Minimal guidance is less effective for novices than explicit instruction¶

Claim · All claims
Evidence · 3 studies · 1 causal r?, 1 quant-synthesis r?, 1 review r? · q2–q4 · i1 small

For learners who lack relevant prior knowledge, discovery- or inquiry-style approaches with minimal instructional support produce weaker learning than approaches that provide explicit structure, modeling, and worked examples.

Subclaims¶

q4 i1 Across 580 comparisons, explicit instruction outperformed unassisted discovery (d = 0.38 in favour of explicit instruction), while discovery that was enhanced with feedback, worked examples, scaffolding or elicited explanations outperformed other forms of instruction. So the deficit belongs to unguided discovery, not to active learning as a whole. → Alfieri et al. 2011

q3 i? With 112 third- and fourth-grade children learning how to design unconfounded experiments, many more learned the procedure from direct instruction than from discovery. Those who learned it by direct instruction did as well on a later authentic transfer task as the few who discovered it on their own. → Klahr & Nigam 2004

q2 i? A narrative review argues that decades of empirical studies show minimally guided instruction to be less effective and less efficient than strongly guided instruction, and that the advantage of guidance recedes only once learners have enough prior knowledge to guide themselves. → Kirschner et al. 2006

q4 i1 Across 580 comparisons, outcomes favored explicit instruction over unassisted discovery, d = -0.38, 95% CI [-.44, -.31]. → Alfieri 2011

Evidence¶

Alfieri et al. 2011¶

Alfieri, L., Brooks, P. J., Aldrich, N. J., & Tenenbaum, H. R. (2011). Does discovery-based instruction enhance learning? Journal of Educational Psychology, 103(1), 1–18. doi:10.1037/a0021017

q4 · meta-analysis · i1 · small effect, d=0.38 favouring explicit instruction over unassisted discovery · n=164 studies · quant-synthesis · r?

Two random-effects meta-analyses drew on 164 studies. The first pooled 580 comparisons of unassisted discovery learning against explicit instruction, and outcomes favoured explicit instruction under most conditions (d = −0.38, negative meaning discovery did worse; the abstract prints the 95% CI as [–.44, .31], and the upper bound's missing minus sign is almost certainly a typesetting slip). The second pooled 360 comparisons of enhanced or assisted discovery against other instruction, and found that enhanced discovery came out ahead (d = 0.30, 95% CI [.23, .36]). Together these support the claim for unguided discovery. They also qualify it: guided forms of discovery (feedback, worked examples, scaffolding, elicited explanations) were beneficial. The abstract does not break results down by learners' prior knowledge.

Klahr & Nigam 2004¶

Klahr, D., & Nigam, M. (2004). The equivalence of learning paths in early science instruction. Psychological Science, 15(10), 661–667. doi:10.1111/j.0956-7976.2004.00737.x

q3 · peer-reviewed experiment · i? · no effect size in the abstract read · n=112 children · causal · r?

In this experiment, 112 third- and fourth-grade children, who were novices at the control-of-variables strategy for experimental design, learned it either by discovery or by direct instruction. Many more children mastered the procedure under direct instruction. On a later transfer task that asked them to evaluate science-fair posters, children taught directly performed as well as the few who had discovered the method on their own. This challenges the assumption that self-discovered knowledge transfers better.

Kirschner et al. 2006¶

Kirschner, P. A., Sweller, J., & Clark, R. E. (2006). Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching. Educational Psychologist, 41(2), 75–86. doi:10.1207/s15326985ep4102_1

q2 · narrative review / theoretical argument · i? · no pooled effect size · n=N/A (review) · review · r?

This narrative review builds its case on human cognitive architecture, expert–novice differences and cognitive load theory. It concludes that empirical studies over the past half-century consistently show minimally guided instruction to be less effective and less efficient than instruction that strongly guides the learning process. It states the claim's novice boundary directly: the advantage of guidance recedes only when learners have enough prior knowledge to provide "internal" guidance (see expertise reversal effect). It is an argument that synthesises other studies, with no pooled estimate of its own.

Studies bearing on this claim¶

Should learners explore a new idea with guidance, rather than be told it or left to find it alone? This table gathers the studies behind this claim and the 34 narrower claims that bear on it: 36 distinct studies (12 causal, 9 review, 6 quant-synthesis, 4 theoretical, 4 qualitative, 1 design), q1–q4, 5 of 36 with an effect size coded. One row per study, codes as each entry records them; no effect is averaged and no verdict is drawn. Built by scripts/synthesize_claim.py from eval/synthesis/; a row that looks wrong is fixed on the claim it comes from.

Discussion¶

Why novices struggle with minimal guidance. Novices lack the schemas needed to recognize productive problem-solving paths, so unguided search imposes heavy extraneous load without yielding retrievable knowledge — the core mechanism proposed in cognitive load theory [+M]. Guidance such as worked examples and direct instruction reduces that load and directs attention to relevant features [+S]. This is the same mechanism by which chunking [+S] and load management [+M] support novice learning: limited working memory is spent on schema construction rather than on fruitless search.

Expertise reversal. The claim is bounded by learner expertise: as competence grows, the same guidance can become redundant and even depress performance [~S] — see expertise reversal effect and Worked examples can become redundant or counterproductive for advanced learners.. Effective design therefore fades guidance over time rather than choosing one extreme [+M]. A sequence that begins with explicit instruction and worked examples, then moves toward independent practice and eventually open-ended inquiry, respects both ends of this boundary.

Constructivist framing. This claim does not imply learners must be passive. Guided approaches can still incorporate activation of prior knowledge [+M], practice, and collaborative learning [~M]; the contrast is with unguided or minimally guided discovery, not with active learning generally [+S]. Well-structured inquiry — with prompts, scaffolds, and clear goals — sits between the extremes and is not contradicted by this claim.

Constraints on scope. The claim fails or weakens under identifiable conditions: (a) learners with substantial prior knowledge, where added guidance becomes redundant and can depress outcomes [~S]; (b) tasks where exploration itself is the learning goal, such as hypothesis generation or epistemic practice, where premature explicit answers may short-circuit productive struggle [~M]; and (c) minimally guided designs that do provide implicit scaffolds — well-designed learning environments labeled "discovery" often embed far more structure than the label suggests, making naive comparisons misleading [~M]. Claims that pure discovery outperforms guided instruction for novices are not supported [-S].

Open questions. The precise threshold at which guidance should fade, and how this varies across domains and learner populations, remain unresolved. Boundary conditions should be documented as evidence entries are added.

Merged from “Minimal guidance is ineffective for novices” (minimal-guidance-ineffective-for-novices): Why novices struggle with minimal guidance. Novices lack the schemas that allow more knowledgeable learners to recognize relevant problem features and constrain search. Without them, unguided or minimally guided activity forces learners to search a large problem space using limited working-memory resources, and the search itself can consume capacity needed for learning — the core mechanism described in Cognitive Load Theory and consistent with the claim that cognitive overload degrades learning [-S]. Guidance such as worked examples reduces this unnecessary search [+S].

Guidance is not the same as lecture. "Minimal guidance" refers to the amount of instructional support during learning, not to passive transmission. Well-structured direct instruction and guided approaches can be highly interactive and activity-based; the contrast is with approaches that expect learners to discover essential content with little or no support. Guided forms of inquiry — with prompts, scaffolds, and advance organizers — can outperform both pure discovery and pure exposition [~M].

Boundary conditions. The claim is scoped to novices. As expertise develops, the benefit of guidance diminishes and can reverse — the expertise reversal effect — so the same scaffolds that help novices can become redundant for advanced learners [-M]. This claim should therefore be read as one end of a fading continuum, not a blanket endorsement of maximal guidance at all levels. Structured sequences that pair examples with practice and fade support over time operationalize this continuum [~S].

Design implications. For novice-facing instruction, prefer explicit structure: clear goals, worked examples, chunking of complex content, and scaffolded practice before independent problem-solving. Where inquiry is pedagogically desirable for engagement or epistemic goals, build in prompts, checkpoints, and advance organizers rather than leaving learners to search unaided. The failure mode to avoid is not activity itself but unsupported search.

Open questions. The precise threshold at which learners have enough prior knowledge to benefit from reduced guidance, and how guidance should be faded over time, remain active design questions.

Merged from “Minimal Guidance Instruction Less Effective” (minimal-guidance-instruction-less-effective): Why minimal guidance fails for novices. The standard explanation is cognitive load: without guidance, novices must search the problem space themselves, consuming limited working-memory resources on search rather than schema construction [-S]. This aligns with the broader account in Cognitive Load Theory and with evidence that cognitive overload degrades learning. Unguided approaches also risk learners encoding misconceptions during unproductive exploration, which then must be unlearned.

The expertise boundary. The claim is scoped to novices. As expertise develops, the advantage of guidance shrinks and can reverse [~S] — more knowledgeable learners may find explicit guidance redundant or interfering, consistent with the expertise reversal effect. Minimal-guidance or exploratory formats are therefore best positioned later in an instructional sequence, after schemas are established through guided practice. This is the same moderator that governs when worked examples stop paying off.

Constructivism vs. guided instruction. A constructivist rationale (learners must construct knowledge actively) is sometimes read as endorsing minimal guidance, but active construction can occur during well-guided instruction; the behavioral activity of the learner is not the same as the cognitive activity required for learning. See Constructivism and Active learning improves exam performance — active learning typically still involves substantial structure and scaffolding, not unguided discovery.

Guidance is a matter of degree. The contrast is not binary. Partially guided formats — worked examples with faded steps, completion problems, prompts and cues — often capture much of the benefit of full guidance while preserving some productive struggle [~M]. The design question is how much search the learner's working memory can productively support, which cognitive load management treats as a core instructional lever.

Open questions. The precise dose and fading schedule of guidance across the novice-to-expert continuum remains an open design question, as does the effectiveness of partially guided formats (e.g., productive failure sequences) relative to fully guided instruction [~W]. Evidence entries supporting this claim still need to be added.

Merged from “Minimally Guided Instruction Less Effective For Novices” (minimally-guided-instruction-less-effective-for-novices): Mechanism. The claim rests on working-memory limits: novices lack the schemas needed to constrain search through a problem space, so unguided or minimally guided tasks impose heavy extraneous load while offering little from which to build schemas. This aligns with Cognitive Load Theory and the claim that cognitive overload degrades learning. Guidance — worked examples, modeling, explicit explanations — offloads search so working memory can be devoted to schema construction. The same logic underlies the value of worked examples and demonstration for novices, and of upfront structuring devices such as advance organizers.

Boundary conditions. This is a claim about novices, not about learners generally. As expertise develops, the advantage of high guidance reverses: highly guided instruction becomes redundant for advanced learners, consistent with the expertise reversal effect. "Minimally guided" should also be distinguished from well-scaffolded inquiry — structured or guided inquiry, where learners explore within a carefully designed task with prompts, feedback, and target concepts made available, is not covered by this claim. The claim targets approaches in which learners must identify essential concepts and procedures largely on their own. Note also that active participation itself is not the problem: active learning improves exam performance, but activity without guidance means learners may actively practice and consolidate errors or inefficient procedures.

Open questions. The precise guidance threshold at which support becomes redundant for a given learner and domain is not well specified, and effect sizes likely vary by domain (well-structured domains such as mathematics versus ill-structured domains). Direct comparisons of guided versus minimally guided designs across longer time scales and retention measures remain an active area of debate. In practice, designers should treat guidance as a fading parameter — begin with high support for novices (cognitive load reduction) and withdraw it as diagnostic evidence of competence accumulates, rather than choosing a fixed point on the guided-to-unguided spectrum.

Merged from “Unguided discovery is less effective than guided instruction for novices” (unguided-discovery-less-effective-than-guided-instruction): This claim is a central tenet of Cognitive Load Theory. The proposed mechanism is that unguided or minimally guided discovery forces novices to search a problem space using limited working-memory resources, generating a heavy extraneous load that leaves little capacity for schema construction — the same overload dynamic captured in Cognitive overload degrades learning. Because novices lack the prior knowledge needed to evaluate candidate solutions during search, they may encode incomplete or incorrect features of the task rather than the target principle. Reducing that search burden through explicit guidance is the rationale behind related findings that cognitive load reduction improves learning.

The claim is bounded by learner expertise. Guidance that benefits novices can become redundant for advanced learners, who suffer interference when forced to process explanations they no longer need — the expertise reversal effect. "Unguided" here should also be distinguished from well-scaffolded inquiry: approaches such as direct instruction and guided forms of constructivism can incorporate learner activity and exploration while still supplying the critical information learners would otherwise have to discover. The contrast is with minimal guidance, not with active learning per se — guided discovery, in which the instructor structures the task, prompts, and provides feedback while learners still generate solutions, sits between the two poles and its effectiveness depends on how much of the critical information is actually left for learners to find.

Scope caution. This claim concerns novices encountering unfamiliar material. It does not license fully expository teaching at every stage: exploration, active learning, and learner-generated solutions retain value once the critical content has been made available. The practical contrast is between leaving learners to find the target knowledge versus helping them apply knowledge they have been given.

Design implications. Treat this as a starting-point decision rather than a fixed prescription: begin with high guidance for novices — explanations, worked examples, structured tasks — then fade support as competence develops, monitoring for signs that guidance has become redundant. Open questions include how quickly guidance should be faded as expertise develops, and how the effect varies across domains where discovery is more or less tractable given learners' prior knowledge.

Terminological caution. Much of the historical disagreement around this claim stems from inconsistent use of "discovery learning," "inquiry learning," and "problem-based learning" to describe interventions differing widely in the amount of guidance provided. Studies labeled "discovery" often embed substantial scaffolding; claims about the superiority of guided approaches should always be checked against how much guidance the comparison condition actually delivered.

Merged from “Unassisted discovery produces less learning than explicit instruction, while discovery enhanced with guidance outperforms other instruction” (guided-discovery-outperforms-pure-discovery): The core mechanism is working-memory limits: unguided search consumes limited cognitive resources on off-task exploration rather than schema construction, so guidance redirects that effort toward the to-be-learned structure — consistent with cognitive load theory and the conditions under which cognitive overload degrades learning. Guidance can take many forms — scaffolds, prompts, feedback, or worked examples — and the amount needed is not fixed: as expertise grows, support should fade, per the expertise reversal effect.

This claim does not imply that pure discovery is useless or that direct telling is always superior. It is a comparative claim about typical conditions: for novices encountering unfamiliar material, some form of guidance beats none. Constructivist activity itself is not discredited — learners can still explore, generate, and test ideas — but the exploration should be structured, not left to chance. Open questions include how much guidance is optimal at different expertise levels, which guidance types are most efficient, and how domain-specific these boundaries are.

Boundary conditions to watch. The claim is strongest for novices in complex, high-element-interactivity domains. With simpler content, or with learners who already hold relevant prior schemas, the guidance advantage shrinks and can reverse — see Expertise reversal effect. Guidance that is too heavy or too persistent can also undercut the generative processing that makes discovery valuable in the first place; the design problem is calibration, not maximization. A practical middle ground is productive failure: let learners generate and test ideas briefly, then provide canonical instruction — the generative phase appears to prepare learners to benefit from guidance rather than replace it.

Evidence status. The evidence recorded above supports the claim for unguided discovery specifically: in Alfieri et al. (2011), explicit instruction beat unassisted discovery (d = 0.38), while discovery enhanced with feedback, worked examples or scaffolding beat other forms of instruction. So the deficit belongs to discovery without guidance, not to discovery as such. That comparison rests on one meta-analysis coded here with a small effect; the other two entries do not add an effect size.

Merged from “Guided Inquiry Outperforms Pure Discovery” (guided-inquiry-outperforms-pure-discovery): The guidance continuum. This claim concerns a comparison along a continuum of support: unguided or minimally guided discovery at one end, fully scaffolded inquiry and direct instruction at the other. Where guided inquiry sits relative to well-designed explicit instruction remains an open question, and the answer likely depends on learner expertise, domain, and the quality of the guidance.

Why pure discovery fails. The mechanism most commonly offered is working-memory limits: unguided search consumes limited working-memory resources on ineffective strategies rather than on schema construction, degrading learning — consistent with cognitive overload degrades learning and cognitive load theory. Guidance such as worked examples, prompts, and feedback redirects effort toward productive processing rather than fruitless search. This is the same mechanism that makes worked examples outperform unguided problem-solving for novices.

Moderators and boundary conditions. Expected moderators include learner prior knowledge (guidance benefits novices most, with a possible expertise reversal for advanced learners — see expertise reversal effect), the structure of the domain, and the fidelity of implementation of the scaffolds. Guided inquiry designs such as the 5E learning cycle embed explanation and elaboration phases that pure discovery omits; constructivism as a theory of learning does not entail minimally guided instruction as a teaching method. Scaffolds must also be faded as competence develops — permanent heavy guidance can become redundant for advanced learners.

What counts as guidance. The scaffolds that distinguish guided inquiry from pure discovery are heterogeneous: procedural prompts, advance organizers, worked examples, timely feedback, structured roles in collaborative learning, and teacher-led explanation phases interleaved with investigation. Designs vary widely in how much and when guidance is delivered, so "guided inquiry" is not a single treatment; effect sizes should be expected to vary with scaffold density, timing, and fade-out.

Constraints on the claim. Pure discovery is most likely to fail where learners lack the prior knowledge needed to recognize relevant patterns during exploration, in high-element-interactivity domains where unguided search quickly overwhelms working memory, and when scaffolds are present but poorly timed or never faded, producing redundancy for more advanced learners. Conversely, brief structured exploration can serve motivational and epistemic purposes — generating cognitive disequilibrium that prepares learners for subsequent explanation — so the claim should not be read as "never let learners explore."

Evidence status. The evidence recorded above supports the claim for unguided discovery specifically: in Alfieri et al. (2011), explicit instruction beat unassisted discovery (d = 0.38), while discovery enhanced with feedback, worked examples or scaffolding beat other forms of instruction. So the deficit belongs to discovery without guidance, not to discovery as such. For this folded guided-inquiry claim, note that Klahr & Nigam (2004) compares direct instruction with discovery, not guided inquiry with discovery, and Kirschner et al. (2006) is a narrative review; no recorded entry tests a guided-inquiry design (such as the 5E cycle) against pure discovery, so where guided inquiry sits on the continuum is untested here.

Learner Variables¶

• Prior Knowledge — moderator: an instructional effect differs with it

Related Claims¶

• Preliminary explicit instruction in scientific reasoning patterns (e.g., IAT statement construction) produced much larger LCTSR gains (g = 0.68) than courses without such intervention (g = 0.11) — the general claim this finding bears on, with the studies behind it

• Example–problem sequences reduce cognitive load and improve learning outcomes — worked examples as the guided alternative to unguided problem-solving

• Worked examples reduce unnecessary search for novices. — the specific guidance mechanism this claim generalizes
• Example-based sequences outperform problem-only practice for novices, and fading support is argued to aid transfer — how guidance should be faded rather than removed
• Worked examples can become redundant or counterproductive for advanced learners. — the expertise-reversal boundary condition in claim form
• Chunking reduces working memory load — mechanism by which guidance eases novice cognitive load
• Cognitive load reduction improves learning — the broader load-reduction principle behind explicit guidance
• Expertise reversal effect — the boundary condition limiting this claim to novices
• Direct instruction — the explicit, structured alternative this claim favors for novices
• Cognitive overload degrades learning. — the working-memory mechanism underlying novice vulnerability to minimal guidance
• Cognitive Load Theory — the theoretical framework from which this claim derives
• Active learning improves exam performance — active formats help, but typically with structure, not minimal guidance
• Advance organizers improve learning. — a form of upfront guidance that structures new material for novices
• Clear structure improves learning. — explicit structure is a low-cost form of the guidance novices need
• Ambiguous tasks with a revision step help students recognize the role of assumptions in mathematical activity — related
• The HPM learning simulation's directed activation mechanism, developed to minimize working memory load, yields a new account of the Zeigarnik effect — related
• Teacher Guided Inquiry Outperforms Student Led — related
• Inquiry Based Teaching Improves Science Achievement — related
• Guided Play Improves Academic Outcomes — related
• Kindergartners struggle to give verbal causal explanations, but model-based tasks let most of them express causal ideas — related
• Productive Failure Improves Conceptual Learning — related
• Rapid prototyping methods can amplify novice designers' tendency to commit to a solution too early. — related
• Some students found the guided-inquiry worksheet activities insufficiently thought-provoking and the data exchange troublesome — related
• Taiwan's constructivist mathematics reform, introduced in 1993, was stopped in 2003 after students taught under it underperformed traditionally educated peers in high school — a narrower finding that bears on this claim
• Verbal and pictorial tools-mediators fail to evoke contradiction discovery without mediation of meaning — related
• A 1966 conceptual paper reports, citing earlier programmed-instruction work, that a rule-example-incomplete-example sequence appeared efficient for introducing a new rule — a narrower finding that bears on this claim
• Self-discovery through errorful exploration is basically incompatible with error minimization in instruction — reports the opposite
• Review synthesis: explicit teaching matters, but its advantage over other approaches varies by context and structure — related
• HQIM teacher guides differ substantially in the depth of educative guidance they provide for the same pedagogical practice — a narrower finding that bears on this claim
• Offloading work to AI costs learning: students given answers did well in practice but scored lower on an unassisted exam — a narrower finding that bears on this claim
• Progressive scaffold fading moved teachers from guided participation toward independent justification, generalization, and task design — a narrower finding that bears on this claim
• Scaffolding must be balanced: too little led to confusion and overload, while too much reduced opportunities for critical thinking — a narrower finding that bears on this claim
• Adult learners perceive conversational AI as thinking partners that support problem exploration without direct instruction — related
• In one randomized working-paper experiment, conceptual-map scaffolding gave LLM-using novices marginally higher scores (p < .10) and lower score variance, with a marginal benefit concentrated among low-AIC novices — a narrower finding that bears on this claim
• Theory-guided five-step prompting of AI interaction improves open-ended explanation performance in construction education review, while unprompted AI access alone does not — a narrower finding that bears on this claim
• Incorporating example problems does not necessarily improve LLM teaching and can degrade it via a shift to example-specific error correction — related
• Without explicit guidance and scaffolding, unguided GenAI use may disproportionately benefit already advantaged students and widen performance gaps — a narrower finding that bears on this claim
• Dependent measure moderates discovery effects: acquisition scores amplify the explicit-instruction advantage, and reaction-time measures reverse the enhanced-discovery advantage — a narrower finding that bears on this claim
• Enhanced discovery learning produces greater learning than other forms of instruction — related
• Expertise reversal: cognitive load effects for novices decrease, disappear, and can reverse as expertise increases — related
• Empirical patterns of the expertise reversal effect fit those found in Aptitude Treatment Interaction studies initiated in the mid-60s — related
• The advantage of guidance begins to recede only when learners have sufficiently high prior knowledge to provide internal guidance — related
• Unguided or minimally guided instructional approaches are popular and intuitively appealing but ignore human cognitive architecture and the empirical evidence — a broader claim this one bears on
• Minimally guided instruction is less effective and less efficient than instruction that places a strong emphasis on guidance of the student learning process — possibly the same claim (merge candidate)
• Recent developments in instructional research and instructional design models support guidance during instruction — a broader claim this one bears on
• Worked examples confer a larger advantage over unassisted discovery than direct teaching does — related
• Constructivism — the learning theory tradition from which discovery learning derives, and which this claim qualifies rather than rejects
• Scaffolding — the most common structural form guidance takes inside discovery activities
• Discovery method in motor skill learning was superior to teacher-directed approach in children's attitudes and social interaction — related
• Case-based learning improves exam performance — a structured, guided case format as an example of scaffolded inquiry
• Collaborative learning improves outcomes — collaboration is another form of support that can structure inquiry
• Conflict-based instruction improves science conceptual learning, though staged contradictions helped only learners who reported being confused, and no study isolates disequilibrium as the mechanism — the motivational rationale for inquiry, which guidance must balance against load
• Discussion promotes more active thinking than lecture, though lecture is no worse for transmitting information — related
• Teacher-guided matching activities and peer scaffolding during grid and path games advanced preschool children's one-to-one correspondence and counting skills — related
• Collaborative tasks should be complex enough that joint effort is worthwhile, engaging the collective working memory effect — related
• Teachers reported redesigning instruction toward learner question formulation, comparison, and inquiry when using AI — related
• Without explicit scaffolding, the two lower-agency paths (Passive Review and Direct Question) accounted for 64% of student-AI interactions in the primary sample — a narrower finding that bears on this claim
• The advantage of explicit instruction over unassisted discovery is largest in the verbal/social skills domain — a narrower finding that bears on this claim
• Formal theory offers advantages over common sense for guiding implementation, but empirical evidence of effectiveness is lacking — related
• The relative effectiveness of instructional methods reverses as levels of learner knowledge in a domain change (the expertise reversal effect) — related
• Self-explanation beats unguided studying for equal time, but instructional explanations, problem solving and retrieval practice can be equally or more effective — related
• For circle geometry problems, guidance in the form of subgoals or worked examples increased cognitive load during problem processing (Experiments 1-3) — related
• Studies comparing pure simulation-based discovery learning with expository teaching show no clear and univocal outcome in favor of simulations — related
• Structuring the discovery process can outperform unstructured environments, but effects depend on learner characteristics and the knowledge test used — a narrower finding that bears on this claim
• Studying worked examples is more effective and more efficient than problem solving for novice learners (the worked example effect) — a narrower finding that bears on this claim
• Relatively guided instructional procedures are more effective than relatively unguided procedures for novice students — a broader claim this one bears on
• Instruction about collaboration rules alone has limited utility; explicit scaffolding via joint reflection tools during collaboration adds unique value — a narrower finding that bears on this claim
• Effects of discovery-related approaches are differential, not uniform across students or learning goals — related
• Explicit scaffolding prompts elicited P1 (exploring) and P2 (planning) engagement at task start that groups did not show on the non-scaffolded task — a narrower finding that bears on this claim
• The worked example effect holds for the schema-acquisition goal, but other instructional goals may be better served by other means — related
• Struggling with a challenging task before receiving instruction prepares learners to benefit from explicit connections — reports the opposite
• Age moderates discovery-learning effects: the explicit-instruction advantage over unassisted discovery is larger for adolescents than adults, and the enhanced-discovery benefit larger for adults than children — related
• Direct instruction and discovery are complementary rather than opposite, linked by repetitive yet explorative practice — related
• Guided discovery outperformed both unguided discovery and direct instruction in classic studies — related
• Cognitive constructivism has strong empirical support and suggests new directions for science education, and must be distinguished from metaphysical constructivism — related
• Minimally guided or unguided inquiry is less effective than explicit instruction, while adequately guided inquiry outperforms expository instruction (review attribution to Alfieri et al., 2011) — possibly the same claim (merge candidate)
• One review reports that more teacher guidance within inquiry is associated with larger advantages over traditional instruction, though the two meta-analyses it cites differ in magnitude — related
• Meta-analysis of 164 studies found assisted inquiry more effective than explicit instruction — related
• Well-controlled comparison studies show clear advantages of inquiry-based instruction over direct instruction for science conceptual learning — related
• Controlled studies comparing exploration-based investigation with explicit instruction overwhelmingly favor explicit instruction for science content learning — related
• Sequencing of exploration and explicit instruction shows inconsistent results, moderated by task complexity per the expertise reversal account — a narrower finding that bears on this claim
• Low variability materials enhanced learning regardless of guidance; expertise reversal found for variability in Experiment 2 — related
• Fading of instructional support in collaborative argumentation has produced mixed results, with effectiveness depending on the fading criterion used — related
• Scaffolding effectiveness depends on knowledge base: high-guidance scaffolding suits less advanced learners, high-self-regulation scaffolding suits advanced learners — a narrower finding that bears on this claim
• Adults can promote young children's STEM skill learning during guided play but may need support to incorporate STEM talk and reduce directive management — related
• The thesis reports, citing Klahr and Nigam (2004), that discovered and explicitly taught science knowledge did not differ in quality — reports the opposite