Primary 5 Science graph and table questions test whether a child can read the correct quantity, understand the units, describe a pattern and draw only the conclusions supported by the data. The hardest part is often not the Science fact; it is interpreting a new representation accurately. A graph of water remaining, for example, falls when water is lost, while a graph of total water evaporated could rise during the same experiment. Read the axis labels before deciding what the line means.
Singapore’s MOE Primary Science syllabus develops interpreting data, scientific reasoning and communication throughout primary school. The national SEAB PSLE is at the end of Primary 6; P5 Science papers are school-based, and their topic scope and exact format are determined by the school. The 20 examples here are original practice tasks, not official past-year questions or promised 2027 questions. Select the parts matching the school’s current syllabus and assessment notice.
- First read: title, horizontal and vertical labels, scale intervals, units, legend and condition names.
- Calculate carefully: identify the starting and ending values and whether the question asks for a change, difference, total, proportion or rate.
- Describe trends: use precise language such as increases, decreases, remains constant or changes then levels off.
- Separate jobs: an observation is not automatically an explanation, and a relationship in a graph is not necessarily proof of one cause.
- Check fairness: use the experimental method to decide whether relevant variables were controlled before making a causal conclusion.
- Answer in context: name the tested material, measured quantity and unit instead of writing only “it went up”.
Why a Science data question can be harder than a content question
A pupil may know that evaporation changes liquid water into water vapour and still struggle with a table showing decreasing water mass. Another may understand plant germination but compare the total number of seeds germinating in groups with different starting sizes. A child can know the correct concept and still make the wrong numerical comparison.
The useful strategy is a data-first routine. Before recalling a chapter, read what each number represents. Is the vertical axis the remaining volume or the volume lost? Is the horizontal axis time in minutes, temperature in degrees Celsius, or number of identical batteries? Does a plotted value represent a single trial or an average of several trials? Answer those questions before making a scientific claim.
Schools may use line graphs, bar charts, pictorial representations, tables, comparative illustrations or written data descriptions. The exact presentation is not standardised at P5 across all schools. The skill that travels across layouts is checking what is measured, how the values compare and what the evidence can justify.
The MOE 2023 Primary Science syllabus emphasises inquiry, analysis and interpretation of data, and evidence-based reasoning. For the related question of whether an experiment was well controlled, read P5 Science Fair Tests and Variables.
How to read a table in six steps
First, read the table title to identify what the data are about. Second, examine column and row headings and note any categories or conditions. Third, check the units and whether all values are recorded in comparable units. Fourth, identify which factor varies across rows. Fifth, compare the correct values, and only then calculate a change or describe a trend. Sixth, return to the actual command word to decide how much to write.
A table might show time, in minutes, and volume of water remaining, in millilitres. A common error is to read “remaining” as “evaporated.” If a row shows 80 mL remaining after starting at 100 mL, the amount lost is 20 mL. If another row shows 60 mL remaining, more water has been lost, not less. The mathematical and scientific interpretation depend on the label.
Some tables list several trials in separate columns. These are not necessarily conditions being deliberately changed; they may be repeat measurements under the same conditions. Ask what the table structure represents. An experimental comparison should not be inferred from a column titled “Trial 2” without checking the method.
When quantities are in different units, make them comparable before judging which is larger. A value of 1.2 litres is not smaller than 700 millilitres simply because 1.2 is less than 700. However, do not introduce conversions beyond what the level and question require. The essential habit is to check units as part of meaning.
How to read a graph in seven steps
Locate the title, then read the horizontal-axis label and vertical-axis label. Read the units. Determine the numerical interval between adjacent marked ticks. Check whether the axes begin at zero or another value. If there are multiple lines or bars, decode the legend before interpreting colours. Finally, identify the specific points required by the question.
A graph whose vertical axis starts at 80 can make a change from 90 to 95 look dramatic. The change is still 5 units. The starting point of the drawn axis affects visual impression but not the numerical difference. This is a useful lesson beyond school: charts can mislead when viewers judge height without reading the scale.
The spacing of axis ticks matters. If ticks are labelled 0, 5, 10, 15 and 20, the third interval does not mean the value is 3. The plotted point must be read using the numerical scale. Some questions use 2, 4 or 10 units per grid line, so the pupil should check rather than assume.
In a line graph, consider both direction and shape. A line can rise steadily, fall, stay level, rise then level off, or rise and later fall. Describe the relevant pattern using the graph’s measured variable. Avoid declaring that a steeper line *must* mean a higher temperature if the vertical axis records something else.
Four questions to ask before forming a conclusion
What is directly observed? Name the measurements or described pattern. What relationship is suggested? Explain how the relevant variables change together in the dataset. Was the comparison designed fairly? Check whether the intended factor was isolated. What is not established? Identify any proposed cause, universal statement or hidden condition that the data cannot support.
These questions prevent the difference between data and conclusion from becoming a mere memorised definition. A graph of a plant’s height next to a sunny window may suggest a pattern, but differences in water, temperature or soil must be considered before claiming light alone was responsible. A water graph can show a change without identifying which environmental factor caused it.
The student should not refuse to make any conclusion. If a controlled experiment shows that evaporation increased with exposed area under comparable conditions, a conclusion about the tested relationship is reasonable. The claim should simply be limited to what was tested.
Worked P5 Science data questions — 20 original examples
1. Water remaining versus water lost
Question: An open tray starts with 120 mL of water. After 30 minutes, 95 mL remains. How much water was lost?
Model answer: 25 mL was lost, because 120 − 95 = 25 mL.
Explanation: The table records water remaining, not water evaporated. Subtract the final reading from the initial one to find the loss. If the question asks for an explanation of the loss, use the given environmental conditions and evaporation, not the calculation alone.
Common trap: Saying 95 mL evaporated just because 95 is the last number in the table.
2. Two trays: compare losses
Question: Tray A starts with 100 mL and ends with 82 mL; Tray B starts with 100 mL and ends with 74 mL after the same interval. Which lost more water, and by how much more?
Model answer: Tray A lost 18 mL; Tray B lost 26 mL. Tray B lost 8 mL more than Tray A.
Explanation: Calculate each loss before comparing. Comparing only the remaining amounts can reverse the result: the tray with less remaining has lost more when starting amounts are equal.
Common trap: Choosing Tray A because 82 is greater than 74 without recognising the measured quantity.
3. A rising graph of evaporation
Question: A graph labelled ‘Cumulative water evaporated (mL)’ rises from 0 to 15 over one hour. Has the amount of liquid water remaining necessarily increased?
Model answer: No. The graph tracks the accumulated amount evaporated, which increases as water leaves the liquid.
Explanation: The label tells you that a rising line describes water loss accumulated over time, not water remaining. Do not transfer the direction of one graph to another measured variable.
Common trap: Assuming every rising water graph shows more liquid water in the container.
4. Read a five-unit scale
Question: A vertical graph axis is marked 0, 5, 10, 15 and 20. A bar reaches halfway between 10 and 15. What approximate reading does it show?
Model answer: Approximately 12.5 units of whatever quantity the axis measures.
Explanation: Halfway between 10 and 15 is 12.5, not 3 or 4. If the vertical label says centimetres, include cm; if it says grams, include g.
Common trap: Counting grid lines rather than reading numerical tick values.
5. A graph that does not start at zero
Question: Two readings are 92 mL and 97 mL. The vertical axis begins at 90 mL. What is their numerical difference?
Model answer: The difference is 5 mL.
Explanation: The axis truncation may make the plotted vertical separation look large, but subtraction is still 97 − 92. Use values, not visual exaggeration.
Common trap: Reporting a much larger change because the bars appear dramatically different.
6. A flat section in a graph
Question: A line graph of the number of germinated seeds stays at 14 between Day 5 and Day 7. Describe the recorded trend during that interval.
Model answer: The number of germinated seeds remained at 14 during Days 5 to 7.
Explanation: A flat line means the plotted quantity remained constant over the indicated interval. It does not prove that no plant processes occurred; it only addresses the number being measured.
Common trap: Saying the seeds stopped growing, although the graph records only the count that germinated.
7. A trend that changes direction
Question: The water temperature in an example rises from 20°C to 32°C and later falls to 26°C. Describe the overall pattern.
Model answer: The temperature increased from 20°C to 32°C, then decreased to 26°C.
Explanation: The description should follow the values and maintain the unit. Do not call it a steady increase across the whole period. A causal explanation would require the heating and cooling conditions.
Common trap: Ignoring the later decrease because the first two points rose.
8. Equal time intervals and rate
Question: A tray loses 12 mL of water in 30 minutes under a stated setup. What is the average volume lost per minute over that interval?
Model answer: 12 ÷ 30 = 0.4 mL per minute.
Explanation: Average rate for a measured interval is quantity changed divided by elapsed time. It is not a claim that the exact same amount evaporated in every individual minute.
Common trap: Writing 0.4 mL as a total loss rather than mL per minute for the average rate.
9. Compare rates with different durations
Question: Trial A loses 20 mL in 40 minutes and Trial B loses 18 mL in 20 minutes. Which has the higher average rate of water loss?
Model answer: A averages 0.5 mL per minute; B averages 0.9 mL per minute. B has the higher average rate.
Explanation: You cannot compare total losses directly when durations differ. Dividing by time gives comparable average rates for the observed intervals.
Common trap: Choosing Trial A simply because 20 is greater than 18.
10. Seed germination proportions
Question: In Group A, 12 of 20 seeds germinate. In Group B, 9 of 10 germinate. Which has the higher proportion?
Model answer: Group B: 90%, compared with 60% for Group A.
Explanation: The denominator matters. Raw counts answer a different question from success proportions. When starting group sizes differ, use an appropriate proportion if the question asks for the higher success rate.
Common trap: Choosing A because twelve germinated seeds is a greater count than nine.
11. Same proportion, different count
Question: Group C has 8 out of 10 germinated seeds, and Group D has 16 out of 20. How do their germination percentages compare?
Model answer: They are equal at 80%.
Explanation: 8 ÷ 10 and 16 ÷ 20 both equal 0.8. The absolute number germinating is larger in Group D, but the proportion is the same.
Common trap: Reporting that D has double the germination rate because it has twice as many germinated seeds.
12. Breathing rate difference
Question: A fictional table records breathing rates of 16 and 28 breaths per minute before and after activity. What increase was recorded?
Model answer: An increase of 12 breaths per minute.
Explanation: Subtract the initial breathing rate from the later one and keep the measured variable. The table does not record heart rate or body temperature.
Common trap: Giving the numerically correct result in heartbeats per minute instead of breaths per minute.
13. Heart rate and breathing rate are distinct
Question: Two graphs have different vertical labels: heartbeats per minute and breaths per minute. Can the same numerical reading be treated as the same physiological quantity?
Model answer: No. The values measure different processes and units.
Explanation: Read the axis label every time even when the horizontal axis is the same time period. Heart and breathing rates may change during activity, but they are not interchangeable measurements.
Common trap: Copying a number from the wrong graph because both refer to exercise.
14. Pollen and seed-production counts
Question: A study records that 18 out of 20 flowers in Group A developed fruit and 12 out of 20 in Group B did. Describe the recorded difference.
Model answer: Six more flowers developed fruit in Group A than in Group B, which corresponds to 90% versus 60%.
Explanation: This is an observation. A causal explanation requires information about what differed between the groups and whether the other relevant conditions were controlled.
Common trap: Claiming insect access definitely caused the difference without seeing the experimental method.
15. A labelled circuit-brightness reading
Question: An example records arbitrary brightness readings of 10, 18 and 25 for three circuit setups. Which setup has the greatest recorded brightness?
Model answer: The setup with reading 25 has the greatest recorded brightness.
Explanation: A comparison is possible if the readings come from the same stated measurement scale. The numbers alone do not prove why the brightness differs; battery and bulb arrangements matter.
Common trap: Assuming the highest brightness reading automatically corresponds to the most batteries without checking the setup.
16. Distinguish observation from prediction
Question: A graph shows that water loss increased as air movement increased in a controlled range. A pupil predicts the result at an untested higher airflow. Is the prediction a direct observation?
Model answer: No. It is a prediction based on the pattern in the measured range, not a directly recorded result.
Explanation: Predicting can be reasonable, but it differs from reporting actual experimental data. Avoid presenting an extrapolated value as if it were measured.
Common trap: Using the word ‘observed’ for a point beyond the graph’s recorded data.
17. A fair-test problem hidden in the method
Question: A table compares plants grown with different water amounts, but one group also receives much more sunlight. Can differences in growth be attributed to water alone?
Model answer: No. Water and sunlight both differed, so the design does not isolate the effect of water.
Explanation: The graph may show a real difference, but the method prevents a confident single-cause conclusion. This question asks about validity rather than the ability to read a numerical value.
Common trap: Believing that a clear trend makes an unfair comparison fair.
18. Repeated measurements and average
Question: A model-seed experiment records landing times of 8, 9 and 10 seconds under comparable conditions. What is the average?
Model answer: (8 + 9 + 10) ÷ 3 = 9 seconds.
Explanation: Add the three readings and divide by the number of trials. The average summarises the measurements; it does not prove every future trial will be exactly 9 seconds.
Common trap: Using the total 27 seconds as the average or omitting the unit.
19. An unusual repeated result
Question: Four trials record 8, 9, 28 and 9 seconds. What should the student do about 28 seconds before deciding on a summary?
Model answer: Check the method, timing and conditions of that trial; repeat it appropriately and report the unusual result rather than silently deleting it.
Explanation: An outlier may reveal measurement error or a real unexpected effect. Scientific reporting investigates anomalous data and avoids changing results to fit expectations.
Common trap: Deleting a legitimate observation solely because it does not match the other numbers.
20. A graph without a proven cause
Question: A graph shows that seedlings at Location A grew more than seedlings at Location B. No information about water, light, soil or temperature is provided. What conclusion is defensible?
Model answer: The observed seedlings at Location A grew more under the recorded comparison; the graph alone does not establish which environmental factor caused the difference.
Explanation: The narrow observation is supported by the data. Attributing the difference to sunlight, fertiliser or water would require information about the setup and controls.
Common trap: Adding a familiar plant-growth explanation that the experiment never tested.
Answer patterns to remember, not numbers to memorise
The correct results in the examples are less important than the methods used to obtain them. For a change in quantity, subtract the relevant readings. For a comparison of rates over different times, divide the amount changed by the elapsed time. For groups of unequal sizes, compare proportions when appropriate. For a trend, describe the measured variable across the relevant range, not a different variable mentioned in the chapter.
These are transferable habits. A student who understands why 12 out of 20 is a lower proportion than 9 out of 10 can solve other germination or classification data problems. A pupil who understands “volume remaining” versus “volume lost” can interpret both a table and a line graph without copying an answer. That flexibility is the point of the exercise.
The 20 tasks are original and intentionally use small, clear datasets. Their numbers should not be taken as real-world predictions about evaporation, plant growth, respiration or circuit brightness. A school examination may present a different pattern, and the learner must read that evidence freshly.
How to describe a trend with scientific precision
A complete description usually includes the variable, the direction of change and, where relevant, supporting readings. For example: “As the tested surface area increased, the measured water loss over the same period increased from 4 mL to 12 mL.” This is more informative than “it went up.” It identifies what changed and what was recorded.
Different graph shapes require different language. A flat line shows a constant recorded quantity over that interval. A line rising then falling has two distinct phases. A steeper section indicates a larger change per unit of the horizontal variable in that section when axes and scales are consistent. That relationship need not be expressed using advanced gradient formulas at P5.
Students should not assume that a line graph is always a time series. The horizontal axis may show surface area, temperature or number of cells. If it shows categories rather than a continuous quantity, the graph may use bars instead. Always describe the specific presentation provided rather than a generic mental image of a rising line.
Units and measurement: mistakes that travel across subjects
Science uses units to give a number meaning. A result of 25 without a label could be millilitres, grams, centimetres, seconds or degrees Celsius. Answering with the wrong unit can turn an otherwise correct calculation into an incorrect scientific statement. Check the table heading before calculating and again before finalising the sentence.
A student should also distinguish mass from volume. They are different quantities. Water loss could be recorded as a decrease in mass or a decrease in volume in different investigations. Do not treat 20 g and 20 mL as universal equivalents for all materials. Use the actual measurement and any conversion information supplied by the question.
Time units require care. A result measured in seconds cannot be compared directly with a reading in minutes without converting or interpreting the intervals appropriately. Average rates must have compound units such as mL per minute or cm per day when those are the relevant quantities.
The most useful quick check is to say the answer aloud as a complete phrase: “The amount of water remaining decreased by twenty-five millilitres.” This can expose a mismatch between the calculation and the question more effectively than staring at a bare number.
Can a graph prove that one factor caused another?
A graph can show how recorded values relate. Whether it supports a causal conclusion depends on the investigation design. If relevant factors were controlled and only one condition was changed, the data may support a conclusion about that factor under the tested conditions. If several conditions changed, the cause may be uncertain.
Suppose one group of seeds receives extra water and more light, while another receives less water and less light. If the first group germinates more successfully, the graph cannot isolate water as the only cause. The student should identify the additional difference and suggest a fairer method rather than choosing the cause that sounds most familiar.
On the other hand, if seed type, number, temperature, duration and other relevant conditions were controlled and water availability was deliberately changed, the results can support a more specific conclusion about water under those conditions. Scientific caution does not mean refusing to infer; it means connecting claims to evidence and method.
See P5 Fair Tests: Changed, Measured and Controlled Variables for dedicated investigation-design practice.
How to check a line graph for misleading presentation
Look for a vertical axis that does not begin at zero, unequal grid spacing, categories that are not in a natural order, or a legend that changes the meaning of colours. The graph may still be mathematically valid, but the appearance can tempt an exaggerated interpretation. Always report the actual numerical difference.
A graph may also show only part of the observation period. A rising trend for the first ten minutes does not guarantee it continues unchanged for the next hour. A pupil should distinguish interpolation within recorded data from prediction outside the measured range, especially when choosing among MCQ options.
If there are two or more curves, identify each curve using the legend and compare readings at the same horizontal-axis value when that is what the question asks. Do not compare a point from minute five in one experiment with minute thirty in another unless the task explicitly requests it.
Missing data should not be silently filled with imagined numbers. A prediction may be reasonable when asked for one, but it should be labelled as a prediction and justified using the observed pattern rather than presented as an official measurement.
What parents can practise without a new assessment book
Use a household chart of laundry drying times, a simple weather record, the growth of a potted plant or a fictional table of seed germination. Ask “What is measured?”, “Which group changed?”, “What is the difference?”, and “What cannot we conclude?” Keep the task about reasoning rather than diagnosing health or making universal claims from a few observations.
For example, write three days of imaginary rainfall readings on paper: 4 mm, 0 mm and 9 mm. Ask the pupil which day recorded the most rainfall, by how much compared with the first day, and why the values alone do not establish the reason for the weather differences. This teaches precise reading without requiring an advanced atmospheric science lesson.
Let the child design one misleading graph as an educational thought experiment: start a vertical axis at 90 and draw bars for 92 and 97. Then ask how the visual impression differs when the axis begins at zero. The ethical lesson is that accurate presentation matters, and a viewer must verify scales.
Use safe observations and teacher-provided data; avoid risky electrical experiments or unnecessarily stressing plants and animals. Scientific thinking can be practised through reading and explaining evidence.
A five-day revision sequence for data questions
Day one: tables. Read five small tables. Identify headings, units, conditions and changes. Practise distinguishing totals, remaining amounts and losses.
Day two: axes and scales. Work on two line graphs and two bar charts with different numerical intervals. Include one graph that begins above zero and one with a legend.
Day three: rates and proportions. Compare water losses over different time intervals and germination counts from unequal starting groups. Use only the arithmetic appropriate to the school’s level.
Day four: conclusions and fairness. Read three experiment descriptions and decide whether the design isolates the factor claimed. Write an observation sentence, then a separate evidence-based conclusion.
Day five: mixed and timed questions. Attempt new data tasks that combine calculation, trend description and explanation. Check units, axes, missing information and the command word. Correct the reasoning, not just the final number.
After the fifth day, retest the three most common error types with unfamiliar datasets. A child who can apply the method to new values is learning the skill rather than memorising a worksheet.
Frequently asked questions
What is the most important step before reading a Science graph?
Read the axis labels, units and scale. Determine what quantity the graph actually measures before interpreting its direction.
Why is “water remaining” different from “water lost”?
One measures the quantity still present and the other measures the quantity removed. In an open container they can change in opposite directions, so the graph label controls the meaning.
How do I compare results from different group sizes?
When the question asks for a relative success rate, use proportions or percentages where appropriate. Raw totals can be misleading.
What if the graph does not start at zero?
Read the actual marked values. A truncated axis can change visual impression but does not change the numerical difference.
Does a steep line always mean faster evaporation?
Only if the axes measure quantities for which the slope represents evaporation-related change, with a consistent scale. Read the variables and units first.
Can a clear pattern prove causation?
A pattern alone does not isolate a cause. Examine the experimental method, controlled variables and alternative explanations before forming a causal conclusion.
Are these official 2027 P5 school examination questions?
No. These are original learning exercises. The school’s current assessment scope, format and exam date remain authoritative.
Sources and next useful reading
Official curriculum source: MOE Primary Science Teaching and Learning Syllabus; national examination distinction: SEAB PSLE. The tables and data above are illustrative and were created as original educational examples, not measurements from an official school experiment.
Further reading: P5 Fair-Test Questions, P5 30 Science MCQs, P5 Common Exam Mistakes, and P5 Science Exam 2027 Syllabus and Revision.