maandag 22 juli 2013

Picking up the Twitter-thread

Last month I have been working on defining, specifying and refining my research proposal. Although there was already a solid research proposal lying ready for me to roll out, getting into the nitty-gritty of my research elicited many questions, some answers.. and then again many more questions. While trying to formulate guiding research questions and an effective methodology I was repeatedly drawn to the world of social media and ICT-tools. What is happening in this seemingly messy world of social media messages and how am I going to find significant patterns?

In my research, a central concern is the interplay between social media and mass media in the construction of meaning and how it ultimately affects decision-making processes in the governance of the agro-food sector. Twitter is put central, not because I presume that it is the dominant factor in the production of social meaning, but because it is a connecting thread of public discourse that displays strong framing processes. Although the short messages on Twitter accommodate only a small part of the public conversation, it incorporates all types of media messages and actors: stakeholders make public statements, mass- and social media messages are posted through hyperlinks and all people interested in the topic attribute meaning to an on-going stream of information. Twitter is a powerful framing tool because it’s the fastest, most open and most inclusive medium.


Catching Tweets that fly by

Picking up and following the Twitter-thread might lead us to social media networks, to the world of mass media production or to the marketing machines of stakeholders.. but following the Twitter-thread will certainly not take us far into the history. Although all tweets are public and printed in 'the Cloud', Twitter does not allow full access to historical data (you can only access data 7 days back in time). This means I have to collect the tweets that fly by. Most organizations interested in Twitter-activity employ a company that delivers customized data on request with some standardized tools for analysis. However, as a social scientist, I want to develop my own methodology that allows me to explore Twitter-activity myself in an on-going cyclical process of observation, data-collection and analysis.

I am now working with TAGS (Twitter Archiving Google Spreadsheet), a little software tool packed in a Google spreadsheet that allows me to archive Twitter activity based on key-words and provides some tools for analyzing and visualizing the data. I found this blog the most easy accessible guidance for starting to work with TAGS.


Exploring a first Twitter Hype 

I now have 15 TAGS running with a combination of Dutch key terms to cover the broad spectrum of messages around the sustainability of agro-food systems. In my Google Spreadsheet on ‘sustainable livestock’, one particular Tweet suddenly stood out as it evoked many replies and got increasingly retweeted:
Up until now, it has more than 1000 RT’s (Retweets) and 70 replies. The tweet got archived within my Google Spreadsheet on 'sustainable livestock', which operates via TAGS on the following Dutch search terms:

landbouw OR veeteelt OR veehouderij OR voedsel OR zuivel OR vlees OR melk OR vee OR koeien OR runderen OR kippen OR pluimvee OR varkens 
AND 
duurzaam OR duurzame OR biologisch OR biologische OR ecologische OR ecologisch OR intensieve OR antibiotica OR knippen

With this link you find a visualization of the collected tweets from 9-19 July (I limited the data-range to make it faster). Following the link brings you into TAGSExplorer in which you can explore the data yourself in an interactive network visualization. If you click on the bottom-right: "Retweets", the tremendous influence of this Tweet in the overall discussion becomes apparent - you are seeing the network of Retweets based on the above search term from 9-19 July. 


Why does this Tweet has such a great impact? 
Does it spread because it puts the news in a different context - linking vaccination and measles with antibiotics and chickens?
Do people retweet the message because it brings up a taboo - an issue that many people agree on but are resistant to ventilate themselves in non-Twitter conversations?

Do people reply to the Tweet because it addresses deep rooted susceptible social identities - Farming culture and Christianity?
Is there a tense conversation
because the Tweet is perceived as an attack on a particular community - the community of Barneveld?

Although @gert_van_dijk seems an influential Twitter-user in some circles, a brief exploration shows that it is not his direct network that got the Tweet go viral. It would be interesting to look at the networks and communities through which the tweet spreads? How many RT’s come from first circle followers, second circle followers, etc. and how are these people interrelated? Are there communities to identify or group dynamics at work? Is there an important hub in the spread of this Tweet? 

I will have to find new ways for collecting, processing and analyzing data to answer such questions. I am currently discovering NodeXL. This free software shows promising for network analysis, but I haven’t found a way to integrate the time-dimension for a conversation analysis.

Although I will probably not study this case in-depth as part of my research, an exploration will certainly guide me in working out my research proposal further.


Geen opmerkingen:

Een reactie posten