🧠 AI Ed Wiki

📄 Research Article

Engagement Assessment in Video Learning

◐ medium · 2026-05-08adaptive-learning, learning-analytics, affective-computing, higher-ed, feedback-loop, self-regulated-learning, affective-tutoring, adaptive-learning-systems, formative-assessment

EduGage (Leng et al., 2026) addresses a core challenge: in online/video-based learning, learners must self-regulate their engagement with instructional materials.

Sensor-based momentary assessment of engagement in self-guided video learning environments.

The Engagement Problem in Video Learning

EduGage (Leng et al., 2026) addresses a core challenge: in online/video-based learning, learners must self-regulate their engagement with instructional materials.

Dimensions of Engagement

DimensionMeasurementRelevance to Learning
AttentionalEye tracking, gaze patternsSustained focus on content
EmotionalFacial expression, sentimentPositive affect supports persistence
CognitivePhysiological signals, task performanceDeep processing vs. superficial viewing

Sensor-Based Momentary Assessment

Traditional engagement measures:

  • Post-hoc surveys: Retrospective bias, low temporal resolution
  • Self-reports: Introspection difficulty, social desirability bias

    • EduGage contributes: Real-time sensor fusion for momentary assessment during video learning.

    Technical Approach

  • Sensors: Webcam (facial analysis), interaction logs (pause, rewind, speed)
  • Assessment: Momentary (in-the-moment) vs. retrospective
  • Feedback loop: Real-time reflection prompts based on engagement state

    • Connection to Adaptive Learning

    This enables adaptive interventions in video learning:

    1. Detect disengagement (gaze diversion, prolonged pauses)

    2. Trigger scaffolds (reflection prompt, content re-summarization)

    3. Close loop: Learner reflects → re-engages → improved outcomes

    This aligns with Adaptive Learning principles: real-time learner modeling → personalized intervention.

    Implications for ITS

    Intelligent tutoring systems increasingly include video components (e.g., worked examples, concept explanations). EduGage's approach enables:

  • Multimodal engagement tracking (cf. Multimodal AI Tutoring, Affective Tutoring)
  • Just-in-time scaffolds when engagement drops
  • Self-regulated learning support (Self Regulated Learning)

    • Connected Concepts

  • Adaptive Learning
  • Affective Tutoring
  • Self Regulated Learning

    • Connected Articles

  • Multimodal AI Tutoring

    • Citation

    al, A.Z.L.E.E.Y.S.E. (2026). Engagement Assessment in Video Learning