“Ethics When You Least Expect It”: A Modular Approach to Data Ethics Instruction. / Shanahan, Hugh; Bezuidenhout, Louise; Quick, Rob.

2018.

Research output: Contribution to conferencePaper

Submitted

Abstract

Modernity is increasingly characterised by the generation and reuse of data. In the wake of this “data deluge” it is increasingly apparent that individuals are critically needed to manage, curate and analyse data online. In particular, programming and computational skills are urgently needed in order to build and maintain data infrastructures and maximize the use of available data.
Educating such individuals is an urgent, and the field of data science pedagogy has rapidly expanded. Within this field, there is a small, but growing, discussion on how best to integrate topics such as ethics and responsible conduct into curricula. Nonetheless, teaching data science ethics poses a wide range of challenges relating to the multi-disciplinary nature of student cohorts, the highly variable array of topics covered by courses, and the practice-oriented nature of these courses pose additional challenges.
This paper describes a novel approach to teaching data ethics first used in the CODATA-RDA School for Research Data Science in 2017. The design of the ethics instruction took to heart the practice-oriented nature of the general course, as well as the modular format of instruction, based on the Software/Data Carpentry model. As a result, ethics content was closely linked to the tool or concept being taught in each module, and disseminated via interactive short “ethics prompts”. It was felt that this approach successfully located ethics within daily data science practice and allowed students - regardless of background - to see how small actions built up into larger ethical concerns.
Original languageEnglish
StateSubmitted - 2018

ID: 30005367