Concept mapping improves learning over most comparison conditions by a moderate average amount, but not over time-matched retrieval practice¶

Claim · All claims
Evidence · 3 studies · 2 quant-synthesis r?, 1 causal r2 · q3–q4 · i2–i3
Currency · newest study 2018 · 1 of 3 entries checked against their sources (automated) · not reviewed by a person

Learners who construct node-and-link diagrams of relationships among concepts show better understanding and retention than learners in most comparison conditions pooled in two meta-analyses, though not better than learners who spend the same time on retrieval practice (Karpicke & Blunt 2011). The claim's scope is construction of maps by learners; studying a supplied map is treated as a distinct, weaker activity (see Discussion).

Subclaims¶

q4 i2 Across 142 effect sizes, learning with concept maps beat comparison conditions by a moderate margin (g = 0.58), and constructing maps (g = 0.72) helped more than studying supplied ones (g = 0.43), in both STEM and non-STEM domains. → Schroeder et al. 2018

q4 i? An earlier meta-analysis of 55 experimental and quasi-experimental studies (Grade 4 to postsecondary) found concept-map use associated with better knowledge retention, with effects ranging from small to large depending on how maps were used and what they were compared against. → Nesbit & Adesope 2006

q3 i3 When learning time is matched, building a concept map while viewing the text produced much less learning than retrieval practice (d = 1.50 in Experiment 1), and was not significantly better than rereading; this sharply qualifies the claim against generative alternatives. → Karpicke & Blunt 2011

Evidence¶

Schroeder et al. 2018¶

Schroeder, N. L., Nesbit, J. C., Anguiano, C. J., & Adesope, O. O. (2018). Studying and constructing concept maps: A meta-analysis. Educational Psychology Review, 30(2), 431–455. doi:10.1007/s10648-017-9403-9

q4 · random-effects meta-analysis · i2 · medium effect, g=0.58 · n=142 effect sizes (11,814 participants) · quant-synthesis · r?

A random-effects meta-analysis of 142 independent effect sizes from studies in which learners used concept or knowledge maps, involving 11,814 participants. Learning with maps produced a moderate overall advantage over comparison conditions (g = 0.58). Learners who created their own maps benefited more (g = 0.72) than learners who studied a supplied map (g = 0.43). The advantage held against other instructional comparison conditions and across STEM and non-STEM domains. Read from the publisher abstract only; the moderator detail on comparison type and time-on-task was not checked.

Nesbit & Adesope 2006¶

Nesbit, J. C., & Adesope, O. O. (2006). Learning with concept and knowledge maps: A meta-analysis. Review of Educational Research, 76(3), 413–448. doi:10.3102/00346543076003413

q4 · meta-analysis of experimental and quasi-experimental studies · i? · no pooled effect size in the abstract read · n=55 studies (67 effect sizes, 5,818 participants) · quant-synthesis · r?

A meta-analysis of experimental and quasi-experimental studies in which students learned by constructing, modifying or viewing node-link diagrams. The pool was 67 standardized mean differences from 55 studies with 5,818 participants, from Grade 4 to postsecondary, in domains such as science, psychology, statistics and nursing. Using concept maps was associated with better knowledge retention across conditions and settings. Mean effects ranged from small to large depending on how maps were used and on the comparison treatment, and most subsets were significantly heterogeneous. Read from the abstract only, so no pooled effect size is recorded here.

Karpicke & Blunt 2011¶

Karpicke, J. D., & Blunt, J. R. (2011). Retrieval practice produces more learning than elaborative studying with concept mapping. Science, 331(6018), 772–775. doi:10.1126/science.1199327

q3 · peer-reviewed randomized experiments (two) · i3 · large effect favoring retrieval practice over concept mapping, d=1.50 · n=80 + 120 undergraduates · causal · r2

In Experiment 1, 80 undergraduates studied a science text in one of four ways: study once, repeated study, study then build a concept map while viewing the text, or study then practice free recall. Learning time was exactly matched between the mapping and retrieval conditions. One week later, every condition beat study-once, but retrieval practice beat concept mapping (M = 0.67 vs 0.45, d = 1.50), and mapping was not significantly better than extra rereading. Experiment 2 (120 students, within-subject) replicated the retrieval advantage on a short-answer test (d = 1.07) and even on a final test that asked students to build a concept map (d = 1.01); 101 of 120 students did better after retrieval practice. The study supports mapping over a single reading, but not over an equally long generative alternative.

Studies bearing on this claim¶

Does having learners build an organization of the content as a concept map help, and against what comparison? This table gathers the studies behind this claim and the 10 narrower claims that bear on it: 10 distinct studies (4 causal, 3 quant-synthesis, 1 review, 1 design, 1 theoretical), q1–q4, 5 of 10 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. Concept mapping is typically explained through cognitive load theory and dual coding: building a map forces learners to identify relationships explicitly, integrate new material with prior knowledge, and represent it both spatially and verbally. The generative, constructive nature of map-building is what distinguishes it from passively reading a supplied map — see Constructive learning activities produce better learning than active or passive ones..

Construct vs. receive. A recurring moderator is whether learners construct the map or merely study one. Constructing a map from scratch is demanding and can overload novices; studying an expert-provided map can be under-challenging for advanced learners. Scaffolded construction (partial maps, fill-in links, node lists) is often the practical middle ground, consistent with cognitive load management and scaffolding.

Domain fit. Concept mapping works best for material with rich relational structure — hierarchical taxonomies, causal systems, interrelated processes — and adds little for content that is genuinely list-like or procedural, where a concept map imposes arbitrary structure.

Open questions. Much of the literature compares mapping to passive or less-structured alternatives; comparisons against equally generative activities (e.g., self-explanation, summarization) show smaller or mixed advantages. Whether the benefit lies in the visual format or in the relational processing it elicits remains contested.

Scoring and implementation. When maps are used as assessment, scoring the quality of links (labeled, directional, hierarchically organized) rather than node count better captures relational understanding. Time-on-task is a persistent confound: mapping is slow, and some reported advantages shrink when matched against other activities given equal time.

Evidence status. Three studies are recorded above: two meta-analyses (Schroeder et al. 2018, g = 0.58 overall, constructing g = 0.72 against studying g = 0.43; Nesbit & Adesope 2006, read from the abstract with no pooled effect recorded) and one pair of randomized experiments (Karpicke & Blunt 2011) in which time-matched retrieval practice beat mapping. Both meta-analyses were read from abstracts only, so their moderator results on comparison type and time-on-task are not recorded here. The strength of support depends heavily on the comparison condition (passive study vs. generative alternatives) and on whether time-on-task is matched — both moderators flagged above. Reading those moderators from the meta-analyses' full texts is what would settle how far the pooled advantage survives a time-matched generative comparison.

Merged from “Concept mapping improves learning outcomes” (concept-maps-improve-learning): Mechanism. Concept mapping is typically explained through cognitive load and schema-building arguments: selecting relevant concepts, organizing them into hierarchical structures, and integrating them with prior knowledge forces generative processing that passive reading or rereading does not. This places it in the family of generative, elaborative strategies alongside advance organizers and chunking, and it depends on learners actively activating prior knowledge to link new concepts to what they already know.

Construction matters more than viewing. The literature generally distinguishes learner-constructed maps from expert-provided maps. Constructing a map is a generative activity; merely studying a supplied map risks the passive-processing problems that undermine other transmission formats. Designers should treat map construction as the active ingredient, with scaffolds (starter nodes, partial maps) faded as learners gain competence.

Moderators and boundary conditions. Expected moderators include learner expertise (novices benefit most; experts may find mapping redundant, consistent with expertise reversal patterns), domain structure (mapping suits hierarchically organized, relational domains better than loosely associated ones), and scoring or feedback on map quality — without feedback, learners may build structurally incorrect maps and consolidate misconceptions. Construction also imposes a high extraneous load on novices unfamiliar with the format, so instruction on how to map is usually necessary first; unconstrained construction can itself produce cognitive overload.

Open questions. The pooled effects and the time-matched comparison are recorded above; the moderators named in this paragraph (learner expertise, domain structure, feedback on map quality) are not tested by any entry here.

Related Claims¶

• Prompting learners to self-explain, one generative strategy, improves learning by a moderate average amount; other generative activities are not tested by the evidence recorded here — the general claim this finding bears on, with the studies behind it

• Constructive learning activities produce better learning than active or passive ones. — mapping is a paradigmatic constructive activity

• Retrieval practice improves long-term retention. — mapping from memory combines generation with retrieval
• Dual coding improves learning. — maps pair verbal and spatial representations
• Advance organizers improve learning. — supplied maps function as graphic advance organizers
• Chunking reduces working memory load. — maps externalize and chunk relational structure that would otherwise burden working memory
• Cognitive load reduction improves learning. — scaffolded map construction manages the load that free-form mapping imposes
• Analogical reasoning improves transfer. — mapping elicits the same relational-structure comparison that drives analogical transfer
• Cognitive overload degrades learning — unconstrained map construction can itself overload novices
• Learner highlighting of text gives a small memory benefit but no reliable comprehension benefit, and evidence on other forms of annotation is not yet recorded here — another generative strategy that forces learners to process and restructure text
• Cognitive flexibility theory: multiple cases — concept mapping is a core tool within cognitive flexibility theory for representing multiple linked perspectives
• Learner-constructed graphic organizers are not shown to outperform provided ones: the one direct test, with college readers, favoured provided organizers on transfer — related
• Graphic Organizers Support Novice Comprehension — related
• An ICAP theory article reports, second-hand and without describing the design, a study in which concept mapping with a peer enhanced learning more than concept mapping alone — related
• Split Attention Effect Degrades Learning — related
• Drawing improves STEM learning by a small-to-moderate amount over not drawing and over other active study strategies, in one meta-analysis — related
• Retrieval practice benefits extend to nonverbal visual materials, spatial map learning, and educational texts with different structures — related
• A review chapter (Karpicke 2017) reports, second-hand, experiments in which retrieval practice produced more learning than concept mapping (Karpicke & Blunt 2011) and than imagery strategies such as the keyword mnemonic (Karpicke & Smith 2012) — related
• Learning is an initiative construction of meanings completed by the interaction of learners' old and new knowledge — a broader claim this one bears on
• A map risks becoming a new metanarrative unless it is continually remapped and readers actively construct their own maps — related
• Multiple representations and explanations of math concepts aid student understanding, but faculty lack time to source or develop them — 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 — related
• Concreteness, imagery, and verbal associative processes play major roles across multiple educational domains — related
• Self-explanation beats unguided studying for equal time, but instructional explanations, problem solving and retrieval practice can be equally or more effective — related
• Self-constructing a complete concept map produces better delayed recall than completing, sorting a map, or writing a summary, with the largest effect in the I+C structure — a narrower finding that bears on this claim
• Self-constructed collaborative concept maps trigger more deep exploratory talk and less non-dialogic talk than completing or sorting maps, which yield mostly superficial exploratory talk — related
• AI shifts the learner's role from producer to evaluator of mathematical meaning across multiple generative strategies — related
• Placing heavier cognitive demands on learners can be counterproductive in mapping tasks — related
• Task type and activity structure interact on recall: self-constructed maps help mainly when individually prepared and then discussed in pairs — a narrower finding that bears on this claim
• The same quizzing or concept-mapping activity produces more learning when done without viewing the material, as retrieval, than while viewing it — related
• Merely adding an expert example concept map does not add to the value of students creating their own concept maps for this population — related
• High cognitive activities were positively associated with essay quality but not with concept maps — related
• No evidence for a mapping between representational format and enhancement of a specific kind of understanding — reports the opposite
• Triads with access to each other's concept maps were more efficient in collaborative problem solving — a narrower finding that bears on this claim
• Concept mapping has a moderate overall positive effect on science achievement for elementary and secondary students (g = 0.776) — related
• Active-processing focus-support techniques such as testing, self-explaining, concept mapping, and drawing show positive effects on learning — a broader claim this one bears on
• Self-generated concept mapping benefits learning from grade 4 onwards with similar effect sizes across grades 4–8, 9–12, and college — a narrower finding that bears on this claim
• Drawing to learn shows positive effects on learning performance compared with non-generative and text-focused learning strategies — related
• Text quality after initial study moderates whether free recall beats generative tasks — related
• Visualizing support—provided diagrams or learner-generated drawings—facilitates higher-quality explanations and better comprehension than explaining text alone — related