How to Read a PSLE Science Investigation When More Than One Outcome Is Measured

Wait, What? Two Measurements Can Come From the Same Experiment and Still Be Answering Different Questions

A plant investigation records both height and number of leaves. After one week, Set-up A is taller than Set-up B, but both plants have the same number of leaves.

A learner looks at the table and says, “The results disagree.”

Not necessarily.

Two measured outcomes can describe different parts of the same system. They do not have to move together.

Height is not leaf count. Temperature is not time. Distance travelled is not speed. Mass remaining is not volume. A question can measure more than one outcome because each measurement gives a different view of what happened.

The difficult part is not reading more numbers. It is keeping each quantity attached to its own meaning.

Quick Answer

When a PSLE Science investigation measures two or more outcomes, label each measured quantity separately before interpreting the result. Ask what scientific question each measurement can answer, whether the outcomes are expected to be connected, and whether the question wants one outcome, both outcomes, or a relationship between them. Never use one measurement as a substitute for another unless the scientific mechanism justifies the connection.

Use this route:

IDENTIFY THE CHANGED CONDITION → NAME EACH MEASURED OUTCOME → KEEP UNITS AND OBJECTS ATTACHED → READ EACH OUTCOME SEPARATELY → ASK WHETHER THE OUTCOMES SHOULD BE CONNECTED → USE THE RELEVANT MECHANISM → ANSWER THE EXACT QUESTION → STATE WHAT THE OTHER OUTCOME ADDS OR DOES NOT ADD.

The Exact PSLE Science Learning Job This Guide Owns

This guide owns one job: how a Primary 5 or Primary 6 learner reasons through an investigation that measures more than one outcome from the same changed condition without mixing the quantities, double-counting evidence or treating different outcomes as interchangeable.

It does not replace fair-test reasoning, graph reading, measurement skills or the scientific concept behind the experiment. It solves a narrower but important problem:

One experiment can produce several legitimate measurements, and each measurement has its own evidence job.

Why This Matters in the 2026 PSLE Science Frame

For examination from 2026, PSLE Science assesses attainment in the 2023 Primary Science syllabus. The official assessment objectives include interpreting and analysing information, evaluating observations and methods, applying scientific concepts, and communicating explanations and reasoning.

Multiple-outcome investigations demand all of those skills. A learner must know what was measured, distinguish the quantities, interpret whether they support the same claim, and resist the temptation to combine numbers simply because they appear in the same table.

First Rule: Every Number Needs a Scientific Identity

Before analysing data, complete this sentence for every column:

This value is the ______ of ______ measured in ______.

For example:

the height of the plant measured in centimetres;
the number of leaves on the same plant measured as a count;
the temperature of the water measured in degrees Celsius;
the mass of water remaining measured in grams.

This prevents a common reasoning error: comparing numbers that refer to different quantities as though they were on one scale.

Changed Condition Versus Measured Outcomes

Role	Question	Example
Changed condition	What did the investigator deliberately vary?	Amount of light received
Measured outcome 1	What first response was recorded?	Increase in plant height
Measured outcome 2	What second response was recorded?	Number of new leaves

The changed condition can be the same for both outcome measures. That does not make the outcome measures the same variable.

Why Scientists Measure More Than One Outcome

One measurement may capture one part of a system while another captures a different part.
One outcome may answer the main question while another checks a related effect.
Two outcomes may help distinguish competing explanations.
One measurement may be an indirect indicator while another is more direct.
Different outcomes may respond on different timescales.
A system can improve in one dimension while remaining unchanged in another.

More measurements do not automatically mean better evidence. Each measurement must have a clear scientific purpose.

Worked Example 1 — Plant Height and Leaf Count

Original practice data:

Set-up	Increase in height / cm	New leaves / count
A	5.2	4
B	3.1	4

What does the first outcome show? Plant A increased in height more than Plant B.

What does the second outcome show? Both plants produced the same number of new leaves during the observation period.

Do the results conflict? No. They measure different features of growth.

A weak conclusion says, “A grew more in every way.” The data do not support that. A better conclusion names the exact outcome: “A showed a greater increase in height, while both produced the same number of new leaves.”

Worked Example 2 — Distance Travelled and Time Taken

A toy vehicle is tested on two surfaces. The investigator records both the distance travelled before stopping and the time taken to stop.

Surface	Distance before stopping / cm	Time before stopping / s
P	140	4.0
Q	80	2.8

The distance and time are related to the same motion, but they are not the same measurement.

You can say the vehicle travelled farther and continued moving for longer on P under the tested conditions.

You cannot call 140 “faster” than 80 because those numbers are distances. If the question asks about speed, you need suitable distance-and-time reasoning rather than relabelling the larger number.

Worked Example 3 — Temperature and Mass Remaining

Two identical open containers begin with equal amounts of warm water. After the same time, an investigation records temperature and mass remaining.

Set-up	Temperature / °C	Mass remaining / g
A	42	91
B	46	94

These outcomes describe different processes or effects that can occur in the same system. Temperature change concerns thermal energy transfer. Mass loss can provide evidence of water leaving the container, such as through evaporation under the stated conditions.

Do not assume the lower-temperature sample must have lost more mass simply because both values are smaller. Use the actual measurements.

Worked Example 4 — One Outcome Changes, the Other Does Not

An investigation changes the number of identical layers placed over a light sensor. It records the light reading and the temperature beside the sensor.

The light reading decreases strongly as layers are added, while the measured temperature remains unchanged within the instrument’s resolution.

That is not automatically a failed experiment. It may mean the changed condition affected one measured outcome under the tested conditions but produced no detectable change in the other.

The conclusion must preserve both results.

Worked Example 5 — Two Outcomes Move in Opposite Directions

A design change causes a toy car to travel a shorter distance but stop in a longer time.

The numbers may feel contradictory only if you expected all outcomes to move in the same direction.

Instead, ask what each outcome means and whether the mechanism can produce that combination. If the science does not support the pair, re-check measurement and method. If it does, preserve the trade-off rather than forcing one result to match the other.

Worked Example 6 — Direct Outcome and Indicator

An investigation measures the mass of water remaining and also records the size of a wet patch on absorbent paper.

The mass measurement directly records one quantity. The wet patch is an indicator that may reflect liquid reaching the paper.

Do not combine the two as if they were duplicate measurements. Decide what each one can support and whether the indicator has a reliable relationship to the process being inferred.

The Three Relationships Between Multiple Outcomes

Relationship	Meaning	What to do
Independent aspects	Outcomes describe different features of the system	Report each separately.
Mechanistically linked	One process can influence both outcomes	Explain the common mechanism but preserve each quantity.
One is an indicator of another process	Outcome provides indirect evidence	State what was measured and what is inferred.

Do Not Average Different Outcomes Together

A height of 12 cm and a leaf count of 5 cannot be averaged into “8.5 growth units”. The quantities have different meanings and units.

Only combine values when the scientific or mathematical operation is meaningful for the same quantity.

Do Not Rank the Whole Set-Up From One Outcome Unless the Question Defines the Job

Suppose Set-up A has a higher temperature but lower mass remaining than B. Which set-up is “better”?

The question has no scientific answer until “better” is defined. Better at staying warm? Better at retaining water? Better at something else?

Attach every comparison to a specific outcome.

One Outcome Can Answer the Main Question While Another Checks the Story

If the investigation asks how a material affects heat retention, final temperature may be the main outcome. A second measurement might help interpret another part of the system, but it should not replace the temperature evidence unless the question asks for that relationship.

Ask:

Which outcome directly answers the stated investigation question?
What does the second outcome add?
Could the second outcome reveal an alternative explanation?
Is it a check, an indicator, or a separate response?

Two Outcomes Can Strengthen an Explanation Without Being the Same Evidence

If two distinct measurements are both predicted by the same mechanism and both move in the expected way, the combined pattern can make that explanation more credible.

But avoid double-counting. “The temperature fell” and “the thermometer reading decreased” are not two independent pieces of evidence if they describe the same measurement.

When the Outcomes Seem to Disagree

Check that you are comparing the same set-up and time point.
Check units and measurement labels.
Ask whether the outcomes are actually expected to move together.
Check whether one response happens faster than the other.
Check whether measurement resolution could hide a small change.
Check for method problems or uncontrolled conditions.
Only then decide whether the scientific explanation needs revision.

Do Not Invent a Relationship Just Because the Two Outcomes Correlate

If plant height and leaf number both increase across several days, that does not prove that increasing height directly causes new leaves to appear, or vice versa. Both may respond to shared growth processes and conditions.

Observed co-change is evidence of a pattern. A causal mechanism needs scientific justification.

Question-Reading Protocol for Multi-Outcome Investigations

What condition changed?
What is Outcome 1? Name object, quantity, unit and time.
What is Outcome 2? Do the same.
Which outcome does the question ask about?
Is the other outcome relevant evidence or background information?
What concept connects the changed condition to each relevant outcome?
Do the outcomes support one explanation, different explanations, or no justified connection?

The Outcome Map

For a difficult investigation, draw a small map:

CHANGED CONDITION → PROCESS A → OUTCOME 1
CHANGED CONDITION → PROCESS B / SAME PROCESS → OUTCOME 2

If you cannot justify the second arrow, do not invent it. The map is a test of causal understanding, not decoration.

Common Failure Mode 1 — Mixing Units

Failure signature: “A changed more because 40 is bigger than 12,” where 40 is seconds and 12 is centimetres.

Repair: compare values only within the same measured quantity.

Common Failure Mode 2 — Treating One Outcome as Proof of Another

Failure signature: “The plant is taller, so it must have more leaves.”

Repair: check the actual leaf-count data. A related process does not make two measurements interchangeable.

Common Failure Mode 3 — Calling Different Patterns a Contradiction

Failure signature: “Height increased but leaf count stayed the same, so the experiment is unreliable.”

Repair: ask whether the two outcomes measure different aspects that can change independently.

Common Failure Mode 4 — Answering With the Wrong Outcome

Failure signature: the question asks which set-up retained more water, but the learner discusses temperature because that column has larger numerical differences.

Repair: underline the requested measured quantity before reading the data.

Common Failure Mode 5 — Counting the Same Measurement Twice

Failure signature: “There are two pieces of evidence: the graph went up and the final value was higher.”

If the final value is simply part of the same graph, those may be two descriptions of one data source rather than two independent measurements.

Common Failure Mode 6 — Building a Mechanism From Numerology

Failure signature: “Outcome 1 doubled, so Outcome 2 should double too.”

Repair: use the scientific mechanism, not a pattern borrowed from another quantity.

The Earliest-Weak-Link Diagnostic

Learner error	Earliest weak link	Repair
Compares 20 cm with 8 leaves	Quantity identity lost	Label each number with object, quantity and unit.
Calls outcomes contradictory	Assumes all responses must co-vary	Ask what each outcome measures.
Uses Outcome 2 to answer an Outcome 1 question	Question target lost	Restate the required measured quantity.
Links outcomes without mechanism	Correlation mistaken for causation	Build separate causal paths.
Ignores second outcome completely	Evidence selection too narrow	Ask whether it tests the explanation or reveals a limit.
Invents a combined score	Different quantities collapsed	Keep measurements separate unless a valid operation is defined.

How This Appears in Tables

Read by column before reading across the row.

Identify the heading and unit for Column 1.
Compare that outcome across set-ups.
Repeat for Column 2.
Only after both are understood should you ask how they relate.

How This Appears in Graphs

A question may use two graphs for the same investigation. Do not assume the vertical axes show the same quantity or scale.

Check:

axis labels;
units;
time intervals;
whether both graphs refer to the same sample;
whether one graph is cumulative and the other interval-based;
whether a trend in one outcome is expected to match the other.

How This Appears in Diagrams

One outcome may be numerical while another is visible in a diagram—for example, measured height plus leaf appearance. Separate what is directly drawn from what is measured and what is inferred.

How This Appears in MCQ

Find which outcome the stem asks about.
Ignore attractive options that use the wrong measurement.
Check whether an option assumes two outcomes must move together.
Reject causal links that are not supported by the scientific relationship.
Use the second outcome only if it helps test the option scientifically.

How This Appears in Open-Ended Answers

A useful reasoning shape is:

The investigation measured ______ and ______. The data show ______ for the first outcome, while ______ for the second. This supports ______ because ______. The second outcome does/does not support the same conclusion because ______.

Use only the parts needed. This is not a compulsory PSLE phrase.

Method Evaluation: Does Each Outcome Have a Suitable Measurement?

If an investigation claims to measure “plant health” using only height, ask whether height alone captures the intended outcome. A broad idea may need a more precise observable measure.

Do not add extra measurements merely to look scientific. Each outcome should be relevant, measurable and interpretable.

When More Outcomes Make the Investigation Worse

when the outcomes are poorly defined;
when different quantities are measured inconsistently;
when the learner cannot tell which outcome answers the question;
when extra measurements introduce uncontrolled changes;
when irrelevant data distract from the main evidence.

A good investigation is not a collection contest. It measures what the question needs.

Evidence and Model Limits

Two outcomes moving together do not prove one causes the other.
One unchanged outcome does not prove nothing happened elsewhere in the system.
Different measurement resolutions can make one outcome look more variable than another.
Outcomes measured at different times may not be directly comparable.
A proxy or indicator must not be confused with the underlying process.
Several outcomes do not rescue a confounded experiment.

Practice Sequence

Take a two-outcome table and label every column with quantity, object and unit.
Write one conclusion for Outcome 1 only.
Write one conclusion for Outcome 2 only.
Decide whether the outcomes are mechanistically connected.
Write one statement that would wrongly mix them and explain why it is wrong.
Change the order of the columns and check whether your reasoning survives.
Change the surface context but keep the same two-outcome structure.
Return several days later with an unfamiliar investigation.

Unfamiliar Transfer Challenge

A mystery material is tested in two designs. The investigation records:

time taken for an object to reach the bottom of a ramp;
distance travelled after leaving the ramp.

Design A gives a shorter ramp time but also a shorter travel distance after leaving the ramp.

Your job is not to decide immediately which design is “better”. First identify what each outcome measures. Then decide what scientific mechanism could connect the design change to each outcome and what additional information would be needed to judge overall performance.

Delayed Independent Return

Three to five days later, take a fresh multi-column investigation and answer without notes:

What condition changed?
What exactly is Outcome 1?
What exactly is Outcome 2?
Which outcome answers the question directly?
Does the second outcome support, limit or simply add a different observation?
What mechanism connects each relevant outcome to the condition?
What claim would mix quantities incorrectly?

The Answer-Checking Receipt

Did I keep every measured quantity separate?
Did I keep units attached?
Did I answer using the outcome the question actually asks about?
Did I avoid assuming two outcomes must move together?
Did I explain any link between outcomes scientifically?
Did I distinguish direct measurement from an indicator?
Did I avoid double-counting one measurement?
Did I state what each outcome can and cannot support?

Useful Internal Routes

Parent and Tutor Teaching Guide

When a learner sees a table with two outcome columns, cover one column first. Ask them to interpret the visible outcome completely before revealing the second.

Then ask:

“What new question does this second measurement answer?”

If the learner immediately mixes the two, return to quantity identity: object, quantity, unit, time.

Use paired examples where the outcomes move together, move differently, and where one stays unchanged. This prevents the learner from forming the false rule that “all good evidence points in the same numerical direction”.

Finally, give a question with an irrelevant extra measurement. A strong learner should be able to say, calmly, “That result is real, but it is not needed for this question.”

Authoritative and Research References

The research references support broader principles of inquiry, measurement and evidence interpretation. They are not PSLE-specific marking schemes.

The Quiet Ending

An experiment can look at the same system through more than one window.

Your job is not to merge the windows.

Keep each measurement clear. Then decide what the views reveal together.

Small Group Tutorials