Open Science

Badges distributed by Zephyr to recognize open and reproducible research will incentivize authors to archive and share data, code and models associated with their research articles and generally enable other authors to more easily build upon that research.

Meetings

2019-01-14 The Open Science Project Management Group met at the TRB Annual Meeting in Washington DC. We discussed the mechanisms to enable and promote data sharing and transparency in transportation research.

Note that seating may be limited and will be prioritized for PMG members

Governance

This project is overseen by a board-approved Project Management Group (PMG) as follows:

  • Greg Erhardt, University of Kentucky (chair)
  • Rolf Moeckel, TU Munich
  • Leah Silen, NumFocus
  • Kay Axhausen, ETH Zurich

More to come! Please let Greg Erhardt([email protected]) know if you are interested in contributing

Approach

Workplan:

  1. Establish a PMG [in process]
  2. Evaluate whether an existing program serves our needs
  3. Agree upon standards
  4. Agree upon review/certification process
  5. Get buy-in and implementation support from academics, journals, Zephyr.

In defining the program, the PMG should address the following questions:

  • Is this a problem that warrants our intervention?
  • What is the unit that we are badging? Our initial assumption is that it would be a research paper.
  • What are our standards? Some examples might be:
  • If model estimation results are presented, the model estimation files should be archived.
  • The model estimation scripts or code should also be provided for the final models presented.
  • If the paper includes figures, the data and associated geographic files underlying those figures should be archived.
  • If the paper presents the results of travel model runs, the model runs themselves should be archived.
  • Metadata should be provided describing files and field names.
  • A knowledgeable reviewer should be able to run the models/scripts/estimations to recreate the tables and results presented in the paper.
  • If there are privacy restrictions or other valid reasons that the data cannot be shared publicly, the authors should offer an explanation, and make a statement about how their results could be independently verified. For example, this might be by contacting the authors directly or signing a limited non-disclosure agreement.
  • Is there a single badge, or is there value in considering different tiers (silver and gold), or dimensions (a la scouts: canoeing, plumbing, etc.)
  • What is the process for awarding badges? An additional stand-alone review? Integrated with the normal peer review process?
  • What strategies can we employ to make a badge a positive incentive?

Feedback + Involvement

Feedback or requests to be involved can be sent to to the Open Science chair: [email protected].