In Conversation with Jeremy Roschelle: Learning Analytics, Measurement, and the Search for What Counts as Evidence

This is presented by the Assessment and Measurement to Support Classroom Learning (AMSCL) SIGIMIE.

This session is a conversation with Jeremy Roschelle, Executive Director of Learning Sciences Research at Digital Promise, exploring the increasingly blurred boundary between learning analytics, psychometric measurement, and AI. Jeremy has a background in learning sciences and research on technology in STEM learning. He’s been increasingly focused on infrastructure for research (leveraging digital learning platforms) and for AI (recruiting educational public goods). Central tensions that emerge in this work include: can the platforms that students really use ever generate data we can trust? can the constructs we can measure in controlled situations be deployed “in the wild” in everyday platforms? How can the strengths of the psychometric community with regard to topics like fairness and validity connect with needs in the AI community for datasets, models, and benchmarks? Underlying these tensions is a driving question that Roschelle has struggled with throughout his career: how can intellectual rigor and large-scale impact be better connected?

This session experiments with a format that breaks from the traditional webinar mold. We’d love your active participation either by raising your hand live in the Zoom meeting, or submitting a question ahead of time using the form below.