Wait, What? Knowing Five Possible Causes Does Not Mean You Should Change All Five
A learner notices that several factors could affect a scientific outcome: amount, distance, time, material, temperature, size or another condition. The learner then designs one investigation in which three of those factors change together.
The experiment may produce a result, but it becomes difficult to tell which changed factor the result belongs to.
WHEN SEVERAL FACTORS COULD MATTER, FIRST CHOOSE THE ONE RELATIONSHIP YOUR INVESTIGATION IS TRYING TO TEST.
Quick Answer
START WITH THE SCIENTIFIC QUESTION → IDENTIFY THE OUTCOME TO OBSERVE OR MEASURE → LIST FACTORS THAT COULD AFFECT IT → CHOOSE THE ONE FACTOR THAT MATCHES THE QUESTION → SET USEFUL VALUES OR CONDITIONS FOR THAT FACTOR → KEEP OTHER RELEVANT CONDITIONS COMPARABLE → COLLECT EVIDENCE → CONCLUDE ONLY ABOUT THE RELATIONSHIP ACTUALLY TESTED.
The Exact PSLE Science Learning Job This Guide Owns
This guide owns one job: turning a multi-factor situation into one bounded PSLE Science investigation by choosing the deliberately changed variable that matches the scientific question.
It does not own all variable theory, all fair-test design or the science concept itself. It also does not mean that real scientific systems have only one cause. The learner job is about designing an interpretable test when several factors could affect the outcome.
Why This Matters in the Current PSLE Science Frame
For examination from 2026, PSLE Science assesses attainment in the 2023 Primary Science syllabus. SEAB’s published objectives include making predictions and formulating hypotheses, interpreting and analysing information, evaluating observations, information and methods, and communicating explanations and reasoning.
The MOE syllabus develops inquiry through asking questions, investigating relationships, observing, measuring and using evidence. This guide applies that official inquiry frame without inventing a compulsory answer template or claiming that every investigation must use the same wording.
Begin With the Outcome, Not With the Apparatus
A common mistake is to look at the available apparatus and ask, “What can I change?” A better scientific route starts with the question:
WHAT OUTCOME AM I TRYING TO EXPLAIN OR MEASURE, AND WHICH ONE FACTOR DO I WANT TO TEST AGAINST IT?
This keeps the investigation from becoming a collection of convenient changes with no clear relationship.
Four Roles to Keep Separate
| Role | Scientific question |
|---|---|
| Deliberately changed factor | What condition am I testing? |
| Measured or observed outcome | What evidence will show the response? |
| Relevant controlled conditions | What else must stay comparable because it could affect the outcome? |
| Context | What identifies the system but does not need to be changed for this question? |
Worked Example 1 — Several Factors Could Affect the Outcome
Suppose a learner wants to investigate what affects how quickly an observable process occurs. The learner thinks distance, amount of material and temperature could all matter.
That list is useful for generating possibilities. It is not yet an investigation.
If the question is, “How does distance affect the measured outcome?”, then distance becomes the deliberately changed factor. Amount of material and temperature are kept comparable if they could also affect the outcome.
A different investigation could later test temperature. Trying to change distance and temperature together would answer a different and harder question.
Worked Example 2 — The Factor Must Match the Question
A question asks whether the type of material affects an outcome. A learner varies both material type and thickness because thicker pieces are easier to handle.
Now the result could be associated with material type, thickness or both. The method has drifted away from the original scientific question.
The repair is not “keep everything the same” as a slogan. The repair is: keep thickness comparable because it is another scientifically relevant factor that could change the measured outcome.
Worked Example 3 — Choose a Factor That Can Actually Be Tested
A learner proposes to investigate “how healthy the specimen is” but has no defined way to change or compare health and no clear measurable outcome linked to the question.
The idea may be scientifically interesting, but the investigation is not yet operational. The learner needs a bounded factor that can be set or compared and an outcome that can be observed or measured consistently.
Worked Example 4 — One Factor Can Have Several Test Values
Changing one variable does not mean using only two set-ups. The same factor can be tested at several values or categories when the question is about a relationship across a range.
For example, testing one factor at low, medium and high values is still one deliberately changed variable. The variable is the scientific quantity or condition; the values are the levels used for it.
Worked Example 5 — When Two Factors Cannot Be Separated Easily
Sometimes changing one physical feature automatically changes another relevant feature. In that case, do not pretend the variables are independent. Either redesign the method, choose a different test, measure the linked change, or narrow the conclusion to what the coupled evidence can support.
A fair test is not created by renaming one of the changes “controlled”. It is created by a method that actually makes the comparison interpretable.
The Factor-Selection Test
- What exactly is the scientific question?
- What outcome would answer it?
- What factors could plausibly affect that outcome?
- Which one of those factors is the question asking me to test?
- Can I set or compare useful values of that factor?
- Which other factors could also affect the outcome and therefore need to stay comparable?
- Will the final evidence let me connect the changed factor to the measured outcome?
Do Not Choose the Variable From the Result You Want
A scientific investigation should not be designed backwards to produce a preferred answer. Choose the factor because it matches the question and produces discriminating evidence, not because you expect one condition to “win”.
Do Not Choose the Easiest Thing to Change if It Answers the Wrong Question
Convenience is not the scientific owner. If the question concerns material type, changing amount because the measuring cup is easier to adjust does not answer the stated relationship.
Variable, Value and Outcome Are Not the Same Thing
| Statement | Role |
|---|---|
| Distance | Possible variable |
| 10 cm, 20 cm, 30 cm | Values of that variable |
| Time taken for the observed change | Measured outcome |
| Same amount of material | Controlled condition if scientifically relevant |
The PSLE Science Reasoning Chain
READ THE SCIENTIFIC QUESTION → IDENTIFY THE OBJECT OR SYSTEM → IDENTIFY THE OUTCOME → GENERATE POSSIBLE FACTORS → SELECT THE FACTOR THAT MATCHES THE QUESTION → KEEP OTHER RELEVANT CONDITIONS COMPARABLE → COLLECT EVIDENCE → INTERPRET THE RELATIONSHIP → SELECT THE RELEVANT CONCEPT → EXPLAIN THE MECHANISM → CHECK THE CONCLUSION AGAINST THE METHOD.
Observable Failure Signatures
| Failure signature | Earliest likely weak link |
|---|---|
| Two or three relevant factors change together | No single tested relationship selected |
| The changed factor does not match the investigation question | Question-method alignment lost |
| The learner names a value such as “20 cm” as the variable | Variable confused with value |
| The learner knows what to change but not what to measure | Outcome not defined |
| Everything is kept the same, including the factor meant to be tested | Controlled and changed roles mixed |
| The easiest apparatus adjustment becomes the tested factor | Convenience replaced scientific purpose |
| Two factors are mechanically linked but the learner claims only one changed | Co-change ignored |
Earliest Weak-Link Diagnosis
- Can I state the investigation question in one sentence?
- Can I name the outcome separately from the factor?
- Can I list more than one possible influence before choosing?
- Can I explain why the selected factor belongs to this question?
- Can I identify the relevant conditions that must stay comparable?
- Can I say what evidence would support or weaken the proposed relationship?
- Would changing a different factor create a different investigation?
Misconception Repair — “One Variable at a Time Means the Whole World Has Only One Cause”
No. Real systems can have many interacting influences. A controlled investigation deliberately narrows the question so one relationship can be interpreted more clearly.
Misconception Repair — “The Variable Is the Number I Used”
The variable is the quantity or condition, such as distance. The numbers are values of that variable, such as 10 cm and 20 cm.
Misconception Repair — “Controlled Means Unimportant”
A controlled condition may be scientifically very important. It is kept comparable precisely because a difference in it could also change the outcome.
Practice Protocol: Question → Factors → One Test
- Write one original observable outcome.
- Generate three factors that could affect it.
- Choose one factor to investigate.
- Write two or more test values or categories for that factor.
- Identify two other relevant conditions to keep comparable.
- Choose the observation or measurement that answers the question.
- Predict what relationship might appear and explain why.
- Now change the chosen factor and show how the investigation question changes with it.
Unfamiliar Transfer Challenge
Take a familiar Science context and strip away the topic name. Describe only an object, three possible conditions and one measurable outcome. Can you design three different investigations by choosing a different factor each time while keeping the outcome clear?
If yes, you are learning the structure of scientific inquiry rather than memorising one experiment.
Delayed Independent Return
Three to five days later, use an unfamiliar investigation scenario. Without notes, identify the possible factors, select the one the question actually tests, define the outcome and name the relevant controls. If you can do this without relying on a familiar apparatus picture, the skill is transferring.
Variable-Selection Receipt
- I started from the scientific question.
- I identified the measured or observed outcome.
- I generated possible factors before choosing one.
- I selected the factor that matches the question.
- I separated the variable from its test values.
- I kept other relevant conditions comparable.
- I noticed if changing one factor automatically changes another.
- My conclusion refers only to the relationship the method actually tested.
Parent and Tutor Teaching Guide
Give the learner a simple outcome and ask for three factors that could affect it. Praise the generation of multiple possibilities first. Then ask, “Which one are we testing today?” This separates scientific creativity from controlled investigation design.
Next, deliberately design a bad test where two factors change together. Ask the learner to produce two different explanations for the result. If both fit, the learner can see why the method is confounded.
Finally, make the learner preserve the same outcome while changing which factor is tested. This reveals whether variable roles are understood structurally rather than memorised from one textbook set-up.
Useful Internal Routes
- PSLE Science Learning Guide
- Tell what the experimenter changed from what changed because of it
- Tell a variable from its values
- Decide which conditions actually need to stay the same
- Spot when changing one variable automatically changes another
- Keep an investigation consistent from question to conclusion
Authoritative References
- Singapore Examinations and Assessment Board — PSLE Science, for examination from 2026
- Ministry of Education, Singapore — Science Teaching & Learning Syllabus, Primary, 2023
Evidence and Boundary Note
This guide teaches inquiry design, not a compulsory PSLE wording formula. Some scientific investigations deliberately study several interacting factors, but a Primary 5/6 fair-test question often becomes more interpretable when the learner clearly identifies the one changed relationship being tested and keeps other relevant conditions comparable.
The Quiet Return
Good investigation design is not about changing as much as possible.
It is about changing exactly what the question needs, measuring the response clearly, and leaving the evidence with one scientific relationship it can genuinely speak about.
