Search

Most personal (egocentric) network studies describe networks using measures that are not structural, opting instead for attribute-based analyses that summarize the relationships of the respondent to network members. Those researchers that used structural measures have done so on networks of less than 10 members who represent the network core. Although much has been learned by focusing on attribute-based analyses of personal network data, the application of structural analyses that are traditionally used on whole (sociocentric) network data may prove fruitful. The utility of this approach becomes apparent when the sample of network members elicited is relatively large. Forty-six respondents free-listed 60 network members and evaluated tie strength between all 1,770 unique pairs of members. Graph-based measures of cohesion and subgroups revealed variability in the personal network structure. Non-hierarchical clustering generated subgroups that were subsequently verified by respondents as meaningful. Further analysis of the correlation between subgroup types and overlap between subgroups demonstrates how the analysis of each network can be summarized across subjects. Four case studies are presented to illustrate the richness of the data and the value of contrasting individual matrix results to the norm as defined by all 45 subjects.

In this paper we compare two methods for estimating the size of personal networks using a nationally representative sample of the United States. Both methods rely on the ability of respondents to estimate the number of people they know in specific subpopulations of the U.S. (e.g., diabetics, Native Americans) and people in particular relation categories (e.g., immediate family, coworkers). The results demonstrate a remarkable similarity between the average network size generated by both methods (approximately 291). Similar results were obtained with a separate national sample. An attempt to corroborate our estimates by replication among a population we suspect has large networks (clergy), yielded a larger average network size.Extensive investigation into the existence of response effects showed some preference for using certain numbers when making estimates, but nothing that would significantly affect the estimate of network size beyond about 6 percent. We conclude that both methods for estimating personal network size yield valid and reliable proxies for actual network size, but questions about accuracy remain.

This study examines the effect of telephone survey dialing patterns on lab productivity and survey responses. Using an original data set of paradata from 2010 to 2017 and a machine learning technique for variable selection, we find that early and late afternoon shifts are as productive as late evening shifts for both landline and cellphone Random Digit Dialing (RDD) samples. Also, early weekdays are more productive than the weekend for the cellphone RDD samples. Most importantly, time of the interview affects survey responses; therefore, survey practitioners and scholars should be cognizant of this effect when scheduling calls.

Social network research often depends on the willingness of respondents to provide personal information about themselves and alters. Survey design strategies that increase willingness to share this information are necessary for social network research to be feasible, especially when name generators are used for sampling because rosters are unavailable. We conducted an experiment in which one group of respondents (n = 94) received an online survey that included an example network map and the other group (n = 100) received one that did not. Results show the map did not increase or decrease provision of network contacts nor influence the types of ties reported. Furthermore, respondents were reluctant to provide names and contact information of alters. Our study demonstrates the difficulty of collecting network information without a previously defined roster and an attempt to improve data collection through strategic survey design. [ABSTRACT FROM AUTHOR]