Did You Know Claim–Evidence–Reasoning Is More Than a Science Writing Frame? | How Observations Become Explanations

Many students first meet Claim–Evidence–Reasoning as a writing frame: write a claim, add evidence, then explain the reasoning. It can be useful in exactly that form. But if CER becomes only a three-box worksheet, students may learn to fill the boxes without learning how scientific explanations actually work.

The deeper idea is more interesting. Science begins with something we want to explain. We observe, measure or compare. We make a claim about what is happening. We decide which observations genuinely count as evidence. Then we connect that evidence to the claim through concepts, mechanisms and models. We ask whether another explanation could fit the same evidence. We revise when better evidence appears.

Seen this way, Claim–Evidence–Reasoning is not a sentence template. It is a miniature architecture for disciplined thinking.

Why students can have the right evidence and still give a weak answer

Consider a simple question: Why did one wet cloth dry faster than another? A student may correctly report that the cloth placed near moving air dried sooner. That is evidence. But “because it was near the fan” is not yet a full scientific explanation. The answer needs a mechanism: moving air removes water vapour near the cloth, maintaining a larger concentration difference and allowing evaporation to continue more quickly.

This gap between observation and explanation appears everywhere. A plant grew taller in brighter light. A circuit bulb became dimmer when another component was added. A metal spoon felt colder than a wooden spoon in the same room. A graph rose after an intervention. Students can describe what happened while remaining unsure why it happened.

CER is valuable because it makes that missing bridge visible. The bridge is reasoning.

Claim: an answer that can be tested, not a guess dressed as certainty

A claim is the answer the learner is prepared to defend. It should be specific enough to be tested against evidence. “Plants need good conditions” is too vague. “Under the conditions of our experiment, seedlings receiving moderate daily light grew taller than seedlings kept in darkness” is testable and bounded.

Claims can describe relationships, identify causes, compare models, classify observations or predict outcomes. The exact form depends on the scientific question. Younger students may make simple comparative claims. Older students may make conditional claims that acknowledge interacting variables and uncertainty.

A strong claim also avoids pretending the evidence is broader than it is. If a class tested two temperatures, the conclusion should not become “all higher temperatures always increase the rate.” If a sample contained ten plants, the claim should not imply certainty about every plant species. Scientific language becomes stronger when it is appropriately limited.

Evidence: not every fact belongs in the answer

Students often believe that more facts mean more evidence. But evidence is not a pile of information. It is information that bears on the claim. Relevance matters.

Suppose students are deciding which material is the best thermal insulator. They measure water temperature after ten minutes in containers wrapped with different materials. The temperature change is directly relevant. The colour of the wrapping is probably not. The fact that one material was easier to cut may matter for a design project but not for the thermal claim.

Good evidence also has quality. Was the measurement repeated? Was the comparison fair? Was there an obvious outlier? Was the instrument precise enough? Were the groups equivalent at the start? Students need to learn that evidence strength depends on how it was produced, not just on whether a number exists.

Reasoning: the part that explains why the evidence matters

Reasoning connects evidence to a claim using scientific ideas. It answers, “Why does this evidence make the claim more believable?”

For the insulation example, students might reason that a smaller temperature decrease means less thermal energy was transferred from the warm water to the surroundings during the same period. If the test conditions were comparable, the material around that container reduced heat transfer more effectively. The scientific concept—energy transfer—turns the number into evidence for insulation.

This is why CER cannot be taught only as writing. The quality of reasoning depends on subject knowledge. A student who does not understand particles, forces, energy, ecosystems or circuits cannot produce a strong explanation simply by knowing the sentence starter “This evidence supports my claim because…”

The hidden fourth move: mechanism

In many scientific explanations, reasoning becomes stronger when students identify a mechanism: the process through which one condition produces another outcome.

If salt dissolves faster in warmer water, “because the water is warmer” simply repeats the observation. A mechanistic explanation refers to particle motion, collisions and interactions. If a population decreases after a food source declines, the mechanism may involve survival and reproduction. If current changes in a circuit, the mechanism may involve potential difference and resistance.

Mechanisms are powerful because they help students transfer. A learner who memorises “fan makes cloth dry faster” owns a fact. A learner who understands evaporation and vapour removal can reason about wind, humidity, surface area and drying conditions they have never seen before.

From CER to CERM: Claim, Evidence, Reasoning, Mechanism

Teachers do not need to rename every worksheet, but it can be useful to ask a separate mechanism question: “What process makes this relationship happen?” This prevents reasoning from collapsing into repetition.

A simple CERM response might look like this:

• Claim: Increasing exposed surface area made the water evaporate faster.
• Evidence: The same volume of water lost more mass in the wide dish than in the narrow container over the same time.
• Reasoning: Greater mass loss indicates more water left the liquid state.
• Mechanism: A larger surface exposes more liquid particles to the air, increasing opportunities for sufficiently energetic particles to escape.

The mechanism does the explanatory work. It tells us why the pattern should occur.

The hidden fifth move: alternatives

Scientific reasoning is stronger when students ask whether the evidence could support another explanation. This does not require advanced philosophy of science. Even Primary and Secondary students can learn to ask, “What else might have caused this?”

A plant near a window grew faster. Was it the light? Perhaps that location was warmer. Perhaps watering differed. Perhaps the starting plants were unequal. A good experiment controls alternatives; a good explanation acknowledges the alternatives that remain.

This habit protects students from overclaiming. It also turns experimental design into reasoning rather than procedure. Fair tests matter because they help rule out rival explanations. Repeated measurements matter because random variation can mimic a pattern. Controls matter because they establish a comparison.

The hidden sixth move: uncertainty

School answers often sound more certain than real science. Students write “This proves…” when the evidence actually suggests, supports or is consistent with a claim. Teaching uncertainty does not weaken science. It teaches proportion.

Useful phrases depend on the level: “Our results suggest…,” “The pattern supports…,” “Within the tested range…,” “One limitation is…,” “The evidence is consistent with…,” “We cannot tell from this experiment whether…” These phrases help students match the strength of language to the strength of evidence.

The goal is not to make every sentence hesitant. If evidence is strong, students can state a clear conclusion. The goal is to avoid false precision and false certainty.

A seven-question CER checkpoint

• What exactly is my claim?
• Which evidence is directly relevant?
• How reliable is that evidence?
• Which scientific concept connects the evidence to the claim?
• What mechanism could produce the pattern?
• What alternative explanation remains possible?
• How certain should my conclusion sound?

These seven questions turn CER from a paragraph frame into a reasoning routine. A student can use them before writing, during a practical, in oral discussion or when evaluating somebody else’s claim.

Why “more evidence” can make an answer worse

Students sometimes add every data point they can remember. The result becomes long but unfocused. Good scientific communication selects evidence strategically.

If a table contains ten measurements, the student may not need to copy all ten into the answer. A comparison of means, ranges or representative values may show the pattern more clearly. If a graph already displays the data, the explanation can identify the key relationship rather than narrate every point.

Selection is a thinking skill. The student asks which observation most strongly discriminates between competing claims. In advanced work, one decisive piece of evidence can matter more than many weakly related facts.

Worked example 1: which paper bridge design is stronger?

Students build paper bridges with the same paper length and test how much mass each supports. One bridge uses a flat sheet; another folds the sheet into repeated ridges.

Claim: The folded design supports a greater load than the flat design.

Evidence: Across three trials, the folded bridge supported substantially more mass before collapsing, while the span length and paper type were kept the same.

Reasoning: The load capacity comparison is relevant because both structures were tested under similar conditions.

Mechanism: Folding changes the cross-sectional geometry and can increase stiffness by distributing forces through the structure rather than allowing the sheet to bend easily.

Alternative: If folds were inconsistent between trials, craftsmanship could partly explain the difference. Repeating with a folding template would strengthen the evidence.

Notice how the final line improves the explanation. The student is not undermining the result. The student is showing an understanding of evidence quality.

Worked example 2: does exercise affect pulse rate?

A class measures pulse rate before and immediately after one minute of stepping exercise.

A shallow answer says, “Exercise increases pulse rate because the pulse was higher after exercise.” That repeats the pattern.

A stronger answer explains that working muscles require increased delivery of oxygen and nutrients and faster removal of carbon dioxide and other products. The cardiovascular system responds by increasing heart rate and blood flow. The measurements support that mechanism because pulse rate rose after the activity.

Students can then consider alternatives: measurement error, differences in effort, fitness levels or delay between stopping exercise and counting pulse. These factors help explain variation between students without overturning the overall physiological model.

Worked example 3: why did the bulb become dimmer?

In a circuit investigation, students observe a change in bulb brightness after circuit configuration changes. A good CER response must be careful because brightness is an indirect observation. Students should connect it to current, potential difference and resistance at the level appropriate to the syllabus.

The evidence may include ammeter readings rather than brightness alone. The reasoning then explains why the changed circuit alters current or the potential difference across components. This illustrates a useful progression: as students mature, their evidence becomes more quantitative and their models more explicit.

Primary Science: keep the reasoning concrete but real

Young learners do not need abstract jargon to reason scientifically. They can make a claim, point to an observation and explain the connection using taught concepts.

For example: “The ice cube on the metal tray melted faster. After five minutes it was smaller than the ice cube on the plastic tray. Metal transfers heat more readily, so thermal energy reached the ice faster.” The answer is short, but the structure is present.

Teachers can use spoken CER before written CER. Ask three oral questions: “What do you think happened? What did you see that makes you think that? Why would that observation support your idea?” Students learn the reasoning pattern through conversation before they are asked to produce formal paragraphs.

Secondary Science: increase the demand on evidence quality

Secondary students can move beyond one observation. They can compare repeated measurements, use graphs, identify anomalies, distinguish correlation from cause and discuss control variables.

The teacher can require students to rank evidence. Which result is strongest? Which is weakest? Would excluding an anomalous trial change the conclusion? Is the sample large enough? Does the graph show a linear relationship or only a general trend? These questions make CER analytical rather than formulaic.

Students can also compare two competing claims. Instead of asking only “Does temperature affect reaction rate?” present two explanations and ask which is better supported. Scientific argument becomes visible.

JC and advanced learners: use CER to test models

At higher levels, scientific explanations become model-dependent. Students may need to decide which theoretical model best explains data, identify assumptions, quantify uncertainty or discuss the limits of a method.

CER can still help, but the reasoning should not be reduced to a paragraph acronym. A claim may compare models. Evidence may involve transformed data or statistical patterns. Reasoning may require equations. Alternatives may arise from competing mechanisms. Uncertainty may be expressed numerically.

The advantage of teaching the core structure early is that students recognise continuity: science keeps asking what we claim, what evidence supports it and why that evidence should count.

CER and practical work: stop treating the lab report as a form

Traditional practical reports can become procedural: aim, apparatus, method, results, conclusion. Students may complete every heading without deeply interpreting the experiment.

Adding CER questions after the results changes the emphasis. What claim can the data support? Which result is most important? What concept explains the relationship? What is one limitation? What further test would discriminate between two explanations?

Students can also make predictions before collecting data. Afterward, compare prediction and evidence. If the prediction was wrong, that is not a failure. It is a model update. Science education improves when being wrong becomes information rather than embarrassment.

CER and graphs: the graph is evidence, not the explanation

Students often describe a graph correctly and stop. “As temperature increased, rate increased.” That is a pattern statement. The reasoning still needs the scientific concept or mechanism.

A good graph-based CER response has three layers: identify the pattern, quantify it where useful, then explain why the relationship makes sense scientifically. If the graph bends, plateaus or contains a threshold, the reasoning should address that shape rather than forcing a simple “increases with” sentence.

Teachers can remove the graph title and ask students to infer what variables might plausibly create such a pattern. This reverses the task and tests whether students understand relationships rather than memorising a particular experiment.

CER and misconceptions: wrong reasoning is more informative than a wrong claim

A student may reach the correct answer for the wrong reason. Multiple-choice questions often hide this. CER exposes it.

Suppose a student correctly predicts that a heavy and light object fall at nearly the same acceleration in ideal conditions but explains it by saying “gravity is stronger on the lighter object.” The claim is right; the model is wrong. The reasoning gives the teacher something to repair.

This diagnostic value is one of CER’s greatest strengths. Teachers can classify errors: weak claim, irrelevant evidence, unreliable evidence, missing mechanism, misconception, overclaim or ignored alternative. Each error requires a different intervention.

A diagnostic table for teachers

• Claim vague: ask the student to specify the relationship, condition or comparison.
• Evidence irrelevant: ask which observation would change the claim.
• Evidence copied but not interpreted: ask what pattern the data show.
• Reasoning repeats evidence: ask for the scientific concept or mechanism.
• Mechanism incorrect: return to a model, worked example or concrete demonstration.
• Alternative ignored: ask what else could explain the result.
• Conclusion too certain: match the wording to sample size, controls and measurement quality.

How to teach CER without creating robotic paragraphs

Sentence starters help novices, but they should be faded. If every answer begins “My claim is…” and “My evidence is…,” students may treat scientific writing as a fill-in-the-blanks task.

Use explicit labels during early learning, then gradually remove them. Next, ask students to highlight claim, evidence and reasoning inside a natural paragraph. Later, give them model answers with the components blended and ask them to identify the structure. Eventually, students should be able to construct a coherent explanation without visible scaffolds.

The structure should become mental rather than merely graphical.

Use bad examples on purpose

Students learn a great deal by comparing explanations. Give three responses to the same question: one with a correct claim but no evidence, one with data but faulty reasoning, and one strong explanation. Ask students to rank them and justify the ranking.

Then make the task harder. Give two answers that are both plausible but use different evidence. Which is stronger? What additional evidence would settle the disagreement? This teaches evaluation, not just production.

CER as an oral discussion routine

Scientific reasoning should not live only in written homework. In class discussion, teachers can respond to an answer with three moves: “What is your claim?” “What makes you say that?” “Why would that evidence support it?”

Peers can be taught to challenge respectfully: “I agree with the claim, but I think the evidence is weak because…” or “Could this result also be explained by…?” This creates a classroom where disagreement is about reasons rather than status.

Small-group tuition is especially suited to this. With three learners, each can take a rotating role: claimant, evidence checker and alternative-explanation challenger. The roles can switch every question.

CER across subjects: useful, but do not flatten disciplinary differences

The general logic of claim and evidence appears in History, English, Geography and even Mathematics, but the nature of evidence differs. A historical source is not evaluated like experimental data. A quotation from a novel does not establish a scientific cause. A mathematical proof is not the same as an empirical generalisation.

Teachers can use the broad structure across subjects while teaching what each discipline accepts as a good reason. This is educationally richer than telling students that “evidence is evidence.” Disciplines have different standards because they ask different kinds of questions.

A Clementi-style learning route for CER

Students benefit when the page lets them enter at the right level. A fast route can show the seven-question checkpoint. A learning route can explain claim, evidence and reasoning with examples. A repair route can diagnose common errors. A challenge route can introduce mechanisms, alternatives and uncertainty. A transfer route can provide unfamiliar scenarios.

This layered structure prevents two opposite problems: weaker students drowning in advanced detail and stronger students being trapped in a three-box worksheet long after they have outgrown it.

A four-week CER teaching sequence

Week 1: separate observation from explanation

Use familiar phenomena. Students list only what they can observe or measure, then separately list possible explanations. This trains the distinction between evidence and inference.

Week 2: strengthen evidence selection

Give students more information than they need. Ask them to select the two most relevant pieces and reject distractions. Discuss why “true” is not the same as “useful evidence for this claim.”

Week 3: teach mechanisms and alternatives

Take common answers that simply restate the pattern and ask “what process would make that happen?” Then ask “what else could produce the same observation?” Students learn to deepen and challenge explanations.

Week 4: remove the scaffold and transfer

Provide unfamiliar data or a new practical scenario. Students write a natural explanation without CER labels, then annotate their own response to show where claim, evidence, reasoning, mechanism and qualification appear.

How to assess CER without rewarding length

A longer answer is not automatically better. Score the quality of the components. Is the claim precise? Is the evidence relevant and trustworthy? Does the reasoning use correct scientific concepts? Is the mechanism plausible? Does the conclusion recognise limitations?

For some questions, two excellent sentences may be enough. For others, students need a paragraph, graph interpretation or extended evaluation. Rubrics should reflect explanatory demand, not word count.

Teachers can also mark one component at a time during practice. On Monday, only evidence selection receives feedback. On Wednesday, reasoning. This reduces marking load and helps students focus on one improvement target.

How parents can help with science explanations

Parents do not need to know the science answer. They can ask structure questions: “What are you claiming?” “Which result supports that?” “Why would that result make the claim true?” “Could anything else explain it?”

If the child cannot explain the scientific mechanism, that is the moment to return to notes, a textbook or the teacher’s model—not for the parent to invent an explanation. The questioning structure helps the learner find the missing piece.

How tutors can diagnose in small groups

Give all three students the same data but assign different tasks. Student A states the strongest claim. Student B selects the best evidence. Student C challenges the reasoning. Then rotate. This exposes whether a student’s difficulty lies in interpreting data, recalling concepts or connecting the two.

Another useful exercise is “repair the answer.” Provide a nearly correct explanation with one hidden flaw. Students must identify and fix it. This is often more revealing than asking for another routine response.

CER and examination performance

Examinations may not label questions “CER,” but the underlying logic appears in explain, justify, account for, evaluate and infer tasks. Students who can distinguish evidence from reasoning are better positioned to answer precisely.

However, exam answers still need syllabus-specific vocabulary, command-word awareness and mark allocation discipline. CER should sharpen thinking, not become a fixed paragraph that students force into every question. A two-mark item does not need a five-part essay. Good students scale the reasoning to the task.

Frequently asked questions

Is CER only for Science?

No. Claim-and-evidence structures appear across many subjects. But teachers should preserve disciplinary differences in what counts as evidence and how claims are justified.

Should every science answer use CER labels?

No. Labels are scaffolds. Mature students should produce natural scientific explanations while retaining the reasoning architecture internally.

What is the most common weakness?

Many students provide evidence and then restate it instead of explaining the mechanism. Asking “why would that evidence support the claim?” and “what process makes it happen?” usually reveals the gap.

Can CER be used with data students did not collect?

Yes. Students can reason from published data, graphs, simulations, case studies or teacher-provided results. The key is that they understand what the evidence represents and its limitations.

How do I know when to fade the scaffold?

Fade it when students can reliably produce the components without prompts. Remove sentence starters first, then labels, then explicit boxes. Continue to use the reasoning questions when new content makes the task harder.

The deeper lesson: science is organised disagreement with reasons

Claim–Evidence–Reasoning matters because it gives students a compact experience of what makes scientific knowledge different from assertion. A claim is not accepted because a confident person says it. Evidence is not useful merely because it is a fact. Reasoning is not decoration after the answer. The pieces constrain one another.

When students learn to ask whether the evidence is relevant, whether the mechanism makes sense, whether alternatives remain and whether the conclusion is too certain, they are learning more than a writing structure. They are learning how explanations earn trust.

That is why the best CER lessons eventually make the acronym disappear. The student no longer needs three boxes because the questions have become part of the way they think: What do I claim? What supports it? Why should that support count? What else could be true? What would change my mind?

---

Further reading

This eduKateSG guide develops an original CER learning architecture with mechanisms, alternatives and uncertainty. For additional classroom perspectives on science inquiry and Claim–Evidence–Reasoning, explore Edutopia’s project-based learning and inquiry resources and broader literacy-and-knowledge resources at Reading Rockets.