Teach

Teach

Telling someone an answer and teaching them a skill are different jobs, and mixing them up produces an illusion of understanding that evaporates the moment the topic changes shape. This skill's defining constraint combines two grounded findings: Bloom's Taxonomy (what's being taught has a cognitive level (Remember, Understand, Apply, Analyze, Evaluate, Create), and the level determines the method, not just the content) and the worked-example effect from cognitive load theory. A novice learns faster from a fully worked solution than from solving the same problem cold, but that advantage fades and reverses as competence grows (the expertise-reversal effect), so support has to be removed as skill builds, not held constant.

Phase 1: Place the target level and the learner's current level

Before teaching anything, decide which Bloom level the goal actually requires; "understand why X happens" is a different job from "be able to do X unaided" or "be able to judge whether X is the right call here." Teaching a Create-level goal with only Remember-level content (definitions and facts, no practice) leaves the learner unable to do the thing; teaching a Remember-level goal with elaborate open-ended practice wastes effort neither the goal nor the learner's current level needs.

Gauge where the learner already stands (what they can already do without help) so the first example lands in their zone of proximal development: challenging enough to be worth teaching, not so far ahead it's unintelligible.

Phase 2: Start from a fully worked example

For anything above pure recall, open with a complete, worked-through instance of the skill, not a partially-blanked problem, not "try it and I'll correct you." Narrate the reasoning at each step, not just the mechanical moves, since the reasoning is what transfers to the next problem and the mechanical steps alone don't. A novice's working memory is the bottleneck here; a worked example spends that budget on understanding the solution, where an unaided attempt spends it on the search for one. Write the explanation without em dashes, using a comma, colon, period, or a reworded sentence instead.

Phase 3: Fade the scaffolding

Move from worked example toward independence in stages, not in one jump:

1. Fully worked: the learner watches the whole thing reasoned through.
2. Partially worked (completion): the learner finishes a step or two of a similar problem, with the rest already solved around it.
3. Independent, with feedback: the learner solves a similar problem alone and gets immediate, specific feedback about what was right, what wasn't, and why.

Skipping straight from fully-worked to fully-independent is the most common failure here: it reintroduces the exact cognitive-load problem worked examples exist to avoid. Equally, leaving a learner on fully-worked examples past the point they've internalized the pattern wastes their time and produces the expertise-reversal effect; passive familiarity that isn't actually the skill.

Phase 4: Match the method to the level

• Remember / Understand: explanation, examples, and analogies; difficulty here is a cost to minimize, since acquiring the knowledge is the whole goal.
• Apply / Analyze: practice problems with a tight, immediate feedback loop; difficulty here is the tool, since effortful retrieval is what builds retention, not recognition.
• Evaluate / Create: open-ended tasks with judgment involved, feedback on the reasoning behind the choice, not just whether the output matches a fixed answer.

Phase 5: Check for retention, not fluency

A learner who can follow along with an explanation often mistakes that fluency for having learned it; the two are different things, and only the second survives until the next session. Before calling a topic taught, have the learner reproduce or apply it without the explanation in front of them; recall from memory, not recognition of something familiar, is the actual signal.

Done when

• The target Bloom level was named before teaching started, and matched against where the learner already stood.
• Teaching opened with a fully worked example, reasoning narrated, not a cold unaided attempt.
• Scaffolding faded in stages (worked, partially worked, independent) rather than jumping straight to unaided practice.
• The teaching method matched the target level (explanation for Remember/Understand, practice with feedback for Apply/Analyze, judgment tasks for Evaluate/Create).
• Retention was checked by unaided recall or application, not by the learner's felt sense of having followed along.