Schedule of Events | Search Abstracts | Invited Symposia | Symposia | Poster Sessions | Data Blitz

Sketchpad Series

Detecting mind wandering in real classrooms using eye-tracking and machine learning

Poster Session B - Sunday, March 8, 2026, 8:00 – 10:00 am PDT, Fairview/Kitsilano Ballroom

Xiaorui Xue1 (), Ido Davidesco1, Na’ama Av-Shalom1, Marina Vasilyeva1, Nanyu Zhang1, Rachael Sabelli1, Jason Geller1; 1Boston College

The dynamic nature of attention means that students often shift their focus from the learning task to unrelated thoughts—a phenomenon known as mind wandering (MW). Although MW can have certain cognitive benefits, it is generally believed to hinder learning. Self-report measures suggest that students spend about 30% of class time MW. To date , most prior cognitive neuroscience research on MW has been conducted in controlled laboratory settings that differ substantially from real-world classroom environments. To address this gap, we are conducting a classroom-based eye-tracking study. We collected data from 15 undergraduates, each of whom attended two regular class sessions while wearing portable eye-tracking glasses (Neon, Pupil Labs). Each class session (N = 6) included 6–8 attention probes, asking students what they were thinking about immediately before the probe appeared and how effortful the prior material was. A short quiz at the end of each session assessed learning outcomes. We plan to examine how self-reports of mind-wandering (probe-caught) and effort correlate with eye-movement measurements (e.g., fixation duration, gaze dispersion, and pupil size) and performance. As data collection continues, ongoing analyses aim to develop machine learning models that predict self-reported MW from continuous eye-tracking data. We will also examine how eye-tracking indicators and self-reports jointly predict learning performance. This work seeks to deepen understanding of attention fluctuations in naturalistic learning environments and improve methods for detecting them.

Topic Area: ATTENTION: Other

CNS Account Login

CNS_2026_Sidebar_4web

March 7 – 10, 2026