How to teach difficult topics: Practical strategies

Every subject has them. The topics where the class goes quiet at the wrong moment, where the same misconception comes back year after year, where you can feel the lesson sliding sideways by the time you get to the second worked example. Moles in chemistry. Algebraic manipulation in maths. Free will and determinism in psychology. The causes of the First World War if you only have one lesson. The list varies by subject, but every teacher has one.

The difficult topic problem is not really about content. It is about cognitive load, prior knowledge, and the gap between what we think the lesson is doing and what students actually experience. Most of the time the topic is not unteachable. It is just being taught at a pace, or in an order, that makes it harder than it needs to be.

This guide walks through a planning approach that tends to work for the topics students struggle with most. It draws on Rosenshine's Principles of Instruction, EEF guidance on effective teaching, and the cognitive load research that sits behind both. None of it is magic. It is the same set of teaching moves you already use, applied a little more deliberately to the parts of the curriculum where the margin for error is smallest.

Before you plan the lesson, it helps to be honest about why students find this particular topic hard. The reasons cluster into a handful of categories, and the right teaching move depends on which one is in play.

Some topics are difficult because they are abstract. Atomic orbitals, market equilibrium, and the concept of variance all live one or two steps away from anything students can see or touch. Others are difficult because they require a stack of prior knowledge that is shakier than the scheme of work assumes. Hardy-Weinberg sits on top of probability, allele frequencies, and the assumption that students remember what a heterozygote is. If any of those layers wobble, the new content will not land.

A third category is procedurally heavy. Long division, balancing redox equations, and structuring an extended essay all require students to hold several steps in mind at once. The content is not deeply abstract, but the working memory cost is high. And a fourth category is counter-intuitive. Inertia. Natural selection without purpose. The idea that correlation is not causation. These topics require students to override an intuitive answer that feels right, and that takes more than a single explanation.

If a topic is genuinely difficult, the worst thing you can do is start teaching it before you know where the class is. Five minutes of diagnostic questioning at the start of the lesson tends to save fifteen minutes of re-teaching in the middle.

The most efficient tool for this is the hinge question. A hinge question is a multiple-choice question where each wrong answer maps to a specific misconception. If you ask a Year 10 class "which of these has the most particles" and offer a mole of helium, a mole of oxygen, a mole of water, and a mole of iron, the pattern of answers tells you exactly what students think a mole is. The students who pick iron think mass and amount are the same thing. The students who pick water are thinking about something else again. You now know who needs what.

Mini-whiteboards do the same job in a slightly different form. Pose a question, give the class 30 seconds, ask everyone to hold up their answer. Scan the room. Decide whether to push on or pause. This kind of diagnostic move feels slow the first few times. It pays for itself within a couple of lessons because you stop teaching ahead of where the class actually is.

Cognitive load theory, developed by John Sweller, makes a fairly blunt prediction: If you try to teach too many new things at once, students will not learn any of them. Working memory is finite. When it overflows, the lesson stops being a learning experience and becomes an exercise in copying down whatever is on the board.

The planning move is to chunk the content into smaller steps than feels necessary. If you can teach a topic in four ten-minute segments rather than one forty-minute block, the cognitive load on any single moment drops sharply. Each chunk should introduce one new idea, give students a chance to practise it in isolation, and check that it landed before adding the next layer.

For most difficult topics, the chunking that works tends to follow a similar pattern. Start with the simplest possible version of the idea, stripped of any complicating factors. Add one variation. Add another. Combine two ideas you have already taught separately. Then, and only then, introduce the full version. Students who learn the components first can hold the whole thing in mind. Students who meet the full version first often cannot.

There is a long history of research, going back to Sweller and Cooper in the 1980s, showing that students learning a new procedure benefit more from studying worked examples than from solving problems from scratch. The reason is straightforward. Solving a brand-new problem requires the student to figure out the method, hold the steps in mind, and execute them, all at the same time. A worked example removes the first two demands and lets the student focus on understanding the procedure.

In practice this means doing more modelling than feels comfortable, especially for procedural topics. Two or three worked examples on the board, with you talking through every decision (including the wrong turns), tends to land better than one example followed by independent practice. Vary the surface features deliberately. Same skill, different numbers. Same skill, different context. Students need to see the procedure applied across enough variations to recognise the pattern.

A useful refinement is the partially completed worked example. Show students the first three steps in full, then leave the next step blank for them to fill in. Then show the fourth step and leave the fifth blank. This bridges the gap between watching you work and working independently. It is sometimes called fading, and it is one of the most reliable teaching moves for procedural content.

For topics that are difficult because they are abstract or counter-intuitive, the issue is rarely the procedure. It is the underlying way of thinking. Natural selection without purpose. The idea that a force is not the same as motion. The argument that a metaphor can carry meaning the literal sentence cannot. In these cases, modelling is not about steps. It is about thought.

This is where think-alouds earn their keep. Instead of just stating the conclusion, narrate the reasoning that gets you there. "My first instinct here would be to say X. But that does not work, because if X were true, we would also expect Y, and we do not see Y. So I am going to try thinking about it differently." Students who watch you reason aloud start to internalise the moves. They begin to do them silently in their own heads.

The second move that works for abstract topics is the dual coding pair: A verbal explanation paired with a visual representation. Diagrams, flowcharts, axes, simple sketches. The visual gives students a place to hang the abstract idea. For topics like genetic drift, electric fields, or the structure of an argument in an essay, a single clear diagram can do more than three paragraphs of explanation.

Almost every difficult-topic lesson has a wobble. There is a moment, usually about 20 to 25 minutes in, where the class hits the first independent task and a meaningful proportion of students cannot do it. This is not a failure of the lesson. It is the point of the lesson. If everyone could do it without effort, you would not be teaching it.

The planning move is to anticipate the wobble and decide in advance what you will do when it happens. Have a fallback explanation ready. Have a partially completed worked example you can pull out. Have a paired task that lets students work with someone who got it. The teachers who handle difficult topics best are not the ones whose lessons run smoothly. They are the ones who plan for the friction and know what to do when it arrives.

A quick decision tree helps. If most of the class is struggling, stop the independent work, pull them back to the board, and do another worked example. If only some students are struggling, work with that group while the rest continue. If nobody is struggling, congratulations: The task is too easy, and you can move to the next chunk. The skill is reading the room quickly and acting on what you see, rather than ploughing through the lesson plan you wrote on Sunday evening.

Difficult topics often have a fade-out problem. Students get to a reasonable understanding in the lesson, but the content does not survive the gap to the next lesson, let alone the gap to the exam. The fix is retrieval, and it has to be built into the planning from the start rather than bolted on at the end of the unit.

A simple routine is the spaced low-stakes quiz. Five quick questions at the start of each lesson covering content from the previous week, the previous fortnight, and the previous half-term. The Education Endowment Foundation guidance on metacognition and the broader cognitive science literature both point to the same conclusion: Regular retrieval practice strengthens long-term memory more reliably than re-reading or re-teaching. For difficult topics specifically, retrieval is the difference between a class that understood it in March and a class that can still do it in May.

The practical version of this in most science and maths departments is a shared bank of retrieval questions that map to the specification. If your department does not have one, building one collaboratively over a half-term is usually a high-return use of meeting time. Cognito's exam-style questions and instant-feedback quizzes are one option some departments use as a ready-made starter bank for GCSE and A-level specs, which can save a chunk of weekly prep time.

Pulling the threads together, here is a short checklist for planning a lesson on a topic the class typically struggles with. It is laid out as a sequence, but in practice most teachers move between these prompts in a few minutes once they have done it a handful of times.