Established in 2013, we are a community of researchers and students with a vision for improving the computational reproducibility of our research, and research produced by other members our fields. On this page you will find details about the community and our recent events. We are supported by the UW eScience Institute, and our current chair is Ben Marwick.

Join our email list

Our Core Activities

Training

We bring Software Carpentry and Data Carpentry workshops to give UW researchers a quick start into efficient, open, and reproducible research

Guest Speakers

We bring technical experts and community leaders to UW to inspire and educate the campus community

Teaching openness and reproducibility

We organise workshops to share tools, strategies, ideas and inspiration for teaching reproducible and open research at all levels.

Sharing information and resources

We discuss the issues around tools and practices to enhance data sharing, preservation, provenance, and reproducibility. Visit the UW Libraries’ Research Guide on Reproducibility to learn more.

Our values, motivtion, and vision

We value improvements in tools to support sharing and reproducibility, the design of repositories for collecting reproducible experiments and observations, techniques to query and analyze shared data, and workflows to facilitate re-use.

Open science is closely related to reproducibility, sharing needs in both technological and software advances as well as in changes to the research culture. However, reproducible research need not be completely open and there are cases where it will not be due to privacy issues, proprietary data, or reluctance on the part of scientists to share software that has taken years to develop or data that was hard to acquire.

We are motivated by the view that reproducibility starts at the level of individual scientists or groups of collaborators being able to reproduce and build upon their own work, as well as to later verify its correctness if necessary. Our vision of open science to work towards broader sharing that typically requires and facilitates reproducibility.

For implementing sustainable policy on reproducible science we are motivated by, and working from, the National Academies of Sciences, Engineering and Medicine reports Open Science By Design: Realizing a Vision for 21st Century Research (2018), Reproducibility and Replicability in Science (2019) and the Center for Open Science’s Transparency and Openness Promotion guidelines.

Recent Events

‘R Markdown for Scientists: A hands-on workshop’, a workshop by Nicholas Tierney (13 Feb 2020)

This workshop will teach participants how to write a reproducible scholarly report or paper using R Markdown. For a scientific report to be completely credible, it must be reproducible. The full computational environment used to derive the results, including the data and code used for statistical analysis should be available for others to reproduce. R Markdown is a tool that allows you to integrate your code, text and figures in a single file in order to make high quality, reproducible reports. A paper published with an included R Markdown file and data sets can be reproduced by anyone with a compute. The workshop is based on Dr Tierney’s book, online at https://rmd4sci.njtierney.com/ Nicholas Tierney, Ph.D. is a Lecturer at Monash University, working with Di Cook and Rob Hyndman in the NUMBAT group. Nick is a central member of the rOpenSci community, a collective that works to make science open using R. He was the lead organiser of the rOpenSci ozunconf for 2016, 2017, and 2018. Dr Tierney also gave a seminar at the eScience Institute on Friday 14 Feb, titled ‘A Realistic Guide to Making Data Available Alongside Code to Improve Reproducibility’, based on his pre-print with Karthik Ram, online at GitHub. Slides are now online!

‘R and friends for better science in less time in big (and small) team collaborations’, a seminar by Julia Stewart Lowndes (19 Nov 2019)

R has demonstrable potential to accelerate scientific research, since it not only provides powerful analytical ‘tools that increase reproducibility but also creates a new frontier for communication when combined with the open web. But thus far, the power of R and friends have largely been harnessed by individuals; how do we harness this power as teams in science, big and small? I will discuss how our Ocean Health Index team dramatically improved how we work by creating an analytical workflow with R – and by peer-learning and peer-teaching the skillsets we needed on-the-job by engaging with the broader #rstats and #openscience communities (Lowndes et al. 2017; Nature Ecology & Evolution). I will also discuss Openscapes, which I developed as a Mozilla Fellow. Openscapes is a mentorship program to engage and empower science teams with open data science tools and practices (openscapes.org). With both the Ocean Health Index and Openscapes, my work aims to help catalyze the fundamental shift needed in scientific culture where we value and prioritize data science, collaboration, and open practices, and provide training and support for our emerging scientific leaders – not only as individuals, but as teams. Julia Stewart Lowndes Ph.D., is a marine ecologist, data scientist, and Mozilla Fellow at the National Center for Ecological Analysis and Synthesis (NCEAS), USA. As founding director of Openscapes and science program lead of the Ocean Health Index, and co-founder of Eco-Data-Science and R-Ladies Santa Barbara, she works to increase the value and practice of environmental open data science. She earned her PhD at Stanford University in 2012 studying drivers and impacts of Humboldt squid in a changing climate. Slides and video are now online!

‘Teaching Integrity in Empirical Research: The Pedagogy of Reproducible Science in Undergraduate Education’, a seminar by Richard Ball (29 Oct 2019)

Richard Ball is the founder of Project TIER (https://www.projecttier.org/), and is here as a guest of the UW eScience Reproducible Research Special Interest Group and the UW Center for Teaching and Learning. Many journals now require authors to submit extensive public-facing documentation to support empirical papers. In this talk, Professor Ball discusses teaching research transparency to students in quantitative fields of study. Project TIER (Teaching Integrity in Empirical Research) integrates transparent and reproducible research methods into instruction by developing standard protocols for conducting and documenting statistical research. Enacting these protocols ensures that all reported results are computationally reproducible and that the methods employed are immediately legible to other researchers. Professor Ball gives an overview of the approaches to research reproducibility that Project TIER promotes, discusses the resulting educational benefits, and considers lessons for professional research practice that emerge from using the Project TIER system. He also identifies potential opportunities for collaboration between Project TIER and UW faculty to promote transparency in education and research. Slides and video are now online!

‘How to improve the value of your research by making it verifiable’, a seminar by Corina Logan (13 Aug 2019)

There is a desperate need to reform the production and dissemination of scholarly outputs to increase transparency, reproducibility, timeliness, and academic rigor. Evidence suggests that open practices actually help researchers rather than hinder them. Logan will discuss what researchers are doing to address these issues by sharing ways to facilitate higher quality research and tackle biases in this rapidly changing world of academia and scholarly publishing. Corina Logan investigates how behavioral flexibility relates to invasion success in grackles (an urban bird) and humans as a Senior Researcher at the Max Planck Institute for Evolutionary Anthropology. She co-leads the #BulliedIntoBadScience campaign where early career researchers are working to change academic culture to adopt open research practices to improve research rigor. You can sign the letter and endorse the campaign at http://bulliedintobadscience.org. Follow along as she learns about grackles, implicit biases, and verifiable research on Twitter @LoganCorina, and at her website (http://corinalogan.com). Slides are now online!

Writing Reproducible & Executable Scientific Papers with R & Python: a Hands-On Workshop (10-11 Jun 2019)

This workshop is co-sponsored by the UW eScience Special Interest Group on Reproducible Research and Open Source Software and the UW Libraries. This workshop is aimed at researchers who write empirical journal articles and want to make it easier for others assess the validity of their work, reuse their work in new research, and enhance public trust in research. We will discuss using Binder, Stencila, Code Ocean, and Whole Tale for writing reproducible research. Slides and video are now online!

Teaching Undergraduates Reproducible Research: A Short Workshop for UW Instructors (19 Apr 2019)

Teaching students integrity when doing empirical research is a high priority for instructors. However, increases in the volume and variety of data means that our students face new challenges in learning how to efficiently and transparently get insights from these data. In response to these changes, we recognise that new approaches to teaching and learning are necessary to ensure our students are equipped with methods and tools to produce credible science. To learn about these new approaches, the UW eScience Institute is organising a short workshop on teaching undergraduates reproducible and transparent research methods. The goal of the workshop is to share information about how to teach reproducible and transparent research, especially in fields outside of computer science. Doing reproducible and transparent research is increasingly becoming a priority in many research communities, and we want UW undergraduates to get the best training on how to do this so they can be at the forefront of their disciplines. We also want to foster a community of instructors who are promoting open science practices on campus, following the recommendations of the National Academies’ 2018 report ‘Open Science by Design: Realizing a Vision for 21st Century Research’ (https://doi.org/10.17226/25116). More details.

Launch of the UW Libraries’ Research Guide on Reproducibility (2 Nov 2018)

With the rapid growth of data science in academia and industry it can be challenging to navigate the numerous tutorials, blog posts, and other sources of information. It can also be hard to know which tools and methods are trustworthy and widely used. To help with this challenge, Liz Bedford led the group to research and write an authoritative official Research Guide on Reproducibility in research and teaching. Liz is the Scholarly Publishing Outreach Librarian with the UW Libraries, and her outstanding new guide is an excellent complement to the UW Libraries’ guides on data management and open access.

Book launch: The Practice of Reproducible Research (27 Jan 2017)

A new collection of papers has been published that includes case studies, lessons learned and the potential future of reproducible research practices. Several members of the UW community contributed chapters, in addition to other researchers affiliated with the other two Moore-Sloan Data Science Environments: the Berkeley Institute for Data Science at UC Berkeley and the Center for Data Science at New York University. The book can be freely accessed online at https://www.practicereproducibleresearch.org/