The Ethical Use of Data

Abstract

Evidence‐based decision‐making requires access to good and reliable data. In recent years, there has been a tremendous increase of data available to institutions, from internal as well as external sources. With this increase in data and its analytics, new challenges and questions arise. One of the most pertinent of which is the ethics and best practices of data collection and usage. This chapter explores the role data have in institutional decision‐making, the need for clear data policies and definitions, and concludes with suggestions for elements of a Data Bill of Rights.

Predictive algorithms and racial bias: a qualitative descriptive study on the perceptions of algorithm accuracy in higher education

“…Responsibility for carrying out an ethical data strategy should include specific stakeholders within the institution with the integration of accountability measures (Tzimas and Demetriadis, 2021). For instance, Mathies (2018) suggested the creation of a data ethics board that functions like an institutional review board (IRB) to provide continuous support for data management and governance. The board would consist of individuals with technical and context expertise and key stakeholders to manage oversight (Mathies, 2018).…”

“…For instance, Mathies (2018) suggested the creation of a data ethics board that functions like an institutional review board (IRB) to provide continuous support for data management and governance. The board would consist of individuals with technical and context expertise and key stakeholders to manage oversight (Mathies, 2018). Prinsloo and Slade(2019) cautioned that a formal data policy addressing the risks of using PAs is essential when creating an ethical data strategy and recommended the creation of a data oversight committee to ensure adherence to the policy throughout the institution.…”

The Good, The Bad and The Ugly: AI in the higher education

“…Ethical and privacy issues are emerging research phenomena and AI implications in higher education (Siemens 2013, Beattie et al 2014Khalil & Ebner 2015;Slade & Prinsloo 2013;Prinsloo & Slade 2016;Mathies 2018a;Mathies 2018b;Reidenberg & Schaub 2018;Klein et al 2020) and deserve special attention. Reidenberg and Schaub (2018, 1) eloquently and reasonably compared education, big data, and student privacy to a combustible mixture.…”

“…The algorithms drive institutional-wide decisions, many higher education institutions use algorithms for admission purposes. Mathies (2018a) suggests that when the admission formula is getting updated, it most likely leads to coding failure. Thus, flawed code may be unnoticed unless a large group of people is affected.…”

“…Mathies (2018a, 89) argues that during a highly competitive admission process, risk assessment algorithm most likely will favour the candidates with advanced placements, hence students who come from low-resourced schools have disadvantage positions and chances are high they will get rejected in comparison with their peers. Mathies (2018a) posits that the concept of ethics in data governance has a degree of relativity, as what is considered ethical in one situation may not be in another, therefore an ethical framework is necessary as a guiding principle. Similarly, Prinsloo & Slade (2016) moot the concept of "Ethics of Care", which is rather value-driven and implies relational understanding in harvesting, analysis, and use of data.…”