Evaluating the Method – Twitter Data

Riots and Twitter: connective politics, social media and framing discourses in the digital public sphere

  1. The researchers show that the connective action does not actually underplay differences between technologies. They also show that connective action sufficiently accounts for cultural and ideological drivers of action. They got their data by analyzing software systems as well as by looking at public issues and discourse to give a fuller account of connective politics during the riot clean-up movements. Their analysis shows that the clean-up movements were both complex and discursive political acts that had much influence from celebrities. Their findings show that the #RiotCleanUp hashtag that is credited with mobilising these groups provides a less compelling explanation of action in comparison to the #OperationCupOfTea hashtag. The researches argue that networks assemble and mobilise through the activation of discourse within a wider media sphere of competing discourses (214).
  2. They use content coding and close textual reading techniques to characterise discourse within major riot-related hashtags (214). They use an empirical method to identify, analyse and compare hashtag-specific discourse critically. This includes three elements: 1) establishing an overview of discourse, which includes identifying the dominant hashtags during the relevant period, 2) differentiating between these hashtags in a way that supports a critical analysis of their relative influence to recognise and allow for the interactive dynamics of the Twitter system, 3) providing a mechanism for interpreting this influence in terms of connective action (revealing clues as to why some discourses energise group mobilisation and others do not). Their methodological approach was an inductive one. They looked at specific hashtags and tried to induce a theory about collective action.
  3. Strengths: They used seven different hashtags, which covered a range of material. Multiple hashtags allow for more representative data on similar topics. Weaknesses: 1000 Tweets is a good number, but it could be increased to more Tweets to provide an even more representative sample. This would be more time-consuming to do so as well.

Understanding a digital movement of opinion: the case of #RefugeesWelcome

  1. The researchers were trying to analyze the digital discussion around the Twitter hashtag #RefugeesWelcome as a case of ‘digital movement of opinion’ (DMO). They got their data by using a triangulation of network, content and metadata analysis. They use a triangulation of quantitative methods (Twitter network and metadata analyses) and more qualitative text-based validations. They found that this DMO was driven primarily by social media elites whose tweets then became echoed by masses of isolated users. Then, they tested the post-DMO status of the hashtag-sphere after the November 2015 Paris terrorist attacks. The researchers argue that the concept of DMO provides a heuristically useful tool for future research on new forms of digital citizen participation because it (1) provides an ideal-type allowing to study empirical cases by observing their adherence and deviations from the theoretical construct; (2) isolates the digital dimension of citizen participation, both as a methodological strategy and an epistemological posture; (3) bridges public opinion and social movement theories and thereby helps apprehend new/future forms – arguably more networked but also more individualized – of collective action.
  2. This study uses three methods: triangulating Twitter network, metadata, and qualitative content analysis. They used a social media analytics tool that was designed for tracking Twitter content. There was a dataset of over 1 million tweets. Their methodological approach was an inductive one. They looked at specific hashtags and tried to understand DMOs.
  3. Strengths: The dataset was large (with over 1 million Tweets). This would provide a representative sample of the population and lead to more generalizable results. Weaknesses: Going through 1 million Tweets can be time-consuming. Were there other hashtags that could be looked at?

2 thoughts to “Evaluating the Method – Twitter Data”

  1. Hi Alice! Nice concise analysis of the two articles. I also agree that going through 1 million Tweets can be time consuming and believe the researchers should have condensed their for optimal performance.

  2. Alice,
    Thanks for your post, and great summaries of both studies. I would agree with the strength you have identified in Study 1- the use of multiple hashtags does help be more representative than using a single hashtag, and the weakness that you mention- the limited amount of Tweets is one that I also mentioned. With the relative ease in extracting Tweets via programming it is easy to pull whatever number you like. In my project for this semester I quickly found that small numbers of Tweets do not provide robust enough data and again with the ease of increasing the volume of Tweets researchers should work backwards from too many versus not enough. In your Study 2 comments I agree also with the strengths and weaknesses you identified- in contrast to Study 2 using over 1 million Tweets. It is true though that a time consuming part of big data with large data sets is the programming that takes place to make sense such large volumes of data. Thanks again for you post- and I agree with Jannie- you did an excellent job of making your summaries concise and effective.

Leave a Reply