By Louis DiPietro
How do we learn to learn effectively? It’s a fundamental question at the heart of education and has particular resonance to leading online learning researchers like René Kizilcec, an assistant professor of Information Science in the Cornell Ann S. Bowers College of Computing and Informaiton Science.
“Learning is a fundamental activity we practice for a lifetime,” said Kizilcec, who also directs Cornell’s Future of Learning Lab. “People don’t often think about learning, especially outside formal educational environments, but being intentional and strategic about learning helps us achieve our learning goals.”
An intentional approach to learning benefits students, especially in online learning environments, where self-direction and agency are critical to student success. What factors hinder learning online, and how to encourage and aid students on their learning journeys are critical and timely questions, as more students than ever shift to online and hybrid learning environments due to the pandemic.
Kizilcec’s research offers answers to these questions. He has examined nearly every facet of tech-assisted learning, from the design and impact of technology in education and the architecture of online classrooms – at major universities and in low-resource countries – to the algorithms used to predict student dropout rates and inclusive marketing strategies used to advertise online courses.
His recent research awards include a Google Award for Inclusion Research and a Jacobs Foundation Research Fellowship. He also launched the Behavioral Intervention Research Infrastructure (BIRI), an educational research infrastructure to increase access for behavioral and psychological scientists to carry out large-scale experiments in online learning environments.
“The way people learn is changing, and my research lab works to advance our understanding of the impact of technology in education and how it can be an engine of equality in society,” Kizilcec said. “The research we conduct informs the design of online learning environments and identifies effective and scalable strategies to support learners.”
AI, Fairness, and Education
Artificial intelligence permeates nearly every field, including education, where algorithmic systems and predictive models provide academic support and insights to students, teachers, and administrators. Improving timely college graduation is one area where predictive modeling can help universities identify students who may be at risk of dropping out, with the goal to intervene and help those students get back on track. However, Kizilcec is leery of a one-size-fits-all approach to such computational models.
“Companies that offer colleges these kinds of services may not train new models for each university due to resource constraints,” he said. “The question is, how well do these models translate across institutions? Will it lead to inaccuracies? More importantly, how might these models affect under-represented and minoritized groups of students?”
Kizilcec, together with Christopher Brooks, an assistant professor of information at the University of Michigan, intend to prod these questions in their project, “Investigating How to Mitigate Bias in Predictive Models of Student Success Across Diverse Institutions,” which received a Google Award for Inclusion Research.
Effective, Affordable Hybrid Learning
Hybrid learning – the mix of traditional, in-class studies paired with supplemental, online learning on platforms – has become a fixture of modern education in light of the pandemic. While hybrid learning’s use and potential are immense, its impact on learning outcomes and academic pathways in low-resource areas in the U.S. and internationally is not well understood. As part of the Jacobs Foundation Fellowship, Kizilcec will explore how hybrid learning can be used to support learning in secondary schools in low-resource areas. In turn, these insights will inform the design and implementation of educational technologies, for all students.
“We aim to explore how teachers are using technology to augment learning in a classroom setting, and the contexts in which students are using online learning platforms for hybrid learning,” he said. “The goal is to understand the real-world use and experiences to improve technology design, marketing, and ultimately learning outcomes.”
Enabling Fellow Researchers
Kizilcec’s extensive understanding of the online learning landscape is also helping fellow education researchers conduct their own research on the effectiveness of interventions. A primary research focus for Kizilcec, these interventions – including brief study planning activities, motivational videos, and self-reflective writing exercises – are embedded into an existing online course. A number of interventions have been shown to boost completion rates among online learners, though, as Kizilcec discovered in a massive research project involving 250,000 students from nearly every country, such interventions are “not a silver bullet.”
Carrying out such large-scale research to test the effectiveness of interventions is resource intensive: researchers need access to large sample sizes with thousands of students; there are student privacy safeguards to respect, and coordinating among collaborators from participating institutions and their respective review boards can further delay an already slow process.
Enter Behavioral Intervention Research Infrastructure (BIRI), a Schmidt Futures-funded research infrastructure Kizilcec and colleague Ryan Baker of the University of Pennsylvania launched last year in collaboration with Realizeit, an adaptive learning platform.
“After running many intervention studies, I realized that there should be a better way to enable more scientists to conduct intervention research to increase the pace of scientific progress in this space,” he said. “Researchers from around the world can come up with great ideas for how to support students in college courses, and BIRI enables them to test them experimentally and at scale.”
Looking forward, Kizilcec will chair the 2022 ACM Conference on Learning at Scale, which will take place on Cornell Tech’s campus in June. In its ninth year, the conference gathers researchers who investigate tech-learning environments with many active learners, like massive open online courses (MOOCs), mobile learning applications, intelligent tutoring systems, and more.
"So much has happened in the education technology space over the last few years. As general conference chair, I want the conference program to engage researchers, technologists, funders, policy makers, and members of the Cornell community," he said. "Cornell may not have an education school, but it no doubt is a thought leader in learning.”