these designs with the classic local D-optimal designs, with regard to robustness and efficiency, for two very well-known academic models (Box–Lucas and Michaelis–Menten models).