This sets of three graphs compare tie weight distribution in the three one-mode networks of co-attendance at the meetings of the Shanghai Rotary Club under President Petit's (1919-20), Fitch's (1930-1) and Harkson's (1938-9) terms. In this network, nodes represent participants (persons who attended meetings), whereas edges represents events. Any edge links two persons who participate in the same events. Tie weight measures how frequently two persons co-participate in meetings. This series of three graphs is based on Newman's method, which presents the advantage of rectifying the negative effects of size variation – by discounting most important events that are more likely to be attended than smaller events, and key officers that are more likely to attend compared to less active members, occasional guests and visitors. The first graph refers to Petit's network, the second to Fitch's and the last one to Harkson's. 

The original datasets and the resulting graphs are available in the "Tables" and "Trees" section, respectively. We used the R package "dplyr" to make these graphs.