Citation: National Academies of Sciences, Engineering, and Medicine. (2018). How People Learn II: Learners, Contexts, and Cultures. Chapter 5: Knowledge and Reasoning. The National Academies Press. URL: https://www.nationalacademies.org/read/24783/chapter/7 (chapter 5 in the read interface; index off-by-one due to front matter) Pulled by: Claude on 2026-04-29 (via WebFetch summary, not the full PDF)
Note on transfer coverage in HPL II
HPL II does not have a dedicated transfer-of-learning chapter the way How People Learn I (2000) did. Coverage is distributed: Ch 4 (Processes That Support Learning) covers metacognition / executive function / memory; Ch 5 (Knowledge and Reasoning) covers learning strategies that produce durable, flexible knowledge use; Ch 6 (Motivation); Ch 7 (Implications for School). The "transfer mechanism" content closest to what we need is in Ch 5, framed as "learning strategies that produce knowledge that can be transformed and applied to new situations."
Five evidence-supported strategies for durable, flexible knowledge
HPL II's Ch 5 identifies five strategies with strong empirical support for producing knowledge that transfers / applies across contexts. All five map directly onto our five "transfer design moves":
| HPL II strategy | Our design move |
|---|---|
| Retrieval practice | Far-transfer assessment (testing in novel surface forms) |
| Spaced practice | Cross-domain task families spread over time |
| Interleaving | Contrasting cases (different problem types in one session) |
| Self-explanation / elaborative interrogation | Metacognitive debrief; explicit pattern naming |
| Summarization / drawing (transformation into different representation) | Pattern naming as re-representation |
This is a strong validation: our five design moves are not idiosyncratic — they're the same evidence-supported mechanisms HPL II names, applied to our domain.
Headline mechanisms (with implications for our co-pilot design)
Retrieval practice
- Active recall during learning (not passive review) enhances later retrieval and transfer.
- Even failed retrieval helps: it "provides feedback signals to learners, signaling that they may not know the information well and should adjust how they encode the material the next time they study it."
- Self-testing is under-used by students — they treat it as assessment, not learning.
- Implication for co-pilot: don't just "tell" — prompt the user to retrieve / try first. The Bastani "refuse-until-effort" pattern is exactly retrieval practice operationalized.
Spaced practice
- Distributing learning sessions across time produces stronger retention than massed practice across vocabulary, grammar, facts, motor skills.
- Mechanism: spacing forces re-engagement of mental operations (some forgetting has occurred → must reconstruct, not passively reprocess).
- Implication: returning users should re-encounter prior patterns in new contexts after spaced intervals — not all in one session.
Interleaving
- Nuanced finding: interleaving improves learning of highly similar categories (forces between-category comparison); blocking is better for low-similarity categories.
- Mechanism: interleaving prevents applying the same solution repeatedly — the learner must identify which approach each problem requires.
- Implication: contrasting cases should be near-neighbors of the target pattern, not wildly different problems.
Self-explanation & elaborative interrogation
- Asking learners to articulate their reasoning ("why does this step work?") improves comprehension, learning, and memory.
- "Why," "how," and "what-if" questions outperform factual questions.
- Caveat: self-explanation depends on prior knowledge. Students with weak foundations may need additional scaffolding before self-explanation works.
- Implication for our adult-novice population: the metacognitive debrief move must be carefully scaffolded — pure "explain why this worked" prompts may fail with our user base. Co-pilot should model self-explanation first, then prompt.
Summarization / drawing / transformation
- Effective when learners are required to transform material into a different representation (verbal → visual, vs. verbatim copy).
- Hand-written notes beat laptop typing because typing speed allows verbatim transcription, which short-circuits the encoding work.
- Implication: pattern naming = transformation into a verbal label = a productive re-representation. Our co-pilot's "you just did decomposition" prompt is doing exactly the work HPL II describes.
Direct quotes worth citing
"The act of retrieval itself enhances learning and... when learners practice retrieval during an initial learning activity, their ability to retrieve and use knowledge again in the future is enhanced." (Ch 5)
"Effective problem-solving requires that retrieved knowledge be adapted and transformed to fit new situations; therefore, memory retrieval must be coordinated with other cognitive processes." (Ch 5)
"Training people to use this skill — and particularly training in asking deep questions — has been shown to have a positive impact on comprehension, learning, and memory." (Ch 5, on elaborative interrogation)
Critical caveat HPL II repeatedly stresses
"Strategy effectiveness depends on learner characteristics, material nature, and learning objectives. The same technique can help or hinder depending on context."
For our population (adult, low digital fluency, low confidence), this means we cannot assume HPL II findings — derived largely from K-12 and college student populations — transfer wholesale. The five strategies are evidence-supported as mechanisms; our specific calibration (timing, scaffolding intensity, vocabulary) must be empirical.
How we'll use this in the docs
- Pedagogy doc — five design moves section: cite HPL II Ch 5 as the umbrella synthesis. Each design move maps to a specific HPL II strategy. This is our strongest single citation for the design moves.
- Pedagogy doc — productive struggle section: retrieval practice + the "even failed retrieval helps" finding directly justifies the Bastani "refuse-until-effort" pattern.
- Pedagogy doc — falsification section: HPL II's own caveat about strategy effectiveness depending on learner characteristics is a research gap we explicitly flag.
- Technical-approach doc — co-pilot section: the self-explanation scaffolding caveat (low-prior-knowledge users need modelled self-explanation first) is a concrete co-pilot design constraint.
What this source doesn't give us
- Adult-learner specific evidence. HPL II's evidence base is primarily K-12 and college.
- Empirical effect sizes / RCT data on these strategies in adult digital-literacy contexts.
- Anything on technology-mediated tutoring beyond passing references.
These are the gaps we fill with Bastani 2026 and the still-to-be-pulled adult-transfer / metacognitive-AI-tutor literature.