One of the questions I’ve found the toughest to answer throughout my study is ‘what do you do when you do Twitter?’ Or, more specifically, ‘what does participant observation look like in the context of your research?’ I’ve previously responded at length, and have been trying to capture a sense of what I do within a single visualisation. I’m still not convinced I’ve quite managed it, but here’s the final version which made it into the thesis:
Rightly or wrongly, I’ve been trying to present what I did in a form that would allow someone else to follow those steps. That’s the tricky bit. What I did might not always be a good indicator of what I (or someone else) would subsequently do. The approach needs to adapt and respond to situations as they unfold, forming as it does, one of the strands within my flanography. That said, within the visualisation, I’ve attempted to portray the five general techniques which might open a session.
Starting in the centre, the quote from Schensul et al (1999) provides an overview of what the purpose of observation might be, although I’m not entirely comfortable with some aspects of it. Some would say for example that ‘collecting data’ implies the data already exist for the researcher to sweep in and hoover them up. With the sociomaterial sensibility I’ve adopted for this study, I’d be more likely to frame it as ‘assembling’ data. The data are brought into being by my actions, working with those of the participants and those of the nonhuman actors like Twitter’s algorithms, tweets, hashtags, Likes and RTs. They do not pre-exist, and become data only when I, my notes, Treeverse, TAGS, MIndView etc enact them as such.
As a participant observer on Twitter, ‘unobtrusivity’ is inevitable for much of the time, at least until you engage with fellow participants. This should not be misunderstood as ‘covert’ research; there was never any intent to deceive or remain hidden. However, there were occasions – such as during hashtag chats – when I chose not to jump in with questions. To do so, in my opinion, would have been ethically unsound, given that I would be interrupting participants who had assembled at a particular time to be involved in a chat on a particular topic, not to be badgered by some researcher. If I needed to ask questions, I waited until the chat was over.
With the quote I chose for the centre of the visualisation providing a somewhat qualified touchstone for subsequent activity, what then unfolded could take one or more of five paths. Moving out from the centre through one of the sectors provides a snapshot of one approach I might take. For example, the cyan sector shows what was usually involved during my daily scanning routine. That would mostly be done on a tablet computer, through the Echofon app . Of the five sub-practices Postill and Pink (2012) describe as ‘everyday routines of digital ethnography practice,’ when in the cyan sector, I’d be ‘catching up, sharing, exploring, interacting,’ with ‘archiving’ taking place later when back at a desktop computer. Only with the full keyboard at my disposal would I then compose jot notes, and subsequently transfer data and notes into MindView for supplementing with annotations and memos.
Through the visualisation, I’ve attempted to provide an overview of what observation involved, but in so doing, I feel I’ve lost the messy nature of day to day activity. Participant observation was never quite that tidy. Some days I might start out in one sector, then move across into another. Having missed a ‘live’ hashtag chat, I might instead pick it up through an archive. Should that have had its own sector too?
I don’t think any single visualisation could adequately describe ‘what I do when I do Twitter,’ but together, all of my different attempts help to paint the picture more (or less) clearly. I have three audiences to which I hope the visualisation will speak. Firstly there’s the examiners of my thesis for whom the vis will act as one strand within an audit trail, hopefully showing that my research was conducted with integrity. Secondly, future researchers who might be looking to critique and extend ways in which research on Twitter has been conducted. And finally there are the good folk who constituted my research participants and willingly or often unwittingly, became involved. The visualisation helps to bring to the fore, a process of which the majority of will have been unaware, but which at least opens itself to scrutiny. Producing the vis at the outset so that it could have been incorporated into participant information would have been helpful, however, the responsive approach I chose meant that some of the most useful activities only emerged as a result of conducting the study.
As I reflect on the aspirations I had for the visualisation, I remain somewhat dissatisfied. There is so much it missed, such as the convoluted paths my experiences inevitably followed. And where on the graphic was the PLDbot? The best I can hope for is that the vis: 1) opens a dialogue with those who view it, and thereby 2) continues to be a work in progress and respond to whatever new developments emerge.
Postill, J., & Pink, S. (2012). Social media ethnography: The digital researcher in a messy web.
Schensul, S. L., Schensul, J. J., & LeCompte, M. D. (1999). Essential ethnographic methods: Observations, interviews, and questionnaires Rowman Altamira.