This Episode will air at 5:30 pm EDT on Saturday June 18th
Big Data in Education with Avriel Epps-Darling, Andrew Ho, and Katina Michael
A large amount of learning happens in digital environments that can preserve data about every click, scroll, and touch. These data can support powerful predictions about a learner’s interests, abilities, and actions. What can we do with these data, and what principles should guide our use of data? In this episode, Avriel Epps-Darling, Andrew Ho, and Katina Michael discuss the promise and peril of “big data” in educational contexts. Each will provide examples where these data have helped learners and improved equity, as well as situations where they have caused harm, even with the best of intentions.
These ideas have implications for data use in Learning Management Systems, Massive Open Online Courses, and Early Warning Indicator Systems, in contexts including K-12, higher education, and online learning. Panelists will also discuss principles for ethical and equitable use of data in education, including beneficence, transparency, informed consent, privacy, bias detection and bias prevention.
More about the topic (references and links) and about our guests below the video.
References
Big Data Analysis in Higher Education: Promises and Pitfalls https://er.educause.edu/-/media/files/articles/2016/8/erm1652.pdf
What All Data Scientists and Business Leaders Need to Know about Data Ethics and AI Bias https://www.correlation-one.com/blog/data-ethics-ai-bias-data-scientists-business-leaders-need-to-know
Just How Much of Higher Education Can Be Automated? https://slate.com/technology/2021/03/trolley-solution-response-essay-automated-higher-education.html
https://www.youtube.com/watch?v=TZpJX_avpeE&feature=emb_logo (panel: Shiv Ramdas, Katina Michael, moderator Punya Mishra)
Detecting and preventing “multiple-account” cheating in massive open online courses https://www.sciencedirect.com/science/article/pii/S0360131516300896
Episode Guests
Avriel Epps-Darling is a computational social scientist, PhD student in Human Development at the Harvard Graduate School of Education, M.S. student in Data Science at Harvard’s School of Engineering and Applied Sciences, Presidential Scholar, and Ford Foundation predoctoral fellow. Her doctoral research focuses on how adolescent racial and gender identity development is influenced by bias in digital products. Her work illuminates the impact of algorithm design and computer-mediated social expectations–communicated through personalized recommendation systems and information filters–on youths’ beliefs and behavior.
Previously, Avriel was a visiting research fellow at Spotify, an ed-tech startup founder, and recording artist. She has spoken at conferences, universities, and tech organizations such as Google and TikTok on the topic of algorithmic bias and fairness. In 2020, Facebook recognized her as an Emerging Scholar Finalist. Avriel has also taught and designed courses for Harvard and EdX on Digital Privacy, Data Science Ethics, and Adolescent Development. Her writing has been featured in academic journals and handbooks as well as popular publications like The Atlantic.
Andrew Ho is the Charles William Eliot Professor of Education at the Harvard Graduate School of Education. He is a psychometrician whose research aims to improve the design, use, and interpretation of test scores in educational policy and practice. Ho is known for his research documenting the misuse of proficiency-based statistics in state and federal policy analysis. He has also clarified properties of student growth models for both technical and general audiences. His scholarship advocates for designing evaluative metrics to achieve multiple criteria: metrics must be accurate, but also transparent to target audiences and resistant to inflation under high stakes.
Ho is a director of the Carnegie Foundation for the Advancement of Teaching and has served on the governing boards for the National Council on Measurement in Education and the National Assessment of Educational Progress. Professor Ho chaired the research committee for the Vice Provost for Advances in Learning (VPAL) at Harvard University, which governed research on “massive open online courses” (MOOCs). He holds his Ph.D. in Educational Psychology and his M.S. in Statistics from Stanford University. Before graduate school, he taught middle school creative writing in his hometown of Honolulu, Hawaii, and high school Physics and AP Physics in Ojai, California.
Katina Michael (BIT, MTransCrimPrev, PhD) is a professor at Arizona State University, a Senior Global Futures Scientist in the Global Futures Laboratory and has a joint appointment in the School for the Future of Innovation in Society and School of Computing and Augmented Intelligence. She is the director of the Society Policy Engineering Collective (SPEC) and the Founding Editor-in-Chief of the IEEE Transactions on Technology and Society. In 2007, Katina was awarded a teaching scholarship at the University of Wollongong toward faculty-based staff development initiatives to disseminate good teaching practice through sharing of expertise within faculties. She championed the concept of total curriculum management.
Katina has continued to innovate within her teaching practice implementing distributive leadership strategies across interdisciplinary fields including: technology/engineering, social sciences, humanities, business and law. Katina brings a unique understanding of the application of AI within higher education, and has much to contribute to the learning analytics space, especially the secondary use of data.