A note on estimating regression coefficients with missing data

In this note, we consider the problem of estimating regression coefficients when there are missing observations of some explanatory variables. Following Dagenais (1973), Gourieroux and Monfort (1981), and Conniffe (1983a, 1983b), we assume auxiliary relationships exist among explanatory varibles. Several estimatprs and their interrelationships are discussed. We begin with the model of Gourieroux and

Monfort (1981)     

Assessing Risk‐Based Upper Limits of Melamine Migration from Food Containers

Melamine contamination of food has become a major food safety issue because of incidents of infant disease caused by exposure to this chemical. This study was aimed at establishing a safety limit in Taiwan for the degree of melamine migration from food containers. Health risk assessment was performed for three exposure groups (preschool children, individuals who dine out, and elderly residents of nursing homes). Selected values of tolerable daily intake (TDI) for melamine were used to calculate the reference migration concentration limit (RMCL) or reference specific migration limit (RSML) for melamine food containers. The only existing values of these limits for international standards today are 1.2 mg/L (0.2 mg/dm²) in China and 30 mg/L (5 mg/dm²) in the European Union. The factors used in the calculations included the specific surface area of food containers, daily food consumption rate, body weight, TDI, and the percentile of the population protected at a given migration concentration limit (MCL). The results indicate that children are indeed at higher risk of melamine exposure at toxic levels than are other groups and that the 95th percentile of MCL (specific surface area = 5) for children aged 1–6 years should be the RMCL (0.07 mg/dm²) for protecting the sensitive and general population.     

A property of coefficients of variation for ordered bivariate log-normal variables

This paper considers the maximum and minimum of a pair of log-normal variables with equal mean. It shows that either order statistic has a smaller coefficient of variation than the two original log-normal variables provided the latter are of equal variance. When the variances are unequal, as the variance ratio increases, the minimum (maximum), has a smaller coefficient of variation if the correlation coefficient of the log-normal variables is small (small) and the variances are large (small).     

Home Modifications among Community-Dwelling Older Adults: A Closer Look at Race and Ethnicity

ABSTRACT

This study investigated racial and ethnic differences in home modification use. Data from the 2011 National Health and Aging Trends Study were used (n=6,764). Compared with non-Hispanic Whites, Hispanics were less likely to have a grab bar (odds ratio (OR) = 0.6), bath seat (OR = 0.8), or raised toilet (OR = 0.6). Non-Hispanic Blacks were less likely to have a grab bar (OR = 0.7) or bath seat (OR = 0.7) than non-Hispanic Whites, but more likely to have a raised toilet (OR = 1.3). English proficiency largely explained ethnic differences, while health status partially accounted for racial disparities.     

Behind the Numbers: Talking Politics with Foreign-born Chinese Americans

This paper discusses results of in‐depth follow‐up interviews conducted with selected individuals of Chinese descent residing in Los Angeles and San Francisco who were previously chosen at random to participate in the 2000 and 2001 Pilot National Asian American Political Survey. A total of 15 male and female informants who had migrated from Taiwan, mainland China, and Hong Kong were interviewed in Mandarin Chinese. They were asked to compare the performance of the US Government with that of the government in their respective ethnic homeland. They were also asked to explain their reasons for supporting a certain US political party, ideology, and type of candidate. In addition, they commented on the state of political participation or the lack of it among Chinese Americans. Finally, they explained the complexity of their ethnic self‐identification and experiences of racial discrimination. Their responses were interpreted within the context of the historical formation of the Chinese American community as well as the summary results of the mass opinion survey. The results help dispel myths about Chinese Americans being politically indifferent and irrational. They help illuminate the possible relationships among ethnic identity, homeland politics, and political participation in the host land. They also provide exciting insights into improving the survey instrument for a majority immigrant and non‐Anglophone population.     

The effects of covariate adjustment in generalized linear models

Results from classical linear regression regarding the effects of covariate adjustment, with respect to the issues of confounding, the precision with which an exposure effect can be estimated, and the efficiency of hypothesis tests for no treatment effect in randomized experiments, are often assumed to apply more generally to other types of regression models. In this paper results pertaining to several generalized linear models involving a dichotomous response variable are given, demonstrating that with respect to the issues of confounding and precision, for models having a linear or log link function the results of classical linear regression do generally apply, whereas for other models, including those having a logit, probit, log-log, complementary log-log, or generalized logistic link function, the results of classical linear regression do not always apply. It is also shown, however, that for any link function, covariate adjustment results in improved efficiency of hypothesis tests for no treatment effect in randomized experiments, and hence that the classical linear regression results regarding efficiency do apply for all models having a dichotomous response variable.     

Individual rights and collective obligation. Compulsory intervention towards substance abusers in Norwegian social law

This article focuses on an important duality in the modern welfare state: the will to help our fellow citizens through institutionalized welfare policies, and the danger of violating individual liberty and integrity. Individual rights of self-determination and integrity are important values of the liberal constitutional government (the Rechtstaat). Our collective ethical obligation to help those in need is an important value of the welfare state. With respect to compulsory intervention towards adult and under-age substance abusers, the possible tensions between these values are particularly visible. The legal foundation for such efforts in Norwegian social law is discussed with regard to different ethical and moral principles. The main questions are how compulsory interventions are justified and how different principles are weighted in positive law. The ethical principles of collective obligation and the moral principle of individual rights are differently balanced in the Child Care Service Act and the Social Service Act. Both acts, however, expose the problematic combination of solving social problems by the use of force and securing individual autonomy and integrity.     

A study on IUD retention by curve-fitting

In order to estimate the demographic impact of IUD, there is a need for estimating IUD retention rates beyond the period of actual observation. This article describes a mixed exponential function for this purpose. With this function of IUD retention, it is possible to estimate the “life expectancy” and “half-life” of an IUD. Another important concept is the “cut-off date”, introduced in estimating the “life expectancy” because in practice no IUD will be worn by a woman forever.     

Gender,power and office politics

Abstract

This interpretative study explores the experiences of Taiwanese female clerical workers as they deal with office politics. A hermeneutical phenomenology is used to describe and interpret the participants' experiences. The study shows that powerlessness and frustration result from: (1) structural barriers; (2) behavioral barriers; and (3) accommodation and rationalization. This study has implications for research and practice in international area of women's studies and for the field of international career development education.     

Moments of order statistics from a non-overlapping mixture model with applications to truncated laplace distribution

We employ two different approaches to derive single and product moments of order statistics from a truncated Laplace distribution. A direct evaluation method establishes recurrence relations whereas the more general non-overlapping mixture model incorporates the truncated Laplace distribution as a special case. The results are thereafter applied to estimate location and scale parameters of such distributions.