Dual coding theory

Allan Paivio proposed in the 1970s that the mind has two related but separate channels for processing information, one for verbal material and one for visual material. Richard Mayer turned the theory into a practical instructional framework over the next two decades.

What it is

Two channels in working memory: a verbal channel for spoken and written language, and a visual channel for images, diagrams, and spatial layout. When learners process the same information through both channels, they build two interconnected mental representations rather than one. Each acts as a retrieval cue for the other.

Why it works

A second channel reduces extraneous load on the first. A diagram next to a definition lets pupils hold less of the explanation in their head and free up capacity for the actual thinking.

Mayer’s multimedia principle says people learn more from words and pictures than from words alone, based on dozens of controlled experiments, mostly with university students. The first meta-analysis restricted to school classrooms (Banks et al., 2026, 56 randomised studies) found a moderate average benefit of about half a standard deviation, so the lab finding does carry into schools, with three qualifications. Secondary pupils gained roughly three times as much as primary pupils. Effects were measured mostly within a week and were not detectable two weeks or more after the lesson. And the benefit varied a great deal between studies, so how the visual is designed matters as much as whether one is used.

What makes a visual work

Three design features separated the visuals that helped from those that did not in the classroom trials.

• Signalling. Link the visual to the words explicitly: point to the part of the diagram as you name it, label the picture, or bold the term the picture illustrates. Visuals with signalling had roughly double the effect of visuals without it.
• Segmenting. Present the content in meaningful chunks, each with its own visual, rather than one continuous block. This mattered most for understanding.
• Coherence. Add nothing decorative. Visuals with irrelevant detail had a small, uncertain effect that could be negative.

Placing the picture right next to its text (contiguity) made much less difference than expected, so it is not where preparation time should go.

Which pupils, and when

Dual coding worked best in whole-class, teacher-paced teaching, and less well in small-group or individual work. Effects for targeted groups, such as pupils with learning difficulties or English as an additional language, were smaller and less certain, and there were too few studies of any one group to say more. Science and technology, and English and literacy, had the clearest benefits; second-language learning was the least certain.

How to use it

Pair definitions with icons or small images. Show timelines for sequences, maps for places, flowcharts for processes. Put labels next to the part of the diagram they refer to, not in a key at the bottom. Refer to the visual as you explain, pointing to the part you are naming. Use the same visual to anchor a concept across multiple lessons so the visual becomes a retrieval cue.

Avoid mixing words and pictures that compete for attention or repeat each other without adding meaning.

When not to use it

If the visual carries no meaning, it is decoration and adds to extraneous load. Stock photos and clip art used to liven up a slide usually do harm, not good.

Do not rely on it for long-term memory or for transfer to new problems. In the classroom trials the benefit faded within two weeks, and for primary pupils there was no measurable effect on transfer. The visual builds understanding in the lesson; retrieval practice and spacing make it stick.

Related Chalk tools

Visual Keywords, Icon Storyboard, Frayer Model, and Graphic Organiser are built on the dual coding principle.