From a network analytical perspective, networks comprise both actors and the links which tie these actors to one another.
Networks by six-month periods
Dividing the material into ten six-month periods creates 5,710 networks. The largest of these comprises 151 individuals, and the smallest, of which there are 3,681, contain just a pair of actors (i.e. a dyad). The individuals in the study are spread over these networks in the following way:
14 per cent are linked into networks comprising six or more actors
20 per cent are linked into networks comprising three to five actors
26 per cent constitute the half of a pair of actors
40 per cent appear as isolates
The single largest networks in each six-month period contain between 23 and 151 actors. As mentioned above, several of these networks are distinctive in the sense that they are in essence comprised of individuals suspected in connection with a single joint offence. The largest network, for example (151 actors) is primarily the result of a single incident where a large group of young people occupied the offices of a newspaper.
The largest networks of 1995
If the time-frame is extended so that the networks depict the pattern of co-offending over the course of a whole year rather than six months, the networks produced become considerably larger. The two largest networks for the spring and autumn of 1995, for example, contain twenty-three and forty-five actors respectively. When the time-frame is extended to cover the whole of 1995, however, the largest network comprises 285 actors. In addition, we find a further seven large networks which comprise between twenty and thirty actors.