How to Decide Whether Two PSLE Science Data Sets Show the Same Relationship

Wait, What? Two Graphs Can Look Different and Still Show the Same Science

Graph A rises gently from left to right. Graph B rises much more steeply. One uses centimetres. The other uses millimetres. One covers five minutes. The other covers twenty.

Do they show different scientific relationships?

Not necessarily.

Visual appearance can change when the scale, units, range, starting value or representation changes. To decide whether two data sets support the same relationship, the learner must compare the scientific structure, not the shape at first glance.

Quick Answer

First identify the changed quantity and measured outcome in each data set. Align units and comparison conditions. Check the direction of the relationship, the tested range, important features such as plateaus or turning points, and what each method actually measured. Only then decide whether the data sets support the same relationship, a partly overlapping relationship, different relationships or an unresolved comparison.

The working chain is:

IDENTIFY VARIABLES → ALIGN UNITS → ALIGN CONDITIONS → COMPARE DIRECTION → COMPARE RANGE → COMPARE PATTERN FEATURES → CHECK METHOD → STATE THE COMMON RELATIONSHIP OR THE LIMIT.

Owned PSLE Science Learning Job

This guide owns one job: deciding whether two separate PSLE Science data sets represent the same underlying scientific relationship.

It does not own graph construction, scale reading, one-data-set trend interpretation, comparing the overall quality of two investigations or translating one data set between table and graph. Those are neighbouring jobs with their own owners.

This page is about cross-data-set alignment: can two different pieces of evidence be compared as evidence about the same relationship at all?

Why This Matters in the Current PSLE Science Frame

The 2023 Primary Science syllabus develops analysis of data and information presented in tables, bar graphs, line graphs, charts and diagrams so learners can infer patterns and relationships or explain findings. The 2026 PSLE Science assessment objectives likewise include interpreting and analysing information and communicating scientific explanations and reasoning.

Comparing relationships across two evidence sets is a natural extension of that work. It should not be treated as a special official question format. It is a durable inquiry skill that helps learners recognise scientific structure when the surface representation changes.

First Distinction: Same Topic Does Not Mean Same Relationship

Two data sets can both be about plants and still test different relationships.

Data Set A: amount of light → plant height increase.
Data Set B: amount of water → plant height increase.

The measured outcome is the same, but the changed condition is different. These are not the same relationship.

Likewise, two data sets can use different objects yet still test the same relationship. A cup of hot water cooling and a metal block cooling may both provide evidence about how temperature changes with time under their respective conditions. Whether they support the same scientific relationship depends on the exact question being compared and the conditions, not whether the objects look alike.

The Six Things That Must Be Aligned

Alignment	Question	Typical failure
Changed quantity	What differs across the conditions?	Comparing light effects with water effects as though they were one relationship
Measured outcome	What result is recorded?	Comparing height with mass without a justified connection
Units	Are the quantities expressed comparably?	Trusting the larger numeral when units differ
Conditions	Are relevant background conditions sufficiently comparable for the intended claim?	Ignoring a different time window or starting state
Range	Over what values was the relationship tested?	Claiming conflict when two studies cover different regions
Method	Did both methods measure the same scientific outcome?	Comparing an indicator with a direct measurement as if identical

Worked Example 1: Same Direction, Different Scale

Data Set A records the distance travelled by a toy car at ramp heights of 5, 10, 15 and 20 cm. Distances are 40, 62, 81 and 101 cm.

Data Set B uses ramp heights of 50, 100, 150 and 200 mm. Distances are 0.39, 0.61, 0.82 and 1.00 m.

At first glance, the numbers look completely different. Convert the units:

50 mm = 5 cm; 0.39 m = 39 cm.
100 mm = 10 cm; 0.61 m = 61 cm.
150 mm = 15 cm; 0.82 m = 82 cm.
200 mm = 20 cm; 1.00 m = 100 cm.

Now the two data sets are closely aligned. Both support the same tested relationship: over this range and under their stated conditions, greater ramp height is associated with greater distance travelled.

The evidence does not prove that the relationship must continue forever beyond the tested range.

Worked Example 2: Same General Direction, Different Range

Data Set A tests a factor from 10 to 30 units and the measured outcome increases throughout.

Data Set B tests the same factor from 30 to 80 units. The outcome rises from 30 to 50 units and then levels off.

Do these data sets disagree?

No. They may describe different regions of one larger relationship. Data Set A never reached the plateau region. Data Set B did.

A careful comparison says:

Both data sets support an increase over their overlapping lower range; Data Set B additionally shows that the increase does not continue across its entire higher range.

This is more accurate than saying “the graphs are different” or “one experiment must be wrong”.

Worked Example 3: Similar Shape, Different Scientific Relationship

Graph P shows temperature decreasing with time. Graph Q shows mass decreasing with time. Both are downward-sloping lines.

They look similar, but they do not automatically show the same scientific relationship. One relates time to temperature. The other relates time to mass.

Visual slope direction is not enough. The axis quantities tell you what the relationship actually is.

Worked Example 4: Same Variables, Different Conditions

Two investigations both measure how much water remains after evaporation. Both vary exposed surface area. But one runs for 10 minutes in still air and the other runs for 30 minutes under moving air.

The two investigations still involve the same broad variables, but the background conditions and observation duration differ. You can compare whether each shows the same direction of relationship, but you should not directly compare raw amounts lost as though the conditions were matched.

One safe statement might be:

Both data sets show greater water loss for the larger exposed surface under their respective conditions. The size of the loss cannot be compared directly because the test conditions and duration differ.

Do Not Compare Slopes Until the Axes Are Comparable

A line can look steeper because the graph uses a compressed horizontal axis or an expanded vertical axis. The actual numerical changes may be identical.

Before comparing apparent steepness:

read both axis quantities;
check both units;
check the numerical interval represented by each grid step;
align the same time or condition interval;
compare actual changes, not visual angle.

If the axes cannot be aligned meaningfully, do not make a slope comparison.

Same Relationship Does Not Mean Same Numbers

Two investigations can support the same relationship while producing different numerical values.

Imagine two types of cloth tested for drying under otherwise different but stable environments. In both data sets, larger exposed area is associated with faster drying. The exact drying times differ because the materials or environments differ.

The shared relationship is about direction or structure, not identical output values.

Different Numbers Can Reveal a Boundary

Suppose Data Set A shows a clear increase while Data Set B shows almost no change. Before declaring contradiction, ask whether Data Set B used:

a much narrower range of test conditions;
a measuring instrument with coarser resolution;
a region where the relationship has plateaued;
a different object or system;
a condition that limits the underlying process.

The mismatch may teach you where the original relationship stops applying as simply as you first thought.

The Cross-Data-Set Comparison Protocol

Name the scientific object or system in each data set.
Name the changed variable in each.
Name the measured outcome in each.
Align units where possible.
Check starting/reference conditions.
Identify the tested range.
Describe each relationship separately before comparing them.
Compare direction, shape and major features such as thresholds, plateaus or reversals.
Check whether method differences change what the evidence means.
State what is common, what differs and what remains unresolved.

Earliest Weak-Link Diagnosis

Failure signature	Earliest weak link	Repair
“Both lines go up, so they show the same Science.”	Variables not identified	Read axis quantities before shape
“Graph B is steeper, so the effect is stronger.”	Scale not aligned	Compare numerical change over the same interval
“The results conflict because the values are different.”	Relationship confused with exact magnitude	Describe direction and conditions separately from values
“One graph has a plateau, so the other must be wrong.”	Tested range ignored	Map where each data set sits on the larger range
“Both are about plants, so I can compare them.”	Topic substituted for variable alignment	Name changed quantity and measured outcome

Misconception Repair: Matching Graph Shape Is Not Proof of the Same Mechanism

Even after two data sets show the same pattern, they may not share the same causal mechanism.

For example, two different systems may both show a rising outcome with time for completely different scientific reasons. The data pattern tells you what changed together. Scientific knowledge and the conditions are needed to explain why.

This keeps three layers separate:

Data: what was measured.
Relationship: how the quantities vary together.
Mechanism: why the relationship occurs.

What If the Two Data Sets Partly Agree?

Partial agreement is a legitimate scientific result.

You might say:

Both show an increase at low values, but only Data Set B reaches a plateau.
Both show the same direction, but the magnitude cannot be compared because the units or methods differ.
They agree over the overlapping range but provide no evidence about each other’s non-overlapping range.
The apparent disagreement cannot be resolved because one data set lacks necessary condition information.

A nuanced comparison is stronger than forcing every pair into “same” or “different”.

How This Helps in MCQ

An answer option may claim that two graphs show the same relationship because both rise, or different relationships because their slopes look different.

Before selecting:

identify variables;
align units;
compare the tested intervals;
look for range boundaries;
test the exact wording of the option.

How This Helps in Open-Ended Reasoning

A useful answer shape is:

Data Set A shows ______ as ______ increases. Data Set B also shows ______ over ______. Therefore, both support ______ under their stated conditions. However, ______ cannot be concluded because ______.

This is a thinking scaffold, not a compulsory marking phrase.

Practice Sequence

Stage 1: Compare two tables using the same variables and units.
Stage 2: Compare a table with a graph of a second data set.
Stage 3: Change the units while preserving the relationship.
Stage 4: Change the tested range so one data set reveals a plateau.
Stage 5: Use the same-looking graph shape for different variables and reject the false match.
Stage 6: Compare two unfamiliar data sets and state a bounded conclusion without hints.

Unfamiliar Transfer Challenge

Data Set X shows a fictional “response score” increasing from 2 to 7 as Condition R increases from 10 to 40 units. Data Set Y shows another response increasing from 20 to 70 as R increases from 0.10 to 0.40 of a different unit scale that is stated to be equivalent.

The names tell you nothing. Your job is to align the variable, unit conversion, direction and range. If those align, the learner can recognise shared structure even without a familiar Science chapter heading.

Delayed Independent Return Test

Several days later, give two new graphs with different visual scales. Ask the learner to write:

the changed variable in each;
the measured outcome in each;
the tested range;
the relationship in each;
whether the relationships can be compared;
what common conclusion is supported;
one limit of that comparison.

Cross-Data-Set Checking Receipt

Did I identify both variables?
Did I align units?
Did I check the reference and time window?
Did I compare actual values rather than graph steepness?
Did I check the tested range?
Did I distinguish relationship from exact magnitude?
Did I distinguish pattern from mechanism?
Did I state partial agreement honestly when appropriate?
Did I keep the conclusion inside the evidence?

Parent and Tutor Teaching Guide

When teaching this skill, do not begin with two complicated graphs. Start with two simple tables that use the same variables. Ask the learner to describe each relationship separately. Only after that should the two be compared.

Then change one surface feature at a time: units, graph scale, object, range or representation. This makes it possible to see exactly which change causes the learner to lose the underlying relationship.

When the learner says two data sets disagree, ask, “Where exactly do they disagree?” A precise answer might be “Data Set B has a plateau above 50 units while Data Set A does not test above 40.” That is much more useful than “the graphs are different”.

Useful Internal Routes

Authoritative References and Evidence Boundary

The comparison protocol here is a teaching scaffold, not an official PSLE response template. Two data sets can be meaningfully compared only to the extent that their variables, conditions and measurements support the comparison. Similarity of pattern does not by itself establish one shared causal mechanism.

The Quiet Return

Two graphs are not the same because their lines look alike. They are not different because their numbers look different.

Strip away the surface. Find the variables. Align the conditions. Follow the relationship. When the same scientific structure survives those checks, you have found something deeper than matching pictures: evidence that two different views may be telling the same scientific story.

Small Group Tutorials