Redundancy Avoidance¶

Strategy · All strategies
Evidence · 4 claims (3 for, 1 mixed) · 12 studies (4 causal, 4 review, 2 quant-synthesis, 2 theoretical), q1–q4 · 2 of 12 report an effect size · 1 claim rests on one study

Description¶

Redundancy avoidance removes simultaneous duplicate presentations of the same information — most commonly, on-screen text that repeats narration word-for-word, or graphics that duplicate what text already states. Learners must reconcile the two streams, spending working memory on coordination rather than construction of understanding. The strategy is to present each piece of information once, in the modality best suited to it.

Design Implications¶

Redundant channels compete for limited working memory: when identical content arrives as both narration and text, learners split attention between them and process the slower stream Redundant on-screen text with narration impairs learning compared with narration alone. [+M]. This is a direct application of Cognitive Load Theory: the coordination cost is extraneous load that yields no learning benefit Cognitive overload degrades learning outcomes. [+S]. The remedy is not less content but non-duplicated content — narration plus a graphic, or text alone, rather than narration plus identical text.

Context¶

Requirements¶

• An analysis of which content belongs in which channel (audio for dynamic or verbal explanation, visuals for structure and spatial relations)
• Willingness to cut or rewrite material so each idea appears once (Coherence Principle [+S])
• Access controls (captions, transcripts) kept available for accessibility, but understood as an accommodation rather than the default presentation

Constraints¶

• Redundancy avoidance fails when learners cannot process the single channel — e.g., non-native speakers or learners with hearing difficulties benefit from captions even when text duplicates narration [~M]
• For learners with low language proficiency or poor listening skills, simultaneous text and narration can support decoding rather than hinder it [~M]
• The effect reverses with expertise: knowledgeable learners may benefit from redundant text because it is easy to ignore or skim Guidance effectiveness reverses as learner expertise increases. [~M]
• Short, essential labels or key terms on screen alongside narration are not harmful redundancy; the effect applies to full verbatim duplication [-S]

Implementation Variability¶

• Narration + graphics replaces narration + verbatim text for multimedia explanations
• Signaling (highlighting key terms on screen) provides partial redundancy without full duplication
• Captions as accommodation: provide transcripts on demand rather than as simultaneous default text
• Print-first environments: when audio is unavailable, text plus graphic is the non-redundant configuration

Target Learners¶

• Novices processing multimedia explanations, who lack the capacity to coordinate duplicate streams Redundant on-screen text with narration impairs learning compared with narration alone. [+M]
• Less applicable to experts, who can selectively attend and may exploit redundancy Guidance effectiveness reverses as learner expertise increases. [~M]
• Second-language learners and learners with auditory processing difficulties need the opposite: deliberate redundancy as an accommodation [~M]

Target Learning Goals¶

• Conceptual understanding from multimedia explanations (science, statistics, procedural how-tos)
• Efficient instructional materials at scale, where load costs compound across many screens
• Not suited to goals requiring verbatim transcription or reading fluency practice, where the "redundant" text is itself the target

Instructions¶

1. Audit each screen or slide for duplicate content across channels (spoken + written + graphic).
2. Choose a single primary channel for each idea: narration for explanation, graphics for structure (Multimedia Learning).
3. Replace verbatim on-screen text with concise labels, key terms, or signaling cues.
4. Move full transcripts to an on-demand location rather than simultaneous display.
5. Test with the target population; add captions deliberately for learners who need them (Accommodations).

Related Strategies¶

• Use Worked Examples — worked examples must themselves avoid redundant explanation of steps already shown
• Chunking — both manage working memory load by controlling what enters it at once

Examples¶

• Khan Academy videos use narration over evolving visuals rather than paragraphs of on-screen text, keeping each idea in one channel.
• 3Blue1Brown (Essence of Linear Algebra) pairs dense visual animation with narration; on-screen text is limited to labels and equations, never a transcript of the narration.
• Duolingo introduces new words in a single modality first (audio with image), rather than simultaneously presenting audio, written word, and translation.

Key Sources¶

• Sweller, J., & Chandler, P. (1994). Why Some Material Is Difficult to Learn. Cognition and Instruction, 12(3), 185–233. doi:10.1207/s1532690xci1203_1
• Mayer, R. E. (2009). Multimedia Learning (2nd ed.). Cambridge University Press. doi:10.1017/CBO9780511811678
• Kalyuga, S., Ayres, P., Chandler, P., & Sweller, J. (2003). The expertise reversal effect. Educational Psychologist, 38(1), 23–31. doi:10.1207/S15326985EP3801_4
• Clark, R. C., & Mayer, R. E. (2016). E-Learning and the Science of Instruction (4th ed.). Wiley. doi:10.1002/9781119239086