Keywords: egocentric networks, complete networks, contact diaries
Social network studies have been divided into two camps that differ sharply in both research issues and analytical tools: complete networks and egocentric networks. Almost all examinations of personal networks rely on the egocentric approach, mainly because the boundaries of one’s networks often shift and their components are constantly restructuring. Although such studies sometimes take advantage of rich self-reported data, especially on subjective well-being, they inevitably miss the sophisticated methods that help untangle the structures of complete networks. In an effort to ease such a limitation, this paper explores how studies of personal networks can advance by analyzing egocentric complete networks, a unique network structure that we reconstruct from comprehensive information about daily contacts. Such network reconstruction employs an innovative approach that involves rigid data collection and detailed contact records on social media in three large projects.
First, 54 individuals kept contact diaries of all one-on-one interpersonal contacts in Taiwan for three months in 2004, with details about alter characteristics, ego-alter tie features, and contact situations. Among these diary keepers, 24 also rated tie strength between each pair of alters within each of their three-month contact networks, which yielded a total of 851,210 alter-alter ties.
Second, another 133 diary keepers recorded similar details about 127,455 contacts with 12,070 alters, in abridged contact diaries, for seven months in 2014. Each diary keeper then rated the strength of 99.97% of alter-alter ties within his or her contact network. Because 74 of these diary keepers also appeared on other keepers’ contact lists, we could cross-check the ego-rated strength of alter-alter ties against alter-rated strength on identical ties among 7,310 persons who were on both contact lists of any two diary keepers who also appeared on each other’s contact diary. Among the ties that we verified, about 95.5% were rated consistently by both parties.
Third, in 2015, a nationally representative sample of 1,866 senior students from 58 university departments in Taiwan completed an online survey and authorized the use of their Facebook records. From this sample, we identify 959 students as members in 26 complete contact networks, as 70-100% of students in each of these 26 departmental classes both completed the survey and provided their Facebook records. To measure the strength of ties between any pair of students within the 26 classes, we calculate the frequency and intensity of interaction from comments or tags involved in 181,544 Facebook exchanges between 2013 and 2015.
Each of the three projects generates comprehensive information about all ego-alter ties and alter-alter ties in egocentric networks. Using these ties as the backbone of network structures, we reconstruct 24, 133, and 959 egocentric complete networks, respectively. Not only do these complete contact networks surrounding particular individuals facilitate a structural analysis of networks in its own right, but they also can help unravel network features that explain how individual behaviors and emotions vary from contact to contact.