Ethical Issues in Big Data and Digital Learning Platforms

• In Event: Data Ethics in Education: The Importance, Issues, and Use Cases

Abstract

Students in PreK-12 and higher education, informal learners, and workers seeking to upskill or reskill access digital learning platforms to prepare for success in school, work, and life. When learners and educators engage with digital learning platforms through immersive simulations, cognitive tutors, interactive videos, massive online open courses, and other technologies, the platforms generate enormous amounts of data that can answer instructional and administrative questions, discover new and non-obvious relationships and patterns, predict outcomes, and automate low-level decisions. Complex and interrelated ethical questions underpin the stages associated with generating, analyzing, and interpreting learning platform data. In this presentation, we discuss the cyclical effects of ethical decisions involving big data and learning platforms and how each stage of the cycle can benefit learning and teaching, how it can introduce bias and further inequity on its own, and, as part of a larger process, how it might increase life outcome disparities among subpopulations. The conditions necessary to support equitable learning through digital learning platforms require ongoing reflection and improvement by educators, educational leaders, designers, and policy makers working together.
The COVID-19 pandemic has heightened the implications and urgency of understanding this cycle and its related ethical questions. As remote and hybrid learning become our new normal and reliance on learning platforms grows, the stakes are high. Inequitable divisions leave already vulnerable populations without access to the resources and rigor they need at the right moment to advance on their learning journey toward more positive and productive life outcomes such as greater proficiency in literacy, numeracy, and 21st century skills, higher wages, better health, and increased civic engagement.
This paper will draw on a cyclical framework developed that examines the divides in data-driven technologies, big data, and learning platforms, and the impact on the ethical use of data in educational practice (Authors, 2021). The divides include: an access divide, a data divide, an algorithmic divide, and interpretation divide, and a citizenship divide. The access divide examines access versus the lack of access to hardware, software, and connectivity; basically a more recent version of the digital divide that has been exposed during the pandemic (see LearnPlatform, 2020). The data divide focuses on representative versus non-representative data. The algorithmic divide examines diverse versus homogeneous teams developing and validating algorithms. An ethical question is whether any single algorithm can make important decisions since educational issues are far more complex than a single algorithm can capture (Daniel, 2019). The interpretation divide focuses on the reasonable interpretations versus misinterpretation, including knowledge of how to make appropriate interpretations and avoiding confirmation bias. The citizenship divide focuses on the skills, economic, health and civic divides that perpetuate structural stigmas. This paper will discuss how the framework with its five divides can be used to inform educational practice in terms of ethical data use.

Author

• Edward Dieterle