An at-risk learner is one whose behavioural pattern predicts non-completion or poor outcome, identified early enough for intervention to be possible.
What it is
Identification models typically combine access frequency, submission timing, assessment performance and — where available — attention signals. Moodle's Analytics API and Blackboard's Retention Centre both ship variants of this.
The better models baseline against each individual rather than a cohort threshold, because working styles vary enormously.
Why it matters
Published work reports 73.3% prediction accuracy with 40% of an activity still remaining. That is enough warning to act, and not enough accuracy to justify consequences for individuals.
A common confusion
Access-based models miss the learner who logs in reliably, opens everything, and absorbs nothing. Attention data catches that case; event data cannot.
Related
See also disengagement prediction, cyberloafing and learner persistence. See the full glossary for the rest, or identifying disengaged learners early for the wider context.
See where your content loses people
Book a walkthrough and we will show you the engagement data on your own content.