A pupil may collect accurate measurements and still make the evidence difficult to understand.
Numbers written randomly across a page do not yet form a scientific data display.
A good table or graph organises evidence so the relationship between conditions and results becomes easier to see, check and explain.
This guide develops data construction inside the Primary 4 Science Learning Hub.
Quick Answer: What Makes a Good Data Display?
A strong table or graph should make clear:
- what was changed;
- what was measured;
- which values belong together;
- the unit for each quantity;
- the order or scale of values;
- the pattern or comparison the learner is meant to see.
A useful eduKate routine is:
VARIABLES → HEADINGS → UNITS → VALUES → SCALE → PLOT → CHECK → INTERPRET
This is a teaching routine, not an official MOE marking formula.
Start With the Variables
Before drawing lines or boxes, identify:
- changed condition;
- measured result.
Example:
Changed: object–torch distance.
Measured: shadow width.
These become the organising structure of the table or graph.
Constructing a Results Table
A clear table can be:
| Distance from torch / cm | Shadow width / cm |
|---|---|
| 10 | 18 |
| 20 | 14 |
| 30 | 11 |
The heading contains the property and unit.
Every row pairs one condition with one result.
Put Units in Headings Where Appropriate
Instead of writing:
Distance | Shadow
write:
Distance from torch / cm | Shadow width / cm
This reduces repeated unit writing and clarifies meaning.
Do Not Mix Units in One Column Without Warning
A distance column should not contain:
10 cm, 0.2 m, 300 mm
unless conversion is part of the task.
Use one consistent unit for easy comparison.
Order Values Logically
Possible orders include:
- increasing time;
- increasing distance;
- trial number;
- named conditions.
Logical ordering helps patterns appear.
Original Heat Table
| Time / min | Water temperature / °C |
|---|---|
| 0 | 72 |
| 5 | 65 |
| 10 | 60 |
| 20 | 52 |
This makes temperature change over time visible.
Original Plant Table
| Root condition | Wilting score after 2 days |
|---|---|
| Healthy | 0 |
| Moderately damaged | 1 |
| Severely damaged | 3 |
The observation scale must be defined elsewhere.
Categories can be used in tables even when the changed condition is not numerical.
When a Graph Helps
A graph is useful when the learner wants to see a relationship or trend visually.
Examples:
- temperature over time;
- shadow width across distance;
- plant height over days.
Not every table needs to become a graph.
Choose the Horizontal Axis
For simple investigations, the changed condition often goes on the horizontal axis.
Example:
x-axis: distance from torch / cm.
y-axis: shadow width / cm.
This convention helps represent how the measured result changes with the condition.
Choose the Vertical Axis
The measured result commonly goes on the vertical axis.
Time → temperature.
Distance → shadow width.
Water amount → plant height, if that is the designed investigation.
Axis Labels Need Units
Weak:
x-axis = distance.
Better:
Distance from torch / cm.
Units convert an abstract number into a scientific measurement.
Scale Intervals
A useful scale should:
- fit all data;
- use regular intervals;
- be easy to read;
- avoid unnecessary empty space;
- not distort the relationship.
Example: values 10, 20, 30 cm can use regular 5 or 10 cm intervals.
Do Not Change Scale Mid-Axis
An axis with intervals 0, 10, 20, 50, 60 is misleading unless a break is clearly shown and taught.
At Primary 4, use simple regular scales whenever possible.
Plot Carefully
Each point must pair the correct x-value and y-value.
Distance 20 cm with shadow width 14 cm should appear at the coordinate corresponding to 20 and 14.
Swapping rows creates a false relationship.
Line Graph or Bar Chart?
A simple line graph is often useful for continuous change such as time or distance.
A bar chart can be useful for separate categories such as wrapping materials.
The school’s expected graph type should be followed.
The key reasoning is to match the display to the data structure.
Original Bar-Chart Case
Materials:
- cloth: 13°C decrease;
- foam: 9°C decrease;
- no wrapping: 18°C decrease.
A bar chart can compare the three categories.
The bars represent temperature decrease, not final temperature.
Title the Display
A useful title can state the relationship:
“Effect of Object–Torch Distance on Shadow Width”
or
“Water Temperature Over Time”
A clear title helps the reader interpret the axes.
Tables Before Graphs
Collect data in a table first.
Then graph the values.
The table preserves the exact numbers.
The graph reveals the visual relationship.
Both representations have value.
Do Not Round Without Reason
If the thermometer reads whole degrees, record whole degrees.
If the ruler provides millimetre precision and the school expects centimetres, follow the instructed reporting method.
Do not invent decimal places.
Original Data Construction Workshop 1: Light
Measurements:
10 cm → 18 cm.
20 cm → 14 cm.
30 cm → 11 cm.
Table: place distance in first column, shadow width in second.
Graph: distance on x-axis, shadow width on y-axis.
Pattern: shadow width decreases as distance increases.
Original Workshop 2: Heat
Times: 0, 5, 10, 20 min.
Temperatures: 72, 65, 60, 52°C.
Best display: time-series table and line graph.
Interpretation: temperature decreases over time.
Original Workshop 3: Materials
Wrapping categories: cloth, foam, none.
Temperature decreases: 13, 9, 18°C.
Best simple visual: bar chart comparing categories.
Original Workshop 4: Plant Observation
Days 1–5 show plant height measurements.
A table preserves exact values.
A line graph can show growth trend over time.
Do not confuse height trend with explanation of why growth occurred.
Misleading Graphs
A graph can mislead if:
- axis labels are missing;
- units are missing;
- scale intervals are irregular;
- bars have unequal widths without reason;
- different conditions are swapped;
- the title claims more than the data show.
Truncated Axes
If an axis begins close to the measured values instead of zero, small differences can look visually large.
Primary 4 pupils can learn:
read numerical values as well as visual height.
Graph Does Not Prove Cause
A beautiful trend line still does not prove one variable caused another if the investigation was poorly controlled.
Data display shows relationship.
Experimental design supports causal interpretation.
Graphing Unexpected Results
Do not remove an unusual point just to make a smooth line.
Plot it if it is part of the recorded data.
Then investigate why it differs.
Missing Data
If one measurement was not taken, mark it clearly rather than inventing a value.
A gap in data is better than fabricated evidence.
Data Integrity
Never change numbers because they “look wrong”.
If a measurement is suspected to be incorrect:
- record the issue;
- repeat if appropriate;
- keep track of original data;
- explain any exclusion.
Primary 4 can begin learning honesty in data handling.
Constructing a Comparison Table
| Set-up | Changed condition | Measured result | Conclusion |
|---|---|---|---|
| A | Cloth wrapping | 13°C decrease | Greater cooling |
| B | Foam wrapping | 9°C decrease | Less cooling |
This table mixes raw result and interpretation, so it is better used as a learning summary than as the original data-recording table.
Know the purpose of the table.
Raw Data vs Summary Table
Raw data table:
records measurements directly.
Summary table:
may include calculated changes or conclusions.
Keep these roles clear.
Original Raw/Summary Example
Raw:
start 70°C, final 57°C.
Summary:
temperature decrease = 13°C.
Calculation transforms raw data into a derived value.
Data Display and Evidence
A graph does not replace the scientific explanation.
It organises evidence.
The learner still needs to connect the pattern to the relevant model.
Data Display and Communication
A good graph makes findings easier for another person to inspect.
Scientific communication depends on readers being able to reconstruct what was measured.
Common Construction Errors
- headings missing units;
- changed and measured variables swapped;
- irregular scale;
- wrong value plotted;
- labels missing;
- bar chart used without categories clearly named;
- line graph connected through invented missing data;
- anomaly removed;
- title overclaims cause;
- raw and calculated values mixed without clarity.
Original Practice Set
Question 1
What usually goes in the first column of a simple investigation table?
Question 2
Why should units appear in headings or axis labels?
Question 3
Why are regular scale intervals important?
Question 4
When might a bar chart be more suitable than a line graph?
Question 5
Why should unexpected values still be plotted?
Question 6
What is the difference between raw data and a calculated summary?
Question 7
Does a graph showing a trend prove cause?
Question 8
Why should missing data not be invented?
Practice Answers
1. Often the changed condition or independent grouping variable.
2. Units define what the numbers mean.
3. Irregular intervals distort visual comparison.
4. When comparing separate categories such as material types.
5. They are part of the evidence and may reveal a method problem or real variation.
6. Raw data are direct measurements; a summary may include differences, averages or conclusions derived from them.
7. No. Causal interpretation depends on investigation design.
8. Invented values create false evidence.
The Data-Construction Diagnostic
| If the learner… | Likely weak link | Repair |
|---|---|---|
| Swaps axes | Variable roles | Name changed/measured first |
| Misses units | Measurement meaning | Include unit in heading |
| Uses odd scale | Graph construction | Choose regular intervals |
| Deletes strange point | Data integrity | Plot, inspect, repeat |
| Cannot interpret own graph | Representation meaning | Describe relationship in words |
A 30-Minute Data-Construction Lesson
Minutes 1–5: identify variables.
Minutes 6–10: construct a results table.
Minutes 11–15: choose graph type and axes.
Minutes 16–20: select scale and plot.
Minutes 21–25: inspect labels and anomalies.
Minutes 26–30: describe the pattern and state one limitation.
What Parents and Tutors Can Ask
- “What does each column mean?”
- “Where are the units?”
- “Which variable belongs on which axis?”
- “Is the scale regular?”
- “Did you plot every recorded value?”
- “What pattern does the graph show?”
- “Does the graph show cause or only relationship?”
Continue the Primary 4 Science Series
- Primary 4 Science Learning Guide | Testable Questions, Hypotheses and Scientific Curiosity
- Primary 4 Science Learning Guide | Claims, Evidence and Reasoning
- Primary 4 Science Learning Guide | Scientific Communication and Presenting Findings
For reading existing displays, use Diagrams, Tables, Data and Patterns.
The Quiet Return
Good evidence deserves a good structure.
Name the variables. Build the table. Keep units visible. Choose a fair scale. Plot honestly. Then let the display reveal the relationship without pretending it proves more than the investigation supports.
