No. 08 Primer
Clean is a learning requirement.
Hierarchy, whitespace and one primary action aren’t only good craft. In a learning product, they decide how much gets learned.
Everything a learner learns passes through a gate that holds about four things at once.
That gate is working memory: the small, temporary space where conscious thinking happens. It empties in seconds. Long-term memory, where learning lives, is effectively unlimited. Everything bound for it has to squeeze through the gate first.
The gate holds about four chunks, not seven
Miller’s famous “seven plus or minus two” was partly rhetorical, and Miller said so. When rehearsal and chunking are prevented, the limit is closer to four. Present too much at once, split attention between two places, or add noise, and very little gets through.
This is why a design system’s ordinary discipline matters more in learning products than anywhere else. Hierarchy, whitespace and one primary action aren’t only good craft. They are a learning intervention.
Small layout choices have some of the largest effects in education
Predict before you look
Which of these design choices has the biggest effect on learning?
1.12
Words beside the graphic they describe (d). Large.
~0.20
Narration over on-screen text, after correcting for publication bias. Once reported as 0.72.
−0.33
Interesting but irrelevant extras. Learning goes down.
Effect sizes are in standard deviations. Roughly: 0.2 is small, 0.5 medium, 0.8 large. In education, anything over 0.4 deserves a designer’s attention.
Three things follow, and every one of them is decided in a component, a layout or a content template:
- Put the label on the thing. A legend the eye has to visit costs working memory every time.
- Cut the seductive details. They are the material a team fights hardest to keep, because they demo well.
- Don’t build an audio-first policy on the modality effect. It exists, and it is much smaller than its reputation.
A screen that shows everything shows nothing well.
Help for a novice is a hindrance for an expert
Worked examples and step-by-step guidance help a learner who has no mental model yet. Once they have one, the same scaffolding gets in the way: it competes with what they already know. This is the expertise reversal effect.
So design scaffolds with an exit. A hint, a worked step or a template should have a defined condition under which it stops appearing. Support that never fades ends up treating experts like novices, forever.
People understand new things only through what they already know
New knowledge sticks to old knowledge. Shown the same physics problems, experts sorted them by the principle underneath (conservation of energy, Newton’s second law). Novices sorted them by what they looked like: inclined planes with inclined planes, pulleys with pulleys. What you already know decides what you can see.
And some of what learners bring is wrong. Misconceptions don’t sit quietly. Presenting the right version beside the wrong one isn’t enough; the wrong one has to be engaged.
✓ Do
Find out what they know before you teach
- Activate prior knowledge before something new
- Offer entry points, not one path
- Design to surface misconceptions
- Reveal complexity in stages
× Don’t
Assume a common starting point
- One onboarding and one pace for a whole country
- Flows where learners only receive right answers
- Pre-tests that change nothing downstream
- Options added “for power users” mid-task
Take this with you
- Working memory holds about four chunks. Every element on screen spends some of them.
- Words beside pictures is a large effect. Decoration is a reliable negative.
- Scaffolds need an exit, and teaching needs to start from what the learner already believes.
Clean isn’t a style. In a learning product, it is part of the lesson.
Try this week
Halve one screen, on paper.
- Screenshot one learning screen and print it.
- Strike out every element that isn’t essential to the task in front of the learner. Aim for half.
- Circle every label that sits away from the thing it describes.
- For one scaffold (a hint, a template, a worked step), write down when it should stop appearing.
To mull over
- If you removed half the elements on your busiest screen, what would the learner actually lose?
- How does your product find out what a learner already believes before it teaches?
- Which of your product’s most “engaging” features would a seductive-details study count as noise?
Sources
- Cowan, N. (2001). The magical number 4 in short-term memory. Behavioral and Brain Sciences, 24(1). Sweller, J. (1988). Cognitive load during problem solving. Cognitive Science, 12(2).
- Ginns, P. (2006). Integrating information: A meta-analysis of the spatial contiguity and temporal contiguity effects. Learning and Instruction, 16(6). Reinwein, J. (2012). Does the modality effect exist? Journal of Psycholinguistic Research, 41.
- Sundararajan, N., & Adesope, O. (2020). Keep it coherent: A meta-analysis of the seductive details effect. Educational Psychology Review, 32.
- Kalyuga, S., Ayres, P., Chandler, P., & Sweller, J. (2003). The expertise reversal effect. Educational Psychologist, 38(1).
- Chi, M. T. H., Feltovich, P. J., & Glaser, R. (1981). Categorization and representation of physics problems. Cognitive Science, 5(2). Clark, R. C., & Mayer, R. E. (2016). e-Learning and the Science of Instruction.