Sub-skill isolation and sequencing

Automate the components of a complex skill one at a time before combining them.

Why it works

A complex skill places multiple processing demands simultaneously on working memory. Attempting to practice the whole skill before any component is automatic guarantees that all components remain capacity-costly. Isolating sub-skills and automating each frees working memory capacity for the next component, so integration at the end faces far lower total load than naive whole-skill practice from the start.

How to do it

1. Decompose the target skill into 3–5 distinct stimulus-response components.
2. Rank them by their load on working memory — highest-load components benefit most from prior automatization.
3. Practice each in isolation to the dual-task criterion before combining.
4. Integrate components gradually: first two, then three, until the full skill runs smoothly.

Evidence

Part-task training research supports the efficiency of isolating high-demand sub-skills before whole-task practice, particularly when sub-skills are separable rather than tightly coupled. (mechanistic)

Part-task training is not universally better than whole-task practice — it depends on sub-skill separability. For tightly coupled skills (e.g., conversation), isolation can decontextualize in ways that hurt transfer.

Common mistake

Practicing the whole task from day one to "get a feel for it," which keeps all sub-skills in the controlled regime and produces experienced novices rather than automatic performers.