
Motivated Students Still Drop Out: Using AI Behavioral Signals to Intervene Before Disengagement Becomes Attrition:
Changing Higher Ed podcast 319 with host Dr. Drumm McNaughton and guest Raz Dar
In this episode of Changing Higher Ed, Dr. Drumm McNaughton speaks with Raz Dar, CEO of Elvee, whose work with U.S. colleges centers on reading behavioral data for early signals of student disengagement. The discussion draws on Elvee’s survey of 1,050 U.S. college students, which found that 86% start college highly motivated, yet 42% came close to dropping out at some point and only 17% of struggling students ever request help. The conversation reframes attrition as a detection failure rather than a motivation deficit.
What you’ll gain from listening to this podcast:
✓ Detection Timing: Why grades and attendance flag struggling students only after many are already halfway out the door, and how behavioral signals like study time, login frequency, and forum participation surface risk before exam period.
✓ The










