Worked examples: why novices learn more from studying solutions than solving

Evidence grade: STRONG. The worked-example effect is one of instructional psychology's most replicated findings, from Sweller and Cooper's originals through decades of extensions — with its boundary condition (the expertise reversal) equally well established. Rare among findings here, it comes with a ready-to-use classroom recipe.

Give novices twenty algebra problems to solve, or ten worked solutions to study plus ten to solve — and the second group reliably learns more, in less time, with less distress. That's the worked-example effect (Sweller & Cooper, 1985), and it offends every instinct that says learning comes from doing. The resolution of the offense is the interesting part.

Why studying beats solving (for novices)

Cognitive load theory's account: a novice attacking a problem has no schema to guide them, so they fall back on means-end search — juggling the goal, the current state, and every operator that might connect them. That search saturates working memory while teaching almost nothing: you can solve a problem and encode no reusable pattern, because all capacity went to the solving. Studying a worked example spends the same capacity on the thing that matters — the solution's structure — building the schema that later makes solving cheap. "Learning by doing" isn't wrong; it's mis-scheduled. Doing is how you strengthen a schema (retrieval, deliberate practice); studying examples is how a novice first gets one.

Two design details decide whether examples work:

Self-explanation multiplies the effect. Learners who explain the example's steps to themselves ("why divide here?") dramatically outperform passive readers (Chi et al., 1989; Renkl's research program). An example read like a novel is nearly worthless; prompts that force explanation ("what will the next line be, and why?") convert reading into processing.
Integrated presentation matters — an example split between diagram-here and explanation-there taxes memory just by its layout (split attention).

The fading sequence, and the reversal that mandates it

Examples aren't a destination; they're an on-ramp with a documented exit. The expertise reversal effect (Kalyuga et al., 2003): as learners build schemas, worked examples lose their advantage and then invert — for competent learners, studying examples is redundant processing, and solving wins. The research-backed bridge between the phases is completion problems (van Merriënboer): partially worked solutions where the learner supplies the missing steps, with the missing fraction growing over time.

The full recipe, compressed into a classroom-ready sentence: watch one, complete one, do one — full example (self-explained), then completion problems, then independent solving, with the schedule per-learner rather than per-class, because expertise reversal arrives at different times for different children (which is what mastery structures track).

What the evidence doesn't say

It doesn't license example-studying forever — that's the reversal; a class still studying examples in week six is being under-challenged into boredom.
It doesn't apply equally to ill-structured tasks — the effect is strongest where solutions have clear step-structure (math, procedures, programming); open-ended composition needs modelling of a different grain.
It doesn't make discovery worthless — it locates discovery's value after schemas exist (the Kirschner-Sweller-Clark argument, read carefully, is about novices).

In the classroom

Open new skills with a worked example, self-explained aloud — "watch me, and tell me why I did that" beats both lecture and struggle-first for genuinely new material.
Use completion problems as the middle gear — most classrooms jump from demo to full problems; the missing middle is where the effect's value concentrates.
Fade per child, not per calendar — the child still needing examples and the child bored by them sit in the same row.
Prompt the why: worked examples on worksheets should carry explanation blanks, or they'll be read like novels.

How Wiz Kids applies this

Our teaching tasks are structured as guided completion: the instruction demonstrates the pattern, early steps scaffold heavily, and the scaffold thins within and across lessons — while review tasks strip it entirely (the recall-mode story is the fading sequence's final stage, mechanized). The expertise reversal is why teaching tasks and review tasks are different artifacts in our system: same skill, different scaffold density, scheduled by demonstrated mastery rather than by the calendar.

References

Sweller, J., & Cooper, G. A. (1985). The use of worked examples as a substitute for problem solving in learning algebra. Cognition and Instruction, 2(1).
Chi, M. T. H., Bassok, M., Lewis, M. W., Reimann, P., & Glaser, R. (1989). Self-explanations: How students study and use examples in learning to solve problems. Cognitive Science, 13(2).
Renkl, A. (2014). Toward an instructionally oriented theory of example-based learning. Cognitive Science, 38(1).
Kalyuga, S., Ayres, P., Chandler, P., & Sweller, J. (2003). The expertise reversal effect. Educational Psychologist, 38(1).
van Merriënboer, J. J. G. — the completion-strategy research program (e.g., Training Complex Cognitive Skills, 1997).

© Glu IO Pty. Ltd. — Wiz Kids (wiz.kids). Link freely; republication requires permission — see terms. Found an error in our reading of the research? We correct fast: tell any teacher piloting Wiz Kids.