A methodological approach to the analysis of PISA microblogs: social media during the release of the PISA 2015 results

ABSTRACT

This paper analyses Twitter microblogs over a 3-day period, during the release of the results of PISA 2015 on 6 December 2016 by the OECD. We document a methodological approach to investigating the social mediatisation of policy and its inclusive potential for enabling the participation of multiple voices. We draw on two data sets from the 3-day period: first, a large data corpus (n = 17,260) of all microblogs about PISA, and second, a smaller cluster of selected participant groups, notably academic, OECD, and teacher union actors (n = 65). Our analytic tools provide a methodological heuristic for scoping social media networks. Our research is guided by questions about ‘Who participates?’ in PISA debates and ‘What is being said?’ and considers the volume, engagement, and content of the microblogs. Of the three participant groups, the academic actors authored the largest volume of microblogs across the corpus and cluster. The content of microblogs from the OECD cluster mainly provided information about and promotion of PISA, while the academic cluster emphasised commentary and critique, with all actors using an analytic tone. The OECD had the largest number of followers and the highest rate of social recommendations and engagement with their microblogs.

KEYWORDS:

• Social mediatisation
• methodology
• twitter
• PISA
• content and sentiment analysis
• key opinion leaders
• policy

Disclosure of potential conflicts of interest

No potential conflict of interest was reported by the author(s).

Data availability statement

Raw data were generated at dnoise. Derived data supporting the findings of this study are not available for distribution due to commercial restrictions.

Notes

1. A microblog is a short curated message to an online audience, potentially consisting of text, images, videos, etc. Microblogs are associated with multiple social media platforms including, Twitter, Instagram, Facebook, Reddit and LinkedIn. Twitter microblogs, also known as ‘tweets’ or ‘posts’ are 280 characters maximum.

2. In respect of the broader social arrangement, we note Zuboff’s (2019) critical account of surveillance capitalism and the monetising of our internet usage.

3. The Twitter data were mined as part of an Australian Research Council Discovery Project, DP150102098, entitled Data in Schools and Systems: An International Study.

4. For example, ‘Teach for Australia’ or ‘Teach for America’ are programs designed to attract and fast-track graduates into the teaching profession.

5. A detailed sentiment analysis of the large data corpus is the subject of another paper. For more information about how the LIWC tool analyses the different categories, see Tausczik and Pennebaker (2010).

6. We note that the larger project obtained ethical clearance to conduct this research. At that time, in 2016, the United States did not consider the extraction of Twitter data to be human research.

Additional information

Funding

This work was supported by the Australian Research Council [DP150102098];

Notes on contributors

Aspa Baroutsis

Dr Aspa Baroutsis is a Postdoctoral Research Fellow at the Griffith Institute for Educational Research at Griffith University. Her most recent book is Education research and the media: Challenges and possibilities (Routledge, 2019).

Bob Lingard

Professor Bob Lingard is a Professorial Fellow in the Institute for Learning Sciences and Teacher Education at Australian Catholic University and Emeritus Professor at The University of Queensland. His most recent books include, Digital Disruption in Teaching and Testing (Routledge, 2021) and Globalisation and Education (Routledge, 2021).