Competition resulting in high social status and maintenance of that status within farmed pigs may have both welfare and economic consequences. Aggressive behaviours, associated with the establishment and maintenance of a dominance hierarchy, may cause poor welfare and skin damage reducing the value of pigs at slaughter. In order to investigate the impact of social status on welfare we first need to establish the status of individuals relative to their peers. In this paper we compare two alternative indicators of pig rank, namely: (i) aggressive interactions between individuals - the occurrence of aggressive behaviour from pig A to pig B is often assumed to imply social dominance of pig A over pig B; and (ii) the order in which pigs stand at the feeder - an advantage of attaining high social status in many populations is priority of access to food. Social rank of individuals based on observed aggressive behaviour can only be assigned once evidence for a linear or quasi-linear hierarchy has been established (Langbein and Puppe, 2004). In a population with a purely hierarchical structure we expect all triads of individuals to be transitive, that is A→B, B→C implies A→C. For observational studies it is unlikely that interactions between all individuals will be observed. Here we present a novel methodology by which we can assess the linearity of the dominance structure that is not affected by missing or null interactions between individuals. By conducting a census of all types of triad in a social network we can assess whether the subset of interactions observed provide evidence of a linear or quasi-linear hierarchy. Transitive triads provide evidence of linear hierarchy whilst intransitive triads, such as A→B, B→C, C→A, contradict the presence of a linear hierarchy.