To address the major environmental issues of the Anthropocene, like global change, sustainability, and biodiversity loss, we have massive amounts of data made available by recent technological advances, including increased computational power, sensor technologies, publicly available software and data, and Internet connectivity. But to weave that jumbled yarn-ball of data into a coherent tapestry of knowledge and wisdom, current (and especially future) scientists need to learn to deal with data on a completely different scale. Moreover, making such data intensive approaches transparent and verifiable requires a revised notion of what it means to make science reproducible.

This sort of data is a different proposition, and it requires that we outfit our students with different skills, computational skills. Not surprisingly, all of the core competencies highlighted in the recent Vision and Change reports from the AAAS are directly related to core computational data science skills (see the figure below). But getting these skills into already crowded biology curricula is really challenging.

Together with my colleagues Sarah Supp (from Denison U.), Matt Aiello-Lammens (from Pace U.), and Susy Echeverria-Londoño (here in my lab at Kenyon), we have been thinking about how to best get these skills to undergraduate students in the life and environmental sciences, and what the main barriers might be for instructors. We think Hampton et al. (2017) summarize the problem really well:

Essentially, we are attempting to fit more material into already-full courses and curricula, which are taught by people who do not feel prepared to address topics relevant to big data and data-intensive research.

To try to help, we have secured a grant from the National Science Foundation Research Coordination Networks in Undergraduate Biology Education (RCN-UBE) program, to bring together researchers and educators interested in solving this important problem. We call our group the Biological and Environmental Data Education Network (BEDE-Network). Bede is also a word for a pickaxe, as in mining data. Sarah even designed a slick logo.

Our hypothesis is that one of the keys is to train the teachers in the basic skills of reproducible computational data science, so that they feel empowered to add these approaches to their courses as they see fit. Over the next year, we will be reaching out to try to better define the key barriers and to convene a group of people interested in developing portable and scalable instructor training workshops along the lines of those offered by The Carpentries. We have applied to offer our pilot instructor training workshop at the 2019 ESA Meeting next August. Comment here or email me if you want more information or if you want to join us!

To give you the flavor of the sort of skills and practices we are talking about, check out this short series of posts by Sonali Roy and Mary Williams over at the Plantae Blog.