Updated on
August 15, 2026
Teacher Cognitive Load: Planning Lessons Without Burnout
Cognitive load theory applied to the teacher, not the learner: the overplanning trap, reducing extraneous load, and designing lessons you can sustain.

Updated on
August 15, 2026
Cognitive load theory applied to the teacher, not the learner: the overplanning trap, reducing extraneous load, and designing lessons you can sustain.
Teachers face many demands daily. Lesson prep and marking take up lots of time. Behaviour and meetings add to pressure on memory. The DfE Working Lives of Teachers and Leaders wave 4 survey keeps workload, wellbeing and career intentions in view as major issues for the profession. The issue is not lack of effort, but unmanageable cognitive load.
Sweller's Cognitive Load Theory explains how learners process information (Sweller, various dates). Teachers also experience cognitive overload, hurting decision-making. Overloaded teachers may use basic strategies and mark less carefully. They might favour lecturing over activities that build deeper thinking. Ironically, workload reduces teaching effectiveness, hindering learner progress.
Cognitive Load Theory informs strategies for teachers' working memory. We can reduce extraneous load and manage intrinsic load. Building automaticity lets you focus on understanding learners (Sweller, 1988; Chandler & Sweller, 1991). This benefits your teaching practice.
Sweller's Cognitive Load Theory (1988) explains working memory. Learners can consciously hold only a small amount of novel information in working memory at one time (Miller, 1956; Cowan, 2001). Learning slows if information overloads this limited capacity.
Sweller identified three types of cognitive load:
CLT-informed design aims to reduce extraneous load and manage intrinsic load so learners can devote working-memory resources to schema construction (Sweller, 1994; Sweller, van Merrienboer and Paas, 1998; Sweller, 2010). Teachers should cut wasteful tasks. This frees mental space to plan better lessons and meet learner needs.
The novel insight is this: teachers are learners too, and CLT applies to teacher working memory as directly as it applies to learner working memory (Feldon, 2007).
During a typical school day, a teacher must simultaneously hold:
The extraneous load items—deadlines, admin, interruptions—don't improve teaching quality. They're pure waste. Yet they consume precious working memory, leaving less capacity for the genuine instructional decisions that matter: How will I explain this concept? Does this learner have a misconception? Should I slow down or accelerate?
Research on expert teacher practice confirms this. Berliner (2001) found that expert teachers show reduced cognitive load during classroom teaching. They don't think harder; they think more efficiently. Routine classroom management is automatic, freeing working memory for responsive teaching. This automaticity took years to develop, but it's the hallmark of mastery.
Many teachers believe that longer, more detailed lesson plans improve teaching quality. The opposite is often true.
An overplanned lesson—with detailed scripts, intricate timing, multiple resources for every contingency—sounds thorough. But it creates extraneous cognitive load both during planning and during teaching. You spend 90 minutes planning a 50-minute lesson, writing scripts you won't use. During the lesson, you're constantly checking your detailed plan, distracted from what learners are actually doing. Your working memory is consumed by "following the plan" rather than responding to learners.
Evidence suggests retrieval practice is a strong routine for strengthening memory. Repeating a small number of lesson structures can also reduce planning decisions: decide what prior knowledge to retrieve, the one key idea to teach, the practice problems learners will attempt and the exit question.
Keep it brief: planning talks should be 15 minutes, not 90. Learners know the lesson routine well, saving your working memory. You can then focus on pedagogy (Sweller, 1988).
Classroom example: A maths teacher planning a Year 5 lesson on fractions. Overplanned version: 12 detailed slides, three different fraction manipulatives, differentiated worksheets for four ability groups, detailed time allocations. Retrieval-based version: Show me yesterday's exit ticket questions (what misconceptions do I see?). Today's retrieval task: three problems on equivalent fractions. Mini-lesson: one key idea and one visual model. Guided practice: five problems together. Independent: four problems. Exit ticket: which representation helps you most? The retrieval-based lesson takes 15 minutes to plan and runs more smoothly because the format is automatic.
The first step is eliminating wasted cognitive effort. Three strategies work:
Create a single lesson template and use it for every lesson in your subject. The template should specify structure but not content. For example:
Planning effort drops because you're not inventing a new structure; you're filling in one variable: the specific content for that lesson. A 50-minute lesson that took 90 minutes to design now takes 15 minutes.
These activities offer formative assessment. Walk around; watch learners. Note any misunderstandings. You now have immediate evidence, not just marked books. This live feedback on misconceptions is more useful (Karpicke & Blunt, 2011; Roediger & Butler, 2011).
Store this live observation data in a simple tracking sheet (one row per learner, columns per key concept). This replaces the marking workload: you see which learners need same-day intervention and which are ready to progress. No marking books required.
When you plan, use dual coding: write in words AND sketch simple visuals for your key idea. For example, a maths teacher teaching area of rectangles might write:
Key idea: Area = length × width. A 5×3 rectangle contains 15 unit squares. [Simple sketch: rectangle, 5 units across, 3 units down, 15 squares inside.]
When you teach, the visual is right there. You're not generating visuals on the spot (which consumes working memory). The visual is already encoded and ready to use. Research on dual coding shows that combining verbal and visual explanations improves learning efficiency without increasing planning time (Paivio, 1986).
Not all cognitive load is extraneous waste. Some is intrinsic: the genuine difficulty of the task. Teaching a new class, a new subject, or learners with complex needs is intrinsically hard. This is different from extraneous load and requires a different response.
Intrinsic load can be managed in two ways:
Your intrinsic load is highest because every decision feels new. The solution is not to reduce planning time (you need it to learn) but to reduce extraneous load ruthlessly.
The goal is to protect enough mental capacity to actually think about pedagogy. In your first year, most of your cognitive effort goes to classroom management and behaviour. That's normal and intrinsic to the role. But if you're also spending 2 hours per night on extraneous admin, you'll burn out before you develop automaticity.
You might be teaching a year group or subject new to you, or working with learners with complex SEND needs. Here, intrinsic load is high because you're learning alongside teaching.
Reduce intrinsic load by:
A secondary English teacher new to KS3 might feel this acutely. You're familiar with GCSE texts but not with teaching Shakespeare to Year 7s who've never read a play. The content knowledge you have doesn't directly map to this age group; their reading level, emotional maturity, and prior experience are different. Rather than designing elaborate lessons on metaphor while simultaneously learning Year 7 pedagogy, use existing schemes of work (such as those published by exam boards). Your cognitive load drops from "design a Shakespeare unit" to "adapt this unit for my learners' pace." You're still thinking pedagogically, but from a clearer starting point.
Teachers find high intrinsic load when autistic learners experience transition anxiety. Complex needs require support, and designing timetables and resources increases cognitive load. Use visual timetable templates, standardised worksheet formats and simple agreed routines to reduce avoidable design decisions. This frees working memory for instructional decisions: Does the learner understand? Are they anxious? What pacing works?
Here's a hard truth: written marking, as taught in teacher training, is not based on evidence. It's based on tradition.
The EEF 2016 review A marked improvement? found that written marking is workload-intensive and that the evidence base for its impact is limited; it did not establish a 0.17 SD effect or compare it with sitting next to a window.17 standard deviations—about the same as having a learner sit next to a window (EEF, 2016). Teachers spend three hours per week on something nearly inert. See also our guide on social-emotional wellbeing.
Why is marking so ineffective? Several reasons. First, feedback delivered days after work is completed is far less impactful than feedback delivered immediately. A learner makes an error on Monday's work and receives written feedback on Wednesday; they've already moved on to Tuesday's and Wednesday's content. The feedback addresses outdated misconceptions. Second, written comments are often generic ("Good work, but check your spellings") and don't address the specific thinking error. Third, learners frequently don't read written feedback at all—or they read it without processing it (a phenomenon called "feedback avoidance"). Finally, even when feedback is read and understood, individual comments don't improve classroom learning for other learners; only the recipient benefits.
By contrast, whole-class feedback (which takes 15 minutes) and retrieval-based assessment (which happens live in lessons) produce larger learning gains with a fraction of the time cost. When you identify that 70% of the class confused area and perimeter and spend 10 minutes re-teaching that distinction the next lesson, every learner benefits. When you observe a misconception live during independent practice and correct it immediately, the learner doesn't embed the error.
So what should you do instead?
During independent retrieval practice (those four problems learners solve alone), circulate. Observe. Note which learners have misconceptions. Which learners are rapid. Write three initials on a tracking sheet: JAC (Jake, Anna, Charlie) struggled with that; RSM got it immediately. That's your feedback data. No books marked.
Next lesson, your intervention is targeted. You re-teach the misconception to JAC while RSM extends. You've used evidence to respond. This takes 5 minutes of observation, not 3 hours of marking.
This saves significant teacher time (Wiliam, 2011). Whole-class feedback identifies common errors for all learners. Address these trends in your next lesson to improve understanding (Sadler, 1989; Shute, 2008). Hattie and Timperley (2007) argue that feedback can strongly influence learning when it clarifies where learners are going, how they are going, and where to go next; effects are not guaranteed for all learners.
This approach produces better outcomes: every learner hears feedback on the most common misconception, and remediation is immediate, not delayed (Wiliam, 2011). It also saves 2 hours per 30 books marked. For related guidance, see our article on AI for reducing teacher workload.
Research suggests expert teachers decide rapidly and accurately (Berliner, 1994). These quick decisions surpass the slower, more considered judgements of new teachers (Ericsson, 2006). How do they achieve this expertise?
Experts make routine classroom decisions automatic (Berliner, 2001). This frees their working memory for complex tasks. A teacher doesn't consciously plan transitions. Instead, routines are automatic. Their working memory notices learners' specific needs. For example, teachers see phoneme confusion or fluency issues.
The contrast with novice teachers is stark. A trainee delivering the same phonics lesson is holding too much consciously: the phoneme I'm teaching, the flashcards I prepared, the transition routine, which learners to check on, the time on the clock. Working memory is maxed out. When a learner becomes challenging, the novice doesn't have spare capacity to diagnose why (Is she frustrated? Is she seeking attention? Is she tired?). The response is reactive and exhausted.
Learners build automaticity through focused, repeated practice. Routines need explicit teaching and repeated practice for best results. Ericsson, Krampe and Tesch-Romer (1993) argued that expert performance is associated with extended deliberate practice, often discussed as roughly 10 years or 10,000 hours, but it is not a simple threshold for all domains. Teachers who keep lesson formats consistent build automaticity faster. Changing formats too often hinders skill development.
Three examples of automaticity-building:
Teach a routine—hand up for attention, line up at the door, pack away—and practise it daily for the first six weeks of the year. Seems tedious. But once that routine is automatic, you spend zero working memory on it. You're not thinking "I need to settle the class" anymore; settling happens. Your working memory is free for teaching.
Use the same lesson structure each time: do-now, mini-lesson, guided practice, independent work, exit ticket. After 30 lessons, it becomes routine. Learners know the schedule, freeing working memory for teaching (Sweller, 1988; Paas et al., 2003).
If you use the same retrieval problems for exit tickets (three per week, one per concept, showing misconceptions and ready-to-progress status), you see patterns. After 20 weeks, you can read an exit ticket and instantly know: "They've confused area and perimeter" or "They need the next concept." That decision, which took deliberation in week 1, is now automatic.
Repeated effort helps learners process routines automatically (Logan, 1988). Practice lowers cognitive load. With expertise and deliberate practice, some task components can become more efficient and automatic over time. Experience helps experts stay calm, unlike overwhelmed beginners (Berliner, 2001).
Here are ten concrete strategies you can implement this week:
Individual teachers can implement these strategies. But schools can amplify impact through systemic change.
Teachers gain time and energy when schools lessen their workload. This improved capacity allows them to focus on lesson quality. Lower workload means more time for each learner (Kirschner, Sweller & Clark, 2006). Staff feel more valued, decreasing turnover (Hattie, 2009).
Instead of allowing every teacher to design their own lesson structure, schools can choose one evidence-based format (retrieval-based, direct instruction, problem-based learning—the choice matters less than the consistency). All teachers in a phase or subject use the same format. Planning time drops, consistency across learners' experience improves, and new staff onboard faster because they're learning one format, not five.
Example: A primary school adopts a consistent maths lesson format: do-now (retrieval), mini-lesson, guided practice, independent practice, exit ticket. Every maths teacher, every day, uses this format. In September, training is on the format. In October, all teachers are designing maths lessons in half the time compared to before. By Christmas, the format is automatic for teachers and learners alike.
Schools can buy or build detailed schemes of work, including concepts and assessments. These also cover misconceptions and teaching notes. Teachers adapt these for learners, instead of starting fresh. Curriculum experts (NCETM) show good schemes cut workload and keep quality high.
Schools can explicitly abolish written marking quotas ("30 books per week per class") and replace them with observation protocols: "Circulate during independent work, note three common misconceptions, deliver whole-class feedback in the next lesson." This is faster, more responsive and reduces cognitive load when leaders protect it from becoming another paperwork exercise.
Many schools erode PPA time by tacking on meetings or duties. Schools committed to reducing teacher cognitive load protect PPA time ruthlessly: two hours per week with no interruption, no ad-hoc duties, no emails. During this time, planning happens. This sounds obvious; in practice, it's revolutionary.
Smaller classes can reduce the number of incidents, interactions and learning needs a teacher has to manage at once (Blatchford, 2003). Where schools cannot reduce class size, teaching assistants or specialist staff can support identified learners. The key claim is practical rather than a specific Kirschner finding: additional adult support can share monitoring, prompting and access work when roles are well defined.
This article rests on Cognitive Load Theory, which has a nuanced evidence base worth understanding.
Sweller (various dates) used lab tasks like algebra to create CLT. Classrooms have interruptions, social factors and varied learners. Kirschner et al. (2006) argue CLT simplifies classroom learning.
CLT began as a learner cognition theory, not for teachers. Using it for teachers makes sense, but lacks evidence. We suggest template reuse and retrieval practice, based on Berliner (2001) and DfE (2019). This article merges findings from several fields. Be aware of the distinction between CLT and its application to teachers.
System-level changes such as smaller classes and more planning time are key. Cognitive tricks cannot fix classes of 35 learners and little marking time. This toolkit works best where school conditions are good, but still matters everywhere.
Expert teachers offer valuable lessons for practice. Berliner (2001) explored how we learn about, and learn from, these experts. The research appeared in the International Journal of Educational Research. The article, volume 35(5), pages 463-482, gives further detail.
Department for Education (2019) published workload challenge findings. Teachers' feedback was analysed (DfE, 2019). You can read it at www.gov.uk/government/publications/workload-challenge.
Education Endowment Foundation. (2016). A marked improvement? A review of the evidence on written marking. EEF Guidance Report. London: EEF.
Feldon, D. F. (2007). Cognitive load and classroom teaching: The double-edged sword of asking productive questions. Journal of Educational Psychology, 99(3), 464–475.
Kirschner, Sweller, and Clark (2006) showed minimal guidance doesn't work. They analysed constructivist and problem-based teaching failures. This research appeared in Educational Psychologist, 41(2), pages 75–86.
Paivio, A. (1986). Mental representations: A dual coding approach. New York: Oxford University Press.
Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257–285.
Cognitive Load Theory (Sweller, 1994) links learning difficulty to instructional design. This theory affects how learners process new information. Good design reduces mental strain.
Wiliam, D. (2011). Embedded formative assessment. Bloomington, IN: Solution Tree Press.