Redesigning Educational Videos for Deaf and Hard-of-Hearing Learners

Paper Title

Redesigning Educational Videos for Deaf and Hard-of-Hearing Learners

Publication Info

• Topic area: Accessibility in educational video design for Deaf and Hard-of-Hearing (DHH) learners.
• Keywords: Accessibility, Deaf and Hard-of-Hearing, educational videos, multimedia learning, motion design, visual attention, cognitive load, captions, video accessibility, inclusive design.

Background and Problem

• Problem / challenge: Current educational video designs primarily focus on captions and transcripts, neglecting the unique visual and cognitive needs of DHH learners. Existing multimedia learning theories assume dual-channel processing (audio and visual), which is not applicable to DHH learners who rely solely on the visual channel.
• Significance: Enhancing video accessibility for DHH learners can improve their educational outcomes, reduce cognitive load, and foster equity in learning environments.
• Motivation and related work: Prior research has explored caption optimization and some visual design elements but has not adequately addressed the integration of motion design to support DHH learners’ distinct visual processing abilities. This paper builds on multimedia learning theory and visual motion research to address these gaps.

Solution

• Proposed approach: Motion-driven design ideas tailored to DHH learners’ visual and cognitive needs, addressing challenges in video lecture delivery.
• Novelty:
  1. Identification of four key challenges in video lecture delivery for DHH learners.
  2. Development of four motion-driven design ideas: D-Illustrate, D-Align, D-Declutter, and D-Slowdown.
  3. Empirical evaluation of these design ideas with DHH learners (n=16) and educators (n=6).
  4. Extension of multimedia learning theory to single-channel (visual-only) contexts.
• Procedure and key techniques:
  1. Phase 1 (Formative Study): Identified challenges and proposed design ideas through thematic analysis of feedback from DHH learners and educators.
  2. Phase 2 (Evaluation with DHH Learners): Conducted user studies with 16 DHH participants to assess the perceived value of revised video clips using NASA Task Load Index (TLX) and Learning Cognitive Demands (LCD) metrics.
  3. Phase 3 (Evaluation with Educators): Collected feedback from six educators on the applicability of design ideas across different video types and contexts.

Results

• Concrete findings:
  • D-Illustrate significantly reduced mental demand (−0.71), physical demand (−0.70), and temporal pressure (−0.68), while increasing learning satisfaction (+1.15).
  • D-Align and D-Declutter improved learning satisfaction but had mixed results for cognitive demand reduction.
  • D-Slowdown showed limited effectiveness, with participants preferring pauses over uniform playback slowdown.
• Advantage over baselines:
  • Revised videos with motion-driven designs were rated higher in learning satisfaction and lower in cognitive demand compared to original versions.
  • D-Illustrate consistently outperformed other design ideas in fostering connections between text and visuals and focusing on essential information.
• Experiments / evaluation:
  • Phase 2: 17 video pairs evaluated by DHH learners using TLX and LCD metrics.
  • Phase 3: Educators reviewed revised clips and assessed applicability across eight video types (e.g., talking-head, screencast, programming/coding).
• Limitations and future work:
  • Limited to a single educational video on augmented reality; broader content and video formats need exploration.
  • Did not measure long-term learning outcomes.
  • Small, predominantly Deaf signer participant sample may not generalize to all DHH learners.
  • Future work should explore combining multiple design ideas, automating implementation with AI, and addressing diverse learner profiles.

Summary

This study introduces motion-driven design ideas to improve the accessibility of educational videos for DHH learners. By addressing challenges such as split attention, visual overload, and rapid pacing, the proposed designs (D-Illustrate, D-Align, D-Declutter, D-Slowdown) enhance clarity, reduce cognitive load, and increase learning satisfaction. Evaluations with DHH learners and educators demonstrate the potential of these strategies to extend multimedia learning theory to visual-only contexts. Future research should refine these ideas, explore broader applications, and leverage AI for scalable, inclusive video production.