Pedagogy Explicit Instruction Worked Examples And Feedback

---
name: pedagogy-explicit-instruction-worked-examples-and-feedback
description: "Use when designing or evaluating instruction: the explicit instruction versus discovery learning debate and what the evidence supports for novices and for experts, worked examples with the fading and expertise-reversal effects, explaining and modelling, questioning and checking understanding, feedback and the conditions under which it helps or actively harms, and formative and summative assessment."
---

# Pedagogy and Learning Science: Explicit Instruction Versus Discovery, Worked Examples, Explaining and Modelling, Questioning, Feedback, and Assessment

> **Part 2 of 5** of the *Pedagogy and the Study of Learning and Teaching* reference (plugin `pedagogy-and-the-study-of-learning`), covering §8–§13. Sibling skills: `pedagogy-reliability-memory-cognitive-load-and-robust-effects` (§0–§7), `pedagogy-subject-instruction-curriculum-and-classroom-practice` (§14–§21), `pedagogy-reading-research-and-zombie-ideas` (§22–§24), `pedagogy-reference` (§25–§30). Section numbers are shared across the set; a reference written as §N → `skill` points into that sibling skill.
>
> **Currency:** The cognitive science is stable and well-replicated. Two areas moved. See §25 → `pedagogy-reference` for AI in education, and post-pandemic achievement data.

> **⚠️ This field has an unusually wide gap between what is well-established and what is
> widely believed and practised.** ⚠️ **Some findings here are among the most robust in
> all of psychology; others in the same textbooks failed replication entirely; and a third
> group were never research findings at all — they were marketed.**
>
> **Complements a psychology reference (§1's replication context) and a family/community
> reference (behaviour genetics, which constrains how much any intervention can do).**
>
> **⚠️ GOTCHA** boxes mark widely-taught ideas that the evidence contradicts.
>
> **⚠️ A disclosure about the political layer:** ⚠️ **education research is unusually
> ideologically loaded, because it touches how children are raised and how resources are
> allocated.** **⚠️ Where a genuine scholarly dispute exists (§8, §14 → `pedagogy-subject-instruction-curriculum-and-classroom-practice`) I set out both
> positions; where the evidence is one-sided I say so and show the evidence.**
>
> **The three ideas that organize this document:**
> 1. **⚠️ Learning is a change in LONG-TERM MEMORY; performance during a lesson is not
>    learning** (§2 → `pedagogy-reliability-memory-cognitive-load-and-robust-effects`, §7 → `pedagogy-reliability-memory-cognitive-load-and-robust-effects`). **The two come apart routinely, and confusing them is the single
>    most consequential error in teaching.**
> 2. **⚠️ Working memory is tiny and prior knowledge is the great multiplier** (§2 → `pedagogy-reliability-memory-cognitive-load-and-robust-effects`, §3 → `pedagogy-reliability-memory-cognitive-load-and-robust-effects`).
>    **Almost every instructional design principle that works is managing that limit.**
> 3. **⚠️ Things that FEEL like good learning usually aren't** (§4 → `pedagogy-reliability-memory-cognitive-load-and-robust-effects`, §7 → `pedagogy-reliability-memory-cognitive-load-and-robust-effects`). **Fluency,
>    familiarity and ease are systematically misleading — for students AND teachers.**

---

## §8. ⚠️ Explicit Instruction vs Discovery

> **⚠️ The field's central and longest-running dispute. I'll set out both cases, then what
> I think the evidence supports.**

**⚠️ The explicit instruction case** (Kirschner, Sweller, Clark; Rosenshine):
⚠️ **minimal guidance during instruction does not work, because it ignores working memory
limits (§3 → `pedagogy-reliability-memory-cognitive-load-and-robust-effects`) and the absence of schemas in novices.** **⚠️ Discovery leaves novices to
search a problem space, which consumes the very capacity needed to LEARN.** **⚠️ The
evidence base — CLT experiments, Project Follow Through, process-product research —
consistently favours guided instruction for novices.**

**⚠️ The constructivist/inquiry case**: ⚠️ **learners build understanding rather than
receive it; motivation, transfer and epistemic understanding of the discipline matter;
and problem-based learning shows benefits in some professional and higher-education
contexts.** **⚠️ Critics of the explicit camp argue it conflates "minimal guidance" with
all inquiry, which is a straw man — well-designed inquiry is heavily scaffolded.**

**⚠️ What I think the evidence actually supports, stated plainly:**
```
⚠️ 1. For NOVICES in a domain, explicit, guided, worked-example-rich
      instruction is clearly superior. This is not seriously contested
      by the cognitive evidence
⚠️ 2. The EXPERTISE REVERSAL EFFECT (§3) means this flips as
      knowledge grows — guidance should FADE
⚠️ 3. Inquiry is better as a LATER activity, or as a way to motivate
      and consolidate, than as an initial acquisition method
⚠️ 4. ⚠️ The framing "which is better" is the actual error. The
      question is WHEN, for WHOM, at WHAT stage
```
**⚠️ Rosenshine's Principles of Instruction is the practical synthesis** — ⚠️ **short
review, small steps, many questions, models and worked examples, guided practice, checking
understanding, high success rate, scaffolding, independent practice, weekly and monthly
review.** **⚠️ It is drawn from three independent evidence streams (cognitive science,
observation of effective teachers, and cognitive-support research) that converge, which is
why it's unusually trustworthy.**

---

## §9. Worked Examples and Fading

**⚠️ Novices learn more from STUDYING worked examples than from solving equivalent
problems** — ⚠️ **because problem-solving search consumes working memory without building
schemas.**
**⚠️ Completion problems and faded guidance**: ⚠️ **full example → partially completed →
independent problem.** **⚠️ Fade based on demonstrated competence, not on a timetable.**
**⚠️ Self-explanation prompts attached to worked examples substantially increase their
value** — ⚠️ **passively reading an example is far weaker than explaining each step.**
**⚠️ Example-problem pairs** (**study one, then solve one**) **are the practical default.**

---

## §10. Explaining and Modelling

**⚠️ Good explanation**: **concrete before abstract, ⚠️ multiple varied examples (⚠️ AND
non-examples, which define the boundary), analogy with its limits made explicit, and
attention to the DEEP structure rather than the surface** (§5 → `pedagogy-reliability-memory-cognitive-load-and-robust-effects`).
**⚠️ Analogies are double-edged**: ⚠️ **they leverage existing schemas and they import
misconceptions.** **State where the analogy breaks.**
**⚠️ Misconceptions are not absences** — ⚠️ **they are existing, often coherent models that
must be confronted directly.** **⚠️ Simply presenting the correct account frequently leaves
the misconception intact alongside it**, **which is why elicit-confront-resolve sequences
exist.**
**⚠️ Modelling with think-aloud** makes expert reasoning visible — ⚠️ **and expertise makes
this hard, because experts have automated the steps they need to narrate (the "curse of
knowledge" / expert blind spot).**

---

## §11. Questioning and Checking Understanding

**⚠️ The purpose of questioning is to find out what students actually know — which means
sampling widely, not accepting the first hand up.**
```
⚠️ COLD CALL / no hands up  ⚠️ samples the class rather than the
   confident minority. ⚠️ Done with warmth it raises participation;
   done as a gotcha it raises anxiety. The implementation matters
⚠️ WAIT TIME  ⚠️ ~3 seconds after a question dramatically improves
   answer length and quality. One of the cheapest available improvements
⚠️ HINGE QUESTIONS  a single diagnostic question whose answer
   determines whether to move on. ⚠️ Well-designed distractors
   correspond to specific misconceptions
⚠️ ALL-STUDENT RESPONSE  mini whiteboards, exit tickets, polling
⚠️ ⚠️ "Does everyone understand?" IS NOT CHECKING UNDERSTANDING.
   It samples confidence, and confidence is miscalibrated (§7)
```

---

## §12. Feedback

> **⚠️ Feedback has one of the largest average effect sizes in the literature AND one of
> the largest variances — including a meaningful proportion of NEGATIVE effects.**
> ⚠️ **Kluger and DeNisi found roughly a third of feedback interventions REDUCED
> performance.** **"Give more feedback" is therefore not sound advice.**
```
⚠️ WHAT MAKES FEEDBACK WORK
   ⚠️ About the TASK or the PROCESS, not the PERSON
   ⚠️ Specific and actionable
   ⚠️ ⚠️ REQUIRES TIME AND OPPORTUNITY TO ACT ON IT — feedback the
      student cannot use is wasted effort by the teacher
   Timing depends: immediate for facts and procedures, delayed can
      be better for transfer
⚠️ WHAT MAKES IT FAIL
   ⚠️ GRADES ALONGSIDE COMMENTS — students read the grade and ignore
      the comments. This is well documented
   ⚠️ Praise directed at the person ("you're clever")
   ⚠️ Too much at once; vague ("be more analytical")
   ⚠️ Ego-involving comparison to peers
```
**⚠️ The efficiency point teachers need**: ⚠️ **whole-class feedback from reviewing a set of
work is often as effective as individual written comments and takes a fraction of the
time** — **and the time saved is better spent on §4 → `pedagogy-reliability-memory-cognitive-load-and-robust-effects`.**

---

## §13. Assessment

**⚠️ Formative (for learning) vs summative (of learning) — ⚠️ and the same instrument can
serve either purpose depending on what happens next.**
**⚠️ Validity is a property of the INFERENCE, not the test** — ⚠️ **"is this test valid?"
is malformed; "is this inference from this score, for this purpose, valid?" is the
question.** **Reliability, and ⚠️ the fact that reliability constrains validity.**
**⚠️ Assessment drives learning through backwash** — ⚠️ **students study what is assessed,
so the assessment IS the operative curriculum** whatever the syllabus says.
**⚠️ Practical cautions**: ⚠️ **multiple choice can test deeply if distractors are
misconception-based; comparative judgement is a genuinely useful technique for
hard-to-mark work; and ⚠️ standard-setting is a judgement, not a measurement — "proficient"
is a decision** (§25.2 → `pedagogy-reference`).