Recall from bar graphs is more accurate when the recall format matches the studied graph format, a small effect that the content's thematic emphasis can override¶

Claim · All claims
Evidence · 1 study · causal r? · q3 · i2 medium · n=157 (Experiment 1) and n≈134 (Experiment 2), German university students
Currency · newest study 2015 · checked against its sources 2026-10-20 (automated, 1 of 1 entry) · not reviewed by a person

The one study recorded here tested bar graphs, not graphic organizers, and did not measure cognitive load. In two experiments, university students recalled election data more accurately when the recall graph matched the format they had studied (η² = .03–.04), and in Experiment 1 cueing the unfamiliar religion perspective degraded recall of party-format items (d = 0.52–0.54). But students also recalled better from whichever perspective was thematically dominant in the material, regardless of the format studied. The authors interpret the cost as load from an incongruent schema; extending it to graphic organizers, and to extraneous load as such, is this page's inference.

Subclaims¶

q3 i2 When the surface structure of a graphic (its visual format) mismatches the structure a task or later recall demands, comprehension/recall accuracy drops relative to a matched format — but this mismatch penalty is modest and can be outweighed by which content is thematically central to the material. → Schnotz & Baadte 2015

Evidence¶

Schnotz & Baadte 2015¶

Schnotz, W., & Baadte, C. (2015). Surface and deep structures in graphics comprehension. Memory & Cognition, 43(4), 605–618. doi:10.3758/s13421-014-0490-2

q3 · peer-reviewed lab experiments (two studies) · i2 · medium effect: d=0.52–0.54 on Experiment 1's instruction (schema-interference) contrasts, d=0.51 on Experiment 2's thematic-dominance contrast; smaller format-match interaction, η²=.03–.04 · n=157 (Experiment 1) and n≈134 (Experiment 2), German university students · causal · r?

Two experiments (graph format and instruction varied between subjects, recall-item format within subjects) had students study bar graphs of election results — the real 1956 and 1960 US presidential elections (adapted from Tufte) in Experiment 1, the same years in a fictitious country, Ustan, in Experiment 2 — presented in one of two informationally-equivalent visual formats ("party" graphs vs. "religion" graphs), then recall the content by filling in blank graphs of either the studied or a differently-formatted item. In both experiments a significant graph-format × recall-item-format interaction showed that, for each item format, participants who had studied that format recalled it more accurately than those who had studied the other format (F(1,151)=6.52, p=.006, η²=.04 in Exp. 1; F(1,128)=4.22, p=.021, η²=.03 in Exp. 2), and in Experiment 1 an instruction cueing the unfamiliar religion perspective significantly degraded recall of party-format items, relative to no instruction, both for learners who had studied party graphs and for those who had studied religion graphs (t(151)=2.03, p=.022, d=0.54, and t(151)=1.99, p=.025, d=0.52; in Experiment 2 the instruction effect tested was only marginal, p=.069) — consistent with the authors' framing that an incongruent schema "impose[s] a higher cognitive load on working memory," producing interference. The important qualifier: participants also recalled content more accurately from whichever perspective was thematically dominant in the material regardless of which graphic format they had studied — in Experiment 2 they recalled the non-studied format's items better than the studied one (t(66)=3.83, p<.001, d=0.51). So format (surface-structure) mismatch imposes a real but modest processing cost, which the content's underlying task-relevance can override.

Discussion¶

What is and is not shown. The paragraphs below apply the bar-graph finding to graphic organizers through cognitive load theory. No study recorded here tests a graphic organizer or measures extraneous load, and the strength tags below predate that reading.

Mechanism. Within Cognitive Load Theory, extraneous load arises from instructional design that requires processing not directed at learning. A well-matched organizer (e.g., a compare/contrast matrix for comparative content, a cycle diagram for cyclical processes) externalizes relationships so learners need not hold them in working memory — consistent with the claim that chunking reduces working-memory load [+M]. A mismatched organizer adds a second, partially inconsistent representation: learners must map organizer slots onto content that does not fit them, splitting attention between the two structures — a condition analogous to the split-attention effect. The result is the same outcome described in cognitive overload degrades learning [+S]: resources spent on reconciliation rather than schema construction.

Constraints. The claim is about mismatched organizers, not organizers generally — the penalty applies only under specific conditions:

• Imposed rather than learner-generated. An organizer assigned without reference to the content's actual structure forces reconciliation work on the learner [-M].
• Unfamiliar content structure. Novices lack the expertise to detect and repair an ill-fitting frame, so they process the mismatch rather than the content [-M].
• Advanced learners. For experts, any imposed organizer may simply be redundant — consistent with the expertise reversal effect [~S].
• High element interactivity. When the content is already near working-memory capacity, the added reconciliation load is most likely to push learners past capacity [-M].

Well-aligned organizers and advance organizers can support learning — see Advance organizers improve learning [+M]. The penalty also parallels the coherence principle [+S]: like decorative but irrelevant material, an ill-fitting organizer is a design feature that invites processing unrelated to the learning goal.

Design implications. Before selecting an organizer, analyze the content's underlying structure — sequence, hierarchy, comparison, cycle, cause–effect — and choose or adapt a frame that mirrors it. Where no standard frame fits, a sparser outline or learner-constructed map is safer than forcing content into a borrowed template. This aligns with the broader guidance in cognitive load reduction and clear structure.

Open questions. How much structural mismatch is tolerable before load costs outweigh organizational benefits, and whether learner-constructed organizers avoid the penalty, remain under-researched. Studies are needed to establish this claim empirically.

Related Claims¶

• Graphic Organizers Support Novice Comprehension — the general claim this finding bears on, with the studies behind it

• Advance organizers produce a small average learning benefit, larger for high-ability than for low-ability learners — the general claim this finding bears on, with the studies behind it

• Advance organizers improve learning — well-matched advance organizers are the constructive counterpart to the mismatch penalty claimed here.

• Chunking reduces working-memory load — matched organizers work by chunking related content; mismatches defeat that chunking.
• Coherence principle: irrelevant material hurts learning — both claims describe design features that add extraneous processing.
• Cognitive overload degrades learning — the consequence when mismatched organizers push learners past working-memory capacity.
• Task difficulty (easy vs difficult simultaneous-equation problems) has significant main effects on intrinsic and extraneous cognitive load, but not on germane load — related
• Typical verbal achievement tests may underestimate students' subject-matter knowledge because of a mismatch between the test's symbolic form and students' symbolic encoding — related
• Subjects with a more developed schema for a body of knowledge show higher recall for materials related to that knowledge — related
• The finding that single-discipline course takers perform as well as peers despite instruction-assessment mismatch runs counter to prevailing alignment assumptions — related
• The webinar presents glyphs, pictographs, bar graphs, and Google Forms as ways for K-2 students to collect and analyze data. — related
• Teams with part-task congruent tools adjusted their representations across phases—broad causal representations first, then more selective algebraic specification—unlike single-tool teams — a narrower finding that bears on this claim
• Cognitive load theory may help identify mismatches between representational schemes and the execution of (sub)tasks in CSCL — related