Shrinkage estimator of regression model under asymmetric loss

This article investigates the performance of the shrinkage estimator (SE) of the parameters of a simple linear regression model under the LINEX loss criterion. The risk function of the estimator under the asymmetric LINEX loss is derived and analyzed. The moment-generating functions and the first two moments of the estimators are also obtained. The risks of the SE have been compared numerically with that of pre-test and least-square estimators (LSEs) under the LINEX loss criterion. The numerical comparison reveals that under certain conditions the LSE is inadmissible, and the SE is the best among the three estimators.     

Simultaneous estimation of several CDF’s: homogeneity constraint

Let {x_ij(1 ? j ? n_i)|i = 1, 2, …, k} be k independent samples of size n_j from respective distributions of functions F_j(x)(1 ? j ? k). A classical statistical problem is to test whether these k samples came from a common distribution function, F(x) whose form may or may not be known. In this paper, we consider the complementary problem of estimating the distribution functions suspected to be homogeneous in order to improve the basic estimator known as “empirical distribution function” (edf), in an asymptotic setup. Accordingly, we consider four additional estimators, namely, the restricted estimator (RE), the preliminary test estimator (PTE), the shrinkage estimator (SE), and the positive rule shrinkage estimator (PRSE) and study their characteristic properties based on the mean squared error (MSE) and relative risk efficiency (RRE) with tables and graphs. We observed that for k ? 4, the positive rule SE performs uniformly better than both shrinkage and the unrestricted estimator, while PTEs works reasonably well for k < 4.     

Women’s status and domestic violence in rural Bangladesh: Individual- and community-level effects

We explore the determinants of domestic violence in two rural areas of Bangladesh. We found increased education, higher socioeconomic status, non-Muslim religion, and extended family residence to be associated with lower risks of violence. The effects of women's status on violence was found to be highly context-specific. In the more culturally conservative area, higher individual-level women's autonomy and short-term membership in savings and credit groups were both associated with significantly elevated risks of violence, and community-level variables were unrelated to violence. In the less culturally conservative area, in contrast, individual-level women's status indicators were unrelated to the risk of violence, and community-level measures of women's status were associated with significantly lower risks of violence, presumably by reinforcing nascent normative changes in gender relations.     

Detection of influential observations in multivariate regression

Several methods have been suggested, in the literature, to detect influential observations from the data fitting usual linear model y=X???+???, ???∽N(0, ???2I). Recently, Chatterjee & Hadi (1986) have reviewed most of these available methods and described the inter-relationships between them. In this article, we extend some of these methods to the case of multivariate regression data. We consider several data sets to illustrate the methods.     

Nonmetropolitan Residential Segregation Revisited

Abstract Nonmetropolitan (nonmetro) residential segregation in 1990 and change in the preceding decade have received insufficient attention. A set of empirical hypotheses are derived and assessed using nonmetro and metropolitan (metro) counties in Texas. Places in nonmetro counties were more segregated than places in metro counties in 1990 as in 1980. Substantial declines in segregation occurred in both nonmetro and metro places but were largest in growing places in nonmetro counties. An analysis controlling for other determinants of segregation supports the premise that population change was a major determinant of 1980–1990 change in segregation. Implications for nonmetro areas in the 1990s are discussed.     

On the structural estimation of certain correlation coefficients

In this paper, the estimation of a real-valued function of the parameter by minimizing the expected value of the quadratic loss function relative to the structural distribution of the parameter is proposed; this is called structural estimation. The general formulae developed have been used to obtain the structural estimate of the bivariate correlation coefficient and of the intraclass correlation coefficient.     

Correlates of Susceptibility to Scams in Older Adults Without Dementia

This study examined correlates of susceptibility to scams in 639 community-dwelling older adults without dementia from a cohort study of aging. Regression models adjusted for age, sex, education, and income were used to examine associations between susceptibility to scams, measured by a five-item self-report measure, and a number of potential correlates. Susceptibility was positively associated with age and negatively associated with income, cognition, psychological well-being, social support, and literacy. Fully adjusted models indicated that older age and lower levels of cognitive function, decreased psychological well-being, and lower literacy in particular may be markers of susceptibility to financial victimization in old age.     

Crack ‘Hos and Skeezers: Traumatic experiences of women crack users

The involvement of women in crack cocaine abuse has had a severe impact on their health, the health of their children and the stability of their communities. Of particular concern has been the development of a system of barter in which crack‐for‐sex exchanges are the means through which women obtain the drug. Earlier studies have suggested that drug abuse may be related to and exacerbated by trauma. In the project described herein, we interviewed women crack users in Harlem to study the relationship between trauma, crack use, and crack‐related sexual behavior. Results suggested the existence of three types of trauma: (1) traumas that predate the respondent's onset of crack use; (2) traumas that were the direct sequelae of crack use; and (3) stigma trauma, that is, trauma that results from membership in a despised or oppressed group. We observed a complex inter‐relationship involving crack use, crack‐for‐sex transactions, and these three types of trauma. Treatment of the eventual co‐morbidity of trauma and addiction is an urgent challenge.     

The effect of undercount on the accuracy of small-area population estimates: Implications for the use of administrative data for improving population enumeration

The effects of underenumeration on the accuracy of alternative methods of population estimation have not been sufficiently analyzed. Although the US Bureau of the Census has decided not to adjust either the counts or its estimates for underenumeration in 1990, the extent to which local population estimates may account for underenumeration is of importance both for those who may wish to adjust existing estimates and in anticipation of future census adjustments. This paper examines the accuracy of small-area population estimation methods with and without adjustment. Mean Percent Errors, Mean Absolute Percent Errors, and Mean Percent Absolute Differences between local estimates for 1990 and 1990 adjusted and unadjusted census counts are computed. Population estimates for 1990 made using housing unit, ratio correlation, and component methods are compared for 451 counties and 2,633 places in the states of California, Florida, Texas, and Wisconsin. An analysis of the data for counties shows little indication that local estimates more accurately estimate the adjusted than the unadjusted population counts. The results for places show clear improvements in accuracy for places in Florida and Texas. Implications of the findings for issues related to undercount adjustment and local population estimates are discussed.