Did you know that the number at the top of a marked test can conceal very different learning stories? Two students can both score 14 out of 20: one may know the ideas but leave explanation marks behind; another may write convincing explanations while making unreliable calculations. The shared total does not tell a parent which lesson should come next.
Reading test scores as skill evidence means looking below the overall mark to see what students were asked to demonstrate, what they actually showed, and what still needs a fresh check. A score can summarise performance on a particular set of questions. It should not become a permanent label such as “good at maths” or “weak at science.” The useful question is: Which next piece of evidence would help us teach or practise better?
Start with the paper, not the percentage
Here is an illustrative, invented 20-mark classroom quiz grouped by the skills its questions were intended to test. A learner earns 4/5 marks on understanding instructions, 5/5 on carrying out calculations, 2/5 on explaining reasoning, and 3/5 on checking answers. The total is 4 + 5 + 2 + 3 = 14/20, or 70%. That arithmetic is correct, but the four rows reveal a question worth investigating: why is the explanation section lower than the calculation section?
It would be too quick to conclude that this student “cannot reason.” Perhaps the student knew the method but ran out of time, misunderstood how much explanation was required, or forgot a prerequisite. Or perhaps the calculations were carried out by memorised procedure without understanding. The section scores are signals. One short, carefully chosen follow-up question can separate competing explanations.
For instance, give the learner this new task without a model answer: Three identical notebooks cost $12. Explain how you know what five notebooks would cost at the same unit price. The method is $12 ÷ 3 = $4 per notebook, then 5 × $4 = $20. A useful explanation identifies the constant unit price and why multiplication by five is appropriate. The teacher can then distinguish someone who finds $20 but cannot justify the price-per-book idea from someone who explains clearly but makes an arithmetic slip.
What a test score measures—and what it does not
A total mark tells us how performance on that test was scored using its particular questions and marking criteria. It does not, by itself, reveal every underlying skill, motivation level or future outcome. Even a clean 5/5 section result may reflect only one or two easy prompts; and a 2/5 section may be especially difficult or unfamiliar. A five-mark slice is a very small sample.
A better diagnosis considers three layers: the task demand (what the student had to do), the response evidence (what the work shows), and the likely explanation (which must still be checked). Keeping these separate prevents a single disappointing mark from becoming an exaggerated story about ability.
Make a useful error-and-evidence map
- Identify the demand: For each missed mark, note whether the task called for reading, retrieving a fact, calculating, explaining a relationship, interpreting evidence or checking an answer. A question may require several skills.
- Circle the first uncertain step: Mark where the actual response stops showing the intended knowledge. Avoid copying the teacher’s finished answer without locating the learner’s divergence.
- Record alternative explanations: A missing paragraph might reflect poor understanding, an unfamiliar command word or insufficient time. Do not announce the cause before you test it.
- Choose one low-stakes check: Ask a fresh, short question that isolates the suspected skill without introducing three new difficulties.
- Review the response: If the learner succeeds independently in a different context, revise the first diagnosis rather than forcing it to fit the original mark.
This map is more useful than a colour-coded report with no teaching decision attached. A parent can write just three columns on paper: What was asked? What happened? What shall we check next? The purpose is to decide the next move, not to manufacture a more complicated report card.
Three common mistakes when parents review a marked paper
Mistake one: treating percentages from unequal question sets as equivalent. An 80% score on a short quiz with familiar items may not represent the same knowledge as 80% on a lengthy, unfamiliar assessment. Compare question demands and conditions before judging a change.
Mistake two: assuming every lost mark needs more drilling. If a child cannot explain why $12 divided by 3 gives the cost of one notebook, another page of multiplication may avoid the missing idea. A brief explanation, picture and independent transfer item may fit better.
Mistake three: confusing marks with identity. “You lost three marks on explaining relationships here” is a fixable observation. “You are bad at explanations” turns limited evidence into a broad claim. The language adults use changes the conversation after a test.
A ten-minute review that students can lead
Give the student the paper and ask for one answer they are confident about and one that surprises them. Have them explain the reasoning behind both. Then choose a single weak skill, model the crucial step if needed, and ask a related new question with the marking guide covered. Finish with the learner writing one specific next action: “I will state the unit price before multiplying,” for example, rather than “I will study harder.”
For younger children, parents can read the instruction aloud and ask the learner to point to evidence on the page. For older students, encourage them to annotate their own marked scripts and take one precise question to a teacher. Either way, the learner should do the explaining where possible. This protects independence rather than shifting the thinking to the adult.
How teachers can check whether progress is genuine
Build a tiny follow-up set that samples the same intended skill using different numbers, wording or contexts. Keep track of whether a hint, prompt or worked model was provided. Across several occasions, compare not only scores but also how accurately the child explains the first step and corrects an error. If question difficulty changes substantially, say so; do not draw a straight progress line between incomparable assessments.
The test may be poorly aligned to the intended learning, and some work samples may not separate literacy from subject knowledge cleanly. A short quiz can guide teaching, but should not be used to diagnose a disability, decide an entire school pathway or promise an exam result. For difficult or persistent concerns, discuss the actual tasks and learning observations with the student’s teacher.
Frequently asked questions
Is 70% a good score? The fraction 14/20 equals 70%, but “good” depends on the task, expectations, support given and purpose of the assessment. The next useful step is to inspect which knowledge and skills that number represents.
Should parents track every subscore? No. Track a few meaningful patterns and revisit them with comparable questions. An elaborate spreadsheet is not necessary for most learners.
What if the mark scheme is not available? Start with the question and the student’s own explanation. Ask the teacher what the task was intended to reveal before guessing why a mark was awarded or lost.
A score should open a conversation, not end it
The core educational move is simple: turn “14 out of 20” into a better question, then a better explanation, then a new independent attempt. See the one-page learning handoff guide for recording useful next actions, the Core Aim of Education learning hub for related skills, and the Clementi teaching reference for an example of diagnosis-led progression.