Collaborative session notes:
https://docs.google.com/document/d/1HoisAStxeLcsHDJwxa3uHzafPyQgZGom/edit#heading=h.jsalll5wuxw

This is a Birds of a Feather (BoF) session organized jointly by the Reproducibility IG and the Metadata IG.

The reproducibility of research results, and most especially computational reproducibility, relies on having sufficiently detailed information about the workflow used. This BoF will discuss the most important information to collect, and explore possibilities for collecting it in a way that is convenient for researchers, while allowing it to be recorded in a structured and actionable manner.

Specific goals for the discussion include:

• gauging the interest in joint work on this topic;
• understanding how the work of the Metadata and Reproducibility IGs informs this topic;
• identifying other current work relevant to this topic;
• identifying outputs that a joint WG could develop that would promote the collection and use of metadata for reproducibility.

• 0:00–0:05. Welcome and introductions
• 0:05–0:15. Brief overview of the Metadata and Reproducibility IGs
• 0:15–0:30. Problem statement
  • What is reproducibility metadata?
  • What should it be?
  • Why is it important?
  • How is it different from other metadata?
• 0:30–1:15. Group discussion – key questions include:
  • There’s a wide reliance on narrative for recording methodology, because research can be messy and anyone can write a narrative. Is it possible to introduce a structure that (a) is flexible enough to cope with the full variety of methodologies, (b) feels intuitive to write, (c) provides benefit to the primary researchers, as well as enabling reproducibility?
  • As a parallel to a machine-actionable Data Management Plan (maDMP) (see RDA Standard for maDMPs), becoming a record of how a dataset was managed: could a framework for planning a methodology/computational workflow evolve into a record of that methodology/workflow?
  • What metadata do we want to capture about the computation itself as distinct from the research methodology;
  • How can various frameworks for research planning popularized by open science and increasingly in use (e.g., pre-registration) be leveraged as metadata.
  • Is there interest in a new “RDA Reproducibility Definition and Checklist” WG?
• 1:15–1:30. Wrap up and next steps