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¶

Claim · All claims
Evidence · 2 studies · 1 quant-synthesis r?, 1 review r2 · q3–q4 · i1 small
Currency · newest study 2022 · not yet checked against its sources · not reviewed by a person

Both studies recorded here concern highlighting and underlining only. A meta-analysis found that learner-generated highlighting improved memory for text with a small effect (0.36) but not comprehension (0.20), and helped college students but not school students; a review of study techniques rated highlighting low utility. Neither tests elaborative annotation such as margin notes, summaries or questions, so the slug's broader "annotating" is not established here.

Subclaims¶

q4 i1 Asking learners to highlight or underline important material while reading improves memory for the text (average effect size 0.36) but not comprehension (0.20), and helps college students but not school students; this covers highlighting only, not elaborative margin notes. → Ponce et al. 2022

q3 i? A review of ten common study techniques rated highlighting and underlining as low utility, which qualifies any general claim that unguided marking of text improves learning. → Dunlosky et al. 2013

Evidence¶

Ponce et al. 2022¶

Ponce, H. R., Mayer, R. E., & Méndez, E. E. (2022). Effects of learner-generated highlighting and instructor-provided highlighting on learning from text: A meta-analysis. Educational Psychology Review, 34(2), 989–1024. doi:10.1007/s10648-021-09654-1

q4 · meta-analysis · i1 · small effect, 0.36 on memory; 0.20 on comprehension · n=36 articles (85 effect sizes) · quant-synthesis · r?

A meta-analysis of experiments, published between 1938 and 2019, in which college or K-12 students read an academic text either with or without being asked to highlight important material (or with or without the important material already highlighted by the instructor), and then took memory or comprehension tests. When learners did the highlighting themselves, memory improved (average effect size 0.36) but comprehension did not (0.20). The benefit held for college students (0.39) but not for school students (0.24). Highlighting supplied by the instructor did better, improving both memory and comprehension (0.44 each). Read from the abstract only. The study covers highlighting and underlining, not other kinds of annotation such as written margin notes.

Dunlosky et al. 2013¶

Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students' learning with effective learning techniques: Promising directions from cognitive and educational psychology. Psychological Science in the Public Interest, 14(1), 4–58. doi:10.1177/1529100612453266

q3 · narrative review of experimental literature · i? · no pooled effect size reported · n=10 techniques reviewed · review · r2

A long review that asks, for each of ten study techniques, whether its benefits hold up across different learning conditions, types of student, materials and tests. Highlighting (or underlining) was one of five techniques rated low utility, along with summarization, the keyword mnemonic, imagery use for text learning and rereading. Practice testing and distributed practice were rated high utility. This counts against the claim for unguided highlighting in particular. It does not test more elaborative annotation. Read from the Crossref-deposited abstract only.

Studies bearing on this claim¶

Does asking learners to highlight or underline while reading improve learning over reading alone, and for which outcomes? This table gathers the studies behind this claim and the 5 narrower claims that bear on it: 4 distinct studies (2 quant-synthesis, 1 review, 1 causal), q2–q4, 2 of 4 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. Annotating is a generative activity: it forces selection of important information and connection of it to prior knowledge, consistent with generative theories of learning. It also produces an external record that supports later review and retrieval practice. Annotation quality likely matters more than quantity — superficial highlighting without elaboration can create an illusion of fluency, and re-reading highlighted text is among the weaker study strategies in the learning-techniques literature.

Moderators to establish. Open questions include whether benefits depend on learner expertise (novices may highlight the wrong things), whether structured prompts or annotation schemes (e.g., genre-based or rhetorical annotation) outperform free-form marking, and whether benefits persist on delayed tests rather than immediate ones. Studies are needed across domains, media (print vs. digital annotation), and age groups before a strength rating can be assigned.

Boundary conditions. Annotation may impose extraneous load when the text is already dense or the learner lacks the background to judge relevance; in such cases, guided or scaffolded annotation is likely preferable to unguided marking. This parallels the expertise-reversal pattern documented for other scaffolds: support that helps novices can become redundant for advanced learners — see Expertise reversal effect. Relatedly, cognitive load management suggests annotation tasks should be matched to text difficulty and learner background.

Design implications. Because unguided marking is unreliable, designers should pair annotation tasks with training in what to mark (e.g., modeling expert annotation), require elaborative annotations (summaries, questions, connections) rather than mere highlighting, and build in opportunities to revisit and use annotations for retrieval rather than re-reading. Structured annotation tasks fit naturally within active learning sequences and can serve as an activation step before discussion or problem-solving. They also align with dual coding when annotations combine verbal notes with visual marks.

Status. Two sources are recorded: a meta-analysis of 36 articles on highlighting (Ponce et al. 2022, read from its abstract) and a narrative review rating highlighting low utility (Dunlosky et al. 2013). Both concern highlighting and underlining. Evidence on elaborative annotation (margin notes, summaries, self-questions) is still missing, so the generative-processing rationale in the Mechanism paragraph is untested here.

Related Claims¶

• Activation improves learning — annotation activates prior knowledge during reading
• Chunking reduces working memory load — annotation segments text into manageable units
• Annotating — the underlying design principle this claim supports
• Von Restorff effect text marking — distinctive marking makes annotated items more memorable
• Expertise reversal effect — scaffolds that help novices can burden advanced learners, including annotation guidance
• Cognitive load reduction improves learning — annotation must not add extraneous load on dense texts
• Relevancy of emphasized text directs attention and influences test performance — related
• Highlighting shows low utility for improving learning outcomes — reports the opposite
• In one experiment, students who generated their own underlining did not score significantly higher on tests than students given experimenter-generated underlining — related
• Strong acquisition tasks explicating the organization produce better internalization than weak tasks asking only for a structured summary — related
• Deep thinking and making connections help students build strong memories and learn more deeply — related
• Summarization, highlighting, the keyword mnemonic, imagery for text, and rereading receive low utility assessments — related
• Ten-year-old readers predominantly underline subordinate passage details rather than superordinate conceptual material, even when conceptual postquestions focus attention on topic sentences — a narrower finding that bears on this claim
• Underlining and adjunct questions may hinder rather than help recall in young readers, reflecting a comprehension rather than a metacomprehension deficit — reports the opposite
• Adjunct aids such as learning objectives and embedded questions show limited effects and may interfere with competent readers' natural strategies — related
• Merely selecting words by clicking on them, without any memorability instructions, enhanced their recall (a choice effect) — related
• Across ten learning techniques, distributed practice and practice testing have the largest effects and underlining and summarization the smallest — related
• Trace-data cognitive learning strategy codes show no correlation with highlighted key words or auxiliary words, while key-word and auxiliary-word highlighting correlate highly — related
• Recall of passage details is most depressed when children receive both underlining and conceptual postquestions — a narrower finding that bears on this claim
• Structuring worked examples to emphasize subgoals, via labels or visual segmentation, enhances transfer to novel problems — related
• Engagement via annotation (underlining, highlighting, notes) shows no significant effect on comprehension — related
• Better note quality and quantity relate to better memory, verbatim overlap relates negatively, and noted propositions are far more likely to be recalled — related