Shifting to Data Savvy: The Future of Data Science In Libraries report

dScholarship@Pitt
Report release - 19 March, 2018

The Data Science in Libraries Project is funded by the Institute for Museum and Library Services (IMLS) and led by Matt Burton and Liz Lyon, School of Computing & Information, University of Pittsburgh; Chris Erdmann, North Carolina State University; and Bonnie Tijerina, Data & Society. The project explores the challenges associated with implementing data science within diverse library environments by examining two specific perspectives framed as ‘the skills gap,’ i.e. where librarians are perceived to lack the technical skills to be effective in a data-rich research environment; and ‘the management gap,’ i.e. the ability of library managers to understand and value the benefits of in-house data science skills and to provide organizational and managerial support.

This report primarily presents a synthesis of the discussions, findings, and reflections from an international, two-day workshop held in May 2017 in Pittsburgh, where community members participated in a program with speakers, group discussions, and activities to drill down into the challenges of successfully implementing data science in libraries. Participants came from funding organizations, academic and public libraries, nonprofits, and commercial organizations with most of the discussions focusing on academic libraries and library schools.

What is Data Savvy? 

A family of data science roles has been identified, which can be characterised by the real-world requirements for actual positions, as described in two related smallscale studies (Lyon & Mattern, 2017; Lyon, Mattern, Acker, & Langmead, 2015). The six roles are: data archivist, data curator, data librarian, data analyst, data engineer, and data journalist. While all of these roles have been framed as data science roles, other framing which comes from the corporate sector has tended to describe only data analyst-type roles as data scientists. However, in reality, there are a wide gamut of roles—‘data savvy’ roles—that orbit within and around the world of data scientists. Data savvy librarians gain familiarity with the datasets, understand technical methods and techniques, and speak multiple disciplinary languages allowing them to work more closely with researchers or the public. Some librarians engage more deeply, becoming technically proficient in data preparation and analysis, allowing them to work with data, automate workflows, and become fully embedded in research teams. In other words, data science exists more or less on a spectrum, depends on an institution’s size and mission, and spans work requiring the deep statistical and software engineering skills, to work that focuses on advocacy, policy, communication, and data management. Being data savvy is an essential ingredient of all of these roles.

http://d-scholarship.pitt.edu/33891/

http://d-scholarship.pitt.edu/33891/1/Shifting%20to%20Data%20Savvy.pdf