Seminar Series

This paper reports on an extension of group-based trajectory modeling to address non-random subject attrition or truncation due to death that varies across trajectory groups. The effects of the model extension are explored in both simulated and real data. The analyses of simulated data establish that estimates of trajectory group size as measured by group membership probabilities can be badly biased by differential attrition rates across groups if the groups are initially not well separated.

There is now consensus that concurrent partnerships increase transmission of infectious diseases in sparse networks -- like HIV in sexual partnership networks. Empirical research is accumulating that supports the hypothesis that concurrent sexual partnerships are one of the key drivers of the hyper-epidemics of HIV in some populations in Eastern and Southern Africa. The next step is intervention developmnet. This presentation reports on an NIH-funded pilot project of a concurrency reduction intervention.

Most demographic research on intergenerational processes focuses on the associations between the numbers and characteristics of individuals in successive generations, and ignores multigenerational (Non-Markovian) aspects of intergenerational mobility. These aspects include both the net effects of grandparents and early ancestors on individuals and also other types of path-dependent sequences of family characteristics.

The Global Health Initiative (GHI), President's Emergency Plan for AIDS Relief (PEPFAR), the President's Malaria Initiative (PMI) and other global health programs are putting increasing emphasis on evaluation and learning to ensure that investments in global health are effective and achieve desired outcomes. At the same time, there is a growing debate in the evaluation literature on the most appropriate methods to assess the effectiveness, efficiency and impact of health interventions in real world situations.

The prevailing model of migration in transitioning countries conceives of a risk-diversifying household in which members share a coherent set of preferences about the departure of one or more members to work elsewhere. Several decades of ethnographic research have questioned the applicability of this model by revealing the importance of gender hierarchies in family decisions. Some scholars argue that, in many contexts, women have little role in determining the migration behavior of spouses and other family members.

The share of the homeless population composed of African Americans and children has grown since the early 1980s, but the causes of these changes remain poorly understood. This article implicates mass imprisonment in these shifts by considering the effects of recent paternal and maternal incarceration on child homelessness using data from the Fragile Families and Child Wellbeing Study.

There are indications that the mean difference in socioeconomic status between homeowners and renters has been increasing in recent decades, and that these trends vary by race and ethnicity. This paper examines the roles of changes in the distribution of household demographic characteristics, the effects of those characteristics on the probability of being a homeowner, and local housing market characteristics in producing this apparent growth in inequality.

Aging is an evolutionarily labile trait that is likely shaped by a diversity of sources of natural selection. Two such mechanisms are sexual selection, or selection caused by competition among members of one sex for reproductive access to members of the other, and kin selection, which can be thought of selection arising from associations between fitness and social interactions. While various models have been used to argue for (or against) their importance in the evolution of aging, strategies for direct measurements of these selective forces are lacking.

Douglas Almond's use of the 1918 U.S. influenza pandemic as a natural experiment led to the seminal works on the subject of in utero health's impact on later life outcomes. The identification strength and clarity of his work, though, is driven by the inherent natural experiment supposition of random assignment. By using data from the 1920 and 1930 U.S. census, this study investigates this keystone assumption and shows that the families of the "treatment" cohort used by Douglas Almond were significantly less literate and economically prosperous than the families of the "control" group.

The results of recent evaluations of genome wide association (GWA) studies of complex phenotypic traits, including age at disease onset or life span, showed that such traits are typically affected by a large number of "small-effect-low-significance" alleles, which were excluded from further analyses in traditional GWA studies. In this talk we show that the joint influence of such genetic variants on human life span can be substantial and highly statistically significant.