Author(s):
Need: Algorithms are a central component of most services we use across a range of domains, including artificial intelligence (AI) applications. These services, platforms, and devices rely on computing and technology professionals – who work as data scientists, programmers, or artificial intelligence (AI) experts – to meet their intended goals. Guiding Question: How do we train future professionals to have an ethical mindset in their understanding, design, and implementation of algorithms? This was the question that prompted the use of a role-playing case study, which we designed, implemented, and studied in an undergraduate technology course. Theoretically, our work is guided by the situated learning paradigm, specifically the need to learn perspectival thinking for decision-making. The ability to make ethical decisions relies to a large extent on the ability of the decision-maker to take context into account – to understand not just the immediate technical need of the work but also larger implications that might even result from unanticipated consequences.Outcomes: Over a three-year period we designed and tested six case studies that encapsulate how a software algorithm shapes decision-making in a complex scenario. Findings from the evaluation of the role-play scenario show that students reported a higher engagement with case study material and a better understanding of the scenario due to taking on a specific role they played. Analysis of pre-and post-discussion assignments shows a shift in their perspective of the case, further supporting the overall goal of developing a more situated understanding of ethical decision-making. Role-play case studies provided students the capability to employperspectival thinking, link micro, meso, and macro level concerns, and use an ethical mindset to examine a case. In practical lessons from our work, we discuss the importance of collaborative learning, designing interest-based situations, and role preparation. Finally, we outline different methods for assessing student learning.Broader Impacts: This work has implications for the design and use of systems that support technological equity and justice. Through our project, we have introduced over 300 students to the case studies and they have also been implemented at other institutions. The cases are available for other to use through the Online Ethics Center repository among other sources.
Coauthors
Ashish Hingle, George Mason University, Fairfax, VA; Huzefa Rangwala, George Mason University, Fairfax, VA; Alex Monea, George Mason University, Fairfax, VA
