Primary 5 Science Learning Guide | Cause, Correlation & Alternative Explanations
When two things change together, the first scientific question is not “Which one caused the other?” It is “What does the evidence actually allow us to say?”
Wait, What? A Pattern Can Be Real Without Proving Cause
Primary 5 students often see a graph or table where two quantities change together and immediately write a causal explanation. Sometimes that conclusion is justified. Sometimes it is not. A strong Science learner checks how the evidence was produced before deciding how strong the claim can be.
If an experiment deliberately changes one important condition while keeping other relevant conditions similar, the evidence can support a causal interpretation more strongly. If the data come from observation alone, several alternative explanations may remain possible.
Quick Answer
Correlation means two quantities or events vary together. Causation means changing one factor produces a change in another through a scientific mechanism. To support causation, look for a controlled comparison, a plausible mechanism, repeated evidence and the absence of strong alternative explanations. When the evidence is limited, use cautious language such as “is associated with”, “supports the idea that”, or “under these conditions”.
The Causal Reasoning Frame
- What changed?
- What outcome changed?
- Was the first factor deliberately changed?
- Were other relevant conditions controlled?
- Is there a scientific mechanism linking the two?
- Are there alternative explanations?
- Does the conclusion match the strength of the evidence?
Worked Example 1: Surface Area and Evaporation
Two dishes contain equal masses of water and are kept side by side for one hour. Dish A exposes a larger water surface than Dish B. Dish A loses more mass.
The design deliberately changes exposed surface area while keeping important conditions similar. There is also a plausible mechanism: evaporation occurs from the liquid surface. This gives stronger support for the claim that larger exposed surface area increased the evaporation rate under the test conditions.
Worked Example 2: Exercise and Pulse
A student observes that pulse rate is higher after running than after sitting quietly.
The relationship is consistent with a causal model: exercise increases the body’s demand for oxygen transport, so the heart pumps faster. But the strength of the conclusion still depends on how the observation was made. Was the same student measured before and after? Was the timing consistent? Was the activity comparable?
Alternative Explanations
An alternative explanation is another plausible reason the outcome could have changed. Scientific investigation becomes stronger when the method reduces these alternatives.
Suppose Dish A loses more water than Dish B, but Dish A is also placed near a fan. The difference might be due to surface area, airflow, or both. The experiment no longer isolates one cause cleanly.
Confounding Variables
A confounding variable is a relevant condition that changes along with the intended changed variable and can also affect the measured outcome. Primary 5 students do not need advanced terminology to understand the logic: if two important things change at once, it becomes difficult to know which caused the result.
Worked Example 3: Number of Cells and Bulb Brightness
A student compares Circuit A with one cell and one bulb against Circuit B with two cells and a different type of bulb. Circuit B is brighter.
Problem: both number of cells and bulb type changed.
Conclusion limit: the result does not isolate the effect of number of cells.
Improvement: use identical bulbs and change only the number of identical cells while keeping the circuit arrangement otherwise comparable.
A Control Is a Challenge to an Alternative Explanation
Controls are not decorative extras. They make rival explanations less plausible. Covering a plant-water container reduces direct evaporation from the water surface. Testing a circuit first with a known conductor checks whether the circuit itself works. Using several similar flowers reduces the chance that one unusual specimen determines the conclusion.
Evidence Strength
| Evidence situation | Claim strength |
|---|---|
| One observation, many uncontrolled differences | Weak causal support |
| Repeated observation of a pattern | Stronger relationship evidence |
| Controlled comparison with plausible mechanism | Stronger causal support |
| Repeated controlled comparisons | Greater confidence, within tested conditions |
Cautious Scientific Language
- Observed relationship: “As X increased, Y also increased.”
- Controlled causal claim: “Increasing X caused Y to increase under these test conditions.”
- Limited evidence: “The result supports the idea that X may affect Y.”
- Uncertain case: “The data are consistent with X affecting Y, but another variable may also have contributed.”
Cause Needs a Mechanism
A causal answer is stronger when it explains the middle. “More leaves caused more water loss” is incomplete. “More leaf area allows more water to leave through the leaves, so more water is drawn through the plant from the container” gives a mechanism connecting condition to outcome.
Worked Example 4: Pollinator Access
Flowers accessible to insects form more fruits than flowers protected from large insects.
A plausible mechanism is that insects transfer pollen between anthers and stigmas, increasing opportunities for later fertilisation. However, if the protective covering also changes humidity and airflow, those changes become alternative explanations that the design should reduce.
Cause Can Run Through Several Steps
In systems questions, the first condition can affect a later outcome through a chain:
exercise → faster oxygen use and carbon dioxide production → increased breathing and blood transport demand → higher breathing and pulse rates.
The longer the chain, the more important it becomes to keep each link scientifically necessary.
Reverse Causation
Sometimes students reverse the direction. A faster pulse does not cause the runner to exercise harder merely because the two occur together in the data. The scientific model and experimental sequence help determine which direction is plausible.
Third-Factor Explanations
If two variables change together, a third factor might influence both. For example, two plants may both lose more water on a hot, windy day. The relationship between temperature and water loss may be influenced by airflow as well. Controlled experiments are designed to separate such effects where possible.
Temporal Order Matters
A cause must occur before its effect. In reproduction, pollination happens before fertilisation. In a circuit, opening a switch happens before the bulb goes out. In an investigation, the changed condition must be applied before the resulting measurement is interpreted.
Worked Example 5: Water Droplets on a Cold Cup
Droplets appear outside a cold sealed cup.
Observation: droplets form on the outside surface.
Possible explanation A: water leaked through the cup.
Possible explanation B: water vapour in surrounding air cooled and condensed.
Evidence from an intact sealed container and the location of the droplets supports the condensation explanation more strongly than leakage.
How to Test Competing Explanations
- State Explanation A.
- State Explanation B.
- Ask what each explanation predicts.
- Design an observation or test where the predictions differ.
- Collect evidence.
- Decide which explanation fits the evidence better.
Prediction as a Causal Test
If larger exposed surface area causes faster evaporation, then increasing exposed area while keeping other relevant conditions similar should produce greater water loss over the same interval. A successful prediction strengthens confidence in the model.
No Difference Is Also Evidence
If changing a condition produces no clear difference, that matters. It may suggest the condition does not affect the outcome within the tested range, or the measurement may be too insensitive, or another limiting factor may dominate. “No difference” should not be forced into a positive causal story.
Small Samples and Biological Variation
One flower, one leaf or one person may not represent a wider group. Biological variation can create differences unrelated to the changed condition. Using several similar specimens can reduce the influence of one unusual individual.
Common Cause-and-Effect Mistakes
- Treating every graph relationship as proof of cause.
- Ignoring uncontrolled variables.
- Using “because” without a mechanism.
- Reversing cause and effect.
- Ignoring third-factor explanations.
- Overgeneralising from one specimen.
- Claiming certainty when evidence only supports association.
- Forgetting that controls challenge alternative explanations.
- Ignoring a result that shows no clear difference.
Answer Surgery: Causal Precision
Weak: “More cells made the bulb brighter because electricity increased.”
Better: “When the circuit used more identical cells while the bulb and arrangement were kept comparable, the bulb was brighter. The controlled comparison supports the conclusion that increasing the number of cells increased the energy supplied to the circuit under these conditions.”
Model Limit: Cause Is Often Conditional
A causal relationship may depend on conditions. Stronger airflow can increase evaporation when water is available and other conditions allow it, but not every environmental change produces the same effect in every system. Scientific statements are strongest when their conditions are visible.
Unfamiliar Transfer Test
Two plants under brighter lamps also receive more airflow and show greater water loss. List at least two possible explanations, identify why the evidence cannot isolate lamp brightness alone, and propose a stronger comparison.
Delayed Return Test
Several days later, inspect four short claims. For each, label it as observation, correlation, causal claim or uncertain interpretation. Then state one alternative explanation or control where appropriate.
Primary 5 Causal Reasoning Receipt
- I distinguish correlation from causation.
- I check whether the changed factor was deliberately controlled.
- I look for alternative explanations.
- I identify confounding variables.
- I use a mechanism to connect cause and effect.
- I respect temporal order.
- I can test competing explanations with different predictions.
- I use cautious language when evidence is limited.
- I keep causal conclusions within the tested conditions.
Parent and Tutor Teaching Guide
Whenever a child says “X caused Y”, ask two questions: “What other explanation is possible?” and “What did the experiment control?” These prompts develop healthy scientific scepticism without teaching the child to distrust every result.
Official Reference Route
Singapore Ministry of Education — Primary Science Teaching & Learning Syllabus 2023
This is an independent eduKate Sengkang learning guide. It uses age-appropriate causal reasoning to support scientific inquiry and evidence interpretation.
Continue the Primary 5 Science System
- Primary 5 Science Learning Hub
- Classification, Patterns & Evidence
- Definitions, Relationships & Mechanisms
- Question Deconstruction, Conditions & Constraints
The Quiet Return
Strong Science does not rush from pattern to certainty. It asks what changed, what was controlled, what else could explain the result and what mechanism connects the cause to the effect. That discipline makes conclusions stronger because it gives evidence the final say.
