Why education must make reasoning visible

A correct answer is useful. It tells us that, on this occasion, the learner arrived at the expected result. It does not always tell us how.

The answer may come from secure understanding, a remembered procedure, imitation of a previous example, elimination of alternatives, or a lucky guess. An incorrect answer can be equally ambiguous: it may reflect a deep misconception, a minor calculation error, a misunderstood instruction or a sound method interrupted at the final step.

If education attends only to the visible output, it loses much of the evidence needed to improve learning.

The answer tells us where the learner arrived. The reasoning helps us understand the route and decide what should happen next.

One answer can come from different thinking

Consider two learners who produce the same correct solution. One can explain why the method works and apply it when the problem looks different. The other remembers a sequence of steps but cannot recognise when to use it.

The mark may be identical. The instructional need is not.

Now consider two incorrect answers. One learner understands the concept but makes an arithmetic slip. The other has applied a rule that does not fit the problem. Correcting both answers without uncovering the difference produces weak feedback.

Reasoning is therefore not an optional performance added after the “real” work. It is part of the evidence that makes teaching responsive.

Recall, reasoning and guessing can produce the same answer but different evidence of understanding.

Visible reasoning does not mean demanding more writing

Calls to “show your working” can become another compliance routine. Learners copy every intermediate calculation or add a sentence because the worksheet requires one, without revealing much about their decisions.

Reasoning can be made visible in several ways:

explaining a choice aloud;
using a diagram or physical representation;
comparing two methods;
identifying and repairing an error;
estimating before calculating;
giving an example and a non-example;
predicting what will change; or
applying the idea in an unfamiliar context.

The method should fit the learning. A lengthy written explanation is not always the best or fairest evidence.

Dialogue can develop reasoning as well as reveal it

Discussion is sometimes treated mainly as a way for teachers to check what pupils already know. High-quality dialogue can also change the thinking itself.

When learners explain a method, hear a contrasting approach and respond to a challenge, they organise ideas and notice relationships that may have remained implicit. But more talk does not automatically mean more learning. Pupils may need explicit teaching about how to participate: how to build on an idea, disagree with a reason, ask for clarification and use subject-specific language.

Sentence stems can help learners enter the discussion: “I chose this because…”, “This will always work when…”, “The methods are similar because…”, or “I disagree at this step because…”.

The scaffold should support the reasoning, not script an empty performance.

Representations reveal structure

In mathematics, a drawing, number line, array or manipulative can show relationships that symbols alone may conceal. Representations can also reveal what the learner believes the symbols mean.

A child who draws equal groups is showing something different from one who records repeated addition without recognising the group structure. A learner who places a fraction correctly on a number line demonstrates a relationship that a rehearsed definition may not expose.

The representation is not merely a route to the answer. It is evidence of conceptual organisation.

Assessment should reward the thinking it claims to value

Education systems often say reasoning, critical thinking, and problem solving matter, yet tasks and marking can still privilege short, easily counted outputs. If only the final answer carries value, learners receive a rational message: the route is secondary. If explanations are requested but never used to shape feedback, reasoning becomes decorative.

Assessment design should ask what evidence the decision requires. Sometimes a short answer is sufficient. At other times, comparison, justification or application is necessary to distinguish secure understanding from surface success.

Not every question needs an essay. A well-designed example, diagram, or hinge question can efficiently make a misconception visible.

Technology can expose or conceal the process

Digital tools can record steps, capture explanations and present contrasting examples. Used well, they can provide teachers and learners with better evidence. They can also make the process less visible. Automated systems may reward task completion or correct responses while the learner relies on repeated hints, pattern matching or generated answers.

Activity data is not the same as evidence of reasoning.

The useful question is not only, “What did the learner complete?” It is, “What decisions did the learner make, and what does that show about their understanding?”

Reasoning needs a safe intellectual culture

Learners will not expose uncertain thinking if every error is treated as a public failure. Making reasoning visible requires classrooms and tutoring relationships in which partial ideas can be examined without lowering expectations.

Teachers can model changing their minds, compare incomplete methods and treat errors as information. That does not mean accepting inaccurate mathematics or vague explanations. It means creating a route for improving ideas.

When adults welcome only polished answers, they see less of the learning they need to support.

Better evidence leads to better next decisions

Visible reasoning helps teachers distinguish a knowledge gap from a procedural slip, choose a more useful example, and decide whether to model, question, practise, or extend.

It helps learners monitor their own understanding. “I got it wrong” becomes “My estimate showed the answer was too large, but I did not check which operation I had used.”

It helps parents and tutors discuss learning without reducing progress to the number of questions completed.

It also helps systems evaluate whether resources and technology produce genuine understanding rather than compliant activity.

The goal is not explanation for its own sake

The purpose of making reasoning visible is not to make every thought public or add narration to every task. It is to gather the evidence required to teach well and help learners take increasing control of their own decisions.

If we want better answers, marking answers alone will not be enough. We need learning environments where the route can be represented, questioned, compared, and improved.

That is how reasoning becomes both visible and stronger.

Sources and further reading

EEF: Improving Mathematics in Key Stages 2 and 3 - Evidence-informed recommendations on representations, misconceptions, problem solving and discussion.
EEF: Scaffolding high-quality talk in maths - Examples of sentence stems and structured dialogue that help pupils articulate reasoning.
EEF: Four principles for high-quality talk in maths - The TOLD principles for participation, opportunities, links and debate.
DfE: Mathematics guidance key stages 1 and 2 - Non-statutory curriculum guidance, including representations and mathematical language.

We are discussing this article on LinkedIn: The answer matters. The reasoning tells us what to teach next.

For parents looking for trusted learning support: Find Tutors

Join the education movement: Teach. Create. Own.

For qualified teachers interested in joining TutorTech: Join as a Tutor

Read other Education Futures blogs: