Analysing Twitter

I have recently been attempting a discourse analysis of a Twitter chat event, The analysis process proved harder than I was expecting. Partly, this was due to the volume of tweets where in a 90 minute event, 922 tweets were made, or 10.2 per minute. No wonder participants often mentioned how hard it was to keep up with the discussions.

Given a structure that is messy, difficult and hard to analyse, the Twitter event appears to exaggerate many of the key problematic features of unstructured discussions identified by Belnap & Withers (2008, p8). These include: sequences extending over many exchanges; overlapping exchanges and sequences; short sequences tending to be cut off prior to a conclusion and sequences re-emerging later in discussions. This suggests that the lack of event coherence and stability should be more problematic, but participants seem to use certain strategies including making lots and lots of retweets to keep sequences going or to restart them as well as making series of statements ‘to’ the Twitter event in the hope that one of these statements will get a response. Direct and traceable exchanges tended to be really short but took place within patterns of longer and often fragmented, sequences as participants attempt to negotiate between exchange, sequence and transcript timeframes.

Many of the features of an online discursive and learning community appear to be present in the event. These include aspects of mutual support for information-seeking activities and the exploration of differences. Discourses in the event can seek to reinforce common understandings and thematic coherence between the participants. There is also strong discursive patterns that appear to seek to differentiate participants from ‘others’ outwith the specific community (Bragd et al 2008). Interestingly, as a group of professionals often with managerial positions, the ‘others’ identified and vilified were “managers” or “the business” or “them”.

Also of note was that there was some evidence of intertextual and retrospective sense-making activities. These were more usually in the form of blog posts but could be as other artefacts including diagrams and mind maps.

The study was undertaken as a sort of proof of concept and it did show that these sorts of Twitter discussion events provide as rich a vein of data as face-to-face interactions or other (dare I say) more traditional forms of computer mediated communication. In other words, it worked and Twitter conversations are meaningful.

Belnap, J. K., & Withers, M. G. (2008). Discourse Analysis : The problematic analysis of unstructured / unfacilitated group discussions. Conference on Research in
Undergraduate Mathematics Education, Feb.

Bragd, A., Christensen, D., Czarniawska, B. & Tullberg, M. (2008) Discourse as the means of community creation. Scandinavian Journal of Management. 24(3):199-208..

Leave a comment

%d bloggers like this: