摘要:

Censored data arise naturally in industrial experiments whose observations are failure times or measurements with mixed continuous-categorical outcomes. When censored data are observed from a fractionated experiment, likelihood-based estimates quite often do not exist, especially when the opportunity for improvement is great. To circumvent this problem, we propose a Bayesian analysis strategy that has a straightforward implementation using the data augmentation and Monte Carlo EM algorithms. For nonregular designs with complex aliasing patterns, a modified analysis strategy is proposed that substantially reduces the computation needed for model selection. The proposed strategy is illustrated using data from three real industrial experiments. For these data sets, the analysis results are fairly insensitive to the specification of the prior.

ISSN:0162-1459    

DOI:10.1080/01621459.1995.10476538

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

Gross flows represent transition counts between a finite number of states for individuals in a population from one point in time to the next. Such flows are important for researchers and policy analysts; for example, gross labour flows for understanding labour market dynamics. Unfortunately, the observed flows are typically subject to classification error. As a result, the problem of adjusting observed flows for classification error has received considerable attention. Currently, three methods for adjustment of classification error are available. All these methods use the key assumption of independent classification errors (ICE) in conjunction with interview-reinterview data. We first give modifications to two of the methods that ensure that margins of the adjusted flow table agree with the published “stocks” without requiring a final margin adjustment. We then propose e-response contamination models and procedures for studying the robustness of ICE under different scenarios of departures from ICE applicable for all the methods. Our empirical results, based on data from the Canadian Labour Force Survey, show that for many scenarios the ICE assumption is fairly robust. There are, however, situations where this is not the case. We thus suggest that users apply the proposed procedures to check the robustness of ICE under scenarios relevant for their own applications. Finally, we provide valid chi-squared tests for modeling flow tables adjusted for classification error under ICE so that the adjusted flows could be further smoothed under the model. These tests are also illustrated using data from the Canadian Labour Force Survey.

ISSN:0162-1459    

DOI:10.1080/01621459.1995.10476539

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

This article presents a random-effects model for multivariate and grouped univariate ordered categorical data. The assumed family of distributions for the random effects adopts a wide variety of forms and shapes. The model's likelihood has a closed expression and can be computed without recourse to numerical integration or Gaussian quadrature. A bivariate and a grouped univariate example are used to illustrate the proposed model.

ISSN:0162-1459    

DOI:10.1080/01621459.1995.10476540

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

An important gap, between the classical mathematical theory and the practice and implementation of nonparametric curve estimation, is due to the fact that the usual norms on function spaces measure something different from what the eye can see visually in a graphical presentation. Mathematical error criteria that more closely follow “visual impression” are developed and analyzed from both graphical and mathematical viewpoints. Examples from wavelet regression and kernel density estimation are considered.

ISSN:0162-1459    

DOI:10.1080/01621459.1995.10476541

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

Nonparametric and semiparametric regression have been suggested as alternatives to parametric models for growth curve analysis. In this article we demonstrate that empirical Bayes estimation can be used to improve linear smoothing estimates when multiple curves are available.

ISSN:0162-1459    

DOI:10.1080/01621459.1995.10476542

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

Berk and Jones described a nonparametric likelihood test of uniformity with greater asymptotic Bahadur efficiency than any weighted Kolmogorov–Smirnov test at any alternative toU[0, 1]. We invert this test to form confidence bands for a distribution function using Noé's recursion. Nonparametric likelihood bands are narrower in the tails and wider in the center than Kolmogorov-Smirnov bands and are asymmetric about the empirical cumulative distribution function.

ISSN:0162-1459    

DOI:10.1080/01621459.1995.10476543

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

The transform-both-sides (TBS) regression model developed by Carroll and Ruppert is applicable when the relationship between the median response and the independent variables has been identified. Several different families of transformations, such as the Box-Cox power family, have been considered in the parametric approach to this model. In this article, we propose a nonparametric estimator of the transformation in the TBS model allowing general smooth monotonic transformations. Asymptotic properties of this estimator are discussed.

ISSN:0162-1459    

DOI:10.1080/01621459.1995.10476544

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

A data-based procedure is introduced for local bandwidth selection for kernel estimation of a regression function at a point. The estimated bandwidth is shown to be consistent and asymptotically normal as an estimator of the (asymptotic) optimal value for minimum mean square estimation. Simulation studies indicate satisfactory behavior of the new bandwidth estimator in finite samples. The findings are improvements over a global bandwidth estimator. The same methodology works for local linear regression and extends easily to weighted local polynomial fits.

ISSN:0162-1459    

DOI:10.1080/01621459.1995.10476545

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

We describe two approaches to instrumental variable estimation in binary regression measurement error models. The methods entail constructing approximate mean models for the binary response as a function of the measured predictor, the instrument, and any covariates in the model. Estimates are obtained by exploiting relationships between regression parameters, just as in linear instrumental variable estimation. In the course of deriving the approximate mean models, we obtain an alternative characterization of instrumental variable estimation in linear measurement error models.

ISSN:0162-1459    

DOI:10.1080/01621459.1995.10476546

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

This article proposes aconsistentnonparametric test of the hypothesis that the disturbance in a linear regression model is distributed symmetrically around zero. Simulation results show that the test has good size and power properties for sample sizes as small as 50. We illustrate the use of the test in a cross-country model of inflation and monetary growth.

ISSN:0162-1459    

DOI:10.1080/01621459.1995.10476547

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor