Schema

Key ideas

Core concepts

• Knowledge is organised in connected networks: individual pieces of knowledge integrate into larger concepts through chunking and recoding.
• Schemas reduce cognitive load because complex information stored as single functional units frees working memory for new learning.
• More connections make retrieval and new related learning easier.
• Instruction should be matched to a student’s current schema development, since strategies that help novices can hinder experts.

A schema is a network of organised knowledge stored in long-term memory that represents complex information as a single, functional unit, reducing cognitive load and enabling more sophisticated thinking as it develops (Bartlett, 1932; Anderson, 1977). Schemas explain how humans store knowledge, why experts and novices process information differently, and why learning becomes easier as knowledge accumulates. Schema theory informs teaching decisions including lesson planning, error correction, and differentiation strategies.

Connected to

Prior Knowledge | Chunking | Cognitive Load Theory | Memory | Learning | Retrieval Practice | Problem-Solving | Fluency | Experts and Novices Think Differently | Misconceptions | Part-Whole Approach

---

How schemas form

Schemas form through recoding or chunking, a process where individual pieces of knowledge become subsumed into larger concepts and assimilated into single, functional units (Miller, 1956; Chase & Simon, 1973). The process operates continuously as students learn. In initial learning, students encounter discrete pieces of information (individual facts, procedures, or concepts) that demand conscious attention. Learning the times tables begins with isolated facts, 3 × 4 = 12, 3 × 5 = 15, each requiring deliberate recall. With exposure and practice, students begin recognising relationships between these isolated pieces: threes always increase by three, patterns emerge in the units digits, multiplication relates to repeated addition.

Eventually, related pieces of information merge into single retrievable units. “Multiplication facts” becomes a schema containing hundreds of individual facts, yet functions as coherent knowledge the student can access and apply as needed. What once required processing dozens of separate elements now operates as a single chunk. Schemas themselves can chunk together into more sophisticated understanding: the multiplication facts schema combines with schemas for place value, addition, and problem-solving strategies to form a higher-level “arithmetic operations” schema. This hierarchical organisation continues developing throughout learning.

Unlike working memory’s severe 4-item limitation (Cowan, 2001), schema development in long-term memory has no practical limits. Research has identified no limit to either the complexity of individual schemas or the total number that can be stored. A mathematics teacher’s schemas for their domain contain tens of thousands of interconnected pieces of information, yet they continue learning and developing new schemas throughout their career. Well-established schemas also impose minimal cognitive load when accessed (Ericsson & Kintsch, 1995): the mathematics teacher can simultaneously access vast content knowledge whilst monitoring student understanding, managing classroom dynamics, and adjusting instruction, because schema retrieval from long-term memory doesn’t burden working memory the way processing new information does. Schemas never stop developing. Even expert teachers continue refining, extending, and connecting their schemas as they encounter new situations, reflect on practice, and engage with new ideas. This unlimited capacity creates the educational opportunity: by systematically building robust schemas, students develop the foundation for more complex learning and expert-like thinking.

Effects on learning and thinking

Existing knowledge facilitates acquisition of new related knowledge (Dochy, Segers, & Buehl, 1999). When a student possesses well-developed schemas related to new content, learning feels like recognition rather than memorisation: new information connects readily to existing knowledge, patterns become visible quickly, and understanding develops efficiently. Students lacking relevant schemas must process new information as genuinely novel and unconnected. Every element demands conscious attention, patterns remain invisible, and understanding develops slowly and incompletely. New content has nowhere to attach, making learning feel like trying to make something stick to a smooth surface.

This principle explains the “Matthew effect” in education (Stanovich, 1986), the pattern where strong students seem to learn more easily over time whilst struggling students fall further behind. Strong students’ accumulated schemas make new learning easier, whilst weak schemas make it harder. The rich get richer in knowledge through accumulated advantage rather than inherent ability. There are equity implications: students from varied backgrounds arrive at school with different schemas, and when curriculum assumes rather than builds foundational schemas, existing knowledge gaps may widen.

Well-developed schemas also change how students interact with problems. Students with weak or absent schemas must analyse problems element by element, consciously processing each component: “This is an equals sign, it means both sides have the same value. This ‘a’ represents a number. The 4 next to it means multiplication. I need to use inverse operations…” Students with developed schemas recognise problem types instantly. They see the same algebra problem and immediately categorise it: “Linear equation, isolated variable.” The solution method activates automatically, and they execute it with minimal conscious processing. What the novice analyses, the expert recognises. This shift from analysis to recognition is one of the core differences between expert and novice thinking (Chi, Feltovich, & Glaser, 1981).

The state of students’ current schemas determines what they can learn next, a principle often underappreciated in curriculum design. New knowledge must connect to existing understanding to transfer into long-term memory. Cognitive Load Theory explains why: working memory can process only about four novel elements simultaneously (Cowan, 2001), so if new content contains more than four elements that have not been chunked into existing schemas, cognitive overload prevents learning (Sweller, van Merriënboer, & Paas, 2019). When students lack prerequisite schemas, teaching advanced content doesn’t challenge them appropriately, it overwhelms them, and transfer to long-term memory becomes impossible regardless of teaching quality or student effort. Whether teaching literature analysis, scientific reasoning, or mathematical procedures, the same constraint operates: students need automated foundational schemas before complex integrated learning becomes possible.

From novice to expert

How schemas develop over time shows where students are in their learning journey and what instruction they need next.

Novices possess schemas that are typically defective, incomplete, or entirely absent for the content being taught. Knowledge exists as disconnected pieces without visible relationships. Students might know individual procedures but not when to apply them or how they relate. Few links between related concepts mean retrieval is slow and uncertain; students must deliberately search memory rather than activating connected knowledge automatically. Novices also often possess incorrect schemas containing systematic errors and misconceptions (Chi, 2008), which actively interfere with learning because they feel intuitively correct. And novices categorise problems by superficial characteristics (context, specific numbers used, diagrams present) rather than underlying structure (Chi, Feltovich, & Glaser, 1981). They see different problems than experts see. Teaching novices means building schemas systematically from foundations through Explicit Teaching of concepts and procedures rather than discovery, Worked Examples showing complete solution paths with explanations, correction of misconceptions through cognitive conflict, extensive guided practice with immediate feedback, and deliberate connection-making. Attempting to accelerate novices through this stage by presenting complex problems or expecting sophisticated reasoning produces frustration and failure, not learning.

As students gain experience with well-designed instruction and practice, their schemas begin maturing. Related concepts start grouping together. Students recognise that certain procedures connect to particular problem types, notice “this looks like that problem from yesterday”, and activate appropriate schemas. Some foundational knowledge begins retrieving automatically, reducing cognitive load: basic arithmetic, common procedures, and frequently used concepts require less conscious processing. Understanding remains fragile, though. Under pressure or in unfamiliar contexts, students struggle to apply knowledge that seemed secure in familiar situations. Instruction at this stage shifts toward consolidation and extension: guided practice with gradually reducing scaffolding, pattern recognition through carefully sequenced minimally different examples, Fluency built through distributed practice, connections strengthened through varied applications, and complexity increased gradually as foundational elements automate. Rushing students through this stage is counterproductive; they need extensive practice at this level to build the automated retrieval that enables expert-like thinking later.

Experts possess qualitatively different knowledge structures. Extensive connections between related concepts mean retrieving one element often activates an entire network of relevant knowledge (Chi, Glaser, & Rees, 1982). Foundational knowledge accesses instantly with minimal working memory load (Ericsson & Kintsch, 1995): experts do not consciously process basics, they automatically retrieve what they need whilst focusing attention on novel aspects of problems. Experts categorise problems by underlying principles and required solution methods, not surface features (Chi, Feltovich, & Glaser, 1981). They see “simultaneous equations requiring substitution” where novices see “a word problem about chickens and rabbits.” Expert schemas support transfer to novel situations: rather than retrieving fixed procedures, experts adapt their knowledge to new contexts fluidly, and they possess sophisticated domain-specific self-monitoring, recognising when solutions seem wrong, when alternative approaches might work better, and when to abandon unproductive paths. Instruction that challenged novices now bores experts. They need complex, high element interactivity problems that require integrating multiple schemas, minimal scaffolding (excessive guidance would slow down automated processing), and novel applications requiring adaptation of existing schemas. Unlike novices, experts learn effectively from wrestling with challenging problems, making problem-solving an effective learning strategy for this group, alongside continued metacognitive development in domain-specific reasoning and evaluation.

The Expertise Reversal Effect captures this: strategies effective for novices become counterproductive for experts, and vice versa (Kalyuga, Ayres, Chandler, & Sweller, 2003). Teachers must continually assess schema development and adjust instruction accordingly.

Schemas and problem-solving

The relationship between schemas and problem-solving capability explains why teaching problem-solving “skills” to novices is ineffective. According to Cognitive Load Theory, effective problem-solving requires three things. First, extensive domain-specific schemas: research examining expert performance across domains (chess, mathematics, science, medicine) finds that expertise requires extensive deliberate practice building schemas (Ericsson, Krampe, & Tesch-Römer, 1993), generally involving thousands of hours of focused practice, though the time varies by domain. Second, automated retrieval: these schemas must be accessible instantly, without conscious effort, because if an expert must deliberately search memory for relevant knowledge, their working memory fills with the search process, leaving no capacity for the novel aspects of the problem. Third, recognition of deep structure: experts must see beyond surface features to underlying principles, which requires schemas organised by deep structure (the actual requirements for solutions), not superficial characteristics.

The notion that novices can learn new content through problem-solving contradicts schema theory and cognitive load research. When novices lack solution schemas, they resort to means-end analysis: trying different approaches, checking if they are getting closer to the goal, backtracking when they are not (Sweller, Mawer, & Ward, 1983). This consumes extensive working memory resources, leaving minimal capacity for learning the solution method (Sweller, 1988). Superficial success also masks learning failure. Novices might solve a problem through trial and error guided by a teacher, but this doesn’t build transferable schemas: they solved this problem but lack the schema needed to solve similar problems independently. Repeatedly struggling with problems they lack the knowledge to solve damages students’ confidence and motivation, harming those from backgrounds where school success feels tenuous.

Decades of research show that novices learn more effectively from studying Worked Examples than from problem-solving (Sweller & Cooper, 1985; Cooper & Sweller, 1987; Atkinson, Derry, Renkl, & Wortham, 2000). Only after developing robust schemas does problem-solving become an effective learning strategy.

Schema rigidity and the Einstellung effect

Whilst schemas enable efficient thinking, they can also create mental rigidity that prevents recognition of simpler solutions (Luchins, 1942). Luchins conducted a seminal experiment examining how established patterns of thinking interfere with finding simpler alternatives. Participants solved problems requiring them to measure specific quantities of water using three jugs of different capacities. After solving several problems using a complex method (B-2C-A, where B represents the capacity of jug B, etc.), participants persisted with this approach even when simpler solutions became available. When a problem could be solved with A-C, participants continued using the complex B-2C-A method because it had worked previously. This demonstrated the Einstellung effect (German for “set” or “attitude”): prior experience with a particular solution method can create mental rigidity, preventing recognition of more efficient alternatives. The established schema for solving these problems blinded participants to simpler approaches.

The research points to risks in over-practising specific procedures without developing conceptual understanding. Students may apply familiar methods mechanically without considering whether simpler approaches exist: after learning the standard algorithm for fraction addition, students might apply it even when adding fractions with the same denominator, failing to recognise the simpler direct approach. Strong procedural schemas can prevent students from recognising when different methods would be more efficient, such as using long multiplication for 25 × 4 rather than recognising this as equivalent to 100. And when procedures are learnt without understanding why they work, students struggle to adapt approaches to new situations; the method becomes the goal rather than a tool for achieving the goal.

To prevent this rigidity, instruction should include varied problem types requiring different approaches, encourage flexibility by explicitly discussing when different methods apply, develop conceptual understanding alongside procedural fluency, and help students understand when and why particular methods are appropriate rather than promoting rote application. The goal is schemas that are both efficient and adaptable: automated procedures free working memory, whilst conceptual understanding enables flexible application.

Defective schemas and misconceptions

Schema theory explains why misconceptions resist correction (Vosniadou & Brewer, 1992). Misconceptions feel intuitively correct: they usually make logical sense to students based on their limited experience. Adding fractions by adding numerators and denominators follows the pattern for whole numbers; students aren’t being careless, they’re applying a schema that seems reasonable. Misconceptions are not isolated errors but integrated into students’ knowledge networks, so correcting them requires not just adding new information but restructuring existing schemas, a far more challenging cognitive task. Many misconceptions also work sometimes, providing intermittent reinforcement: “multiplication makes bigger” works for whole numbers greater than one, making the misconception seem valid until encountering fractions or decimals. And often students do not realise their thinking is incorrect. The defective schema represents their understanding of how things work, so they do not seek correction or recognise their errors.

Simply telling students they are wrong is ineffective because their schemas feel correct to them. Stronger approaches address the schema directly. Create cognitive conflict by demonstrating situations where the defective schema produces absurd or impossible results (Limón, 2001), letting the contradiction do the work of persuasion: “If adding $\frac{1}{3}+\frac{1}{5}=\frac{2}{8}=\frac{1}{4}$, we have made the number smaller by adding to it, is that possible?” After exposing the deficiency, explicitly teach the correct schema; do not assume students will construct correct understanding from the failure of their misconception. The new correct schema then needs extensive practice with feedback to become established and automatic, since brief correction followed by moving on allows the defective schema to persist and resurface under pressure. When possible, prevent misconceptions in the first place by teaching in ways that avoid creating them: use representations and language that will remain valid across contexts, and anticipate common errors and address them proactively.

Curriculum planning

Before teaching new content, identify the required schemas explicitly during planning (what must students already know?), assess schema development genuinely rather than assuming knowledge, build deficient schemas before proceeding even if it means departing from the planned curriculum, and accept that schema building takes time that cannot be rushed without undermining learning. Running out of curriculum time is preferable to teaching content students cannot learn because prerequisite schemas are missing or defective. The latter wastes time entirely whilst the former simply means prioritising what matters most.

Curriculum sequences should teach components before integration when complexity exceeds working memory capacity (see Part-Whole Approach), return to core concepts repeatedly at increasing levels of sophistication so schemas deepen over time, provide extensive practice at each level before advancing (schemas need consolidation), make connections explicit rather than assuming students will see relationships, and respect the gradual nature of schema development. Expertise cannot be rushed.

Differentiation follows schema development, not simply “easier” or “harder” work. Students with novice schemas need explicit teaching, extensive worked examples, heavy scaffolding, and a focus on accurate schema construction. Students with developing schemas need guided practice with reducing scaffolding, deliberate pattern recognition work, fluency building, and gradual complexity increase. Students with expert schemas need minimal scaffolding, complex problem-solving, novel applications, and metacognitive development. The same student might have expert schemas in one domain and novice schemas in another, requiring completely different instructional approaches.

Everyday teaching decisions

During explanations, make schema organisation explicit: “This concept connects to what we learnt last week about…”, “Notice how this pattern appears in several different contexts…”, “This is a specific example of the general principle we discussed…”. Help students organise knowledge rather than assuming they’ll spontaneously see connections that are obvious only to experts with developed schemas.

During practice, recognise that practice serves schema building. Provide sufficient practice to automate retrieval, not just accurate performance, distributed over time to allow consolidation, in varied contexts to build flexible and transferable schemas, and with feedback to prevent defective schema formation. Students with weaker schemas need substantially more practice than those with stronger schemas, not because they’re slower but because they’re building more from scratch.

During assessment, assess schema quality, not just factual recall: whether students can apply knowledge in novel contexts, whether they recognise deep structure or only surface features, whether they can retrieve knowledge flexibly or only in specific familiar contexts, and whether schemas are robust enough to remain accessible under pressure. Performance on isolated questions may mask schema deficiencies that become apparent only in complex applications.

During error correction, recognise errors as schema evidence. Systematic errors reveal defective schemas requiring reconstruction, random errors suggest incomplete schemas needing consolidation, and context-dependent errors indicate schemas too narrow for transfer. Address the schema, not just the mistake: correcting individual errors without addressing underlying schema deficiencies guarantees the error will recur.

Elaboration theory and sequencing

The order in which content is presented affects how schemas develop. Elaboration theory proposes that instruction should be organised from simple to complex, presenting an overview of content before exploring details (Reigeluth & Stein, 1983). This approach mirrors a zoom lens: first showing the big picture, then zooming in for details, then zooming out again to maintain perspective.

The theory recommends using an epitome (a simple, representative version of the content) as the starting point, then elaborating on details in subsequent lessons whilst maintaining connections to the overall structure (Reigeluth & Stein, 1983). This supports meaningful learning by providing a conceptual framework before introducing complexity. The epitome captures essential relationships and main ideas without overwhelming detail, giving students a mental scaffold on which to attach subsequent learning.

This approach uses subsumptive sequencing (general to specific), where broad concepts are introduced before detailed specifics. When teaching algebra, teachers might first present the general concept of “solving for an unknown” with simple examples before introducing specific techniques for different equation types. This contrasts with linear sequencing that presents content in order of occurrence without first establishing the overall framework. Regular synthesis activities help students maintain connections between details and the overall structure: after introducing new details, teachers explicitly link back to the initial framework, showing how the new information fits into the bigger picture. This prevents students from accumulating disconnected facts without understanding how they relate to the broader topic (Reigeluth & Stein, 1983).

Elaboration theory connects to other sequencing principles that support schema development. Advance organisers introduce general ideas before specific content, providing a mental structure for new learning. Spiral curriculum revisits topics at increasing complexity across years. Learning prerequisites ensures foundational schemas exist before building more complex ones. All these approaches recognise that schemas develop most effectively when students first understand the overall structure before encountering detailed complexity.

Summary

Schemas are networks of organised knowledge in long-term memory that explain how learning occurs within cognitive constraints. Long-term memory has unlimited capacity whilst working memory has severe limitations; schemas address this by chunking knowledge to reduce cognitive load.

Schema theory informs multiple aspects of instruction. Prior knowledge assessment matters because existing schemas determine what can be learnt next. Worked examples are effective for novices because schemas must be built before problem-solving becomes productive. Misconceptions persist because they are embedded in schemas rather than existing as isolated errors. Expertise development requires time because building robust schemas requires extensive practice. Students arrive with different schemas, making uniform instruction differentially effective; systematic schema building, matched to students’ current knowledge state, supports learning across varied starting points.

References

Reigeluth, C. M., & Stein, F. S. (1983). The elaboration theory of instruction. In C. M. Reigeluth (Ed.), Instructional design theories and models: An overview of their current status (pp. 335-381). Lawrence Erlbaum Associates.

Anderson, R. C. (1977). The notion of schemata and the educational enterprise: General discussion of the conference. In R. C. Anderson, R. J. Spiro, & W. E. Montague (Eds.), Schooling and the acquisition of knowledge (pp. 415-431). Lawrence Erlbaum.

Atkinson, R. K., Derry, S. J., Renkl, A., & Wortham, D. (2000). Learning from examples: Instructional principles from the worked examples research. Review of Educational Research, 70(2), 181-214. https://doi.org/10.3102/00346543070002181

Bartlett, F. C. (1932). Remembering: A study in experimental and social psychology. Cambridge University Press.

Chase, W. G., & Simon, H. A. (1973). Perception in chess. Cognitive Psychology, 4(1), 55-81. https://doi.org/10.1016/0010-0285(73)90004-2

Chi, M. T. H. (2008). Three types of conceptual change: Belief revision, mental model transformation, and categorical shift. In S. Vosniadou (Ed.), International handbook of research on conceptual change (pp. 61-82). Routledge.

Chi, M. T. H., Feltovich, P. J., & Glaser, R. (1981). Categorization and representation of physics problems by experts and novices. Cognitive Science, 5(2), 121-152. https://doi.org/10.1207/s15516709cog0502_2

Chi, M. T. H., Glaser, R., & Rees, E. (1982). Expertise in problem solving. In R. J. Sternberg (Ed.), Advances in the psychology of human intelligence (Vol. 1, pp. 7-75). Lawrence Erlbaum.

Cooper, G., & Sweller, J. (1987). Effects of schema acquisition and rule automation on mathematical problem-solving transfer. Journal of Educational Psychology, 79(4), 347-362. https://doi.org/10.1037/0022-0663.79.4.347

Cowan, N. (2001). The magical number 4 in short-term memory: A reconsideration of mental storage capacity. Behavioral and Brain Sciences, 24(1), 87-114. https://doi.org/10.1017/S0140525X01003922

Dochy, F., Segers, M., & Buehl, M. M. (1999). The relation between assessment practices and outcomes of studies: The case of research on prior knowledge. Review of Educational Research, 69(2), 145-186. https://doi.org/10.3102/00346543069002145

Ericsson, K. A., & Kintsch, W. (1995). Long-term working memory. Psychological Review, 102(2), 211-245. https://doi.org/10.1037/0033-295X.102.2.211

Ericsson, K. A., Krampe, R. T., & Tesch-Römer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review, 100(3), 363-406. https://doi.org/10.1037/0033-295X.100.3.363

Kalyuga, S., Ayres, P., Chandler, P., & Sweller, J. (2003). The expertise reversal effect. Educational Psychologist, 38(1), 23-31. https://doi.org/10.1207/S15326985EP3801_4

Limón, M. (2001). On the cognitive conflict as an instructional strategy for conceptual change: A critical appraisal. Learning and Instruction, 11(4-5), 357-380. https://doi.org/10.1016/S0959-4752(00)00037-2

Luchins, A. S. (1942). Mechanization in problem solving: The effect of Einstellung. Psychological Monographs, 54(6), i-95. https://doi.org/10.1037/h0093502

Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63(2), 81-97. https://doi.org/10.1037/h0043158

Stanovich, K. E. (1986). Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy. Reading Research Quarterly, 21(4), 360-407. https://doi.org/10.1598/RRQ.21.4.1

Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257-285. https://doi.org/10.1207/s15516709cog1202_4

Sweller, J., & Cooper, G. A. (1985). The use of worked examples as a substitute for problem solving in learning algebra. Cognition and Instruction, 2(1), 59-89. https://doi.org/10.1207/s1532690xci0201_3

Sweller, J., Mawer, R. F., & Ward, M. R. (1983). Development of expertise in mathematical problem solving. Journal of Experimental Psychology: General, 112(4), 639-661. https://doi.org/10.1037/0096-3445.112.4.639

Sweller, J., van Merriënboer, J. J. G., & Paas, F. (2019). Cognitive architecture and instructional design: 20 years later. Educational Psychology Review, 31(2), 261-292. https://doi.org/10.1007/s10648-019-09465-5

Vosniadou, S., & Brewer, W. F. (1992). Mental models of the earth: A study of conceptual change in childhood. Cognitive Psychology, 24(4), 535-585. https://doi.org/10.1016/0010-0285(92)90018-W