Maximum entropy sampling and optimal Bayesian experimental design

When Shannon entropy is used as a criterion in the optimal design of experiments, advantage can be taken of the classical identity representing the joint entropy of parameters and observations as the sum of the marginal entropy of the observations and the preposterior conditional entropy of the parameters. Following previous work in which this idea was used in spatial sampling, the method is applied to standard parameterized Bayesian optimal experimental design. Under suitable conditions, which include non-linear as well as linear regression models, it is shown in a few steps that maximizing the marginal entropy of the sample is equivalent to minimizing the preposterior entropy, the usual Bayesian criterion, thus avoiding the use of conditional distributions. It is shown using this marginal formulation that under normality assumptions every standard model which has a two-point prior distribution on the parameters gives an optimal design supported on a single point. Other results include a new asymptotic formula which applies as the error variance is large and bounds on support size.     

Classroom sex composition and first-grade school outcomes: The role of classroom behavior

Using nationally representative data from the Early Childhood Longitudinal Study-Kindergarten Cohort (N = 21,409; 10,452 girls and 10,957 boys; mean age = 7.24 years), the association between first-grade classroom sex composition (CSC), measured as the percentage of female students, and end of the year academic (reading, mathematics) and socio-emotional (externalizing problems, internalizing problems, self-control, interpersonal skills) outcomes was examined. Using multilevel modeling techniques and controlling for prior achievement levels, CSC was positively associated with children’s reading achievement at the end of first grade; students performed better in reading in classes with a higher percentage of female students. CSC was also associated with three of the socio-emotional outcomes; controlling for prior levels, students in classrooms with a higher percentage of girls had better self-control and interpersonal skills and fewer internalizing problems. Classroom behavior mediated the effects of CSC on reading achievement and the socio-emotional outcomes. Implications for the composition of first-grade classrooms are discussed.     

找准加快发展的突破口 推动阿勒泰经济社会又好又快发展

按照科学发展观的要求,从阿勒泰地区的实际出发,为了实现经济发展方式的根本转变,推动地区经济社会实现又好又快发展,应努力实现“新区建设、牧民定居、优势资源深加工”三大突破,以增加财政收入和农牧民人均纯收入为主线,努力做好“工业强区、投资拉动、项目先行、引大扶强”等重点工作。     

Residualization is not the answer: Rethinking how to address multicollinearity

Here I show that a commonly used procedure to address problems stemming from collinearity and multicollinearity among independent variables in regression analysis, “residualization”, leads to biased coefficient and standard error estimates and does not address the fundamental problem of collinearity, which is a lack of information. I demonstrate this using visual representations of collinearity, hypothetical experimental designs, and analyses of both artificial and real world data. I conclude by noting the importance of examining methodological practices to ensure that their validity can be established based on rational criteria.     

Portmanteau test for conditional heteroscedasticity, using ranks of squared residuals

Squared residual autocorrelations have been found useful in detecting departures from linearity in time series models. This is especially the case with data exhibiting heterogeneity in variances. A rank test is proposed which is much more robust than its parametric counterpart.     

Assessment of influence in non-linear measurement error models

We consider the assessment of outliers and influential observation in non-linear measurement error models. Residuals, leverage measures and case-deletiondiagonostics are examined. The method of local influence is also applied to the models. In particular, the perturbation of measurement error variances has been found useful in assessing the adequancy of the model assumptions. A numerical example is given to illustrate the application of the diagonostics.     

Child maltreatment and foster care

In “Child Maltreatment and Foster Care,” Richard J. Gelles, a professor at the University of Pennsylvania School of Social Work, reports what is known about the incidence of child maltreatment and foster care placement rates since welfare reform. He relies on four major sources of data: Current Trends in Child Abuse Reporting and Fatalities: Results of the 1998 Annual Fifty State Survey, the National Child Abuse and Neglect Data System (NCANDS), the National Incidence Survey of Reported and Recognized Child Maltreatment (NIS), and the Voluntary Cooperative Information System (VCIS).     

Recent fertility trends in industrialized countries: Toward a fluctuating or a stable pattern?

By the mid-1980s, fertility in most of the world's developed countries had declined to unprecedentedly low levels. Since then, it has declined still further in some, increased slightly in others, and fluctuated in still others. Irrespective of cause, these changes could not have occurred in the absence of substantial control over childbearing.While future increases and decreases are both possible, it is argued that, contrary to the usual demographic expectations for populations exercising substantial control over fertility, fertility in most of these countries will increase to approximate replacement levels and then undergo only minor fluctuations around these levels thereafter.