Open, Data-driven Science for Decision Makers

Goal: Enable critical evaluation of research proposals according to the best practices of open science.

Rationale

"Open Science" and "Big Data" are popular buzzwords in biological research. A laundry list of publications [REFFS] indicate that many researchers are working on building the skills needed to be computationally savvy biologists. While significant resources are available to help them come up to speed, we are unaware of efforts that target the audience of decision makers who need to be informed about the progress and value of open science and computation. These decision makers need to be able to understand the wider implications of open science, in order to influence and adjust policy in line with new developments from data science communities.

We reason that if cutting-edge investigators are only recently navigating through new technologies and practices, the audience of 'decision makers' are in a much earlier stage of knowledge acquisition and require input from technology and best practice leaders.

However, the need of this decision-maker audience is not deep technical knowledge but a broad understanding of relevant topics. An increased understanding of the value and reasoning behind practices and technologies can empower this audience to make more helpful suggestions and critical evaluations of proposals for data-intensive research and/or learning and training strategies.

Audience

Funding agencies/program managers, grant reviewers, investigators, and educators who are/will be reviewing research proposals or training curricula that substantially make use of data integration, software development, and large-scale computation in the biological and biomedical sciences.

We make no assumption of familiarity with the technical tools or practices of open science, bioinformatics, computational biology, etc. This proposed work structure uses a base model of "raising the bar", not training the audience to be experts

Format and development

When first deployed, we will teach this as a one-day workshop to a live audience. As the content progresses and we have feedback from an initial presentation, we can evolve the format to be suited for a self-guided course, decentralised training through virtual webinars, or additional in-person workshops.

Each workshop module will involve:

• Useful background reading materials and references
• A minimal introductory lecture
• Problem-based hands-on exercises
• Additional learning resources including best practice reference guides

Workshop objectives

At the completion of this workshop, an attendee should:

• Understand key principles of open science
  • FAIR principles
  • Open access and publication
  • Reproducible science
• Understand the relationship between technology and open science
  • How software enables open science
  • 'Responsible' software development and use
  • Tools of open science
    • Automation
    • Version control
    • Open-source software for statistical and biological analysis
    • Large-scale computing
• Be able to write and critique a data management plan
  • Identify common, fixable pitfalls
  • Spot 'red-flags' and closed terminology
  • Provide actionable feedback
• Be aware of learning resources
  • Resources for improving computational skills
  • Resources for improving statistical skills
  • Resources for improving software project management skills
  • Resources for designing curriculua
• Be able to explain the value of open science
  • Explain the benefits of open science
  • Deliver balanced and realistic feedback
  • Overcome objections to open science
  • Use open science to fulfil your values and mission