Multimedia Learning Principles for Technical Training

A training module on a new sensor family has everything: an animated cutaway, a narrator, full on-screen captions of the narration, a legend in the corner, background music and a quiz at the end. Engineers who take it remember the music. The material wasn’t wrong; there was just too much of it competing for the same limited attention. Multimedia learning principles exist to prevent this. They come from decades of experiments on how people learn from words and pictures together, and they translate directly into decisions about technical training: what to show, what to say, what to cut and when to let the learner control the pace.

Key Takeaways

• Working memory has limited capacity, and learning suffers when a lesson spends that capacity on material that doesn’t help understanding. Cognitive load theory, starting with Sweller (1988), frames instructional design around this limit.
• Mayer and Moreno (2003) describe nine research-based ways to reduce cognitive load in multimedia learning, including narrating animations, segmenting, pre-training, removing extraneous material, signaling and aligning words with pictures.
• Meta-analyses by Ginns support two of the most practical rules: present words as narration rather than on-screen text with animations (the modality effect), and put related words and pictures together in space and time (the contiguity effects).
• These effects are strongest for novices and complex material. For experienced learners, extra explanation can become redundant and hinder learning: the expertise reversal effect (Kalyuga et al., 2003).
• In technical training, the practical version is: decide who the learner is, cut everything that doesn’t serve the explanation, and design for the learner’s level rather than adding every possible support.

The problem: technical material is already heavy

Technical training asks people to hold several interacting ideas at once: a signal path, the role of each component, the conditions that change behavior. Cognitive load researchers call this element interactivity, and it’s high for most engineering content. A learner watching an explanation of a sensor’s signal chain has to track the physical quantity, the transducer, the amplifier, the filter and the ADC, and understand how they affect each other.

When material is intrinsically demanding, every extra demand matters. Text that duplicates the narration, a legend on the other side of the screen, an animation that runs faster than anyone can follow, decorative graphics: each one takes some of the same capacity the learner needs for the actual content.

What the research says

Cognitive load and working memory

Sweller’s 1988 paper in Cognitive Science, “Cognitive Load During Problem Solving: Effects on Learning,” argued that some common instructional approaches impose a heavy load on working memory that interferes with learning. Cognitive load theory grew from that work, and it distinguishes load that comes from the material itself, load that comes from how it’s presented, and the effort that goes into building understanding. Instructional design can’t change the first much, but it can reduce the second and leave room for the third.

Nine ways to reduce load

Mayer and Moreno’s “Nine Ways to Reduce Cognitive Load in Multimedia Learning” (Educational Psychologist, 2003) summarizes a program of experiments into practical methods. In their framing, they are:

• Off-loading: move some words from on-screen text to narration, so the eyes and ears share the work.
• Segmenting: break a continuous presentation into learner-paced pieces.
• Pre-training: teach the names and characteristics of the parts before showing how the system works.
• Weeding: remove interesting but extraneous material.
• Signaling: add cues that highlight the organization and the essential material.
• Aligning: place printed words next to the parts of the graphic they describe.
• Eliminating redundancy: don’t present identical narration and on-screen text together with an animation.
• Synchronizing: present narration and the corresponding animation at the same time, not one after the other.
• Individualizing: account for learner differences; the authors note that some of these effects depend on characteristics such as prior knowledge and spatial ability.

Mayer’s book Multimedia Learning (Cambridge University Press) develops these into a set of named principles that are widely used in instructional design.

What the meta-analyses add

Single experiments can overstate effects, so meta-analyses are useful. Ginns’s “Meta-analysis of the modality effect” (Learning and Instruction, 2005) pooled studies comparing pictures with spoken text against pictures with written text and found a reliable advantage for the spoken version, with the benefit depending on conditions such as the complexity of the material and whether the presentation was system-paced. Ginns’s 2006 meta-analysis of spatial and temporal contiguity, in the same journal, found support for integrating related words and pictures in space and in time.

The practical point of “depending on conditions” matters. A narrated animation helps most when the material is complex and the learner can’t pause to read; a self-paced diagram with labels placed directly on it may not need narration at all.

When help becomes a hindrance

Kalyuga, Ayres, Chandler and Sweller’s “The Expertise Reversal Effect” (Educational Psychologist, 2003) reviewed evidence that instructional techniques that help novices can lose their benefit, or reverse it, for more experienced learners. An expert who already understands the components has to process redundant explanation anyway, and that costs effort. For technical training, which often has a mixed audience, this is the reason to segment by level or let experienced learners skip.

Principles and what they mean in technical training

Principle (research term)	What the research says	Applied to technical training
Coherence / weeding	Remove extraneous material	Cut decorative motion, background music and tangents
Signaling	Highlight structure and essentials	Highlight the active component; use consistent color for each signal
Spatial contiguity / aligning	Put words next to what they describe	Label parts on the diagram, not in a separate legend
Temporal contiguity / synchronizing	Present narration with the matching visual	Narrate each step as it happens on screen
Redundancy	Avoid identical narration and on-screen text with animation	Captions for accessibility, but don’t design them as the main channel
Modality / off-loading	Narration with visuals can beat on-screen text	Narrate fast-moving animations; keep text for self-paced diagrams
Segmenting	Learner-paced pieces	Pause points or chapters after each stage of a process
Pre-training	Teach parts before process	A short “meet the components” section before the full system
Expertise reversal	Supports can hinder experienced learners	Offer a fast path or skip option for experienced engineers

A note on redundancy and accessibility: the redundancy research concerns learners who can hear the narration. Captions remain essential for deaf and hard-of-hearing learners and for anyone watching without sound. The design lesson is to avoid putting a full paragraph of duplicate text on screen as part of the visual design, while providing captions as a user-controllable accessibility feature.

What to actually do

Decide who the learner is first

A training module for field technicians new to a product needs pre-training, segmenting and signaling. A module for application engineers who know the product category needs less. If both audiences use the same module, give experienced learners a way through quickly.

Weed before you add

Start each module by listing what the learner must be able to do or explain at the end, then remove anything on screen or in the narration that doesn’t serve that. Brand animation, ambient motion and music are the usual first cuts.

Put words where the eyes are

Label components directly. If a waveform and a schematic need to be compared, put them close together and use the same color for the same signal.

Narrate motion, write for static

For a moving animation, use narration with short labels. For a static diagram the learner studies at their own pace, concise text placed next to the relevant part works well.

Segment and pace

Break processes into stages with natural pause points. Let the learner control playback, and avoid auto-advancing slides in complex material.

Teach the vocabulary first

Before the full system animation, show each component once with its name and its job. Then the system explanation has something to attach to.

How to put it to work

If you have an hour

Take one existing training video and list everything on screen in a typical 30-second section. Mark each item as essential, supporting or decorative, and note any on-screen text that duplicates the narration.

If you have a sprint

Rebuild one module with weeding, direct labeling, signaling and segmenting. If you can, run a short comparison with a few learners from the target audience and ask them to explain the system back.

If you want it to last the product’s lifetime

Write the principles into your training style guide and storyboard template: a pre-training section, a labeling rule, a segment-length guideline and a fast path for experienced learners. In our own work on radiology-physics animations for Sybil Digital Learning, the educators supplied the subject expertise and reviewed the explanations, and that expert review is the right place to check each principle against the real audience.

The question to ask yourself: “Is everything on this screen helping the learner understand the system, or is some of it just there?”

Frequently asked questions

What are the multimedia learning principles?

They’re research-based guidelines for designing instruction that combines words and pictures, developed mainly by Richard Mayer and colleagues. They include coherence, signaling, redundancy, spatial and temporal contiguity, segmenting, pre-training and modality, all aimed at reducing unnecessary load on working memory.

Should technical training videos have on-screen text?

Short labels on the relevant parts help. Full paragraphs that duplicate the narration usually don’t, according to the redundancy research. Captions should still be available as an accessibility feature that viewers can turn on.

Is narration better than text?

For animations and system-paced presentations, the meta-analytic evidence favors narration with visuals over written text with visuals. For self-paced static diagrams, well-placed text can work just as well.

Do these principles apply to experienced engineers?

Partly. The expertise reversal effect shows that supports that help novices can become redundant for experts. Experienced engineers often benefit from shorter, denser material and the ability to skip.

How long should a training segment be?

The research supports learner-paced segments rather than a specific length. In practice, a segment should cover one stage of a process or one idea, with a natural pause point at the end.

Related reading: do animations teach better than static diagrams?, animation, article or interactive demo and how we review technical animations for accuracy.

Sources

1. Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257–285. https://doi.org/10.1207/s15516709cog1202_4
2. Mayer, R. E., & Moreno, R. (2003). Nine ways to reduce cognitive load in multimedia learning. Educational Psychologist, 38(1), 43–52. https://doi.org/10.1207/S15326985EP3801_6
3. Kalyuga, S., Ayres, P., Chandler, P., & Sweller, J. (2003). The expertise reversal effect. Educational Psychologist, 38(1), 23–31. https://doi.org/10.1207/S15326985EP3801_4
4. Ginns, P. (2005). Meta-analysis of the modality effect. Learning and Instruction, 15(4), 313–331. https://doi.org/10.1016/j.learninstruc.2005.07.001
5. Ginns, P. (2006). Integrating information: A meta-analysis of the spatial contiguity and temporal contiguity effects. Learning and Instruction, 16(6), 511–525. https://doi.org/10.1016/j.learninstruc.2006.10.001
6. Mayer, R. E. (2009). Multimedia Learning (2nd ed.). Cambridge University Press.