Protocol
Reduce simultaneous complexity when overload appears
More struggle is not always more learning.
When it fits
- A practice task has so many moving parts that errors multiply but you cannot tell which part is actually weak.
When to avoid it
- Do not simplify until the assessment demand disappears. Difficulty is useful when it is informative and recoverable.
Why it matters
When the learner cannot keep the task state coherent, remove one layer of complexity without removing the target skill. Shorten the case, expose one worked step, split the explanation at a meaningful boundary, or hold one variable constant. Restore complexity after the missing part becomes executable.
Steps
- The simplified task still trains the target skill and has a clear condition for restoring complexity.
An example
If a full architecture case collapses because the integration path is unclear, hold business context constant and solve only the integration decision first.
Check your result
The simplified task still trains the target skill and has a clear condition for restoring complexity.
Keep this limit in mind
- Do not simplify until the assessment demand disappears. Difficulty is useful when it is informative and recoverable.
Connected ideas
Useful withInterleave confusable cases, not unrelated workstreams
Evidence and sources
A meta-analysis found benefits when complex multimedia instruction was divided into meaningful learner-paced segments rather than presented as one continuous unit.
The evidence concerns multimedia instruction; applying it to study blocks is a cautious design analogy.
A Meta-analysis of the Segmenting Effect · Abstract