Evaluating mathematics learning, and the problems of teaching it

What this heading covers

Two headings set together, and they belong together: most of the problems of teaching mathematics are made worse by how it is assessed, and several are caused by it.

The examinable position, which the whole heading rests on: evaluation is part of teaching, not an event after it. A question offering assessment as something that happens at the end, to award a mark, is offering the position the syllabus rejects.

Formal and informal evaluation

	Formal	Informal
What it is	Planned, announced, recorded — tests, written exams, assignments	Unplanned, continuous — observation, questioning, listening, looking at work in progress
When	At set points	All the time
Produces	A mark, a record	An impression, a note, an immediate change of course
Good at	Comparability, reporting, covering a syllabus	Catching a misconception the same day it appears
Blind to	How the answer was reached; the child who freezes in a test	Precision; comparability; anything the teacher did not happen to see

The syllabus wants both, and wants you to know what each cannot see.

A formal test shows that a child got 1/2 + 1/3 wrong. Only informal assessment — asking, watching, listening — shows that they added across, which is the thing a teacher can act on. Conversely, informal assessment alone is unsystematic: the quiet child who is never asked is the child whose difficulty is never found.

Informal methods named in the syllabus and worth listing in an answer:

• Observation during work, especially of how a child works rather than what they produce.
• Oral questioning, including asking for the method rather than the answer.
• Listening to children explain to each other, which is where misconceptions surface most freely.
• Looking at rough work — the margin, the crossings-out, the finger counting.
• Self- and peer assessment, which the syllabus treats as assessment rather than as a time-saving device.
• Portfolios of work over time, which show change that a single test cannot.

The three purposes

Purpose (its name)	Question it answers	When
Diagnostic — Diagnostic test	Where exactly is the difficulty, and why?	Before or during
Formative — Assessment for learning	How is learning going, and what should I change?	During
Summative — Assessment of learning	How much has been learned?	At the end

The distinction the paper tests most: assessment for learning against assessment of learning. Formative assessment feeds back into teaching and is useless unless something changes as a result. Summative assessment measures and reports.

The same instrument can serve either purpose — a test given mid-unit and used to reteach is formative; the same test given at the end for a grade is summative. It is the use, not the instrument, that decides, and a question asking whether a particular test is formative is asking what was done with the results.

CCE — Continuous and Comprehensive Evaluation. Continuous means throughout, not once; and comprehensive means across more than the written answer — understanding, reasoning, participation, attitude. It is part of the RTE framework, and a question asking what the two words mean wants exactly that pair.

What a good mathematics question looks like

The syllabus expects a view on this, and it is assessable in both papers.

A question that only asks for an answer can be got right by a remembered procedure and wrong by a slip, and neither outcome tells the teacher much. Better questions ask for:

• The method — "how did you work it out?"
• A justification — "why does that work?"
• A representation — "draw a picture that shows this."
• An estimate before computing — which reveals number sense that an exact answer hides.
• More than one route — "can you find another way?"
• A constructed example — "write a subtraction whose answer is 27", which is far more revealing than asking for the answer to one.
• An error spotted — "this child wrote 1/2 + 1/3 = 2/5. What did they do?" — which tests understanding at a level a correct answer does not reach.

Open-ended questions admit several correct responses and several routes; closed ones have one. Both are needed; a diet of only closed questions trains procedure and measures recall.

Marking

Marking only the final answer loses the information. A child who sets the problem up correctly and slips in the arithmetic, and a child who computes flawlessly from a wrong setup, both score zero and need completely different teaching. Marking the method separates them.

The problems of teaching mathematics

A named heading in both CTET papers, and an answer is expected to go beyond "children find it hard".

1. Mathematics anxiety. A real and specific phenomenon: fear of the subject that blocks working memory, so that a child who can do the mathematics cannot do it under the feeling. It is produced by public failure, by speed pressure, by ridicule for wrong answers, and — the syllabus is blunt about this — by teachers who are themselves anxious about mathematics and transmit it. It is strongly gendered in reporting, and the belief that girls are weaker at mathematics is named as a stereotype to be countered, not accommodated.

2. Abstraction and sequence. The subject is abstract and strictly cumulative, so a gap early on blocks everything resting on it. A child who never secured place value will keep failing at regrouping, decimals and long division, and practice at those will not fix it. This is why diagnosis matters more in mathematics than in most subjects.

3. The "I can't do maths" belief. Treated by the syllabus as a learned belief rather than a fact about the child, and as something a teacher's response to error either installs or prevents. The fixed view — that mathematical ability is a thing you have or lack — is the one to argue against.

4. Pace against understanding. A syllabus to be covered pushes towards telling rather than building, and towards moving on before a concept is secure. The conflict is real and a good answer acknowledges it rather than pretending the ideal is cost-free.

5. Language. Covered fully in its own lesson: the register, the word problem's double load, and the language of instruction.

6. Large and mixed-ability classes. One teacher, fifty children, several levels at once — which makes individual diagnosis hard precisely where it matters most. Mitigations the syllabus accepts: group work, peer explanation, graded tasks on the same context, and using the children's own methods rather than requiring one.

7. Materials and resources. Concrete materials are needed and are often absent. The expected response is improvisation — bottle caps, seeds, sticks, folded paper, string — rather than treating the absence of a kit as the end of the argument.

8. Rote and the textbook. Mathematics taught as procedures to memorise, assessed by speed and accuracy, produces children who can compute and cannot reason. This is the root the other problems grow from, and it is what the whole pedagogy syllabus is written against.

A question of the kind the paper sets

A teacher gives a short test mid-unit, finds that most children have added fractions by adding across, and reteaches using paper folding. This is an example of: (a) summative assessment (b) formative assessment (c) a diagnostic test (d) grading

(b) formative assessment.

It is worth seeing why (c) is the tempting wrong answer. Something was diagnosed, so "diagnostic" feels right — but a diagnostic test is a specific instrument, built narrow and graded to locate a difficulty. What is described is an ordinary short test whose results were used to change the teaching, and that use is what makes assessment formative.

The deciding words are mid-unit and reteaches. Assessment during, feeding back into teaching, is formative; the same test at the end for a grade would be summative.

Mistakes that cost marks here

• Calling assessment an end-of-teaching event. The syllabus's position is that it is part of teaching.
• Treating the instrument as deciding the purpose. The same test is formative or summative according to what is done with the results.
• Swapping assessment for and of learning. For is formative.
• Reading "comprehensive" in CCE as "covering the whole syllabus". It means across more than written answers.
• Offering only formal methods when asked how to evaluate mathematics learning.
• Marking only the final answer, which cannot distinguish a wrong setup from a slip.
• Explaining mathematics anxiety as low ability. It is a learned fear that blocks demonstrable ability.
• Listing "children find it difficult" as a problem of teaching. The heading asks about causes.

In one line

Formal methods give comparability and informal ones give the reason, the purpose rather than the instrument makes assessment formative or summative, a good mathematics question asks for the method rather than the answer, and the real problems of the subject are anxiety, its cumulative abstraction, pace, language and class size — most of them made worse by assessment that counts only right answers.

In the classroom

The single change with the largest effect here is marking the method as well as the answer, because it alters what children think they are being judged on. A class that knows the working is read will show the working, and a teacher who can see the working can teach.

For anxiety, the practical move is removing speed from the public part of the lesson. Mental arithmetic races reward the already-fluent and teach everyone else that mathematics is a thing done fast or not at all. The same facts can be practised without an audience and without a clock.