New, Forthcoming, and In-Progress Work
Noting the Ties After Tying the Knot: Photo-Based Elicitation of Personal Networks
Alexandra Marin and Chang Lin
Forthcoming, Network Science
We report on an interview method using photo-based network elicitation and a landmark anchoring event to collect data on relationship change in ego networks retrospectively. Using the wedding albums of married or formerly married respondents, we populate a network from many years ago with persons pictured in the album and then collect data about each respondent’s relationship with each person at the time of the wedding and at the time of the interview. This data collection method mitigates many of the problems associated with retrospective data collection and does not have the logistical difficulties associated with panel studies. Our findings show that this method is successful in collecting significant numbers of connections that have undergone change and especially reductions in activity along various dimensions.
Network Instability in Times of Stability
Alexandra Marin and Keith Hampton
Forthcoming, Sociological Forum
Personal networks undergo change in response to major life course events. Individual, relational, and network characteristics that influence network instability in the absence of a significant life transition/crisis are less understood. We focus on those ties that transition from active to dormant. Because the shift to dormancy is often interpreted as a reduction in support or social capital, it is considered problematic. This study is based on longitudinal survey data of middle‐class adults who did not undergo life changes. Even in this context of relative stability, support networks experience rates of dormancy similar to those observed during periods of major upheaval. Tie dormancy is unrelated to individual characteristics, network size and density, or homophily along dimensions other than sex. Frequency and medium of communication are particularly notable as factors that were not related to tie dormancy. Ties were less likely to become dormant if they were geographically or emotionally close, immediate kin or neighbors, highly supportive, the same sex, or more embedded in the network. These findings provide context for how support networks operate when not buffeted by exogenous forces. They provide a baseline for understanding the impact on networks of transitions, trauma, new media, and difficult life circumstances.
The Relationship Between the Position of Name Generator Questions and Responsiveness in Multiple Name Generator Surveys
Reza Yousefi-Nooraie, Alexandra Marin, Robert Hanneman, Eleanor Pullenayegum, Lynne Lohfeld, Maureen Dobbins
Forthcoming, Sociological Methods and Research
Using randomly ordered name generators, we tested the effect of name generators’ relative position on the likelihood of respondents’ declining to respond or satisficing in their response. An online survey of public health staff elicited names of information sources, information seekers, perceived experts, and friends. Results show that when name generators are asked later, they are more likely to go unanswered and respondents are more likely to respond that they do not know anyone or list fewer names. The effect of sequence was not consistent in different question types, which could be the result of the moderating effect of willingness to answer and question sensitivity.
Social Network Analysis: An Example of Fusion Between Quantitative and Qualitative Methods
Reza Yousefi Nooraie, Joanna E. M. Sale, Alexandra Marin, Lori E. Ross
Forthcoming. Journal of Mixed Methods Research
A quantitative approach to social network analysis involves the application of mathematical and statistical techniques and graphical presentation of results. Nonetheless—as with all sciences—subjectivity is an integral aspect of network analysis, manifested in the selection of measures to describe connection patterns and actors’ positions (e.g., choosing a centrality indicator), in the visualization of social structure in graphs, and in translating numbers into words (telling the story). Here, we use network research as an example to illustrate how quantitative and qualitative approaches, techniques, and data are mixed along a continuum of fusion between quantitative and qualitative realms.
The Role of Skin Color in Latino Social Networks: Color Homophily in Sending and Receiving Societies
Wendy Roth and Alexandra Marin
How does skin color shape the social networks and integration pathways of phenotypically diverse immigrant groups? Focusing on Dominicans and Puerto Ricans, groups with considerable diversity across the Black-White color line, we explore whether migrants to the United States have greater color homophily in their primary social networks than non-migrants in the sending societies. We analyze egocentric network data, including unique skin color measures for both 114 respondents and 1,702 alters. We test hypotheses derived from ethnic unifier theory and color line racialization theory. The data show evidence of color homophily among Dominicans, but suggest that these patterns may be imported from the sending society rather than fostered by the U.S. context. Further, we find that migrants’ skin color is associated with having ties to White or Black Americans, but with different patterns for each ethnic group. We discuss the implications of these findings for economic mobility and U.S. racial hierarchies.
Relationship Change, Network Change, and the Use of Single Name Generators in Longitudinal Research on Social Support
As relationships change and people change the kinds of support they provide, name generators that collect information about ties that provide particular kinds of support in a repeated points of time may not effectively capture ties that are active but whose roles have changed. This paper shows that a significant minority of network members change the kinds of support they provide over time. They either discontinue a support previously provided or provide a new type of support. I examine the implications of this finding for longitudinal studies of support networks based on single name generators and show that this practice can result in frequent misperceptions of network membership change.
Towards Multidimensional Understandings of Ego Networks
Alexandra Marin, Chang Lin, Xiaowei Li, and Soli Dubash
Ego network researchers commonly collect rich data on the nature of relationships between respondents and alters including support exchanged, the nature and frequency of interactions, affect and conflict. Yet the common approaches to analyzing ego data do not engage with the complexity of these data. Typically, researchers take one of three approaches: combining ties that provide any kind of support or interaction into a single network, thus using the combined dimensions to create a network boundary; analyzing dimensions separately; or combining multiple dimensions to create a measure of multiplexity. Following these approaches tend to reduce relationships on which detailed data are available to the presence or absence of a tie, or at best, to ties with varying levels of intensity.
When we treat ties that provide any kind of support as equal and equivalent members of a social support network we ignore and mask variation in the character of relationships. Considering multiplexity or other measures of relationship intensity adds some nuance but can still treat very different relationships as equivalent. For example, a neighbour who socializes with ego and does small favours and a sibling who discusses important matters and would lend ego a large amount of money would both have a multiplexity of two despite having very different relationships.
We argue that researchers must dive into multidimensional understandings of ego networks to investigate how the character of relationships varies and changes. Understanding ego-centric social support networks as multi-dimensional involves considering how different dimensions of relationships intersect and combine. Which dimensions commonly vary together? What combinations of activity characterize different kinds of relationships? Are relationships fluid in shifting their forms of activity or do they display strong inertia? We draw on examples from the literature as well as from our own work to highlight the kinds of questions and findings that are possible when we rigorously engage the multi-dimensional nature of relationships.
We highlight in detail one example of multidimensional approaches to ego networks. Using detailed data on 607 relationships at two points in time at least 15 years apart, we use latent class analysis to develop a typology of ties based 12 dimensions of tie activity. Our model reveals 6 classes that vary, not only in their levels of activity, but in the nature of the relationship implied by the particular dimensions active among ties of each type. We show how relationships evolve and change categories over time and discuss the implications of these results. In particular, we highlight how ties that could be classified as dormant by a less multi-faceted analysis actually vary: While some may well and truly be gone, some are latent pools of social support and social capital ripe for reactivation.
Over-Connected Bridges: Cognitive Bias in Respondent Estimation of Alter Embeddedness
Xiaowei Li and Alexandra Marin
Research into cognitive biases in the memory of network structures has focused primarily on whole networks and on the ways in which the perceived overall structure varies from structure measured in other ways. We draw on whole network research to consider how cognitive biases in remembering network structure shape respondents’ abilities to report structural characteristics of their ego networks.
Remembering connections between people is a cognitively difficult task because even in relatively small networks there are many relationships. In a network of ten people there are 45 connections to be recalled. Research shows that people simplify this task by using schema – organizing rules about what networks typically look like. One important schema is triadic closure, in which people who know the same others are remembered as knowing each other, as well.
We hypothesize that when estimating the embeddedness of alters, triadic closure schemas will result in inflated reports of embeddedness for alters who are bridges. For example, if an alter, X, is connected to Y who is an alter within a cluster, then triadic closure schema could result in X being thought of as connected to the entire cluster to which Y belongs, rather than just to Y.
In this study we examine whether and how egos’ ability to estimate the embeddedness of their alters is shaped by cognitive biases. In a name generator survey of 568 respondents who named 5202 alters, respondents were asked for each alter “please estimate how many of the other people listed that person knows” and chose from the following options, “none or a few,” “less than half,” “about half,” “more than half,” or “all or almost all.” Following this, a full dyad census was completed, in which respondents indicated for all pairs of alters whether those alters know one another.
Using multi-level models to account for the alters nested within egos, we then regress egos’ reports of alters’ embeddedness on their bridging status and controlling for the number of ties the alter is reported to have and the respondent’s ego network size. We show that alters who are more bridging within the network are reported to be more embedded than similarly embedded alters who do not act as bridges. Further, the bias is greater for respondents who completed name generators that primed them to think of their alters in terms of clusters.
Results suggest that cognitive biases matter for ego network data collection not only because they influence which alters are named as part of the network, but also because they shape respondent reports of network structure. We discuss how this may influence data collected using existing data collection methods, how this shapes our own development of new data collection methods, as well as limitations of our study.
I’m Not Here to Make Friends: Intentional and Incidental Interaction Contexts and Alter Embeddedness Profiles
Soli Dubash and Alexandra Marin
Personal networks are lived out in the various contexts in which people interact: workplaces, restaurants, homes, and voluntary associations. Some—workplaces, educational institutions, and clubs—are what Feld (1981) termed foci. Others, like homes and bars, are simply settings where relationships are enacted. We argue there is an overlooked but theoretically significant distinction between different interaction contexts. Some, such as homes and restaurants, are contexts in which people interact with the express purpose of interacting with the specific others whom they encounter there. Others are settings entered for other purposes and where social contact is incidental to the primary purpose. We argue that these interaction contexts nurture relationships that play different roles in egos’ networks.
In this paper, we consider the embeddedness profile of alters in ego networks, a concept that includes not only their level of embeddedness, but the ways in which they are embedded in different parts of egos’ networks and their structural position in those networks. We seek to understand: (1) What kinds of interaction contexts foster domain-specific ties? (2) Does type of interaction context predict the extent to which an alter is embedded in other parts of the network? (3) Does type of interaction context predict the extent to which ties hold bridging or bonding positions within the network?
We analyze data from a web-based survey in which 622 respondents, reporting on 6,081, ties completed a name generator and were subsequently asked to indicate the contexts in which they interacted with each alter and to complete a dyad census, providing the full structure of the ego network. We find that different interaction contexts are associated with different embeddedness profiles. Alters who interact in incidental-interaction contexts are more commonly domain-specific ties, limited to single contexts of interaction. Conversely, ties enacted in purposeful-interaction contexts are more likely to bridge across clusters and those in incidental-interaction contexts more likely to be located within clusters. We develop a typology of embeddedness profiles and examine how they are associated with contexts of interaction, contexts of tie formation, and tie content.
Our findings suggest a potentially theoretically fruitful approach for broadening our understanding of the ways in which foci and interaction contexts structure ego networks.
Density Without Drudgery? Measuring Alter Embeddedness and Ego Network Density Without the Dyad Census
Alexandra Marin, Chang Lin, Soli Dubash, and Xiaowei Li
Ego network density is surely the most widely-used structural measure used by ego network researchers. Density is commonly used to conceptualize cohesion and has been shown empirically to be conducive to creating high levels of social support in networks, and creating and reinforcing consistent expectations.
While important, ego network density is a pain to measure, for all involved. Whole network density can be calculated using data in which each respondent is asked only to report on their own ties, but data for ego network density require that respondents complete a full dyad census, reporting on the ties between all alters in the their networks. For a network of twelve alters this is 66 ties. This is tedium for respondents. For researchers it represents precious survey minutes that could be used for other questions, and can also cost us the goodwill of our respondents, which is why these questions are typically placed at the end of surveys.
A single item measure of network density in which respondents simply report whether their network is highly or minimally connected, would be ideal. However, such a measure has already been tested and shown to be unreliable (Burt 1984). We test here an alternative that requires one survey question per alter. Respondents report on the embeddedness of each alter, a cognitively simpler task, rather than the density of the overall network.
622 respondents completed one of three name generator surveys with respondents randomly-assigned to one of the name generator options. For each of the 6081 alters listed, we asked “Please estimate how many of the other people listed this person knows.” Respondents selected from the options “None or a few,” “Less than half,” “about half,” “more than half,” or “all or almost all.” We correlated measures based on the dyad census to those based on the name interpreter item. Our results show that the reliability of this measurement method depends on the type of name generator used to collect the data and on network size. For smaller networks and those that do not prime respondents to think about different network domains, the measure correlates acceptably well with the dyad-census-based measures. Respondents are less able to reliably estimate alter-embeddeness when networks are large or when they have been elicited using multiple domain-clustering name generators.
Our results suggest that while the dyad census is more reliable, where it is not feasible and where network sizes are not expected to be large, alter-level respondent reported-embeddedness may be an acceptable substitute.