Signaling Principle

Principle · All principles
Evidence · 32 claims (15 for, 16 mixed, 1 against) · 26 studies (11 review, 10 causal, 4 quant-synthesis, 1 theoretical), q1–q4 · 8 of 26 report an effect size · 28 claims rest on one study

Conditional relationship

This page owns one decision inside Multimedia Learning: whether, and how heavily, to cue the essential material and its organisation. When learners study a lesson whose essential elements and structure they cannot yet pick out for themselves (a novice on this material), adding cues that point to those elements and to how they connect is expected to improve learning over the same lesson without the cues. Cues include headings, an overview sentence or outline, highlighting or bold, arrows, colour that links a word to the part of the picture it names, and spoken emphasis. Signals add no new content; they change where attention goes.

The claims in this wiki test the relationship directly, in two meta-analyses, an overview of meta-analyses and a series of experiments. They also report where it bends:

• The size. Across 29 experiments (2,726 participants) signaling was associated with better learning, d = 0.38 (Alpizar et al. 2020). The largest signaling meta-analysis, as reported in Noetel et al.'s (2022) overview, gives g = 0.43 (k = 209) on learning and g = 0.25 on reduced cognitive load. Mayer's (2008) programme found signaled lessons better on transfer in 6 of 6 experiments, median d = 0.52.
• Prior knowledge. In an eighth-grade field study, signals linking text to pictures helped learners with low prior knowledge and harmed those with high prior knowledge, with higher extraneous load (Richter et al. 2018). In an eye-tracking study with middle-school students, revised visuals (labels, colour coding, call-out boxes) reduced processing for low-prior-knowledge students and increased it for middle-prior-knowledge students, with no difference on the post-lesson test (Clinton et al. 2017). Noetel et al. found no consistent prior-knowledge moderation across multimedia principles as a whole, so the reversal is recorded for signaling in two studies, not as a general rule.
• Pacing and complexity. Good multimedia design, signaling among it, mattered more for complex material and for system-paced lessons (Noetel et al. 2022). Alpizar et al. report that pacing and other presentation features moderated the effect.
• What is cued. Cues move attention to what is marked, whether or not it matters: underlining low-relevance text raised scores on low-relevance questions (Nist & Hogrebe 1985), and learners shown a cued object noticed less elsewhere on the screen (Mutlu-Bayraktar & Bayram 2018). A signal helps the outcome it points at.
• Outcome. In expository text, organizational signals often changed what college readers recalled more than how much (Lorch & Lorch 1996). Most of the effects were measured soon after a short lesson.
• Representation and age. In a middle-school electricity study, signaling the written text helped with an abstract animation (d = .35) but not significantly with a concrete one (Kutbay & Akpınar 2020).

How this sits beside its neighbours: Multimedia Learning holds the umbrella model (words with relevant pictures, and the principles that qualify it) and its pattern the reusable design; this page goes deeper on the cueing decision only and does not repeat that model. Clear Structure covers making a lesson's or course's organisation explicit more broadly (organizers, graphic organizers, course-level structure); signaling is its cue-level instance within a presentation, and the two share the meta-analysis behind Clear structure improves learning [+M]. Annotating is about marks the learner makes; signals are made by the designer. Cognitive Load Management gives the shared explanation: a cue saves a novice the search for what matters.

Default design, while the dose is untested

The claims establish the direction, the size on average and the prior-knowledge boundary. None in this wiki sets a dose (how many cues per screen, which kind of cue, when to fade them). This default is a labelled proposal: the kinds of cue and the fade follow the claims named, the numbers are untested choices.

1. Check the condition. Is the material complex enough that a newcomer could miss what matters (a busy diagram, an animation, a long expository text, a system-paced video)? Are the learners new to this content? If both are no, signal lightly or not at all (Signaling improves learning [~M], the Richter et al. reversal).
2. List the essential elements first, from the learning goal and the assessment the brief uses: the three to five ideas, parts or steps the test or task will ask about. Cue only these. Cues move attention to whatever is marked (Relevancy of emphasized text directs attention [+M]). Three to five is an untested proposal.
3. Signal the organisation. Open with a one- or two-sentence overview of the steps or parts, give each section a heading that names its idea, and close with a summary (Signaling principle, Mayer 2008 [+M]: overview sentences, headings and vocal emphasis).
4. Signal within the display. Mark the one element each screen, slide or paragraph is about (bold, highlight, an arrow), and use the same colour for a term and the part of the picture it names. In narration, stress the key words. On a busy screen, pair a highlight with a pointing cue such as an arrow (The signalled and cued change was most detected [+W]). One cued element per screen is an untested proposal.
5. Early sessions (first one or two on a topic): full signals as above, including the colour or label links between text and picture.
6. Later sessions on the same topic, once a short probe shows learners can name the parts and steps unaided: drop the text–picture link cues first, keep headings and the overview. Added link cues harmed learners with high prior knowledge and raised processing for middle prior knowledge (Signaling improves learning [~M]; Revised visuals reduce processing of lesson text for low prior knowledge students but increase it for middle [~M]). The probe (two or three "name the parts" items) and the order of fading are untested proposals.
7. Move on when the learner explains the material without the cues. If performance drops with cues removed, restore them for one more session (untested proposal).
8. Judge it on the brief's own test, and check that the test asks about the cued elements; a cue that points at what the test does not ask about cannot show a gain.

Fitting the design to a situation

Three facts change the decision most. Ask for them when the brief omits them: how much do the learners already know about this material? (a newcomer or someone who could explain it already); how complex is the display and who sets the pace? (a busy system-paced animation, or a short learner-paced text); and what will the learners be tested or asked to do? (so the cues point at it). Most of the evidence is from lab lessons with adult (postsecondary) learners on immediate tests; rows below that carry it to children, other media or longer horizons say so.

If the brief says…	Then change…	Basis
Learners are new to the content (novice) and the material is a complex multimedia lesson	Use the full default design: overview, headings, one cued element per screen, colour links between words and picture	Signaling improves learning [+S] (d = 0.38 over 29 experiments; g = 0.43, k = 209)
Learners already know much of the material (advanced)	Drop the text–picture link cues; keep at most headings; let learners turn highlighting off	Signaling improves learning [~M] (signals harmed high-prior-knowledge eighth graders in one field study)
A class with mixed prior knowledge	Give a short prior-knowledge probe and offer the cued version to those who score low; expect no difference on the test for the rest	Revised visuals reduce processing of lesson text for low prior knowledge students but increase it for middle [~M]; Revisions to visuals had no effect on post-lesson test scores [~M]
Learners are children in middle school learning from animation	Signal the written text where the animation is abstract (symbols, schematic); a concrete, realistic animation may need less	The signaling effect holds only for abstract animation [~M] (one quasi-experiment, d = .35 abstract; not significant concrete)
Pre-readers or early readers with a read-aloud e-book	Highlight or track each word as the narration reads it, to bring eyes to the print; expect attention to drift once the animation stops, and do not expect a reliable comprehension gain	Narration with finger-tracking animation directs preschoolers' attention to print [+M]; Attention drops after the animation ends [~M]; Comprehension benefit only marginal [~M]
Adults in a short system-paced video	Signal heavily: spoken overview, on-screen headings for each segment, a pointer or highlight on the part being described	Signaling improves learning [+S] (design mattered more for system-paced lessons)
A learner-paced text or print reading	Use headings, a preview and a summary; test what is recalled as well as how much	Clear structure improves learning [~S] (signals shifted what readers recalled more than the amount)
Slides for a lecture	One heading that states the slide's point, one highlighted element per slide, and an agenda slide at the start	Signaling principle, Mayer 2008 [+M]; the one-per-slide dose is untested
The brief chooses between spoken narration and on-screen text for young learners	Signaled written text did as well as spoken text in one middle-school study; if text must stay, signal it	Spoken text does not differ from signaled written text [~M] (a null, not a test of equivalence)
A VR scene, simulation or busy animated display	Pair each highlight with a pointing cue (arrow), keep few objects on screen, and expect learners to miss changes outside the cued area	The signalled and cued change was most detected [+W]; Signalling focuses attention and reduces change detection elsewhere [~W]; carrying it to VR is untested
Phone-only, small screen	Fewer cues fit: keep the heading and one highlighted element per screen, and put the overview first	untested
A single session	Use the early-session design only; there is no later session to fade in	untested proposal
A fixed short time budget (for example ten minutes a day)	Signals add no content, so keep them; spend the minutes on the cued elements, not on extra decoration	Signaling (cueing) increases learning and reduces cognitive load [+M] (reduced load g = 0.25); its time cost is untested
One teacher for many learners	Build the cues into the shared materials once (slides, handouts, video) rather than relying on the teacher to point in the moment	untested proposal
Low literacy, or learners reading in a second language	Label illustrations with short text and use colour or arrows to link label and part; pair the written key word with narration for young bilingual readers	Signaling with labeled illustrations is associated with increased learning (review for emergent bilinguals) [+W]; Narration with finger-tracking directs attention to the target-language print [+M]
Learners who are deaf or hard of hearing	Carry every cue visually (captions, on-screen headings, highlight, arrow); do not rely on vocal emphasis	untested
Learners who are blind or have low vision	Carry cues in structure a screen reader announces (real headings, a spoken overview, stressed key words); colour alone signals nothing	untested
The brief supplies its own materials, routine or rubric	Keep them as given; add cues to them (headings, highlight, an overview) and point the cues at what the rubric scores; run the brief's own assessment, and add no test it did not ask for	Relevancy of emphasized text directs attention [+M]; keeping the materials is untested
Private or sensitive material	No change: signals need no learner data; a prior-knowledge probe, if used, stays with the teacher	untested
The task must be kept whole (a full case, a whole procedure)	Keep it whole and cue its parts and their order inside it, rather than splitting it	untested proposal
High-stakes assessment follows	Cue only what the assessment covers; marking the wrong material raises scores on that material instead	Relevancy of emphasized text directs attention [~M]
The plan is to have learners highlight their own copies instead	Supply designer cues; learner highlighting alone has low utility, and in one experiment learners' own underlining did no better than supplied underlining	Highlighting shows low utility [-M]; Students' own underlining did not beat experimenter underlining [~M]
The test that matters is delayed, or is use on a real task	Keep the cues, but do not count on the immediate-test effects to carry; check at the delay	untested; the recorded effects are mostly immediate

Observation, state and explanation

Record what the learner did with the cued material under stated conditions: which cues were present, who set the pace, and what they were asked to do afterwards. Where a learner looked, how long they took and what they rated as effort are observations; an organised representation of the material is an inferred state.

Response	Plausible explanations	Observation that could change the interpretation
Recalls the cued items but not how they connect	Cues marked elements but not their organisation; the test asked only for the items	Add an overview and connecting cues; ask for an explanation of how the parts act on each other.
Learner who knows the topic slows down or does worse with cues	Cues are redundant with what they already know and add material to process	Give the uncued version and compare on the same test; probe prior knowledge first.
Scores well on cued content and poorly on the rest	Attention went to the cues, as intended or too much; the cues missed something the test asks	Check the cue list against the assessment; ask about an uncued but essential item.
No difference with or without cues	Material simple enough to need none; learners already knew it; cues too many to stand out	Count the cued elements per screen; try a version with one cue per screen; check prior knowledge.
Looks at the cued print during narration, then drifts	The cue holds attention only while it moves	Check whether the drift matters for the outcome; repeat the cue at the point that matters.

These rows are proposals, untested with learners as diagnostics. A single contrast shifts confidence among the explanations; it does not identify a cause.

Evidence and qualifications

• Signaling improves learning [+S]: Noetel et al. (2022), an overview of 29 reviews, ranks signaling among the multimedia principles with the largest benefits and gives g = 0.43 (k = 209) from the largest signaling meta-analysis; Alpizar et al. (2020), a meta-analysis of 29 studies (44 effect sizes, 2,726 participants), d = 0.38, with study, participant, presentation and method features moderating it and clearer benefits in higher-quality studies. The third entry qualifies it: Richter et al. (2018), a quasi-experimental field study with eighth graders, found text–picture integration signals helped low-prior-knowledge learners and harmed high-prior-knowledge ones. All three were read from abstracts or a preprint.
• Signaling principle: highlighting essential words improves transfer from multimedia lessons [+M]: 6 of 6 experiments (computer lessons on airplanes, paper lessons on lightning and biology), median d = 0.52 on transfer. One research programme, reported in a review; read its median as that programme's result.
• Signaling (cueing) increases learning and reduces cognitive load in multimedia [+M]: the same Noetel et al. overview, adding that signaling reduced cognitive load (g = 0.25, k = 27), increased learning-relevant fixations (g = 0.39, k = 14) and slightly raised motivation (g = 0.13, k = 13).
• Clear structure improves learning [~S]: Alpizar et al. again, and Lorch & Lorch (1996), two experiments with 139 college readers in which organizational signals often did not raise the amount recalled but shifted what was recalled.
• Relevancy of emphasized text directs attention and influences test performance [+M]: 67 college freshmen; underlining high- or low-relevance text raised scores on questions about whichever was underlined. One experiment, immediate test.
• The signaling effect holds only for abstract animation representations [~M]: middle-school students learning electricity; signaled written text beat unsignaled written text with abstract animation (d = .35) but not significantly with concrete animation (p = .074). The entry calls the design a quasi-experiment.
• Revised visuals reduce cognitive processing of lesson text for low prior knowledge students but increase it for middle prior knowledge students [~M] and Revisions to visuals had no effect on post-lesson test scores [~M]: Clinton et al. (2017), 62 middle-school students reading an algebra lesson with original or revised visuals (labels, colour coding, call-outs, applying signaling, contiguity and coherence together). The eye-tracking measures moved in opposite directions by prior knowledge; the test score did not move. A non-significant difference is not equivalence.
• The signalled and cued change was the most detected change [+W] and Signalling focuses attention and reduces change detection elsewhere [~W]: Mutlu-Bayraktar & Bayram (2018), an eye-tracking case study with a small group of undergraduates (12 in the detection counts); 10 of 12 detected the change on the signalled and cued object. Descriptive counts, no comparison group.
• Narration with finger-tracking animation directs bilingual preschoolers' attention to the target-language print [+M], Attention drops after the animation ends [~M] and Comprehension benefit only marginal [~M]: Yow & Priyashri (2019), eye-tracking with preschoolers; the cue moved looking to the print while it ran (partial η² = .74), less after, and comprehension rose only marginally (p = .060) in dual-language books and not at all in single-language ones.
• Spoken text modality does not significantly differ from signaled written text modality [~M]: Kutbay & Akpınar again; a null result, not a test of equivalence.
• Signaling in multimedia (labeled text with illustrations) is associated with increased learning outcomes [+W]: a narrative review for dual-language instruction, attributing the result to three cited studies; no effect sizes.
• Highlighting shows low utility for improving learning outcomes [-M]: Dunlosky et al. (2013); learner-made highlighting, a different act from designer signals, gave little benefit over reading.

Taken together: designer-made cues that point at the essential material improve learning from complex multimedia lessons by roughly a third to a half of a standard deviation on average, mostly on immediate tests with university students. Cues help what they point at; for learners who already know the material they can add load, and in two studies they did. No claim in this wiki tests the number of cues, a fading schedule, or delayed outcomes.

The evidence by question

Each question below links a general claim that gathers the studies behind it into one table, one row per distinct study, with each study's design and effect as recorded and no averaged verdict (eval/synthesis/signaling.json, scripts/synthesize_claim.py).

• Do cues that highlight the essential material improve learning, and by how much? Signaling Improves Learning [+S]
• Do the cues still help learners who already know the material? Instructional guidance that helps novices can become redundant or counterproductive as expertise grows [~S]
• Does it matter what is cued, and what happens to attention elsewhere? Relevancy of emphasized text directs attention and influences test performance [+M]

Further evidence, not yet read against this model

• Change in the most-focused central object was detected at a low rate unless it was signalled and cued [+W]: the same case study as the change-detection claims above.
• Among two objects with the same signalling, participants focused more on the arrow-cued object [+W]: same study; a pointing cue beat a highlight alone for attention.
• Changes were detected more easily on screens presenting fewer objects [+W]: same study.
• Revised visuals reduce cognitive load while reading lesson text for low prior knowledge students but increase it for middle [~M], Revisions to math-relevant visuals affect cognitive load differently by prior knowledge group [~M] and Revised visuals reduce processing of math-relevant visuals for low prior knowledge students [~M]: further measures from Clinton et al. (2017).
• Multimedia enhancing features increase attention to print in single-language e-books [+M]: Yow & Priyashri, Study 2.
• Multimedia design principles produce a moderate average improvement in learning, with large differences between principles [+M]: Noetel et al.; signaling is among the larger ones.
• Multimedia principles benefit novices [~S] and Instructional guidance that helps novices can become redundant as expertise grows [~S]: Kalyuga et al.'s expertise-reversal account, from other cognitive-load paradigms than signaling.
• Advance organizers produce a small average learning benefit, larger for high-ability learners [~S]: the organizer, a neighbour of the overview signal, owned by Clear Structure.
• In one experiment, students who generated their own underlining did not score significantly higher than students given experimenter underlining [~M].

Objective and learner-valued goal

The designer's objective is that learners select the essential material and see how it fits together. Ask separately what the learner wants from it: a learner revising for a test values cues that match the test; a learner using the material on a job values cues on what the job needs, which may differ. Cues that serve the designer's list and not the learner's purpose will move attention, and the learner may still miss what they came for. Record both aims and point the cues at the outcome that will be checked.

What would revise this model?

The expectation should weaken if comparisons with comparable novice learners, holding content and time equal, find no learning advantage for cued over uncued material on aligned tests, or if the advantage seen on an immediate test is absent at a delay. It should be narrowed if more studies like Richter et al. and Clinton et al. find cues harming learners with moderate prior knowledge, which would make a prior-knowledge probe a requirement rather than a refinement. If the gains turn out to come from the cues matching the test items rather than from organising the material, the explanation should shift from organisation to attention alone. Do not protect the model by calling every null result a case of too many cues after the fact.

Claims

• In 5-year-olds the transparent-cue benefit appears only with finger tracking, producing a significant Cue × Tracking interaction [~W]
• Expert revisers' improvements are inconsistent across texts, with only signaling-structure strategies reliably corresponding to improved retention [+W]
• Signaling with on-screen text, color, contrast, or symbols improves retention and transfer from animations and video [+W]
• Combined verbal and nonverbal cueing most often draws attention to landmarks and clarifies tool use [+W]
• In a three-group eye-tracking experiment, the pointing group significantly outperformed both the tracing group and the no-touch group on comprehension performance [+W]
• Realism can serve as signaling when a drastic contrast in shading and rendering highlights relevant parts [~W]

Related Principles

• Modality Principle
• Redundancy Principle
• Segmenting Principle
• Contiguity Principle (Spatial and Temporal)
• Coherence Principle

• Pre-training Principle

• Multimedia Learning — the umbrella model this principle sits inside

• Clear Structure — making organisation explicit more broadly; shares the signaling meta-analysis
• Cognitive Load Management — why a cue helps a novice and can burden an expert
• Annotating — marks the learner makes, as distinct from designer cues
• Graphic Organizers — a picture of the organisation, rather than a cue on the material
• Scaffolding and Fading — the fade from full cues to few

Examples

• Signaling, Signaling and Highlighting and Visual Cues — strategies that carry the principle into materials
• "Headings and Highlight" Strategy
• Pedagogical Agent Signaling for Feedback Salience and Attention — building blocks
• Prior Knowledge and Attention — the learner variables the claims measure
• Multimedia Learning pattern
• Clear Structure Presentation
• Amplifying Key Ideas
• Highlight and Explore Patterns, Critical Features, Big Ideas, and Relationships
• Brown-Peterson working memory task with animated goal cue tracking for preschoolers

Key Sources

• Noetel, M., Griffith, S., Delaney, O., Harris, N. R., Sanders, T., Parker, P., del Pozo Cruz, B., & Lonsdale, C. (2022). Multimedia design for learning: An overview of reviews with meta-meta-analysis. Review of Educational Research, 92(3), 413–454. doi:10.3102/00346543211052329
• Alpizar, D., Adesope, O. O., & Wong, R. M. (2020). A meta-analysis of signaling principle in multimedia learning environments. Educational Technology Research and Development, 68(5), 2095–2119. doi:10.1007/s11423-020-09748-7
• Mayer, R. E. (2008). Applying the science of learning: Evidence-based principles for the design of multimedia instruction. American Psychologist. doi:10.1037/0003-066x.63.8.760
• Richter, J., Scheiter, K., & Eitel, A. (2018). Signaling text–picture relations in multimedia learning: The influence of prior knowledge. Journal of Educational Psychology, 110(4), 544–560. doi:10.1037/edu0000220
• Lorch, R. F., & Lorch, E. P. (1996). Effects of organizational signals on free recall of expository text. Journal of Educational Psychology, 88(1), 38–48. doi:10.1037/0022-0663.88.1.38
• Clinton, V., Cooper, J. L., Michaelis, J., Alibali, M. W., & Nathan, M. J. (2017). How revisions to mathematical visuals affect cognition: Evidence from eye tracking. In Eye-tracking technology applications in educational research (pp. 195–218). IGI Global. ERIC ED574984. doi:10.4018/978-1-5225-1005-5.ch010
• Kutbay, E., & Akpınar, Y. (2020). Investigating modality, redundancy and signaling principles with abstract and concrete representation. International Journal of Education in Mathematics, Science and Technology, 8(2), 131–145. ERIC EJ1255527
• Nist, S. L., & Hogrebe, M. C. (1985). The effects of high and low relevant text underlining on test performance. ERIC ED265520
• Mutlu-Bayraktar, D., & Bayram, S. (2018). Effects of cueing and signaling on change blindness in multimedia learning environment. World Journal on Educational Technology: Current Issues, 11(1), 128–139. ERIC EJ1205386
• Yow, W. Q., & Priyashri, S. (2019). Computerized electronic features direct children's attention to print in single- and dual-language e-books. AERA Open. doi:10.1177/2332858419878126
• Kalyuga, S., Ayres, P., Chandler, P., & Sweller, J. (2003). The expertise reversal effect. Educational Psychologist, 38(1), 23–31. doi:10.1207/s15326985ep3801_4
• 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
• Ceballos, M., & Nutta, J. W. (2022). Emergent bilinguals and multimedia instructional design: Applying the science of learning principles to dual language instruction. Education Leadership Review, 23(1). ERIC EJ1380081

Page metadata

Field	Value
type	principle
id	signaling
title	Signaling Principle
description	When learners study material whose essential elements and organisation they cannot yet pick out for themselves, adding cues that highlight those elements and how they connect (headings, an outline, highlighting, arrows, colour links between words and picture, spoken emphasis) is expected to improve learning over the same material without cues, by a small-to-medium amount in the meta-analyses the wiki records; the claims test this directly and report it weakening or reversing for learners who already know the material, and helping only for what is cued.
status	review
canonical	True
generated	by claude/unspecified on 2026-10-14
trust	unverified — no one has confirmed this page's content
sources	13 — noetel-2022, alpizar-2020, mayer-2008, richter-2018, lorch-1996, clinton-2017, kutbay-2020, nist-1985, mutlu-bayraktar-2018, yow-2019, kalyuga-2003, dunlosky-2013 … and 1 more

Frontmatter is the page's machine-readable record — see the Schema & Guide.