Survival Convergence and the Preceding Mortality Crossover for Two Population Subgroups

In this paper, we present and develop the argument that if the survival functions for two population subgroups converge in later life, a mortality crossover must precede the occurrence of this convergence. Specifically, two survival curves, S ₁(x) and S ₂(x), associated with two distinct population subgroups, G₁ and G₂, tend to converge before all members die out, as often observed and anticipated. This convergence leads to an increased mortality acceleration for the “advantaged” group, and eventually fosters the occurrence of a mortality crossover. We present a mathematical proof for this relationship and offer several explanations for the mechanisms involved in the process of survival convergence and the preceding mortality crossover. This new presentation demonstrates that mortality crossover is a highly observable demographic event given the trend of survival convergence in later life.     

Three faces of international inequality

Recent studies of international inequality have focused mostly on the trend in international income inequality. This article extends the analysis of international inequality to also include inequalities in education and health. Analyses of time-series data for more than 100 countries show that international income inequality declined from 1980 to 2003 as several large, poor Asian countries outpaced many Western countries in national income growth. By contrast, international health inequality followed a U-shaped trend, falling in the 1980s before rising in the 1990s. The turnaround in health inequality coincides with a trend of declining life expectancy in sub-Saharan Africa. International educational inequality experienced the sharpest recent decline, spurred by the global expansion of formal schooling. These findings confirm that there is more to international inequality than income inequality alone and suggest that patterns of inequality in the current era of globalization are likely more complex than many leading theories suggest.     

为中国人民服务

我来自加拿大安大略省的多伦多市,相信大家对这个地方并不陌生,它是大山的家乡,白求恩医生也在此出生并接受医疗培训。我所在的约克大学就有一个本科生学院是以白求恩医生的名字命名的。白求恩医生的事迹感召着我们．我们每一位获奖外国专家都将永远为我们有机会“为人民服务”,为中国人民服务而心存感激,这是我们无上的荣耀。     

USING DUMMY VARIABLES TO CHECK FOR ROUNDING ERROR IN COMPUTERIZED REGRESSION PROGRAMS*

After a brief review of the role of dummy variables in regression analysis and the current state-of-the art in rounding/truncation error detection in computerized least squares programs, this paper presents a theorem that can be used to detect this type of error whenever an analyst is running a regression program that has one (or more) dummy variables as independent variables.     

REGRESSION METHODOLOGY WITH GROSS OBSERVATION ERRORS IN THE EXPLANATORY VARIABLES***

The robustness of linear programming regression estimators is examined where the disturbance terms are normally distributed and there are observation errors in the explanatory variables. These errors are occasional gross biases between one set of observations and another. The simulation of short series data offers preliminary evidence that when these biases have a non-zero mean, MSAE estimation is more robust than least squares.     

THE EFFECTS OF INTERDISCIPLINARY EDUCATIONAL FORMATS ON CONCEPT RETENTION AND PROBLEM–SOLVING ABILITY*

The purpose of this study was to examine the effects of different interdisciplinary problem-oriented formats on a student's knowledge of and retention of disciplinary concepts and principles, his application of these concepts and principles to disciplinary problems, and his use of a discipline in his analysis of a complex problem. Performances of students enrolled in the experimental sections were compared to those of students enrolled in standard lecture-discussion sections. Regression analysis was used to analyze students' performance in order to control the effects of differences in student backgrounds. It was found that altering the pedagogical format and reward system within the experimental sections had no significant effect upon student performances in the experimental, problem-oriented program. In general, this study indicates that students taught in the standard lecture-discussion format in which grades are determined by examination over course material retain concepts and principles, apply these concepts and principles to disciplinary problems, and integrate disciplinary concepts into their analysis of complex problems as well as, if not better than, students taught in an interdisciplinary, problem-solving format.     

ALL-BIAS DESIGNS FOR POLYNOMIAL SPLINE REGRESSION MODELS

Polynomial spline regression models of low degree have proved useful in modeling responses from designed experiments in science and engineering when simple polynomial models are inadequate. Where there is uncertainty in the number and location of the knots, or breakpoints, of the spline, then designs that minimize the systematic errors resulting from model misspecification may be appropriate. This paper gives a method for constructing such all‐bias designs for a single variable spline when the distinct knots in the assumed and true models come from some specified set. A class of designs is defined in terms of the inter‐knot intervals and sufficient conditions are obtained for a design within this class to be all‐bias under linear, quadratic and cubic spline models. An example of the construction of all‐bias designs is given.     

BAYESIAN SUBSET SELECTION AND MODEL AVERAGING USING A CENTRED AND DISPERSED PRIOR FOR THE ERROR VARIANCE

This article proposes a new data‐based prior distribution for the error variance in a Gaussian linear regression model, when the model is used for Bayesian variable selection and model averaging. For a given subset of variables in the model, this prior has a mode that is an unbiased estimator of the error variance but is suitably dispersed to make it uninformative relative to the marginal likelihood. The advantage of this empirical Bayes prior for the error variance is that it is centred and dispersed sensibly and avoids the arbitrary specification of hyperparameters. The performance of the new prior is compared to that of a prior proposed previously in the literature using several simulated examples and two loss functions. For each example our paper also reports results for the model that orthogonalizes the predictor variables before performing subset selection. A real example is also investigated. The empirical results suggest that for both the simulated and real data, the performance of the estimators based on the prior proposed in our article compares favourably with that of a prior used previously in the literature.