Identifying positions from affiliation networks: Preserving the duality of people and events

Abstract

Frank's [Frank, K.A., 1995. Identifying cohesive subgroups. Social Networks 17, 27–56] clustering technique for one-mode social network data is adapted to identify positions in affiliation networks by drawing on recent extensions of p * models to two-mode data. The algorithm is applied to the classic Deep South data on southern women and the social events in which they participated with results comparable to other algorithms. Monte Carlo simulations are used to generate sampling distributions to test for the presence of clustering in new data sets and to evaluate the performance of the algorithm. The algorithm and simulation results are then applied to high school students’ transcripts from one school from the Adolescent Health and Academic Achievement (AHAA) extension of the National Longitudinal Study of Adolescent Health.