Word usage patterns and social roles in online communities

The recent adoption of applications like Facebook, Twitter and Instagram has led to a staggering increase in the amount of people using these tools to share digital items (e.g., photos, videos) and talk. This explorative study investigated the feasibility to assign particular roles to community partners based on word usage patterns that are associated with particular partners. Different roles can be discerned in social communities, and for this study it is assumed that the role of ‘Leader’ is a prominent role in a community. For example, this person could be designated as an opinion leader. These social roles are helpful in processing community partners’ contributions as they allow attributing utterances to particular people which would aid discourse comprehension. Twenty-seven participants were asked to identify the ‘Leader’ in three online conversation transcripts and to answer some questions unrelated to this study. The transcripts were presented counterbalanced. A pilot study determined which of the dialogue partners was considered to be the ‘Leader’. All three transcripts were analysed with Linguistic Inquiry and Word Count (LIWC) 2007 software (Tausczik & Pennebaker, 2009) to investigate whether the ‘Leader’ role is characterised by certain linguistic signatures. LIWC allows the categorisation of 80 word pattern categories to psychological processes, such as positive emotion words, social relationships and social status. The results seem to indicate that a ‘Leader’ role in a dialogue could feasibly be identified by focusing on word usage patterns from the Positive emotion category and, to a lesser extent, the Personal pronouns category.