Solène Lellinger et Christian Bonah : analyse d'une collecte Twitter autour du procès du Mediator

Value of the Information & Strength of the Argument

The presentation offers valuable insights into the practicalities of using Twitter data for historical research, particularly for a non-viral event. The speakers provide a transparent account of their methodology, including the rationale for a broad collection strategy and the iterative cleaning process. They argue for the importance of audiovisual content in social media, a dimension often overlooked. The argumentation is solid, grounded in their extensive prior research on the Mediator scandal and supported by concrete examples from the collected data. They effectively demonstrate how Twitter activity reflects and diverges from traditional media coverage, contributing to a nuanced understanding of intermedia circulation.

Scientific Rigor, Source Quality, Title Accuracy

The presentation demonstrates scientific rigor through a well-defined research protocol, collaboration with INA, and careful data analysis. The speakers acknowledge limitations, such as data noise and missing tweets, and describe their efforts to address them. They cite their own previous work and the BodyCapital project, but do not provide external sources beyond the INA resources. The title accurately reflects the content, focusing on the analysis of a Twitter collection. The presentation is based on original research, though it is presented as an expert opinion rather than a peer-reviewed study.

The title accurately reflects the content: a detailed analysis of a Twitter collection around the Mediator trial, presented by the researchers involved.

8/10

The presentation is based on a well-documented case study (Mediator scandal) and a rigorous methodological approach, with collaboration from INA. The speakers are established academics in health history. The analysis is transparent about limitations and noise in data collection.

Markers derived by PSI from the transcript: the creator did not define chapters.

• 02:00Introduction: origins of the project, collaboration with INA, and the BodyCapital project.
• 05:00Background on the Mediator scandal: drug history, Irène Frachon's alert, and regulatory changes.
• 10:00Methodology of the Twitter collection: broad strategy, account selection, and hashtags.
• 15:00Data cleaning process: dealing with noise, filtering tweets, and reconstructing missing tweets.
• 20:00Initial analysis: tweet volume, main actors (media, journalists, Servier), and temporal distribution.
• 25:00Comparison with television coverage: intermedia circulation and differences in focus.
• 30:00Discussion of peaks: Irène Frachon's testimony and the verdict, and the shift to broader health issues.
• 35:00Challenges and lessons learned: noise, missing data, and the value of interdisciplinary collaboration.
• 40:00Conclusion: implications for historical research using web archives and future directions.

• Inathèque — Mentioned as the INA archive platform used for the project.
• INA le lab blog — Mentioned as the seminar's blog for further information.

• BodyCapital project — The project that funded the research, aligning with the presentation's goals.