Exam technique · 25 September 2026 · 6 min read
Physics command words: how to describe, explain, compare and evaluate
A student can know the topic and still answer a different question from the one asked. Describing that a liquid cools does not explain why its temperature decreases. Listing limitations does not necessarily evaluate whether an experiment supports its conclusion.
Command words tell you what to do with your physics knowledge. The examples below are original teaching examples, not official model answers or promises of particular marks. Compare their structure with the wording and mark scheme of your own qualification, including the additional instructions around the command word.
Read the whole task before choosing a response
Underline the requested action, identify the object of that action, and note any evidence you must use. ‘Describe the graph’ and ‘describe how to obtain the graph experimentally’ share a command word but require very different content. One asks for observed features; the other asks for a reproducible method. A learned definition of ‘describe’ cannot replace reading the rest of the sentence.
Use the mark allocation as a clue to the expected development, not a rule that every mark equals one sentence. A calculation may reward several stages; a written response may require linked reasoning. Do not fill space because a question has many marks. Plan the necessary physical ideas and relationships, then express them clearly enough for the reader to follow.
Describe what changes, with useful evidence
Imagine a cooling graph that falls steeply at first and then becomes less steep as the liquid approaches room temperature. A description could state that the temperature decreases, the rate of decrease becomes smaller, and the curve approaches the surrounding temperature. If the task supplies values and asks you to use them, quote appropriate coordinates with units to support the pattern.
Avoid replacing a description with unexplained jargon. ‘It is exponential’ may be less useful than a clear account of how the gradient changes, especially if that mathematical model has not been established. Equally, ‘it goes down’ leaves the quantity unnamed. Refer to temperature, speed, current or the actual variable shown, and distinguish a decreasing value from a decreasing rate of change.
Explain by linking the cause to the outcome
For the same cooling example, an explanation needs a mechanism. Energy is transferred from the warmer liquid to cooler surroundings; the liquid’s internal energy decreases and its temperature falls, assuming no change of state. As the temperature difference becomes smaller, the net rate of energy transfer usually decreases under otherwise comparable conditions. The falling temperature alone was the observation; these links explain it.
Try a second example: a fixed amount of ideal gas is heated in a rigid sealed container. Its mean molecular kinetic energy increases. The changed molecular motion produces a greater rate of momentum transfer to the walls, so the pressure increases. Mentioning ‘more collisions’ without the molecular and force links may leave the reasoning incomplete. Choose the model and detail appropriate to your course.
Compare like with like and support your judgement
A useful comparison uses the same criterion for both items. If comparing two energy-transfer methods, discuss their efficiencies under stated conditions, their practical limitations or another requested property on both sides. Do not write an unrelated fact about method A and a different unrelated fact about method B and assume that two facts make a comparison.
For numerical comparisons, a difference, ratio or percentage can make the relationship precise when relevant. If one measurement is 4.0 units and another is 6.0 units, the second is 1.5 times the first. Whether that matters depends on the context and uncertainty. Avoid calling a difference ‘significant’ in a statistical sense unless the data and method support that conclusion.
Evaluate the claim, not just the apparatus
Suppose a student concludes that doubling a wire’s length doubles its resistance. An evaluation might consider whether the graph supports direct proportion, whether the range is broad enough, whether repeated readings agree, and whether temperature and wire diameter were controlled. A final judgement should reflect that evidence: the data may support the model over the tested range while leaving particular limitations unresolved.
A list such as ‘reaction time, parallax, heat loss’ is not automatically an evaluation. Some items may be irrelevant to the experiment. Connect each important limitation to its likely effect on the evidence, then explain whether it weakens the conclusion and how it could be addressed. You do not need to invent equal numbers of advantages and disadvantages when the task does not require that structure.
Practise changing the task while keeping the physics
Choose one familiar diagram or dataset and write three short responses: describe its features, explain the mechanism and evaluate a conclusion drawn from it. This isolates the command-word skill from learning a new topic. Then check whether each response actually fulfils its task, using your teacher’s feedback or a relevant official mark scheme for comparable questions.
Before submitting an answer, check that every sentence contributes. Replace vague nouns with the named physical quantities, add a missing causal link, and remove unsupported claims. Precise meaning matters more than inserting a supposed magic phrase. Keep a record when the real problem was misunderstanding the command word so that your revision targets response technique instead of unnecessarily relearning the entire chapter.
Questions, explained
Choose a question for a direct answer, then explore the explanation and supporting resources. Each answer has its own link to save or share.
What is the difference between describe and explain in physics?
Describe sets out the relevant features, observations or pattern. Explain gives the physical reasons and connects them to the outcome. For cooling, ‘the temperature falls more slowly later’ describes the graph; relating the smaller temperature difference to a lower net rate of energy transfer explains that behaviour. Always check the full question, because it may ask for a method rather than a graph description.
How do I evaluate a physics experiment?
Judge how well the evidence supports the investigation’s conclusion. Consider relevant features such as the range, repeated measurements, uncertainty and control of other influences. Explain why important limitations matter and suggest realistic changes where requested. Finish with a judgement that matches the evidence; a memorised list of possible errors does not show whether this particular conclusion is justified.
Should I write one sentence for every mark in a physics question?
There is no universal sentence-per-mark rule. Mark allocation indicates the expected substance, but some marks depend on linked reasoning, calculations or interpretation. Plan the necessary ideas and show their relationships without repeating yourself. Use official mark schemes and teacher feedback to learn how your board assesses the task, rather than stretching a response to an arbitrary sentence count.