An alternate version of the conceptual predictive statistic based on a symmetrized discrepancy measure

The conceptual predictive statistic, C_p, is a widely used criterion for model selection in linear regression. C_p serves as an estimator of a discrepancy, a measure that reflects the disparity between the generating model and a fitted candidate model. This discrepancy, based on scaled squared error loss, is asymmetric: an alternate measure is obtained by reversing the roles of the two models in the definition of the measure. We propose a variant of the C_p statistic based on estimating a symmetrized version of the discrepancy targeted by C_p. We claim that the resulting criterion provides better protection against overfitting than C_p, since the symmetric discrepancy is more sensitive towards detecting overspecification than its asymmetric counterpart. We illustrate our claim by presenting simulation results. Finally, we demonstrate the practical utility of the new criterion by discussing a modeling application based on data collected in a cardiac rehabilitation program at University of Iowa Hospitals and Clinics.     

Multistage sampling for latent variable models

I consider the design of multistage sampling schemes for epidemiologic studies involving latent variable models, with surrogate measurements of the latent variables on a subset of subjects. Such models arise in various situations: when detailed exposure measurements are combined with variables that can be used to assign exposures to unmeasured subjects; when biomarkers are obtained to assess an unobserved pathophysiologic process; or when additional information is to be obtained on confounding or modifying variables. In such situations, it may be possible to stratify the subsample on data available for all subjects in the main study, such as outcomes, exposure predictors, or geographic locations. Three circumstances where analytic calculations of the optimal design are possible are considered: (i) when all variables are binary; (ii) when all are normally distributed; and (iii) when the latent variable and its measurement are normally distributed, but the outcome is binary. In each of these cases, it is often possible to considerably improve the cost efficiency of the design by appropriate selection of the sampling fractions. More complex situations arise when the data are spatially distributed: the spatial correlation can be exploited to improve exposure assignment for unmeasured locations using available measurements on neighboring locations; some approaches for informative selection of the measurement sample using location and/or exposure predictor data are considered.     

The stratification of universities: Structural inequality in Canada and the United States

This paper conceives national systems of higher education as stratified populations of organizations. This stratification is a structural component of ‘horizontal inequality’ in higher education, and may be exacerbated by current pressures for colleges and universities to compete for resources and status. To explore this structural inequality, we compare the level of stratification in financial resources across four-year institutions in Canada and the United States over a 35-year period (1971–2006). Our analyses provide a first-look at this form of stratification, employing Gini coefficients, Lorenz curves, and boxplots. Our results provide new and compelling evidence of increasing structural stratification, even in Canada's predominantly publicly funded postsecondary system. Findings indicate that the distribution of resources is far more stratified in the American system, and that both systems have become more stratified over time. We conclude by situating structural inequality within more general processes of stratification in education.     

Mothers and child care: policies,values and theories

Government expansion of child care services is based on the assumption that both parents are employed (the adult worker model) and make cost‐benefit calculations in choosing child care (the rational economic choice model). This paper addresses this assumption, based on research examining mothers' assessments of appropriate child care. These assessments involve complex moral and emotional decisions around their own and their children's needs, and differ between social groups. On this basis, we conclude that the assumptions underlying current child care expansion policy are inadequate, and that the mere expansion of services is not enough.     

An EM-type algorithm for multivariate mixture models

This paper introduces a new approach, based on dependent univariate GLMs, for fitting multivariate mixture models. This approach is a multivariate generalization of the method for univariate mixtures presented by Hinde (1982). Its accuracy and efficiency are compared with direct maximization of the log-likelihood. Using a simulation study, we also compare the efficiency of Monte Carlo and Gaussian quadrature methods for approximating the mixture distribution. The new approach with Gaussian quadrature outperforms the alternative methods considered. The work is motivated by the multivariate mixture models which have been proposed for modelling changes of employment states at an individual level. Similar formulations are of interest for modelling movement between other social and economic states and multivariate mixture models also occur in biostatistics and epidemiology.     

Educating Women: Gender Inequalities among Canadian University Graduates*

Dans le présent article, les auteurs proposent une analyse comparative des revenus et du chomage dliommes et de femmes diplômés universitaires entre 1978 et 1988. Les résultats montrent que le sexe continue d'exercer une influence importante sur ces données; le rapport global hommes-femmes est en effet passé de 84% a 91%. Seule une petite partie de l'écart des revenus peut s'expliquer par des différences entre les hommes et les femmes au regard du recrutement sur le marché du travail. Les auteurs considèrent ces résultats révélateurs de la ségrégation entre les hommes et les femmes, et que celle-ci mène à des différences dans les cheminements de carrière. De plus, pour l'avenir, ils prévoient que, tandis que l'écart des revenus continuera probablement de diminuer, les formes d'inégalité entre les diplômés seront peut-ôtre de plus en plus marquées, surtout que l'écart des revenus est en déclin parmi les cohortes récentes. This paper compares male versus female income and unemployment for three cohorts of university graduates between 1978 and 1988. Results show that gender became a weaker predictor of these outcomes, with the female-male earning ratio growing from 84% to 91% over this period. Nonetheless, a notable wage gap remained in later cohorts, and for one cohort income disparities widened with time in the labour market. Moreover, only a relatively small portion of the earnings gap could be attributed to male-female differences in market endowments. The authors interpret these results as illustrating gender segregation within occupations that result in disparate earning trajectories. Also, it is suggested that while the wage gap will likely continue to shrink, future patterns of inequality among graduates may become increasingly uneven, especially since real wages are declining among recent cohorts.     

'Til death do us part: intimacy and sexuality in the marriages of Alzheimer's patients.

1. Alzheimer's disease (AD) is likely to have a significant effect on sexual behavior, but both patient and partner will still have sexual feelings and needs. 2. Research has shown that a high proportion of men with AD develop erection problems, but causes of their erection difficulties are not understood. 3. Research has shown that inappropriate sexual behavior in AD patients is uncommon, although it can be very troubling to the family and health-care provider if it occurs. 4. More professionals need to be trained to discuss sexual issues openly and sensitively with AD patients and partners and to offer useful clinical suggestions.