ON SOME OPTIMALITY PROPERTIES OF FISHER-RAO SCORE FUNCTION IN TESTING AND ESTIMATION

The score function is associated with some optimality features in statistical inference. This review article looks on the central role of the score in testing and estimation. The maximization of the power in testing and the quest for efficiency in estimation lead to score as a guiding principle. In hypothesis testing, the locally most powerful test statistic is the score test or a transformation of it. In estimation, the optimal estimating function is the score. The same link can be made in the case of nuisance parameters: the optimal test function should have maximum correlation with the score of the parameter of primary interest. We complement this result by showing that the same criterion should be satisfied in the estimation problem as well.     

A Test to Detect Within-family Infectivity when the Whole Epidemic Process is Observed

An epidemic model for the spread of an infectious disease in a population of families is considered. The score test of the hypothesis that there is no higher infectivity between family members is constructed under the assumption that the epidemic process is observed continuously up to some time t . The score process is a martingale as a function of t and by letting the number of families tend to infinity, a central limit theorem for the process can be proved. The central limit theorem not only justifies a normal approximation of the test statistic—it also suggests a smaller variance estimator than expected.     

60年代调整后农村人民公社个人收入分配制度

个人收入分配制度是人民公社体制下重要的制度安排.本文从公社个人收入分配的构成和分配依据,以及与收入分配密切相关的公社劳动管理制度和工分制几个方面,说明了人民公社体制调整后到改革前的个人收入分配的实况,由此从一个侧面证明了以大包干为特征的分配制度创新,并最终取代公社平均主义分配制度所具有的历史必然性.     

Testing the Normality Assumption in the Tobit Model

This paper examines a number of statistics that have been proposed to test the normality assumption in the tobit (censored regression) model. It argues that a number of commonly proposed statistics can be interpreted as different versions of the Lagrange multiplier, or score, test for a common null hypothesis. This observation is useful in examining the Monte Carlo results presented in the paper. The Monte Carlo results suggest that the computational convenience of a number of statistics is obtained at the cost of poor finite sample performance under the null hypothesis.     

Young Fatherhood and Subsequent Disadvantage in the United Kingdom

Although there has been increasing attention to the importance of fathers and their relationships with their children, few studies have examined young parenthood and its consequences for fathers’ life chances. In recent years, this has begun to change, and research is examining, to a far greater extent, the experiences of young fathers. Using data from a cohort of British men born in 1970, this paper uses a propensity score–matching technique to compare the well‐being of 344 men who reported becoming fathers before the age of 22 with men from similar backgrounds who did not. The findings suggest that selection into young fatherhood is substantial but, for some outcomes, significant differences remain.     

基于数据仓库的大学生诚信评估系统设计

结合金融系统广泛采用的FICO个人信用评估思想,提出了大学生诚信评价的加减分模型,构建了大学生诚信评估系统(CESS)指标体系及其数据仓库,实现了诚信数据校内整合与共享,并分别以ASP和VB.NET为工具开发了Web及C/S两种风格的CESS系统,介绍了系统结构与功能设计,讨论了包括直接利用Analysis Services多维分析工具实现的丰富的信用查询与统计方式,通过实际数据测试表明,CESS系统能够较好地满足高校需要.     

MORE INFORMATIVE TESTING FOR BIVARIATE SYMMETRY

In testing for bivariate symmetry against arbitrary alternatives the well‐known test developed by Bowker in 1948 is shown to be a score test, and to have useful components. These components are asymptotically independent and asymptotically have the standard normal distribution. Moreover they assess particular pairs of cells for symmetry. These components can also be used in a data analytic manner to complement a test for bivariate symmetry against ordered alternatives.     

Marital disruption and economic well-being: a comparative analysis

Summary.  Though there is a considerable literature concerned with the economic consequences of marital breakdown, there is still substantial disagreement in terms of its magnitude. One of the major problems underlying this debate is how economic well-being is defined. We implement several measures of well-being of monetary and multidimensional nature by using data from the European Community Household Panel. Another issue in this literature concerns selection bias of divorcing couples. We tackle this issue by using a propensity score matching technique combined with a difference-in-differences estimator. Results confirm the importance of the definition of well-being. We find a strong gender bias when using monetary measures but a considerably lower bias, and for some countries non-existent, when using non-monetary indices.     

Third-grade retention and reading achievement in Texas: A nine year panel study

The academic performance of over 38,000 Texas students who failed the state’s 1994 reading test was examined through their sophomore year in high school. Propensity score matching resulted in strata with retained and promoted students of comparable observed characteristics. Reading scores were analyzed using a two-level hierarchical linear model. Same grade comparisons show that third graders failing the state-mandated reading test who repeated the grade consistently outperformed in later grades the socially promoted children who also failed the third grade test. Additional analyses indicate that alternative explanations for the findings such as omitted variables, regression to the mean, differential panel attrition and cohort effects are not supported. The results are consistent with findings from other recent studies which suggest that grade retention in third grade may help increase student achievement.     

An improved comorbidity summary score for measuring disease burden and predicting mortality with applications to two national cohorts

Abstract

Research involving administrative healthcare data to study patient outcomes requires the investigator to account for the patient’s disease burden in order to reduce the potential for biased results. Here we develop a comorbidity summary score based on variable importance measures derived from several statistical and machine learning methods and show it has superior predictive performance to the Elixhauser and Charlson indices when used to predict 1-year, 5-year, and 10-year mortality. We used two large Veterans Administration cohorts to develop and validate the summary score and compared predictive performance using the area under ROC curve (AUC) and the Brier score.