Enjeux et méthodes de la reproductibilité scientifique | UMI-SOURCE (UVSQ)

Value of the Information & Strength of the Argument

The value of the information is high, as it provides concrete, actionable advice for improving reproducibility in empirical research. Galiana’s demonstration of using linters and formatters, and his emphasis on project structure, are practical and directly applicable. Espinosa’s discussion of the replication crisis and pre-registration is well-argued, drawing on established literature and personal experience. The argumentation is solid, though it relies more on anecdotal evidence and general knowledge than on specific data or citations. The speakers effectively convey the importance of reproducibility and offer a clear hierarchy of practices, from basic code hygiene to advanced containerization.

Scientific Rigor, Source Quality, Title Accuracy

The scientific rigor is moderate: the speakers are credible experts, but they do not provide formal citations or references to specific studies. The quality of sources is therefore based on their authority and the general acceptance of the tools and concepts discussed. The title accurately reflects the content, and the presentation is well-structured. No comments were provided for analysis.

The title accurately reflects the content: a conference on the challenges and methods of scientific reproducibility, covering both technical tools and pre-registration practices.

8/10

The speakers are established researchers (INSEE, CNRS) with practical experience in reproducibility. The content is grounded in established practices and tools (Git, Quarto, Docker) and addresses the replication crisis with references to psychology. However, no formal citations or data are provided, and the presentation is largely based on personal experience and opinion.

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

• 00:00Introduction by the moderator, presenting the speakers and the topic.
• 02:00Lino Galiana begins his talk on technical practices for reproducibility, introducing Git and code readability.
• 10:00Galiana discusses project structure and the importance of organizing files for reproducibility.
• 20:00Galiana demonstrates the use of Quarto for reproducible reports and discusses virtual environments and Docker.
• 30:00Galiana concludes his talk and hands over to Romain Espinosa.
• 40:00Espinosa introduces the replication crisis and its implications for scientific credibility.
• 50:00Espinosa discusses pre-registration and registered reports as solutions, and the resistance in economics.
• 1:00:00Espinosa concludes and opens the floor for questions.

• No specific sources cited in the video — The speakers did not explicitly cite any external sources during the talk.

• Open Science Collaboration — The talk references the replication crisis in psychology, which is documented in this large-scale replication project.