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¶

Claim · All claims
Evidence · 3 studies · 2 quant-synthesis r?, 1 causal r1 · q3–q4 · i3 large
Currency · newest study 2024 · 1 of 3 entries checked against their sources (automated) · not reviewed by a person

Instruction that confronts learners with evidence or ideas conflicting with their existing conceptions produces large gains in science learning in meta-analyses, and a small experiment found staged contradictions helped only when learners reported being confused. The recorded studies measure learning outcomes, not the disequilibrium itself, so the Piagetian account that disequilibrium is what drives the reorganisation (see Discussion) is the claim's proposed mechanism, not something its evidence isolates.

Subclaims¶

q4 i3 Across 218 science-education studies (18,051 students), conceptual change strategies produced a large improvement in learning; strategies built on cognitive conflict did as well as the whole set (g = 1.10, k = 150). Heterogeneity was high, and the studies measure learning outcomes, not the disequilibrium itself. → Pacaci et al. 2024

q3 i? An earlier meta-analysis of reading- and science-education interventions on misconceptions found that the strategies that worked had one thing in common: they produced conceptual conflict. → Guzzetti et al. 1993

q3 i? Contradictions staged between tutoring agents did not raise learning on their own; learners gained only when the contradiction actually made them confused. This is a heavy qualification: conflict alone is not enough. → Lehman et al. 2013

Evidence¶

Pacaci et al. 2024¶

Pacaci, C., Ustun, U., & Ozdemir, O. F. (2024). Effectiveness of conceptual change strategies in science education: A meta-analysis. Journal of Research in Science Teaching, 61(6), 1263–1325. doi:10.1002/tea.21887

q4 · random-effects meta-analysis with robust Bayesian bias adjustment · i3 · large effect, g=1.10 (adjusted g=0.93) · n=218 studies, 18,051 students · quant-synthesis · r?

A random-effects meta-analysis of 218 primary studies testing three families of conceptual change strategy in science learning: cognitive conflict, cognitive bridging and ontological category shift. The overall effect was large (g = 1.10), and still large after a robust Bayesian adjustment (g = 0.93). Cognitive-conflict strategies, the family that most directly works through cognitive disequilibrium, showed the same effect (g = 1.10, k = 150). Effect sizes varied widely, with a prediction interval of 0.19 to 2.38. The studies measure learning outcomes, not whether disequilibrium was felt, so they support conflict-based instruction rather than isolating disequilibrium as the mechanism. The meta-analysis was read as an abstract only.

Guzzetti et al. 1993¶

Guzzetti, B. J., Snyder, T. E., Glass, G. V., & Gamas, W. S. (1993). Promoting Conceptual Change in Science: A Comparative Meta-Analysis of Instructional Interventions from Reading Education and Science Education. Reading Research Quarterly, 28(2), 116. doi:10.2307/747886

q3 · meta-analysis of experimental and quasi-experimental studies · i? · no effect size in the abstract read · n=? (not stated in the abstract read) · quant-synthesis · r?

This meta-analysis combined experimental and quasi-experimental studies from reading education and science education that tested instructional interventions aimed at students' misconceptions. It identified which strategies produced conceptual change and reported that the effective ones shared an element of producing conceptual conflict. Only the brief ERIC abstract was read, so its effect sizes and study counts are not recorded here.

Lehman et al. 2013¶

Lehman, B., D'Mello, S., Strain, A., Mills, C., Gross, M., Dobbins, A., Wallace, P., Millis, K., & Graesser, A. (2013). Inducing and Tracking Confusion with Contradictions during Complex Learning. International Journal of Artificial Intelligence in Education, 22(1–2), 85–105. doi:10.3233/JAI-130025

q3 · peer-reviewed within-subjects laboratory experiment · i? · no standardized effect size reported; interaction F(3, 240) = 2.56, p = .028 · n=32 · causal · r1

Thirty-two undergraduates took part in eight computer-based "trialogues" about research-methods case studies with an animated tutor agent and an animated student agent. The researchers induced disequilibrium by having the agents disagree or state false information, then asking the learner to decide between them. The contradictions increased self-reported confusion, but on their own they did not improve posttest scores (p = 1.00). In a follow-up analysis that split cases by how confused learners reported being (a median split), learners scored higher after contradictions only in the high-confusion cases. Disequilibrium helped only when it was actually experienced, and the study gave no scaffolding to help learners resolve it.

Studies bearing on this claim¶

Does conflict-based instruction improve conceptual learning more than standard exposition? This table gathers the studies behind this claim and the 9 narrower claims that bear on it: 8 distinct studies (2 quant-synthesis, 2 causal, 2 design, 1 review, 1 qualitative), q2–q4, 2 of 8 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¶

Mechanism. The claim rests on a Piagetian account of learning: existing schemas are stable until perturbed, and it is the discomfort of disequilibrium — not the conflicting information alone — that drives accommodation. This places the claim within constructivism and connects it to the cognitive disequilibrium principle as a design lever: instruction must first surface the learner's prior conception, then present a conflict it cannot explain. Classic conceptual-change designs operationalize this as a "predict–observe–explain" sequence, where learners commit to a prediction before seeing disconfirming evidence.

Boundary conditions. Disequilibrium is motivating only within a window. If the conflict is too weak, learners assimilate the anomaly to their existing conception and change nothing [~M]; if it is too strong, or if the conflicting material imposes excessive processing demands, learners may disengage or experience the overload that degrades learning [-M] rather than productive restructuring. Learners also need sufficient prior knowledge to notice the conflict at all — without activation of the existing conception, there is nothing to bring into disequilibrium [~M]. Learners with low epistemic tolerance or low self-efficacy in the domain may respond to conflict defensively rather than with restructuring [~W].

Design implications. Effective disequilibrium designs share three features: (1) elicit and externalize the prior conception (a committed prediction, an initial explanation) so the conflict has something to attach to; (2) present the disconfirming evidence immediately after the commitment, while the expectation is still active; (3) support resolution — through analogies, contrasting cases, or guided discussion — rather than leaving learners to resolve the conflict unaided. Sequences such as the 5E learning cycle build this elicit–conflict–resolve structure into their engage and explain phases.

Open questions. The strength of this claim depends on how "conceptual change" is measured: self-reported satisfaction of curiosity, post-test performance, and durable restructuring of misconceptions are different outcomes, and evidence supporting the claim in one does not automatically support it in others. Three studies are recorded above: two meta-analyses of conflict-based instruction (Pacaci et al. 2024 and Guzzetti et al. 1993, both read from abstracts), which measure learning outcomes rather than disequilibrium itself, and one small within-subjects experiment (Lehman et al. 2013) in which staged contradictions helped only when learners reported being confused. None measures durable restructuring of misconceptions, and none isolates disequilibrium as the mechanism.

Related Claims¶

• Some classroom discussion approaches improve text comprehension, but gains are not uniform across approaches and more student talk alone does not produce them — the general claim this finding bears on, with the studies behind it

• Conceptual change strategies in cooperative groups reduced misconceptions in community college chemistry, but poor group leadership undermined discussion — the general claim this finding bears on, with the studies behind it

• Activation improves learning. — activating prior conceptions is the precondition for creating a productive conflict with them

• Cognitive overload degrades learning. — disequilibrium that exceeds working-memory capacity tips from motivating to overwhelming
• Analogical reasoning improves transfer. — analogy is one route for resolving disequilibrium by mapping a familiar conception onto the anomalous case
• Cognitive flexibility theory: multiple cases. — multiple representations and cases help learners restructure knowledge after conflict
• Coherence principle: irrelevant material hurts learning. — conflict should come from the target conception, not from extraneous material adding load
• Contradictions within and across activity systems are vital forces for change and provide a rationale for CSL — related
• Instruction that confronts science misconceptions (conceptual-conflict interventions, refutation text) produces conceptual change more often than standard exposition — related
• Children in Constructivist kindergarten classrooms showed greater effort resolving interpersonal conflicts than children in eclectic or didactic classrooms — related
• Higher internal disequilibrium (Level Mixture) predicts Modal Level advance but not Map Drawing Mean improvement — a narrower finding that bears on this claim
• Change in new teachers requires a restructuring of their prior beliefs — related
• Peer conflicts did not always produce conceptual change; they appeared to work only for students prepared to reflect on and reconstruct their conceptions — a narrower finding that bears on this claim
• Conflict with a machine-learning decision support system's output induces deeper reflection than confirmation — a narrower finding that bears on this claim
• Immersive learning tools should surface misconceptions and track learner difficulties so instructors can adapt teaching and assessment — related
• Dialogue evidence shows coach feedback induces genuine conceptual change, shifting the agent from misconception to conceptually correct explanation rather than superficial answer alignment — a narrower finding that bears on this claim
• Error correction supports learning when it requires substantive cognitive work such as explanation and reconstruction — related
• Material difficulty is not determined solely by element number and interactivity — a broader claim this one bears on
• Only the interactive antagonistic agent, not stepwise guidance, significantly shifted novice designers' design thinking relative to self-reflection — a narrower finding that bears on this claim
• Alternative conceptions in force and motion were very difficult to change: of 14 students, 6 showed substantial conceptual change, 1 some change, and 7 no change — a narrower finding that bears on this claim
• Science-classroom studies of cooperative learning show mixed results, with several null or negative comparisons — related
• The learning cycle is reported as an effective means for learning science concepts and processes — related
• Perturbation type determines the interaction types needed: conflicting evidence elicits argumentation — related
• A striking number of biology conceptual change studies measured only superficial learning outcomes such as knowledge enrichment rather than knowledge restructuring — related
• Cognitive conflict induced by refutational texts and conflicting perspectives may promote epistemic belief change in student teachers — related