Session #6 – Social Networks through Time

Changes in Relational Ties constituting Personal Social Networks Across the Life Course in the USA: A Latent Class Approach

Yoosik Youm
Yonsei University, South Korea

Edward O. Laumann
University of Chicago

Keun Bok Lee
University of California, Berkeley

Using data from the 2002 General Social Survey (GSS), our latent class analysis revealed six distinct configurations of social networks over the life course based on seven relational ties constituting personal networks, including spouse, friends, neighbors, coworkers, parents, children, and siblings. In particular, we identified a “middle-aged, poor” class with very limited relational ties, only 12% of whom were married and all of whom had fewer ties with all the other possible relational ties on average when compared to other groups.  Even after controlling for other individual socio-demographic characteristics including sex, age, and race, this group reported the statistically significant lowest level of happiness. In sharp contrast to this class, we identified a well-to-do, midlife class (median age of 41) with a robust complement of all the relational ties present excepting children. This class was the happiest among the six latent classes even after controlling for the above socio-demographic factors. In addition, race differences among the six classes are of interest. While only 19% of whites belonged to ‘single, with parental ties,’ more than a third of African Americans belonged to this class. In other words, about one third of African Americans were not married and were tied\ to their parents. Because of this pattern, only 8% of African Americans were in ‘married, embedded in work’ class while 20% of whites belonged to this class. A definite limitation in the analysis is that it is based on cross-sectional data that rely on using age in years as a crude proxy for approximating stage in the life cycle. To the extent that different socio-demographic groups have different life course trajectories tied to different ages at life course transitions, we distort the association of social network patterns with the life course for the population as a whole.

Paper and presentation coming soon.


Seasonal Variation in Social Network Resources among Older Adults

Markus H. Schafer, Jason Settels & Laura Upenieks
University of Toronto

There is a general consensus that informal social networks convey many of the resources that protect and enhance older adults’ health and well-being.  Interestingly, most studies contributing to this conclusion have ignored the time of year when their social network and social support data have been collected—in effect assuming a seasonally- secure vision of social capital. A wealth of research, however, attests to the importance of seasonal variation in a wide variety of social behaviors, noting that temperatures, institutional calendars, and holidays shape the accessibility and nature of social interaction, among other important outcomes. Using the National Social Life, Health, and Aging Project (NSHAP), we identify 13 issues related to social networks and their conferral of resources—network structure, network content, social support, and general social connectedness—that may be sensitive to seasonal influence. Baseline NSHAP interviews were conducted between July 2005 and March 2006, providing 9 months for comparison and the inclusion of months with significant American holidays. Also of note, the date of participants’ interviews was exogenous, not systematically patterned by region of country or other variables. Our analysis first tests for overall monthly variation in each of the 13 outcome variables; second, we investigate several factors which may moderate observed monthly variation. Results reveal that core networks shrink by more than half a confidant in February and March relative to the maximum sizes observed during July and November. This late-winter contraction, moreover, is most pronounced for respondents living in the shabbiest homes. Similarly, older adults see far less of their neighbours in February and March relative to summertime, but this pattern is especially characteristic of those in lower socioeconomic standing and of those who live in decrepit neighbourhoods. Finally, we observe that both the average levels of closeness and the likelihood of talking to confidants about health is higher in winter months than during the summer, but this boost holds only for the youngest and the healthiest NSHAP participants. We conclude by noting some of the practical considerations raised by these findings and urge more research to carefully consider seasonal in/security of social network resources for older populations.

Paper and presentation coming soon.


Ego-Network Recruitment in the Wake of Network Mortality

Benjamin Cornwell
Cornell University

Edward O. Laumann
University of Chicago

Older adults are disproportionately likely to experience the death of a close network member. An important question is how their social networks are shaped by this experience. We study this question using unique longitudinal data on egocentric networks from the National Social Life, Health, and Aging Project (NSHAP). First, our analysis shows that the loss of confidants due to any cause increases the likelihood that older adults will add new confidants to their networks. How confidant mortality specifically is associated with network recruitment, however, depends on age. Confidant mortality is more positively linked to subsequent network recruitment at older ages. We consider some potential explanations for this, including the possibility that the oldest adults are more used to dealing with confidant death. We also discuss some implications of these findings for theories of social-network-related behavior in later life.

Please click each link to view the PDF: Presentation