Dynamic network analysis of contact diaries

Analyzing two-mode networks linking actors to events they attend may help uncover the structure and evolution of social networks. This classic network insight is particularly valuable in analysis data extracted from contact diaries where produce — at same time are product relations among participants. Contact comprise any number meeting a specific point time. In this paper we recall correspondence between actor–event hypergraphs, propose relational hyperevent models (RHEM) as general modeling framework for time-stamped multi-actor which diarist (“ego”) simultaneously meets several her alters. RHEM can estimate event intensities associated with each possible subset that jointly participate events, test effects be theoretical or empirical interest. Examples such include preferential attachment, prior shared activity (familiarity), closure, covariate explaining propensity co-attend events. Statistical tests these processes govern formation informal groups diarist’s We illustrate value using comprising almost 2000 former British Prime Minister Margaret Thatcher cabinet ministers, transcribed covering first term office (1979–1983).