By Robin Brooks
Ben Marwick, eScience data science fellow and member of the steering committee, has co-authored a commentary piece published May 25, 2017, in Nature Neuroscience. The article “Toward standard practices for sharing computer code and programs in neuroscience” provides tools and guidelines to follow to encourage neuroscientists to practice open science.
Marwick calls the concept “revolutionary” as there can be multiple benefits to sharing code, including discovering potential collaborators, locating flaws early on and correcting them, and allowing others to benefit from using work that has already been done. “Code is the expression of many different decisions,” he said, and he and his co-authors are “trying to normalize” this type of sharing for the benefit of all.
Thirteen other co-authors, along with Marwick, drafted the paper quickly after attending a conference back in 2014. In fact, Marwick was at the conference as a result of his posting an eScience slide-deck presentation on GitHub which the conference organizers found. They felt he would be a good fit with the other collaborators and invited him to join. It was a “good return on the investment” he says of his attendance, which, after months of collaboration, has resulted in the publication. He and the other authors drafted the commentary together in a private repository on GitHub, using the same tools they advocate in their article.
Marwick championed the Institute’s embrasure of transparency and data sharing, and its collaborative nature, saying “an undervalued element [of the Institute] is the creation of cultures and communities” which help to foster relationships, provide resources and incentives, spark different research interests and encourage participants to tackle challenging projects.
“Reproducibility is literally the definition of science, and as science moves from the lab to the computer, code sharing must be at the core of how we conduct research and train students” said Bill Howe, associate director of the eScience Institute. “The reproducibility working group at UW [which Marwick is a part of] has been actively working to develop standards, policies, and tools to support reproducibility and code sharing.”
“My research interests would never have come this far without the opportunity to explore these tools and interests. … Broadly speaking my interest is reproducibility … and everything eScience kind of stands for” Marwick said. “Drafting the commentary was a great experience” he said, adding that “many of the concepts and ideas from my work with eScience went into the article”.
Learn more about Marwick and the paper in UW Today’s article “UW anthropologist: Why researchers should share computer code”.