A graph can be mathematically correct and still be read incorrectly.
A table can contain every number you need and still hide the relationship the question is testing.
A pie chart can show a fraction of a whole, but only if the learner remembers what the whole actually is.
Data questions are rarely just about arithmetic. They test whether a learner can extract, connect and interpret information from a representation before deciding what calculation is justified.
That is why a student can know percentages, averages and ratios but still lose marks on a data question.
The calculation may be familiar. The representation is doing the difficult work.
The quick answer: read the representation before the question numbers
Before calculating, identify five things:
- What is being measured?
- What does each row, column, bar, point, sector or symbol represent?
- What are the units?
- What scale is being used?
- What is the whole or reference quantity?
If one of those is wrong, correct arithmetic can produce a wrong answer.
Start with the title, labels and legend
Students often jump directly to the tallest bar or largest number.
Instead, read the representation like a sentence.
- The title tells you the subject of the data.
- Axis labels tell you what varies and what is measured.
- Units tell you the size of each quantity.
- A legend tells you which colour, line or symbol belongs to which series.
- Footnotes may define exclusions, totals or categories.
Only after these are clear should the learner treat the marks on the page as numbers to calculate with.
Worked example: a table that requires a relationship, not a lookup
A school records the number of books borrowed from four categories in one week.
| Category | Books borrowed |
|---|---|
| Adventure | 48 |
| Science | 36 |
| Biography | 24 |
| Comics | 72 |
Question: What fraction of all borrowed books were Science books?
The Science value is 36, but the denominator is not written directly.
Total = 48 + 36 + 24 + 72 = 180.
Fraction = 36/180 = 1/5.
The key reading move was recognising that “of all borrowed books” defines the whole.
A graph scale is part of the data
If a vertical axis is marked 0, 20, 40, 60, one small interval may represent 5, 10 or 20 depending on how many spaces lie between labels.
Do not assume one square means one unit.
A reliable learner asks:
How much does one smallest visible interval represent?
This prevents a common graph error: reading visual height instead of numerical scale.
Line graphs show ordered change, not just separate heights
A bar graph often compares categories.
A line graph often shows how a quantity changes across ordered positions such as time.
That means the relationship between neighbouring points matters.
- A rising line means the measured value increased.
- A horizontal line means the measured value stayed constant.
- A falling line means the measured value decreased.
- A steeper segment means a larger change per horizontal interval when the scales are comparable.
But do not automatically call a steeper line “faster” unless the graph actually represents distance against time or another rate context where that interpretation is justified.
Worked example: reading change from two points
A line graph shows a tank containing 120 L at 9:00 a.m. and 180 L at 10:30 a.m.
Increase = 180 − 120 = 60 L.
Time elapsed = 1.5 hours.
Average increase over that interval = 60 ÷ 1.5 = 40 L per hour.
That does not prove the tank filled at exactly 40 L per hour at every moment unless the graph or context supports a constant rate.
Data questions require both calculation and restraint.
Pie charts require a known whole
A pie chart divides one whole into sectors.
If a sector is 90°, it represents 90/360 = 1/4 of the whole.
If the whole contains 240 pupils, that sector represents 240×1/4 = 60 pupils.
But if the total number of pupils is not given, the sector angle alone cannot tell you an absolute number of pupils.
This distinction between proportion and quantity is central.
Mixed-representation questions require translation
A PSLE-style problem may provide a table for Monday to Wednesday, a graph for Thursday to Saturday, and a sentence giving Sunday’s total.
The learner must combine them into one mathematical system.
A useful method is to create a temporary common table:
| Day | Value | Source |
|---|---|---|
| Mon | … | table |
| Tue | … | table |
| Wed | … | table |
| Thu | … | graph |
| Fri | … | graph |
| Sat | … | graph |
| Sun | … | sentence |
This reduces representation switching during the calculation itself.
Do not compare unlike quantities
A table may show kilograms.
A graph may show grams.
A percentage may refer to a subgroup rather than the full sample.
Before combining values, standardise units and reference groups.
For example:
2.4 kg = 2400 g.
Only then should 2.4 kg be compared or added to a value stated in grams.
A percentage needs a base quantity
Suppose 40% of Primary 6 pupils chose football and 30% of Primary 5 pupils chose football.
You cannot conclude that more Primary 6 pupils chose football unless the cohort sizes are known.
40% of 80 = 32.
30% of 120 = 36.
A larger percentage can correspond to a smaller count.
Data interpretation fails when a proportion is treated as if it were already an absolute quantity.
Read what the graph does not say
A graph may show association across time without explaining cause.
A table may show a sample without proving that every larger group behaves the same way.
A truncated axis may exaggerate visual differences.
A missing value may be genuinely unknown rather than zero.
Strong data reading separates what is displayed, what can be calculated and what cannot be inferred safely.
Common misconception 1: the tallest mark means the largest value
Only if the scale and baseline support that interpretation.
Repair: read the axis values, not visual height alone.
Common misconception 2: every line segment represents constant change
A line between plotted observations may connect data points without claiming what happened at every instant.
Repair: distinguish displayed observations from assumptions about what occurred between them.
Common misconception 3: a missing category has value zero
Missing data and zero data are different.
Repair: check whether the representation explicitly records zero or simply omits information.
Common misconception 4: the same colour means the same quantity across every chart
Legends are local to the representation unless stated otherwise.
Repair: reread the legend whenever the representation changes.
A diagnostic ladder for data questions
- Can the learner state what the representation measures?
- Can the learner identify units and scale?
- Can the learner read one value accurately?
- Can the learner calculate a difference or total from several values?
- Can the learner identify the correct whole for a fraction or percentage?
- Can the learner translate a pie-chart angle into a fraction of the whole?
- Can the learner combine values from a table and a graph?
- Can the learner standardise units before combining data?
- Can the learner distinguish a count from a percentage?
- Can the learner explain what cannot be concluded from the representation?
How this fits the 2026 PSLE Mathematics framework
The MOE Primary Mathematics syllabus includes Statistics as one of its three major content strands alongside Number and Algebra, and Measurement and Geometry. The 2021 syllabus applies to Primary 6 from 2026 onwards.
SEAB’s 2026 PSLE Mathematics assessment objectives include interpreting information, applying mathematical concepts and skills in varied contexts, analysing information, making inferences and selecting appropriate strategies. Data questions sit directly at the intersection of those demands.
This article does not reproduce copyrighted examination questions. It teaches the transferable reading and reasoning moves that data representations require.
A three-pass method for exam conditions
- Representation pass: title, labels, scale, legend, units, whole.
- Question pass: what exact quantity or relationship is being requested?
- Calculation pass: extract only the necessary values, calculate, then check the unit and reasonableness.
This order prevents arithmetic from beginning before the learner knows what the numbers mean.
What parents and tutors should ask
- “What does one small interval on this axis represent?”
- “What is the whole in this percentage?”
- “Are those two values in the same units?”
- “Which representation gave you that number?”
- “Does the graph prove that, or are you adding an assumption?”
- “Could you rewrite all the information into one table before solving?”
The deeper lesson: representation is part of the mathematics
A table organises exact entries.
A graph makes change and comparison visible.
A pie chart compresses part-whole relationships.
None is automatically superior.
The question is what relationship each representation makes easy to see—and what it hides.
Strong data reasoning begins when the learner stops seeing a chart as decoration and starts treating it as a mathematical language.